From 9615cfdcc32cb4f45569328c2aef41345bb19b25 Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Wed, 26 Feb 2020 11:03:03 +0100 Subject: [PATCH 01/43] Update README.md --- travis/tests/2_travis_likelihood_test/README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/travis/tests/2_travis_likelihood_test/README.md b/travis/tests/2_travis_likelihood_test/README.md index 7d5d13d..ee4090e 100755 --- a/travis/tests/2_travis_likelihood_test/README.md +++ b/travis/tests/2_travis_likelihood_test/README.md @@ -1,3 +1,3 @@ -# Travis CI Accuracy test №2 +# Travis CI Likelihood/Resources test №2 -This is a Travis CI test of the likelihood pipeline. It is a smaller version of the example №2. \ No newline at end of file +This folder contains data required for Travis CI tests of the likelihood and resources pipeline. It is a smaller version of the example №2. From 4a6d5f7f5292e3b4eb0683cc47906fc19c4c165d Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Wed, 26 Feb 2020 11:03:40 +0100 Subject: [PATCH 02/43] Update README.md --- travis/tests/1_travis_accuracy_test/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/travis/tests/1_travis_accuracy_test/README.md b/travis/tests/1_travis_accuracy_test/README.md index 8623755..cc31317 100755 --- a/travis/tests/1_travis_accuracy_test/README.md +++ b/travis/tests/1_travis_accuracy_test/README.md @@ -1,3 +1,3 @@ # Travis CI Accuracy test №1 -This is a Travis CI test of the accuracy pipeline. It is a smaller version of the example №1. \ No newline at end of file +This folder contains data required for Travis CI test of the accuracy workflow. It is a smaller version of the example №1. From ea15833ad0ac6fcef76b69739a5fc44ec40f3cfb Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Wed, 26 Feb 2020 11:09:13 +0100 Subject: [PATCH 03/43] [skip travis] Update README / Test skip Travis --- travis/tests/2_travis_likelihood_test/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/travis/tests/2_travis_likelihood_test/README.md b/travis/tests/2_travis_likelihood_test/README.md index ee4090e..16fb20f 100755 --- a/travis/tests/2_travis_likelihood_test/README.md +++ b/travis/tests/2_travis_likelihood_test/README.md @@ -1,3 +1,3 @@ # Travis CI Likelihood/Resources test №2 -This folder contains data required for Travis CI tests of the likelihood and resources pipeline. It is a smaller version of the example №2. +This folder contains data required for Travis CI tests of the likelihood and resources workflows. It is a smaller version of the example №2. From 88a4e75b718976109dda8dddff51ea109e29d838 Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Tue, 3 Mar 2020 13:31:07 +0100 Subject: [PATCH 04/43] [skip travis] Update README --- README.md | 66 +++++++++++++++++++++++++++---------------------------- 1 file changed, 33 insertions(+), 33 deletions(-) diff --git a/README.md b/README.md index 343ef81..d8bc91f 100755 --- a/README.md +++ b/README.md @@ -6,34 +6,34 @@ ## Overview -PEWO compiles a set of worflows dedicated to the evaluation of phylogenetic placement algorithms and their software implementation. It focuses on reporting placement accuracy under different conditions and associated computational costs. +PEWO compiles a set of workflows dedicated to the evaluation of phylogenetic placement algorithms and their software implementation. It focuses on reporting placement accuracy under different conditions and associated computational costs. -It is built on *Snakemake* + *Miniconda3* and rely extensively on the *Bioconda* and *Conda-forge* repositories, making it portable to most OS supporting Python 3. A specific phylogenetic placement software can be evaluated by PEWO as long as it can be installed on said OS via conda. +It is built on [Snakemake](https://snakemake.readthedocs.io/en/stable/) and [Miniconda3](https://docs.conda.io/en/latest/miniconda.html) and relies extensively on the [Bioconda](https://bioconda.github.io/) and [Conda-forge](https://conda-forge.org/) repositories, making it portable to most OS supporting Python 3. A specific phylogenetic placement software can be evaluated by PEWO as long as it can be installed on said OS via conda. -**Main purposes of PEWO :** +**Main purposes of PEWO:** -1. Evaluate phylogenetic placement accuracy, given a particular reference species tree (and corresponding reference alignment). For instance, one can run PEWO on different trees built for different taxonomic marker genes and explore which markers produce better placements. +1. Evaluate phylogenetic placement accuracy, given a particular reference species tree and corresponding reference alignment. For instance, one can run PEWO on different trees built for different taxonomic marker genes and explore which markers produce better placements. -2. Given a particular software/algorithm, empirically explore which sets of parameters produces the most accurate placements and for which computational costs. This can be particularly useful when huges volume of sequences are to be placed and one may need to consider a balance between accuracy and scalability (CPU/RAM limitations). +2. Given a particular software/algorithm, empirically explore which sets of parameters produces the most accurate placements and for which computational costs. This can be particularly useful when a huge volume of sequences are to be placed and one may need to consider a balance between accuracy and scalability (CPU/RAM limitations). -3. For developers, provide a basis to standardize phylogenetic placement evaluation and the establishement of benchmarks. PEWO aims to remove the hassle of re-implementing evaluation protocols that were described in anterior studies. In this regards, any phylogenetic placement developper is welcome to pull request new modules in PEWO repository or contact us for future support of their new productions. +3. For developers, provide a basis to standardize phylogenetic placement evaluation and the establishment of benchmarks. PEWO aims to remove the hassle of re-implementing evaluation protocols that were described in anterior studies. In this regard, any phylogenetic placement developer is welcome to pull request new modules in the PEWO repository or contact us for future support of their new productions. -**Procedures currently implemented in PEWO :** +**Implemented procedures:** * *Node Distance (ND)* : -This standard procedure was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. The reference tree is pruned randomly. For each pruning the pruned leaves are placed and accuracy is evaluated as the number of nodes separating expected and observed placements. +This standard procedure was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. The reference tree is pruned randomly. For each pruning, the pruned leaves are placed and accuracy is evaluated as the number of nodes separating expected and observed placements. * *Expected Node Distance (eND)* : An improved version of ND, which takes into account placement weights (e.g. Likelihood Weight Ratios, see documentation). * *Likelihood Improvement (LI)* : -Rapid evaluation of phylogenetic placements designed for developers and rapid evaluation of small changes in the code/algorithms. Following placement, a reoptimisation simply highlights better of worse results, in terms of likelihood changes. +Rapid evaluation of phylogenetic placements designed for developers and rapid evaluation of changes in the code and algorithms. Following placement, a re-optimization simply highlights better or worse results, in terms of likelihood changes. * *Ressources (RESS)* : -CPU and peek RAM consumption are measured for every steps required to operate a pylogenetic placement (including alignment in alignment-based methods and ancestral state reconstruction + database build in alignment-free methods). This procedure mostly intend to evaluate the scalability of the methods, as punctual analyses or routine placement of large sequence volumes do not induce the same constraints. +CPU and peek RAM consumption are measured for every step required to operate phylogenetic placement (including alignment in alignment-based methods and ancestral state reconstruction + database build in alignment-free methods). This procedure mostly intends to evaluate the scalability of the methods, as punctual analyses or routine placement of large sequence volumes do not induce the same constraints. -**Software currently supported by PEWO.** +**Supported software:** * **EPA**(RAxML) (Berger et al, 2010) * **PPlacer** (Matsen et al, 2011) @@ -41,23 +41,23 @@ CPU and peek RAM consumption are measured for every steps required to operate a * **RAPPAS** (Linard et al, 2019) * **APPLES** (Balaban et al, 2019) -Currently, (october 2019) there are no other implementations of phylogenetic placement algorithms. If you implement a new method, you are welcome to contact us for requesting future support or you can directly code a new snakemake module and contribute to PEWO via pull requests (see documentation for contribution rules). +Currently (March 20202) there are no other implementations of phylogenetic placement algorithms. If you implement a new method, you are welcome to contact us for requesting future support. You can also implement a new snakemake module and contribute to PEWO via pull requests (see the [documentation](https://github.com/phylo42/PEWO/wiki/Developer-instructions) for contribution guidelines). ## Wiki documentation -**An complete documentation, including a tutorial is available in the wiki section of this github repository.** +**An complete documentation, including a tutorial is available in the [wiki section](https://github.com/phylo42/PEWO/wiki) of this github repository.** ## Installation ### Requirements -Prior to installation, the following packages should be available on your system must be installed on your system: +Before installation, the following packages should be available on your system must be installed on your system: * Python >=3.5 -* Miniconda3. Please choose the installer corresponding to your OS : [Miniconda dowloads](https://docs.conda.io/en/latest/miniconda.html) +* Miniconda3. Please choose the installer corresponding to your OS: [Miniconda dowloads](https://docs.conda.io/en/latest/miniconda.html) * GIT -PEWO will basically look for the commands 'git' and 'conda'. Not finding these commands will cancel PEWO installation. +PEWO will look for the commands 'git' and 'conda'. Not finding these commands will cancel the PEWO installation. ### Installation @@ -73,7 +73,7 @@ chmod u+x INSTALL.sh ./INSTALL.sh ``` -After installation, load environement: +After installation, load environment: ``` conda activate PEWO ``` @@ -101,27 +101,27 @@ If the test is successful, you should produce the following statistics and image * summary_plot_eND_pplacer.svg * summary_plot_eND_rappas.svg -The content and interpretation of these files is detailed in the wiki documentation. +The content and interpretation of these files are detailed in the wiki documentation. Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/tree/master/examples/1_fast_test_of_accuracy_procedure). ## Setup your own PEWO analyses -**1. Activate PEWO environement :** +**1. Activate PEWO environment:** ``` conda activate PEWO ``` -By default, the latest version of every phylogenetic placement software is installed in PEWO environement. +By default, the latest version of every phylogenetic placement software is installed in PEWO environment. If you intend to evaluate anterior versions, you need to manually downgrade the corresponding package. -For instance, downgrading to anterior versions of PPlacer can be done with : +For instance, downgrading to anterior versions of PPlacer can be done with: ``` conda uninstall pplacer conda install pplacer=1.1.alpha17 ``` -**2. Select a procedure :** +**2. Select a procedure:** PEWO proposes several procedures aiming to evaluate different aspects of phylogenetic placement. Each procedure is coding as a Snakemake workflow, which can be loaded via a dedicated Snakefile (PEWO_workflow/\*.smk). @@ -134,30 +134,30 @@ Ressources | eval_ressources.smk | Given a reference tree/alignment and a set of Likelihood Improvement | eval_likelihood.smk | Given a reference tree/alignment, compute Node Distance and expected Node Distance for a set of software and a set of conditions. -**3. Setup the workflow by editing config.yaml :** +**3. Setup the workflow by editing config.yaml:** The file config.yaml is where you setup the workflow. It contains 4 sections: * *Workflow configuration* The most important section: set the working directory, the input tree/alignment on which to evaluate placements, the number of pruning experiments or experiment repeats (see procedures). * *Per software configuration* -Select a panel of parameters and parameters values for each software. Measurements will be operated for every parameter combinations. +Select a panel of parameters and parameter values for each software. Measurements will be operated for every parameter combination. * *Options common to all software* Mostly related to the formatting of the jplace outputs. Note that these options will impact expected Node Distances. * *Evolutionary model* -Very basic definition of the evolutionary model used in the procedures. Currently, only GTR+G (nucleotides), JTT+G, WAG+G and LG+G (amino acids) are supported. +A very basic definition of the evolutionary model used in the procedures. Currently, only GTR+G (nucleotides), JTT+G, WAG+G and LG+G (amino acids) are supported. -**4. Test your workflow :** +**4. Test your workflow:** -It is strongly recommended to test if your configuration is valid and matches the nalayses you intended. -To do so, launch a dry run of the pipeline using the command : +It is strongly recommended to test if your configuration is valid and matches the analyses you intended. +To do so, launch a dry run of the pipeline using the command: ``` snakemake --snakefile [snakefile].smk -np ``` -where \[snakefile\] is one of the subworkflow snakefiles listed in the table above. +where \[snakefile\] is one of the sub-workflow snakefiles listed in the table above. -This will list the operations that will be run by the workflow. It is also recommended to export a graph detailling the different steps of the workflow (to avoid very large graphs in "Accuracy" subworkflow, we force a single pruning). +This will list the operations that will be run by the workflow. It is also recommended to export a graph detailing the different steps of the workflow (to avoid very large graphs in "Accuracy" sub-workflow, we force a single pruning). ``` # to display the graph in a window @@ -167,13 +167,13 @@ snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot | dis snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot -Tsvg > graph.svg ``` -**5. Launch the analysis :** +**5. Launch the analysis:** ``` snakemake --snakefile [snakefile].smk -p --core [#cores] ``` -Note that the workflow can be launched on a grid environement such as qsub. -Refer to the snakemake documentation to learn how to configure the snakemake workflow for such environement. +Note that the workflow can be launched on a grid environment such as qsub. +Refer to the snakemake documentation to learn how to configure the snakemake workflow for such an environment. ## Contacts *B Linard, N Romashchenko, F Pardi, E Rivals* From bcad6c5cfdfbf1bc2238c5d54aedd367b9060699 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Wed, 4 Mar 2020 13:08:47 +0100 Subject: [PATCH 05/43] Update README.md --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index d8bc91f..7ba2c48 100755 --- a/README.md +++ b/README.md @@ -63,8 +63,8 @@ PEWO will look for the commands 'git' and 'conda'. Not finding these commands wi Download PEWO: ``` -git clone --recursive https://github.com/blinard-BIOINFO/PEWO_workflow.git -cd PEWO_workflow +git clone --recursive https://github.com/phylo42/PEWO.git +cd PEWO ``` Execute installation script: From 02b7d9ffc2d2c4383c763d2208065cdc8946f74f Mon Sep 17 00:00:00 2001 From: rivals Date: Fri, 27 Mar 2020 17:39:35 +0100 Subject: [PATCH 06/43] update of main README.md file --- README.md | 58 +++++++++++++++++++++++++++++++------------------------ 1 file changed, 33 insertions(+), 25 deletions(-) diff --git a/README.md b/README.md index f71d24d..848cdd8 100755 --- a/README.md +++ b/README.md @@ -14,13 +14,13 @@ It is built on [Snakemake](https://snakemake.readthedocs.io/en/stable/) and [Min 1. Evaluate phylogenetic placement accuracy, given a particular reference species tree and corresponding reference alignment. For instance, one can run PEWO on different trees built for different taxonomic marker genes and explore which markers produce better placements. -2. Given a particular software/algorithm, empirically explore which sets of parameters produces the most accurate placements and for which computational costs. This can be particularly useful when a huge volume of sequences are to be placed and one may need to consider a balance between accuracy and scalability (CPU/RAM limitations). +2. Given a particular software/algorithm, empirically explore which sets of parameters produces the most accurate placements, and at which computational costs. This can be particularly useful when a huge volume of sequences are to be placed and one may need to consider a balance between accuracy and scalability (CPU/RAM limitations). 3. For developers, provide a basis to standardize phylogenetic placement evaluation and the establishment of benchmarks. PEWO aims to remove the hassle of re-implementing evaluation protocols that were described in anterior studies. In this regard, any phylogenetic placement developer is welcome to pull request new modules in the PEWO repository or contact us for future support of their new productions. ## Wiki documentation -**An complete documentation, including a tutorial is available in the [wiki section](https://github.com/phylo42/PEWO/wiki) of this github repository.** +**An complete documentation, including a tutorial for each workflow, is available in the [wiki section](https://github.com/phylo42/PEWO/wiki) of this github repository.** ## Installation @@ -45,24 +45,24 @@ chmod u+x Miniconda3-latest-Linux-x86_64.sh ### Installation Download PEWO: -``` +``` bash git clone --recursive https://github.com/phylo42/PEWO.git cd PEWO ``` Execute installation script: -``` +``` bash chmod u+x INSTALL.sh ./INSTALL.sh ``` After installation, load environment: -``` +``` bash conda activate PEWO ``` You can launch a dry-run, if no error is throwed, PEWO is correctly installed: -``` +``` bash snakemake -np \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/1_fast_test_of_accuracy_procedure/run \ @@ -71,19 +71,20 @@ snakemake -np \ You can launch a 20 minutes test, using 2 CPU cores. -``` +``` bash snakemake -p --cores 2 \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/1_fast_test_of_accuracy_procedure/run \ --configfile examples/1_fast_test_of_accuracy_procedure/config.yaml ``` -If the test is successful, you should produce csv and svg files in the PEWO_workflow directory, for instance: -* results.csv -* summary_plot_eND_epang_h1.svg -* summary_plot_eND_pplacer.svg -* summary_plot_eND_rappas.svg +If the test is successful, you should produce the following statistics and image files in the `PEWO_workflow` directory: +* `results.csv` +* `summary_plot_eND_epang_h1.svg` +* `summary_plot_eND_pplacer.svg` +* `summary_plot_eND_rappas.svg` + The content and interpretation of these files are detailed in the wiki documentation. Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/tree/master/examples/1_fast_test_of_accuracy_procedure). @@ -119,7 +120,7 @@ Currently, (october 2019) there are no other implementations of phylogenetic pla **1. Activate PEWO environment:** -``` +``` bash conda activate PEWO ``` By default, the latest version of every phylogenetic placement software is installed in PEWO environment. @@ -127,14 +128,14 @@ If you intend to evaluate anterior versions, you need to manually downgrade the For instance, downgrading to anterior versions of PPlacer can be done with: -``` +``` bash conda uninstall pplacer conda install pplacer=1.1.alpha17 ``` **2. Select a procedure :** -PEWO proposes several procedures aiming to evaluate different aspects of phylogenetic placement. Each procedure is coding as a Snakemake workflow, which can be loaded via a dedicated Snakefile (PEWO_workflow/\*.smk). +PEWO proposes several procedures aiming to evaluate different aspects of phylogenetic placement. Each procedure is coding as a Snakemake workflow, which can be loaded via a dedicated Snakefile (`PEWO_workflow/\*.smk`). Identify the Snakefile corresponding to your needs. @@ -145,9 +146,10 @@ Ressources | eval_ressources.smk | Given a reference tree/alignment and a set of Likelihood Improvement | eval_likelihood.smk | Given a reference tree/alignment, compute tree likelihoods induced by placements under a set of conditions, with higher likelihood reflecting better placements. -**3. Setup the workflow by editing config.yaml:** +**3. Setup the workflow by editing `config.yaml`:** -The file config.yaml is where you setup the workflow. It contains 4 sections: + +The file `config.yaml` is where you setup the workflow. It contains 4 sections: * *Workflow configuration* The most important section: set the working directory, the input tree/alignment on which to evaluate placements, the number of pruning experiments or experiment repeats (see procedures). * *Per software configuration* @@ -155,39 +157,45 @@ Select a panel of parameters and parameter values for each software. Measurement * *Options common to all software* Mostly related to the formatting of the jplace outputs. Note that these options will impact expected Node Distances. * *Evolutionary model* -A very basic definition of the evolutionary model used in the procedures. Currently, only GTR+G (nucleotides), JTT+G, WAG+G and LG+G (amino acids) are supported. +A very basic definition of the evolutionary model used in the procedures. Currently, only GTR+G (for nucleotides), JTT+G, WAG+G and LG+G (for amino acids) are supported. **4. Test your workflow:** It is strongly recommended to test if your configuration is valid and matches the analyses you intended. To do so, launch a dry run of the pipeline using the command: -``` +``` bash snakemake --snakefile [snakefile].smk -np ``` -where \[snakefile\] is one of the sub-workflow snakefiles listed in the table above. +where `\[snakefile\]` is one of the sub-workflow snakefiles listed in the table above. This will list the operations that will be run by the workflow. It is also recommended to export a graph detailing the different steps of the workflow (to avoid very large graphs in "Accuracy" sub-workflow, we force a single pruning). -``` # to display the graph in a window +``` bash snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot | display - +``` # to produce an image of the graph +``` bash snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot -Tsvg > graph.svg ``` **5. Launch the analysis:** -``` +``` bash snakemake --snakefile [snakefile].smk -p --core [#cores] ``` -Note that the workflow can be launched on a grid environment such as qsub. -Refer to the snakemake documentation to learn how to configure the snakemake workflow for such an environment. + + +Note that the workflow can be launched on a grid environment such as [SunGridEngine](https://en.wikipedia.org/wiki/Oracle_Grid_Engine) or [SLURM](https://en.wikipedia.org/wiki/Slurm_Workload_Manager) (i.e., with `qsub` command). +Refer to the snakemake [documentation](https://snakemake.readthedocs.io/en/stable/executing/cluster-cloud.html#cluster-execution) to learn how to configure a workflow for such environments. + + ## Contacts *B Linard, N Romashchenko, F Pardi, E Rivals* + MAB team (Methods and Algorithms in Bioifnormatics), LIRMM, Montpellier, France. ## Licence From 59d67c48fb1fb3bcc5970d468a35d9eac0187c06 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Mon, 30 Mar 2020 17:41:39 +0200 Subject: [PATCH 07/43] Update README.md --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 848cdd8..a97bb91 100755 --- a/README.md +++ b/README.md @@ -78,7 +78,7 @@ snakemake -p --cores 2 \ --configfile examples/1_fast_test_of_accuracy_procedure/config.yaml ``` -If the test is successful, you should produce the following statistics and image files in the `PEWO_workflow` directory: +If the test is successful, you should produce the following statistics and image files in the `examples/1_fast_test_of_accuracy_procedure/run` directory: * `results.csv` * `summary_plot_eND_epang_h1.svg` * `summary_plot_eND_pplacer.svg` @@ -172,11 +172,11 @@ where `\[snakefile\]` is one of the sub-workflow snakefiles listed in the table This will list the operations that will be run by the workflow. It is also recommended to export a graph detailing the different steps of the workflow (to avoid very large graphs in "Accuracy" sub-workflow, we force a single pruning). -# to display the graph in a window +**to display the graph in a window** ``` bash snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot | display ``` -# to produce an image of the graph +**to produce an image of the graph** ``` bash snakemake --snakefile [snakefile].smk --config pruning_count=1 --dag | dot -Tsvg > graph.svg ``` From 0a4504e00740fddf79da45d201fb1da64706a47f Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Mon, 30 Mar 2020 17:42:48 +0200 Subject: [PATCH 08/43] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index a97bb91..5f85fa3 100755 --- a/README.md +++ b/README.md @@ -86,7 +86,7 @@ If the test is successful, you should produce the following statistics and image The content and interpretation of these files are detailed in the wiki documentation. -Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/tree/master/examples/1_fast_test_of_accuracy_procedure). +Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/wiki/IV.-Tutorials-and-results-interpretation). ## Setup your own analyses From 59dfdf5859256a304574b8d5a4463885172a4beb Mon Sep 17 00:00:00 2001 From: rivals Date: Thu, 2 Apr 2020 23:35:59 +0200 Subject: [PATCH 09/43] some more Readme updates --- README.md | 31 +++++++++++----- .../README.md | 23 ++++++------ .../README.md | 37 +++++++++---------- 3 files changed, 51 insertions(+), 40 deletions(-) diff --git a/README.md b/README.md index 5f85fa3..a8ee57c 100755 --- a/README.md +++ b/README.md @@ -32,7 +32,7 @@ Before installation, the following packages should be available on your system m * Miniconda3. Please choose the installer corresponding to your OS: [Miniconda dowloads](https://docs.conda.io/en/latest/miniconda.html) * GIT -PEWO will look for the commands 'git' and 'conda'. Not finding these commands will cancel the PEWO installation. +PEWO will look for the commands `git` and `conda`. Not finding these commands will cancel the PEWO installation. Below are debian commands to rapidly install them: ``` @@ -94,16 +94,22 @@ Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/wiki/IV.-T ### PEWO procedures * *Node Distance (ND)* : -This standard procedure was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. The reference tree is pruned randomly. For each pruning the pruned leaves are placed and accuracy is evaluated as the number of nodes separating expected and observed placements. +This standard procedure was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. The reference tree is pruned randomly. For each pruning, the pruned leaves are placed and accuracy is evaluated as the number of nodes separating expected and observed placements. + + +This distance measure between two placements was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. It is computed as follows: The reference tree is pruned randomly, which removes one sequence of the tree. The pruned sequence serves as query for placement against the pruned tree, but one knows the true solution (ie. the position of that sequence in the original tree). The resulting placement is compared to the true solution by evaluating the accuracy as the number of nodes separating the true and observed placements. This procedure is repeated with numerous prunings. One drawback: the running time. * *Expected Node Distance (eND)* : An improved version of ND, which takes into account placement weights (e.g. Likelihood Weight Ratios, see documentation). + + * *Likelihood Improvement (LI)* : -Rapid evaluation of phylogenetic placements designed for developers and rapid evaluation of small changes in the code/algorithms. Following placement, a reoptimisation simply highlights better of worse results, in terms of likelihood changes. +Rapid evaluation of phylogenetic placements designed for developers and rapid evaluation of changes in the code and algorithms. Following placement, a re-optimization simply highlights better or worse results, in terms of likelihood changes. * *Ressources (RESS)* : -CPU and peek RAM consumption are measured for every steps required to operate a pylogenetic placement (including alignment in alignment-based methods and ancestral state reconstruction + database build in alignment-free methods). This procedure mostly intend to evaluate the scalability of the methods, as punctual analyses or routine placement of large sequence volumes do not induce the same constraints. +CPU and peek RAM consumptions are measured for every step required to operate phylogenetic placement (including alignment in alignment-based methods and ancestral state reconstruction + database build in alignment-free methods). This procedure mostly intends to evaluate the scalability of the methods, as punctual analyses or routine placement of large sequence volumes do not induce the same constraints. + **Software currently supported by PEWO.** @@ -113,8 +119,12 @@ CPU and peek RAM consumption are measured for every steps required to operate a * **RAPPAS** (Linard et al, 2019) * **APPLES** (Balaban et al, 2019) -Currently, (october 2019) there are no other implementations of phylogenetic placement algorithms. If you implement a new method, you are welcome to contact us for requesting future support or you can directly code a new snakemake module and contribute to PEWO via pull requests (see documentation for contribution rules). + +PEWO can easily be extended to integrate new tools for phylogenetic placement, and new tools are welcome. +As of March 2020, these tools are the main software for phylogenetic placement. To the best of your knowledge, no other implementation of phylogenetic placement algorithms are available (with a conda package). + +If you implement a new method, you are welcome to contact us for requesting future support. You can also implement a new snakemake module and contribute to PEWO via pull requests (see the [documentation](https://github.com/phylo42/PEWO/wiki/Developer-instructions) for contribution guidelines). ## Analysis configuration @@ -133,17 +143,20 @@ conda uninstall pplacer conda install pplacer=1.1.alpha17 ``` -**2. Select a procedure :** +**2. Select a procedure:** PEWO proposes several procedures aiming to evaluate different aspects of phylogenetic placement. Each procedure is coding as a Snakemake workflow, which can be loaded via a dedicated Snakefile (`PEWO_workflow/\*.smk`). Identify the Snakefile corresponding to your needs. + + Procedure | Snakefile | Description --- | --- | --- -Accuracy (ND + eND) | eval_accuracy.smk | Given a reference tree/alignment, compute both the "Node Distance" and "expected Node Distance" for a set of software and a set of conditions. This procedure is based on a pruning approach and an important parameter is the number of prunings that is run (see documentation). -Ressources | eval_ressources.smk | Given a reference tree/alignment and a set of query reads, measures CPU/RAM consumptions for a set of software and a set of conditions. An important parameter is the number of repeats from which mean consumptions will be deduced (see documentation). -Likelihood Improvement | eval_likelihood.smk | Given a reference tree/alignment, compute tree likelihoods induced by placements under a set of conditions, with higher likelihood reflecting better placements. +Accuracy (ND + eND) | `eval_accuracy.smk` | Given a reference tree/alignment, compute both the "Node Distance" and "expected Node Distance" for a set of software and a set of conditions. This procedure is based on a pruning approach and an important parameter is the number of prunings that is run (see documentation). +Ressources | `eval_ressources.smk` | Given a reference tree/alignment and a set of query reads, measures CPU/RAM consumptions for a set of software and a set of conditions. An important parameter is the number of repeats from which mean consumptions will be deduced (see documentation). +Likelihood Improvement | `eval_likelihood.smk` | Given a reference tree and alignment, compute tree likelihoods induced by placements under a set of conditions, with a lower likelihood reflecting better placements. + **3. Setup the workflow by editing `config.yaml`:** diff --git a/examples/1_fast_test_of_accuracy_procedure/README.md b/examples/1_fast_test_of_accuracy_procedure/README.md index 90cbe80..fff857b 100755 --- a/examples/1_fast_test_of_accuracy_procedure/README.md +++ b/examples/1_fast_test_of_accuracy_procedure/README.md @@ -3,12 +3,11 @@ ## Overview This demo measures placement accuracy in terms of Node Distance (ND) -and expected Node Distance (eND)for a reference dataset +and expected Node Distance (eND) for a reference dataset of 150 16S-rRNA barcodes. -EPA-ng, PPlacer and RAPPAS are run using only their default parameters. -Only 2 pruning are launched, to rapidly produce results in less than -20 minutes. +EPA-ng, PPlacer and RAPPAS are run using their default parameters. +Only two prunings are launched to yield results in less than 20 minutes. A better analysis would require for >50 prunings to generate a wide range of topologies (1 leaf pruned, large clades pruned, ...). @@ -17,24 +16,24 @@ range of topologies (1 leaf pruned, large clades pruned, ...). ## How to launch Download pipeline. -``` +``` bash git clone --recursive https://github.com/phylo42/PEWO.git cd PEWO ``` Execute installation script -``` +``` bash chmod u+x INSTALL.sh ./INSTALL.sh ``` After installation, load environement. -``` +``` bash conda activate PEWO ``` Test workflow before launch. -``` +``` bash snakemake -np \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/1_fast_test_of_accuracy_procedure/run \ @@ -42,7 +41,7 @@ snakemake -np \ ``` Execute workflow, using 1 CPU core. -``` +``` bash snakemake -p --cores 1 \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/1_fast_test_of_accuracy_procedure/run \ @@ -51,16 +50,16 @@ snakemake -p --cores 1 \ ## Comments -In this example, 'workdir' and 'query_user' config flags are set +In this example, `workdir` and `query_user` config flags are set dynamically, as it is required they are passed as absolute paths. You could also set them manually by editing the config.yaml file before launch. Raw results will be written in -'examples/1_fast_test_of_accuracy_procedure/run'. +`examples/1_fast_test_of_accuracy_procedure/run`. Results summaries and plots will be written in -'examples/1_fast_test_of_accuracy_procedure/run'. +`examples/1_fast_test_of_accuracy_procedure/run`. See PEWO wiki for a more detailed explanation of the results: https://github.com/phylo42/PEWO/wiki/IV.-Tutorials-and-results-interpretation diff --git a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/README.md b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/README.md index 39476b3..6813634 100755 --- a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/README.md +++ b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/README.md @@ -3,47 +3,46 @@ ## Overview This demo measures placement accuracy in terms of Node Distance (ND) -and expected Node Distance (eND)for a reference dataset +and expected Node Distance (eND) for a reference dataset of 150 16S-rRNA barcodes. EPA-ng, PPlacer, RAPPAS and Apples are tested. -Only 10 pruning are launched and for a set of parameters in each program. +Only 10 prunings are executeed and for a set of parameters in each program. This analysis will require around 2 hours of computation. A better analysis would require for >50 prunings to generate a wide range of topologies (1 leaf pruned, large clades pruned, ...). -## How to launch - -Download pipeline. -``` +## How to run the pipeline +Download the pipeline. +``` bash git clone --recursive https://github.com/phylo42/PEWO.git cd PEWO ``` -Execute installation script -``` +Execute the installation script +``` bash chmod u+x INSTALL.sh ./INSTALL.sh ``` -After installation, load environement. -``` +After installation, load the environment. +``` bash conda activate PEWO ``` -Test workflow before launch. -``` +Test workflow before execution. +``` bash snakemake -np \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/2_placement_accuracy_for_a_bacterial_taxonomy/run \ --configfile examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml ``` -Execute workflow, using 2 CPU cores. -``` +Execute workflow using 2 CPU cores. +``` bash snakemake -p --cores 2 \ --snakefile eval_accuracy.smk \ --config workdir=`pwd`/examples/2_placement_accuracy_for_a_bacterial_taxonomy/run \ @@ -52,16 +51,16 @@ snakemake -p --cores 2 \ ## Comments -In this example, 'workdir' and 'query_user' config flags are set +In this example, `workdir` and `query_user` config flags are set dynamically, as it is required they are passed as absolute paths. -You could also set them manually by editing the config.yaml file -before launch. +You could also set them manually by editing the `config.yaml` file +before execution. Raw results will be written in -'examples/2_placement_accuracy_for_a_bacterial_taxonomy/run'. +`examples/2_placement_accuracy_for_a_bacterial_taxonomy/run`. Results summaries and plots will be written in -'examples/2_placement_accuracy_for_a_bacterial_taxonomy/run'. +`examples/2_placement_accuracy_for_a_bacterial_taxonomy/run`. See PEWO wiki for a more detailed explanation of the results: https://github.com/phylo42/PEWO/wiki/IV.-Tutorials-and-results-interpretation From 9dbadad357aaaf27392ef92aebf2111c81ca25bb Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Fri, 1 May 2020 17:19:27 +0200 Subject: [PATCH 10/43] Update README.md --- README.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/README.md b/README.md index a8ee57c..f3a37a1 100755 --- a/README.md +++ b/README.md @@ -40,6 +40,9 @@ sudo apt-get install git wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh chmod u+x Miniconda3-latest-Linux-x86_64.sh ./Miniconda3-latest-Linux-x86_64.sh +# when installation ask if you want to run conda init, answer yes +# after installation ends, reload bash so that conda belongs to your PATH +bash ``` ### Installation From 3a195989962c15517287077585128bc27c266965 Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Sat, 2 May 2020 17:48:30 +0200 Subject: [PATCH 11/43] Added IPython notebook to the example 6 --- .../ll_difference.ipynb | 1270 +++++++++++++++++ 1 file changed, 1270 insertions(+) create mode 100644 examples/6_placement_likelihood/ll_difference.ipynb diff --git a/examples/6_placement_likelihood/ll_difference.ipynb b/examples/6_placement_likelihood/ll_difference.ipynb new file mode 100644 index 0000000..0a98a90 --- /dev/null +++ b/examples/6_placement_likelihood/ll_difference.ipynb @@ -0,0 +1,1270 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# LogLikelihood difference\n", + "\n", + "This examples demonstrates how visualize the difference in LogLikelihood (LL) values obtained by LAC procedure for different software and parameter combinations." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Get the data\n", + "\n", + "First, run the example as described in the README. Then load the produced .csv file:" + ] + }, + { + "cell_type": "code", + "execution_count": 198, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
pruninggquerylengthLLsoftwaremax-strikesstrike-boxmax-pitchesredarkomega
000.018f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39708.145904epaNaNNaNNaNNaNNaNNaNNaN
100.108f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39705.898393epaNaNNaNNaNNaNNaNNaNNaN
200.01ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39779.279980epaNaNNaNNaNNaNNaNNaNNaN
300.10ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39778.547272epaNaNNaNNaNNaNNaNNaNNaN
400.0164463bb954837427bbaa8bff6de57ba4bbaccb8b-97172820-39791.397634epaNaNNaNNaNNaNNaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " pruning g query length \\\n", + "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", + "\n", + " LL software max-strikes strike-box max-pitches red ar k \\\n", + "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", + "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", + "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", + "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", + "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + "\n", + " omega \n", + "0 NaN \n", + "1 NaN \n", + "2 NaN \n", + "3 NaN \n", + "4 NaN " + ] + }, + "execution_count": 198, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "import pandas as pd\n", + "import matplotlib.pyplot as plt\n", + "import seaborn as sns\n", + "sns.set_style(\"whitegrid\")\n", + "\n", + "\n", + "df = pd.read_csv(\"run/likelihood.csv\", sep=\";\")\n", + "df.rename(columns={'likelihood': 'LL', 'ms' : 'max-strikes', 'sb' : 'strike-box', 'mp': 'max-pitches', 'o': 'omega'}, inplace=True)\n", + "df.head()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "What software was tested:" + ] + }, + { + "cell_type": "code", + "execution_count": 199, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "['epa' 'epang' 'pplacer' 'rappas']\n" + ] + } + ], + "source": [ + "print(df[\"software\"].unique())" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Calculate LL-difference\n", + "\n", + "Let's take a look at the LL values, produced by EPA: " + ] + }, + { + "cell_type": "code", + "execution_count": 200, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
pruninggquerylengthLLsoftwaremax-strikesstrike-boxmax-pitchesredarkomega
000.018f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39708.145904epaNaNNaNNaNNaNNaNNaNNaN
100.108f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39705.898393epaNaNNaNNaNNaNNaNNaNNaN
200.01ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39779.279980epaNaNNaNNaNNaNNaNNaNNaN
300.10ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39778.547272epaNaNNaNNaNNaNNaNNaNNaN
400.0164463bb954837427bbaa8bff6de57ba4bbaccb8b-97172820-39791.397634epaNaNNaNNaNNaNNaNNaNNaN
\n", + "
" + ], + "text/plain": [ + " pruning g query length \\\n", + "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", + "\n", + " LL software max-strikes strike-box max-pitches red ar k \\\n", + "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", + "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", + "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", + "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", + "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + "\n", + " omega \n", + "0 NaN \n", + "1 NaN \n", + "2 NaN \n", + "3 NaN \n", + "4 NaN " + ] + }, + "execution_count": 200, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df_epa = df[df[\"software\"] == \"epa\"]\n", + "df_epa.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 201, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "Text(0.5, 0.98, 'Log Likelihood of extended trees')" + ] + }, + "execution_count": 201, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "ax = sns.catplot(data=df_epa, x=\"g\", y=\"LL\", palette=\"muted\")\n", + "ax.set_axis_labels(\"g\", \"LogLikelihood\")\n", + "plt.subplots_adjust(left=0.1, top=0.9)\n", + "ax.fig.suptitle(\"Log Likelihood of extended trees\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Let's compare LL values query-by-query, taking the [EPA, g=0.1] run as baseline." + ] + }, + { + "cell_type": "code", + "execution_count": 202, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
queryLL
730425c2f49866ff47d2d6678dfad027644b20dc0d-1708795-39792.713826
12107a16500d8ec9d9ff772b852d5b12fcc67185e0c-1296378-39706.944825
907c4ff61b0b704a6e3e2aee860c9d10447abe20f-5648814-39715.289683
8107db0d6f4e199b4015f4c81eace27063926865ad-1635504-39707.965122
130dbcba127f79ff6ec9aa56486685d702f1da0ca7-4235956-39793.188953
\n", + "
" + ], + "text/plain": [ + " query LL\n", + "73 0425c2f49866ff47d2d6678dfad027644b20dc0d-1708795 -39792.713826\n", + "121 07a16500d8ec9d9ff772b852d5b12fcc67185e0c-1296378 -39706.944825\n", + "9 07c4ff61b0b704a6e3e2aee860c9d10447abe20f-5648814 -39715.289683\n", + "81 07db0d6f4e199b4015f4c81eace27063926865ad-1635504 -39707.965122\n", + "13 0dbcba127f79ff6ec9aa56486685d702f1da0ca7-4235956 -39793.188953" + ] + }, + "execution_count": 202, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "baseline = df_epa[df_epa[\"g\"] == 0.1][[\"query\", \"LL\"]].copy()\n", + "baseline.sort_values(\"query\", inplace=True)\n", + "baseline.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 203, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
pruninggquerylengthLLsoftwaremax-strikesstrike-boxmax-pitchesredarkomegaLL_baseline
000.018f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39708.145904epaNaNNaNNaNNaNNaNNaNNaN-39705.898393
100.108f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39705.898393epaNaNNaNNaNNaNNaNNaNNaN-39705.898393
200.01ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39779.279980epaNaNNaNNaNNaNNaNNaNNaN-39778.547272
300.10ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39778.547272epaNaNNaNNaNNaNNaNNaNNaN-39778.547272
400.0164463bb954837427bbaa8bff6de57ba4bbaccb8b-97172820-39791.397634epaNaNNaNNaNNaNNaNNaNNaN-39791.397634
\n", + "
" + ], + "text/plain": [ + " pruning g query length \\\n", + "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", + "\n", + " LL software max-strikes strike-box max-pitches red ar k \\\n", + "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", + "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", + "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", + "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", + "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + "\n", + " omega LL_baseline \n", + "0 NaN -39705.898393 \n", + "1 NaN -39705.898393 \n", + "2 NaN -39778.547272 \n", + "3 NaN -39778.547272 \n", + "4 NaN -39791.397634 " + ] + }, + "execution_count": 203, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df_epa = df_epa.merge(baseline, left_on=\"query\", right_on=\"query\", suffixes=('', '_baseline'))\n", + "df_epa.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 204, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
pruninggquerylengthLLsoftwaremax-strikesstrike-boxmax-pitchesredarkomegaLL_baselineLL difference
000.018f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39708.145904epaNaNNaNNaNNaNNaNNaNNaN-39705.898393-2.247511
100.108f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39705.898393epaNaNNaNNaNNaNNaNNaNNaN-39705.8983930.000000
200.01ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39779.279980epaNaNNaNNaNNaNNaNNaNNaN-39778.547272-0.732708
300.10ea4a29b8b59fb031b08cf1a7575658013dd45baa-98967690-39778.547272epaNaNNaNNaNNaNNaNNaNNaN-39778.5472720.000000
400.0164463bb954837427bbaa8bff6de57ba4bbaccb8b-97172820-39791.397634epaNaNNaNNaNNaNNaNNaNNaN-39791.3976340.000000
\n", + "
" + ], + "text/plain": [ + " pruning g query length \\\n", + "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", + "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", + "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", + "\n", + " LL software max-strikes strike-box max-pitches red ar k \\\n", + "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", + "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", + "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", + "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", + "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + "\n", + " omega LL_baseline LL difference \n", + "0 NaN -39705.898393 -2.247511 \n", + "1 NaN -39705.898393 0.000000 \n", + "2 NaN -39778.547272 -0.732708 \n", + "3 NaN -39778.547272 0.000000 \n", + "4 NaN -39791.397634 0.000000 " + ] + }, + "execution_count": 204, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df_epa[\"LL difference\"] = df_epa[\"LL\"] - df_epa[\"LL_baseline\"]\n", + "df_epa.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 205, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "Text(0.5, 0.98, 'Log Likelihood difference (vs. EPA, g=0.1)')" + ] + }, + "execution_count": 205, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "ax = sns.catplot(data=df_epa, x=\"g\", y=\"LL difference\", palette=\"muted\")\n", + "ax.set_axis_labels(\"g\", \"LogLikelihood\")\n", + "plt.subplots_adjust(left=0.1, top=0.9)\n", + "ax.fig.suptitle(\"Log Likelihood difference (vs. EPA, g=0.1)\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Other software\n", + "\n", + "Let's build this plot it for every tested software." + ] + }, + { + "cell_type": "code", + "execution_count": 206, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
pruninggquerylengthLLsoftwaremax-strikesstrike-boxmax-pitchesredarkomegaLL_baselineLL difference
000.010008f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39708.145904epaNaNNaNNaNNaNNaNNaNNaN-39705.898393-2.247511
100.100008f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39705.898393epaNaNNaNNaNNaNNaNNaNNaN-39705.8983930.000000
200.999008f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39709.165734epangNaNNaNNaNNaNNaNNaNNaN-39705.898393-3.267341
300.999998f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39707.929337epangNaNNaNNaNNaNNaNNaNNaN-39705.898393-2.030944
40NaN8f43808640b0e835dfd588c7c30a955c78cf2252-170840500-39707.929337pplacer6.03.040.0NaNNaNNaNNaN-39705.898393-2.030944
\n", + "
" + ], + "text/plain": [ + " pruning g query \\\n", + "0 0 0.01000 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "1 0 0.10000 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "2 0 0.99900 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "3 0 0.99999 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "4 0 NaN 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "\n", + " length LL software max-strikes strike-box max-pitches red \\\n", + "0 0 -39708.145904 epa NaN NaN NaN NaN \n", + "1 0 -39705.898393 epa NaN NaN NaN NaN \n", + "2 0 -39709.165734 epang NaN NaN NaN NaN \n", + "3 0 -39707.929337 epang NaN NaN NaN NaN \n", + "4 0 -39707.929337 pplacer 6.0 3.0 40.0 NaN \n", + "\n", + " ar k omega LL_baseline LL difference \n", + "0 NaN NaN NaN -39705.898393 -2.247511 \n", + "1 NaN NaN NaN -39705.898393 0.000000 \n", + "2 NaN NaN NaN -39705.898393 -3.267341 \n", + "3 NaN NaN NaN -39705.898393 -2.030944 \n", + "4 NaN NaN NaN -39705.898393 -2.030944 " + ] + }, + "execution_count": 206, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df2 = df.merge(baseline, left_on=\"query\", right_on=\"query\", suffixes=('', '_baseline'))\n", + "df2[\"LL difference\"] = df2[\"LL\"] - df2[\"LL_baseline\"]\n", + "df2.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 207, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 207, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "df_epa = df2[df2[\"software\"] == \"epa\"]\n", + "sns.catplot(data=df_epa, x=\"g\", y=\"LL difference\", palette=\"muted\")" + ] + }, + { + "cell_type": "code", + "execution_count": 208, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 208, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAFgCAYAAACFYaNMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3dd3wUdf748df29AosLYSWkNBBQDQUCe1oIsUTz4bKoZ6nwlmuqHjCeZ7nT887vVPBUyynfvUAFVEE6aAnAkooAWmBULKQ3rfN/P6ILiybkADZTHb3/Xw8fMjMfmbmvTD73s9+5lN0qqqqCCGEaHJ6rQMQQohQJQlYCCE0IglYCCE0IglYCCE0IglYCCE0EpQJ+MCBA1qHIIQQ9QrKBOxyubQOQQgh6hWUCVgIIQKBJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIJGAhhNCIUesARONzuVWWbT7D7iPldEuKZPqwlphN8l0r/M95NAv7js/QmSyEDZqCoVVHrUNq1iQBB6GFn55g+f8KANi6v4yTBXYe+nkHjaMSwc516gDl7z0GihsAx/6viL17EfrIOI0ja76kWhSE1n5f5LW9IasYRZHFr4V/ObI3eZIvAPZKnAe3ahdQAJAEHIQSY0xe2/FRRvR6nUbRiFChj0po0D5xliTgIDR7QlvCzDX/tGajjrsmttU4IhEKLH3GYGib6tk2dcvA2Lm/hhE1fzpVVYPut2l2djbp6elah6Gp8io3B09W0rl1ODGR0tQvmoaqqrhP7kdnCpMHcA0gn8wgFRVuoG+XaK3DECFGp9NhbJemdRgBQ5oghBBCI5KAhRBCI5KAhRBCI5om4I0bNzJ27FhGjx7NwoULfV5funQpgwcPZvLkyUyePJkPP/xQgyiFEMI/NHsI53a7mT9/Pm+88QZWq5Xp06eTmZlJ165dvcqNHz+eefPmaRSlEEL4j2Y14KysLJKTk0lKSsJsNjNhwgTWrFmjVThCCNHkNKsB22w2Wrdu7dm2Wq1kZWX5lFu1ahXffvstnTp14ve//z1t2rSp99x2u53s7OxGjVcIIS5VXeMSmnU/4BEjRjBx4kTMZjPvv/8+v/3tb3nrrbfqPc5isYT8QAwhRPOnWROE1WolLy/Ps22z2bBarV5l4uPjMZvNAFx//fXs2bOnSWMUQgh/0iwB9+rVi5ycHHJzc3E4HKxYsYLMzEyvMqdPn/b8ee3atXTp0qWpwxRCCL/RrAnCaDQyb948Zs2ahdvtZtq0aaSkpPD3v/+dnj17MnLkSN5++23Wrl2LwWAgNjaWp59+WqtwhRCi0clkPEIIoREZCSeEEBqRBCyEEBqRBCyEEBqRBCyEEBqRBCyEEBqRBCyEEBqRBCyEEBqRBCyEEBqRBByi9uRU8NnWAk4V2rUORQQ5pfQM9u9W4szZqXUozU6zng1N+MfiL07xf+tr5tkwGnQ8fnNHBqXFaByVCEbOY7spf/9xcDkAMPcbT+S4ezWOqvmQGnCIqah2s2TTGc+2y63y/jqbhhGJYFb99Yee5Avg+H4lSlmBhhE1L1IDDjGKoqIo3tN/2J2KRtGIQKE67VR+uQjnD1+jj29LxOjZGNuk1H+g87wmLlVBdTv9E2QAkhpwiImOMDK8T5zXvmuvaqFRNCJQVG1+F8d3n6NWFOM+vpfy/y5AVdz1HmcZMAnQebZNXQdiiGtd9wEhRmrAIeg30zvQp0sUR/OqGdAthv4p0VqHJJo5V473cmFqWQFKfi6GVh0veJw5LQPdzX/BeeB/GOLbYu49yo9RBh5JwEGkqMzJiXw7Ke0jsJjq/nFjNOgYOyCxCSMTgc7QuivuUz94tnVhkejirLhO7EMXHo0hoV2dx5o69MTUoWdThBlwJAEHiZXfFvDPj0/gcqvERBr40+2dSWkXoXVYIkiED78FpTgP15Ed6KJbEJ55B+VvPYz79BEAzH3HEjn+fo2jDDySgIOA3amwaMVJXO6ah2ulFW7eWHmKP9/ZOEs4ZR0u58sdhcRGGpl8dUtaxJoa5byi+XGfOYp9+6eoKoRdMcHTxKCPiCH6xgWozmowmqne9J4n+QI4vv8CS9+fYWybqlHkgUkScBCorHZTaffuyZBf2jhPmrMOl/P71w7xU8eJjVnFLHowDbNRnt8GG6X0DKVvPgiOKgAcu9cS88t/ej0005nCasqW5fseX5YPSAK+GPIpCgLx0Sb6dI7y2ndN7/hGOffq7YWc22vtdLGTnYfKG+XconlxZG/2JF8AnNU4szfVWtbcfZjXti48BlPHvv4MLyhJDThIPHpTMv+3/jRHbdUMSI1mUiN1LYuJ9L1FYmvZJwKfLsJ3NKQuIrbWsqZO/Yi8fh6OnavQhUURNng6Oos8c7hY8kkKEtERRmaNb9vo570uowWbdxVzurimSeOaPnGktpcPWjAypw/FvuMz3Cf2AWBom+pT0/Uqn3Il5pQrmyq8oCSrIot6OZwK3x8qJzbSSLckSb7BTFUVXMd2g6piTO6FTlfTSqmUF6GU5WOwdkanN3jKK6Vn0IXHoDNZtAo5oEkNWNTLbNLLZD0hQqfTY0ru7bWv+ptlVK1bDIoLfXwbomYsQGeyUP7fBbhP/gCWCCJG34VFBllcNEnAQog6KRUlnuQLoBSdonrze2A01yRfAHsllSv/hSnlSvThMqryYkgvCNFgpRUuTubL/MHBzl14AqWyFAClvMCTfH+ilJxGyT/mfZDLjlIis+pdLKkBhxBFUVnzXRH7cyvp2SmSa/rEY3cqnmHLhaVOlv8vn4oqN6OuSPB62PbeWhvvrrXhcqt0T47gj7d1Ijpcbp9golSVUf7BH2sewumNhA29kbCrf44+sT1KwXFPOXd5IWpFkdexupiWGFp1qvcaqsuBzmhu9NgDlaYP4TZu3MhTTz2Foihcf/31zJ492+t1h8PBI488wp49e4iLi+Nvf/sb7du3r/e8ofIQbu/RCr4/VE6nVhbeW2/jdLGTlHZhdGkTgVOBghI7reLMdLKGsz6rmB+OV1BccXbARnyUgaJyN1FhOqYNa8mHG/OprK55Xa+HUf3jiTAbiAjT8+7a017XnjGiFbeNadOk71f4h+v0ERz7vsKVdwj3wW+8XtO3644hoS1KiQ130UmobS7fyHgsAyahM0fgOpaFIb4tpo69MST1pHrbchy716K6XFBeCM4qiGuDqW0q5rQMzGkZTfQumyfNErDb7Wbs2LG88cYbWK1Wpk+fzvPPP0/Xrl09Zf7zn/+wf/9+5s+fz4oVK1i9ejUvvPBCvecOhQS84J0jfLWnVLPrd0+O4Lm7GzAfrGjWKlcvxP7tx5pdX98imdjZ/9Ls+lrTrA04KyuL5ORkkpKSMJvNTJgwgTVr1niVWbt2LVOmTAFg7NixfP311wRhr7mLduRUlabJF+BkvqP+QqJZUx1V2Lct1zQGJf8ojkPbNY1BS5o14tlsNlq3PjvG3Gq1kpWV5VOmTZuan7lGo5Ho6GiKiopISEi44LntdjvZ2dmNH3Qz8cNJ7VewqHK4gvrvOBToq0tppWp/L53cv5MqR3D3L6/rF3lQPkWxWCxB3QTRNUXh3a/2UlpZ/4oE/jJuUAvS0+ueA1YEhpKtSSgFudoFoDeSPPoXnkl+Qo1mTRBWq5W8vDzPts1mw2q1+pQ5deoUAC6Xi7KyMuLjG2eSmUBmMup56f5UBnWLJjrC95/QbATDj7t1OoiN0Hu2zxUfZSAmwkC4WU+rOCMRlppyZqOOmAg9EWE6n+N0wC2jrX4Z9iyaXszM5zGkDAKd4bxXzi4jhN5Y89/5TGG176+LTldzTFgUmMLQt+pE9B0vhGzyBQ1rwL169SInJ4fc3FysVisrVqzgueee8yqTmZnJsmXL6NevH1988QWDBw9Gp9PVccbQ0jLWzJMzO1NU5mTe4iMcPHl2Fqsws4HXHkwjOsL7n3fv0QoefOWgZ7uo3M2tY1pz44iaLz6HS2Huvw5w+FQ1DpdKYoyRBTM78sKSXHLP2Imw6LlvSnuu6SNfgsFAVRWq1r+J+9AOMBgxd8/E3HsUxqTuniHI5yt7/3Fch3fUbDiriZjyO1yHt+PIWgP1NGfoEzug5B8FZ83saVHTHwv59eE0qwEbjUbmzZvHrFmzGD9+POPGjSMlJYW///3vnodx06dPp7i4mNGjR/PGG2/w0EMPaRVus7V4VZ5X8gUorXTz0ZYzPmXzS3znCD533zfZpRw+Ve3ZLih1sfNQOa/O7cZrD6bxnz90l+QbRJzZW7Bv/7RmoIXLjiNrNahqncnXXXTqbPL9kWPHZ0ROmEP0rBfRRV743lDyj3r+rFaVYt/x2eW/iQCnaRvw8OHDGT58uNe+Bx54wPNni8XCP/7xj6YOK6AcPi/5/iT7WKXPvitSo4mJMHjajnU6GN777ArJTpdvDxOnW0Wn09GuhUy2Emzcpw/77rMdxpTcq9byOqMZdHrvmq6iUPrv+3DbfM/lxRzuPdcwgNtVe9kQIkORA1zfrlG17u/QyrddLTLMwLN3dWX0FfFc3SOWJ2/rRO8fJ3K3FTmodrhJjDaeU17PmCukxhusjOdPoK7TY+rYp87y+uhELP3Gnd1hsqCUnqkz+Zr7jsWUehXm3qOJuvmZmrZfz7UMmHuPuZzwg0JQ9oIIJbeMak21w83Kb4s8a8LFRBiYnFH7hOwdWoXxm+kdvPZ9taeEP7+bg/vHis2VaTF0S4ogs1881ngZNhqsTB37EPGzX1O99SN0BiNhV/+83mXmI372K0zdh6EUncLYPp3SV++qtZw+1kr4NTPR/zjJu/PQNqg+ZyUV1Y1SmAvWC18v2EkCDnBmk557Jydx18T2fJNdQnmVm6u6x9a6kkVd3vkyz5N8AbKOlPPoTcmYZN23oGfpPw5L/3H1FzyHqUNP+HGZeYO1s1cN2NAmFcuAiZhTr/JaIcN95qjPedxnjkHw9hZtEPmEBQmjQUdGzzjGDky8qOQLUHXegp4Op+KVkIWoS+TkRzAm9QCTBVPKIKKun4el10if5YlMna+oaT/20GHqOqBpg22GpAYcwBxOhd05FbSMNZFUS5vvhSiKyncHy8gvdZLZL85rsp1R/RMIM8t3swBVcePK3YPOEomxdRfPfnexDdeR7zC0SCL6lr/Wex5Dq45ETv0D1f/7LygKYYOuw9i2mz9DDwiSgAPUyXw7jyw6SEFpzZPkaUNbXtTgiD+/e5Qte0qAmodtvxzfhpMFDrq0DWfMFRce6i1Cg1JZStk7v/XM/WtKG0LU1N/jPPId5R/80dOLwTJ4GhGZd9R7PnO3qzB3u8qfIQccqeYEqP/bcNqTfAGWbj6DrahhE+QcOVXlSb4AFdUKObZqfn1de8YNSsRgkMEuAuw7PvOaeN25bzPOo7uo3vy+Vxcy+9aPUarKtAgx4EkCDlBFZd6DKlQVissb1q+y2unbwGt3yixzwtv5k67/tE91nbcqiuKWPr2XSBJwgGoV5909zGjQ0blNw9qB05IiSG0f7tnW62HClYmNGp8IfOYe13g9ONNFxGLqMgDLFRO9ypnSMtBHSX/xSyFtwAHqZIF3LcTlVjl8qrpBy8brdDqentWFL74tpKDUyfA+caS0C+7pAMXFM7ZPJ+oXT+H4/gt0lggsg6bU/L/3KPTRLXAe+hZDYhLm3iMv6rzuguM4D23H0KI9xk79Q3p+F0nAAajaoXDghPewTr0OEmNMDT5HhMXAlCEtGzs0EWRMyb19lqkHMHXqi6lT31qOuDDnoW2Ufzi/ptkCsPQfT8TP7r3sOAOVNEEEoK/21Ay4OFefLlG0iG14AhZCC9Vf/9eTfAHs361EqSi5wBHBTWrAAcit+D4w69Wp9jkhANbvLOKtVXlU2hXGD0rgltGtQ/pnn7gwtbqCipUv4Tz4LYaWyUSM/ZVXH+DLOvd5S9yjqqBqt7CA1qQGHICu7hFLq7iztd3ocAOj+tf+EORUoZ1nPzjGqUIHJRUu3lt3mnXfFzdVqCIAVa59HefejeCown1iHxVLn0JtpKWLwgZO5tzJ3k3dh6GPCt1+51IDDkCRYQb+8etU1uwowulWyOwXT8vY2ifN2ZtTgXLeZ2d3TjmZ/XwTttutcqbEQcs4Mwa91JBDlevYbq9tpdiGUpqPIbbVZZ/bnD4UXVQCjj0bMLRLw9JjeP0HBTFJwAEqNtLI1KEXfojmVlR2Hi732d+tvW+Ph+yjFTz17lEKSp20jDXx6E0dG9SjQgQffXQiSuHxszuiEtBHN043RefRLCo++itqRRH6w9swtuqI0dq5Uc4diKQJIoh9sa2Q1du9O9OP6h/HqFqGGv/jo+MUlNYM7jhT4uTFj477lBHBT6kqw3Vin9c+c8pgdPrz14wDpbIE+/YV2LO+RHVW+7xem8rP/uEZ4KEU26j64pXLDzqASQ04iO3NqfDZNygtttbmhdzT3h+gY6cb9oESwcWddwjOG+mmluX7lFNK8yl94wHUiprnCfZvlhF9+99qVs2og+pyohSd8r5ewbE6SocGqQEHsbQO3k0Iej2k1tL8ADAoLeaC2yI0GKyd4bwkamiX5lPOvnOVJ/kCuM/k4DzwzQXPrTOafFbhMHUddBnRBj6pAQexcYMSOXa6mlXbCokKN3LHz9rUucLF3GlJxEaeYu/RCronR3LHuDZNHK1oDvQRMURe9whVqxehlBdg7jGCsCun+hZUauk6Vtu+80ROfpiqta/jOnUAU3Jvwq+5rRGiDlw6VVWDbhaW7Oxs0tNDfKp9IfzIXZxH2esPoP64zJA+vi0xs15EZ7q4ealDndSAhRAXzRDXmphZL+HYvQ5MYZh7jpDkewkkAQeJ/BInW/YUkxBt4qrusRhlTl/hZ/qYloRd/XPPtqq4cR7YilJyGlPqYAxxVg2jCwySgIPAkbwqHnzloGdtt75donh6VuMMHRWioSqW/QXn/q8AqFq/mOibnsZYywM8cZb0gggCn3yV77Ww5veHytl3zLcLmhD+4s7P9SRfAFwOqr9Zpl1AAUIScBCobXIeVy37hPCbS+wVEeo0aYIoLi5m7ty5nDhxgnbt2vHCCy8QGxvrUy49PZ3U1FQA2rRpwyuvhPaombpMuLIF63cW43TVJN1uSRH0SI7UOCoRSgytOmLs1A/Xke9qdugNWAZM0jaoAKBJN7S//vWvxMXFMXv2bBYuXEhJSQkPP/ywT7l+/frx3XffXfT5Q7EbWu7pajbuKiY+2kRm3zjCzL5DR4XwJ9XlwLFnA0rJacxpGRhaddQ6pGZPkyaINWvWcN111wFw3XXX8eWXX2oRRlBJahXGTSNbM35QoiRfoQmd0Yylz2jCh90kybeBNGmCKCgooFWrmqntWrZsSUFBQa3l7HY7U6dOxWg0Mnv2bEaNGtWUYQaNM8UOXl95ipy8aq5IjebW0a0xm6T5Xwit+S0Bz5w5k/x830k85syZ47Wt0+nqXJ1h3bp1WK1WcnNzue2220hNTaVDhw71Xttut5OdnX1pgQehv61wkltQ09KUY6vmTH4B1w2UHohCNJW6mkT99ilcvHhxna8lJiZy+vRpWrVqxenTp0lIqH1GfKu1piN3UlISgwYNYu/evQ1KwBaLJeTagOtSWOYkt2Cv176Dp43y9yNEM6DJ79DMzEw++ugjAD766CNGjvRd1rqkpASHwwFAYWEhO3bsoGvXrk0aZzCIiTASG+n9PZvUSoaMCtEcaJKAZ8+ezZYtWxgzZgxfffUVs2fPBmDXrl08+uijABw6dIhp06Zx7bXXctttt/HLX/5SEvAlMBp0zJnanujwmgdz7VtamDWurcZRCSFAZkMLGp9vLeDdtTZcbpXJV7dgxgjvcfgOp0JhmRNrvFlWRBaimZAnMUHg4MlK/rHs7BJCb67Ko1PrcK5MPzuputmkp3WCRYvwhBB1kL5IQWDPEd95H3bn+C7GKYRoXiQBBzhVVYmLMvnsT5MVjYVo9qQJIoDZihw8/sZhcs/YMRp0GPU1/aqvvboFGT3jtA5PCFEPScAB7M1Vp8g9U7OCrcutYtDreed36USGX94/a9bhcj7acgaA6zJa0rtz1GXHKoKf6nJQ/dUHuI7txtCuG+EZM9CZw7UOq1mTBBzATuY7vLbtToWCMtdlJeBjp6t59PXDuNw1nWO27ivlH79OpXMb+SAJcBfbUAqPY2yXjs7i3cxVueoVHN9/AYDr2C6UYhtRU37neV1VFdx5h9BHxqOPadGkcTdXkoAD2FU9Yth/vNKzndTSQlLLy+vp8PXeEk/yBXAr8PCrB3nxvlTaJkovilBW/c0yqta+DqqCLiyKqBv/hLFNiud1R/Ymr/LOfVtQVQWdTo9Slk/Zu4+iFBwHnZ6wjBsIH3ZzU7+FZkcewgWw6cNacfvY1nRrH8GIvnHMn9npsvv4WuN8l62vtCt8uOH0ZZ1XBDbVXknVhrdBrVl5Ra0ur9k+hyHWu++5PqYlOl1Niqn++r81yRdAVaje/D7uYpv/A2/mJAEHMINex8+vsfLCvSk8ckNyo/TzHdIrjh7Jvj0oCstcl31uEbhUeyW47N77Koq8tsNHz0YX9uNCAOZwIsbe7XlNKTn/C1xFKT3jj1ADiiRg4cVo0PHM7K60jvfu2jaqf7xGEYnmQB/TAmNyb6995l6ZXtum5N7E3vcW0bc9R+w9r6G6XThzdqKqKqb0oeedr6Us2Im0AYvz7M+tZOu+UqYNbcWJAjsFpU6G94kno4fvklEitERNe4zqb5bizs/FlDIQS+/RPmV0pjD0sa0oWzzXU+s1dupP1Iz5oCo4dq9HH51IWMYN6AySfuRvQHhs3l3Mn989yk+zg1zTJ44//KKjpjGJ5kMXFkn48FvqLWffvsKrycF1ZAeunJ1Yeo3E0st35sNQJk0QwmPZ5jOcOzXThqxiCkud2gUkApJq9x0aX9s+IQlYnMOg9+5BoQP0coeIi+DOP4a74LjXPl1MS0xdrtAoouZNPl7CY/qwVhjOuSPGDkyodZ4JIWqjuhyU/ecPZ5emB0xdBxFz23PoTLIIQG2kDVh4DEqL4eUHuvHt/jLat7QwsFu01iGJAOI6sc+na5oK6CJlXpK6SAIWXpJahcmSReKS6ONag07vGawB4Dq0neLnbyDsyqmED/2FhtE1T9IEIYRoFIbYVoQNvwX059TrVDc4qqje9B+cR7O0C66ZkgQshGg04Vf/nNj73sLcb5zPa+5TBzSIqHmTBCyEaFT6yFjM5418A3xG0okGJuBt27axZMkSoGaJ+NzcXL8GJYQIbKaOfQgfczf6WCv6+LZETHjAa+Y0UaPeVZFfeukldu/ezZEjR/jiiy+w2Ww88MADvP/++00V40ULxVWRhRCBp94a8OrVq3n55ZcJD6+ZkNtqtVJRIaNaAs3Bk5V8vrWA42fs9RcW4gLcRaewf/8FrhP7tA4l4NXbDc1kMqHT6TzzzFZWVtZzhGhulmw6zWufnQJAr4NHZnRgeG+Z3UxcPMeBrVQs+RMobgDCht4k3csuQ7014HHjxjFv3jxKS0v54IMPuP322/n5z3/eFLGJRuB2q7y75uzE14oK//lSJsIWl6Z683ue5AtQ/fWHqI5qDSMKbPXWgO+88062bNlCZGQkR44c4f777ycjI6MpYhONQFFVHC7vZv5qh1JHaSEuTDk/2bpdqIqLy1uHJXTVm4Bzc3MZMGCAJ+lWV1dz/Phx2rdv7/fgxOUzGfWMG5jA8v8VePZNukoWRBQXz5mzE8rzvfaZe45AHyarZl+qepsgHnjgAa91xvR6PQ888MBlXfTzzz9nwoQJpKWlsWvXrjrLbdy4kbFjxzJ69GgWLlx4WdcMJRXVbv6zJo9n3j/Kxqxi7p7UjoeuT2Ly1S14/OaOXD+8ldYhigCjqgqVn/6tZmmin4RFgU4vD+MuQ701YLfbjdl8dqFGs9mM03l5c8Smpqby4osv8sQTT1zwuvPnz+eNN97AarUyffp0MjMz6dq162VdOxQ8+dYRdh2p6amyfmcx901pz/hBiYzsf2nnO3KqCodLJbV9+GUv+ikClL3Kdw236nIcWatx7F5H9O3PY7R20Sa2AFZvDTghIYE1a9Z4tr/88kvi4y/vCXqXLl3o3LnzBctkZWWRnJxMUlISZrOZCRMmeMUhapdXaPck35+s2lZ4SedSFJUF7+Twq3/8wJx/HeA3Lx+kyu6u/0ARdHRhkRja19G3XnHh2L2uaQMKEvXWgJ988kkeeughFixYgKqqtGnThmeeecbvgdlsNlq3bu3ZtlqtZGU1bDIPu91Odna2v0Jr1irsKgY9uM95zmZQqy7p72PvcYWv9pxdDXlfbiVvf5bN0DRDY4QqAoy+53Si1eWYC45gqC7xei2/3E5FiH7mGqKugWH1JuAOHTrwwQcfeAZfREZGNuiCM2fOJD8/32f/nDlzGDVqVIPOcaksFktIj4S7sSCPd37sahYZpueuyV1Ibe+71Hx9csoKAO/VDYzhCaSnt22MMEUg6j8YVVUo/78/4jq8HQB9iw60H3Mr+nCZP/pi1ZuAHQ4HX3zxBSdOnMDlOlsb+vWvf33B4xYvXnxZgVmtVvLy8jzbNpsNq9V6WecMFTeNbM2wXnGcyLfTq3MUkWGXVmO9Mj2G8M/0VNlrqtMGPQzrLZNrhzqdTk/0jPm4ju9FddoxJvdGp5dfRZei3gR8zz33EB0dTY8ePbwexvlbr169yMnJITc3F6vVyooVK3juueea7PqBrjEmVk+INvH/7urK0k1ncLgUJg5uQde2F1+TFsHJ2L671iEEvHon45k4cSKffvppo1509erVLFiwgMLCQmJiYkhPT+ff//43NpuNxx57jEWLFgGwYcMG/vznP+N2u5k2bRr33HNPg84vk/EIIQJBvQn48ccf5+abb6Zbt25NFdNlkwQshAgE9TZBbN++nWXLltGuXTuvJojly5f7NTAhhAh29Sbgn5oDhBBCNK56B2K0a9eOU6dO8b///Y927doRHh6OoshkLkIIcbnqTcAvvfQSr732mmcuBqfTycMPP+z3wIQQItjJihhCCKERWREjhOTkVbF6RxhLl50AAB1rSURBVBERFj3jByUSH23SOiQhQlq9Cfj8FTGWLFkiK2IEoCN5Vcz55wHP5OyrthXy6tw0wswNWhhbCOEHsiJGiFi9rdBrZYzTxU6+3V/K0F61Dy1e/nU+H2w4jaqqTBvaiilDWjZVqEKEjAsmYLfbzcyZM3n77bcl6Qa42mq64XXUfvcereBfn5zwbC9ccZLObcLp00VWPhCiMV3w96fBYECv11NWVtZU8Qg/mTC4BS1iz7b59uoUSb+U2mev2n2kvEH7hBCXp94miIiICCZNmsTVV19NRMTZiVgee+wxvwYmGldijImFc7vxzb5SIiwGrkiNxqCvfXWLtA6+U452S5JJeIRobPUm4DFjxjBmzJimiEX4WbjFwDV96l/NpHfnKG4b05oPN5xGBaYNacmAbjH+D1CIEFPvZDxQsxLyyZMn611GqLmQyXgax0+3hqwDJ4R/1NsHae3atUyePJlZs2YBNcnt7rvv9ntgQnvn9v8WQjS+Bg1F/u9//0tMTM1P0PT0dI4fP17PUSJQ2J0Km3YVs2VPCU6XzPEhRFOqtw3YaDQSHe39tFxqRcGhrMrF3H8e4ESBA4DObcJ4/p4ULCYZnCEaj+p24Tq+F31EHIaWHbQOp1mpNwF37dqV5cuX43a7ycnJ4e2336Zfv35NEZvwszU7ijzJF+DwqWq27C4hs1/9D+qEaAilLJ+yd36HUnQKAHO/8USOu1fjqJqPeqs6jz/+OAcPHsRsNvPggw8SFRXFo48+2hSxCT9SVZUT+Xaf/dUOaYYINarixnXqAEpFSf2FL1L1Nx95ki+A47vPcJ/OafTrBKo6a8APP/wwzz77LB988AFz585l7ty5TRmX8KPCMiePvn6YnLxqr/1xUUYyesZqFJXQgrvwBOXvPY5SYgO9kfBRswgbMKnRzq+UF/ruqyjCQMdGu0Ygq7MGvGfPHmw2G0uWLKGkpITi4mKv/0TgWrLxjE/ynTQ4kX/cm0JsZL2tUiKIVG18pyb5Aiguqta+jlLVeCNfLT1HeG3rY1piTOrZaOcPdHV+2mbMmMHMmTPJzc1l6tSpnNtdWKfTsWbNmiYJUDQ+W5HDZ99VPWJpGWeupfRZBaVOThbY6dY+ArM8qAsKSslp7x0uB2pFMYTXPkxdqS7HbTuCwdoJfVj9c4OYug4k8voncOxagz4yDsvgaeiMMg3qT+pMwCNGjODWW2/liSee4Mknn2zKmISfDe0dy5Y9Z9v7EmOM9OjoO/z4XJ98lc/CFSdwKxAfZeRPd3Smc5twf4cq/MycPoSqE/s824ZWnTC0SKq1rOPAVio++gs47WAKI2rq7zF1GVD/NVIGYU4Z1GgxB5M6E/ADDzzA0qVLycnJacJwRFMY3jsel0tlzXdFJESbmDHCitlYd4220u7m9ZUncf/4fK6o3MVbq/P4462dmihi4S+WgdeBzoDzh6/Rx7clfOiNdZatWv1qTfIFcFZTuXohsQ1IwKJudSZgRVF45ZVXyMnJ4Y033vB5/fbbb/drYMI/CkqdRIYZGNk/gZH9Exp0TFmlG7vTe8R6fonTH+GJJqbT6QgbeC1hA6+tt6xSVnDedr6/wgoZdVZ7nn/+efR6PW63m4qKCp//RGApq3Tx8MKD3Pz0Xn7x1B4++6ag/oN+ZI030z3Zeza0a/rUPpG7CF7mHsPP275Gm0CCSL2T8WzYsIHhw4dfqEizI5Px+Fr02UmWbjrj2TYadLz523QSGrguXEmFiw/Wnyb3TDVXpscwflCijIgMMarLQfX/luA6no2xfXfC5IHaZauzCeLjjz9m8uTJHDp0iMOHD/u8fjlNEJ9//jkvvfQShw4d4sMPP6RXr161lsvMzCQyMhK9Xo/BYGDp0qWXfM1Qd8zm3e3M5a4ZiNHQBBwbaeSXE9r6IzQRIHRGM+FD6m4jFhevzgRcVVUF+GcV5NTUVF588UWeeOKJesu++eabJCQ0rK1S1G1gtxi2/XC2f2dspJFu7WWSdSG0dMF+wAC//vWvG/2iXbp0afRzigubdFUilXY3G3YW0yLWxG1jW0tfXiE0VmcC/tOf/nTBA5tqSaI777wTnU7HDTfcwA033NCgY+x2O9nZ2X6OLPD0aV3zH9hxlhwlu/GH/gshalHXM6k6E3CPHj0A2LFjBwcPHmT8+PEArFy5skE12JkzZ5Kf79tNZc6cOYwaNapBQb/33ntYrVYKCgq4/fbb6dy5MwMHDqz3OIvFIg/hhBDNXp0JeMqUKUBNEnz33XcxGmuKzpgxg5tuuqneEy9evPiyg7NarQAkJiYyevRosrKyGpSARcOdzLezYVcxMREGMvvGE24xaB2SCFLusgKqt7yPzhxB2KDJ6KPk2U69M6+UlJRQXl5OXFxNv8/KykpKSvz/27WyshJFUYiKiqKyspItW7bwq1/9yu/XDSVHTlUx9+WD2J01Q9w+31rI3+9NqXO1ZCEulSvvIGVvzAW15l6zb/+U2HsWhXwSrjcBz549mylTpnDllVeiqirffvst991332VddPXq1SxYsIDCwkLuuusu0tPT+fe//43NZuOxxx5j0aJFFBQUcO+9NRM3u91uJk6cyLBhwy7rusLbim8KPMkX4NDJKnYeKqd/Su0TsQhxqSpXveJJvgA4q7FvW074NbdpF1Qz0KBVkc+cOcPOnTsB6NOnDy1btvR7YJdDBmI0zMufnOCTr73b6f8yqwt9utQ/y5UQF6Nk0a9Qzhz12mcZOJmI0bM1iqh5aFA/pJYtWzJq1ChGjRrV7JOvaLhJVyUSGXb2FuieHEGvTheeFU2ISxE2cLL3Dr0By9XXaxNMM9KgGnCgkRowOJwKb63O47uDZXRqHc4dP2tDQozvqLeCUidbdpcQE2Hg6p6xF5wVTYjLYd/5JfZtn6ALiyJizN2yQCeSgIPWvz45zvKvz06406NjJP/vrq4aRiSEON8lVXeuueaaRg5DNLZvsku9tvfkVFBW6dIoGiFEbS4pAQdhpTnotGth8dpOiDYSESZ9fIVoTi4pAcs0hM3fXRPb0TqhZo236HADD0xNkv69QjQzdfYDrm0VDKip/fpjhjTRuJKtYfz7wTROFTpoGWuSiXeEaIbqTMAXWvXi1ltv9UswonHp9TqfpgiXWyXrcDkRFj1pHaTLmfCmup24ju5CFxaFsW2q1uEEvToT8IWmoWyMeR5E0ysud/HQqwc5kV+zsOJV3WN4/OaO0qQkAFDKCyl7+7coRScBMKVlEDX1DxpHFdwu6XepJODAtOJ/+Z7kC/D13lJ2HZH1/UQN+7ZPPckXwLlvC67cvRpGFPykF0QIKanw7YZWXC5d00QNpdJ3ki2lsliDSEKH9IIIIZn94jGc8y8eH21kQLeGT7xTXuWWhB3ELL1Ggu7sDaKLSsTUub+GEQW/OkfC9evXr9ZEq6oqdrudvXub708TGQlXt9055azaVkiExcB1GS1onWCp/yDg35+f5KMt+bgVleG943jw+g4YDfJFHGycx3bj2LkKXXg0loGTMcS20jqkoCZDkUW9dh0p55GFh7z2zZ2WxJgBoT2XqxCXSzqHinodO13ts+9oLfuEEBdHErCoV7+u0T7NDQMvou1YCFG7elfEEKJtooV5t3Tk/XU2HE6Va69uQd8ukoCFuFySgEWDDOwWw8BuMVqHIURQkSYIIYTQiNSAhY/icifrvq/pgJ/ZL57YSLlNhPAH+WQJL8XlLn794g8UlNYMuFi6+Qz/uj+V6Ai5VYRobNIEIbys31nkSb4A+SVONuyU4ahC+INUa0KM263yzpo8Nu0qwRpv4s5xbencJtzzem2jzHXyNS2EX8hHK8Qs2XyG99ed5kS+nR0Hypm3+DBOl+J5fUSfeFrFnV09uXW8meG947UIVQQQpbocZ87OWif0EXWTGnCI2bbfe7HOglIXR/KqSW0fAUBMpJGX7k9lU1YJ6GB47zgiZS25kKJUleE89C36qESMyb3rnXzLeeQ7ypc8BY4qMJiIvPZBzOlDmyjawCYJOMR0tIZ5zQFsMelok2j2KhMdbmT8lYlNHZpoBtz5xyh7+xHUqjIATOlDiZryuwseU/XlazXJF8DtpHL1IkxpQ1BKz+DYuQoAS58xYLJg3/oR7mIb5rQMzGkZfn0vgUCTBPzMM8+wbt06TCYTHTp04OmnnyYmxreT/8aNG3nqqadQFIXrr7+e2bNnaxBtcPnFyNYcyatmd04FUWEG7p7Uluhw+R4WNaq/WeZJvgDO7E24Mm7A2KpTncco5YVe22pFEUrpGcpev99zLvv2Fegi41Hyj9acd+8G1GsfwtJzhB/eReDQpA04IyODTz/9lOXLl9OxY0deffVVnzJut5v58+fz2muvsWLFCj799FMOHjyoQbTBJS7KyLN3deXdP3TnP492Z2R/mdFMnKX+VJM9l72Wfecw97jGe7v7MJz7NnslcrWq1JN8f+LI+vKS4wwWmlR9hgwZ4vlz3759WblypU+ZrKwskpOTSUpKAmDChAmsWbOGrl27NlmcgeTAiUq2ZpfSroWFob3iMJwzec7pYgfrvy8iwmJgSK9YVn5bSPaxStq3MFNW5aa8ys2YAQkMTo/V8B0ILahuF859m3EXnsSUciWWfuNw7tsCas2DWYO1M4b2aRc8R/jIO9HHtMB1bBf6xPboLFG4TuyvpaQOODv7rT4yrhHfSWDS/LfnkiVLGDdunM9+m81G69atPdtWq5WsrKymDC1gfL23hD+9k4Py4739zb5SfjsjGYAT+Xbuf+kHKu01H6g3V+VRXu0GYKvXOUr50+2duSJVJtkJJRUfP4tz32YAqje/S+T0eUTf+iyOvRvRRyVg7vczdPX0Q9QZjIQNnoY7fahXswN6Iyg1fcr1LTpg7Nwfx9aPal4zmMAcjlJVhj48dO85vyXgmTNnkp+f77N/zpw5jBo1CoCXX34Zg8HAtdde26jXttvtZGdnN+o5m7P/rHJ6ki/Ahp3FDEupIC5Cx/LtLk/yBTzJtzYfb8whwq35d7JoIvrKIlr9mHwBUFXKPvorp8fNg/Y/9mI4ktvg80Vlf0HUOc0OKC4qOmfgTOxEdZseNT0k0quJzl4JbieO7z6n/OheCoff10jvqPmqa4EIv33a6ls5eenSpaxfv57FixfX2s3FarWSl5fn2bbZbFit1gZd22KxhNSKGFGbDwHlZ3fooFtqCgnRJr4+ehI406DzdOnQgvT0Nn6JUTQ/Smk+JV9479M7q+jsOoWl59iLPl9VwfdU7/Pe16r3EMzdh3m2y7a9wbmrCpoLj5LaKhpDYvuLvl4w0OQh3MaNG3nttdd4+eWXCQ8Pr7VMr169yMnJITc3F4fDwYoVK8jMzGziSAPD9GEtvRbbHH1FAgnRNYMpxg9KJNxy9sWocAPmH7929ecck2wNY/LVLZsiXNFM6GNaYGiT6rNfKTl9Seez9BmN7px2XX2LZEwpV3qV0UWc95xBb0AXFnVJ1wsGmqwJN3r0aBwOB3FxNf9Yffr0Yf78+dhsNh577DEWLVoEwIYNG/jzn/+M2+1m2rRp3HPPPQ06fyiuCXfsdDXf7q95CDeoWwx6fc2vilOFdu78f/s49195ztT2tIo307VtOCUVbsoqXXRLivAcI0KHqziPslfu8rTVAkRMfgRLj+GXdD6lsgRH9iZ0Rgvm9CHozN4VLJftMOXv/sHTThw25EbCh9186W8gwMminEFu9fZCnv+vdzve+CsTue+6hv/k27a/lC+2FRIdYWDa0Fa0a9GwlZRFYCh7fx6uw9s92/r4tsTcvbDeEXCXSrVX4srdjT6uDYYWSX65RqCQJy5B7qchxl772tXe7FObnYfKmffmEU8N+uu9pbz+UBrhFhmeHCyUwuPe20UnUSuK0EX5p4+4zhKBqesgv5w70MhkPEEu2RrG3RPbEhVmwGjQMX5QIqOuaPgHa933RV7NF8XlLnYcLK/7ABFwjO28fy3q49ugi5QJmJqC1IBDwOSMlky8qgWKomIyXtx3bkKMyWdfYrTcNsEkfNRsVHsFzsM7MLTsSMSE+/3W/CC8yScpRBj0OgyX8JBt8tUt2LK7mGOn7QCMviKetA6RjR2e0JA+Mpaon/9R6zBCkjyEE/VyKyrZxyqIDjeSbA3TOhwhgobUgEW9DHodPTuGbl9NIfxFHsIJIYRGJAELIYRGJAELIYRGJAELIYRGJAELIYRGJAELIYRGJAELIYRGJAELIYRGZCBGiHO5VbIOl1Ne5aZPlyhiI+WWEKKpyKcthNmKHPzm5QMUlv24cKIOHry+A5n9ZCYsIZqCNEGEsP9bb/MkXwBFhX99fByXO+imBxGiWZIEHMIKSpw++yrsClWOuldOFkI0HknAIWxEX9+mhv5do4gOl5YpIZqCfNJC2DV941GB99fZqLQrDOwWzR0/a6t1WEKEDJkPWAghNCJNEEIIoRFJwEIIoRFJwEIIoRFJwEIIoRFJwEIIoRFJwEIIoRFN+gE/88wzrFu3DpPJRIcOHXj66aeJiYnxKZeZmUlkZCR6vR6DwcDSpUs1iFYIIfxDkxpwRkYGn376KcuXL6djx468+uqrdZZ98803+fjjjyX5CiGCjiYJeMiQIRiNNZXvvn37kpeXp0UYQgihKc2HIi9ZsoRx48bV+fqdd96JTqfjhhtu4IYbbmjQOe12O9nZ2Y0VohBCXJa6Rub6LQHPnDmT/Px8n/1z5sxh1KhRALz88ssYDAauvfbaWs/x3nvvYbVaKSgo4Pbbb6dz584MHDiw3mtbLBYZiiyEaPb8loAXL158wdeXLl3K+vXrWbx4MTqdrtYyVqsVgMTEREaPHk1WVlaDErAQQgQCTdqAN27cyGuvvcbLL79MeHh4rWUqKyspLy/3/HnLli2kpKQ0ZZhCCOFXmsyGNnr0aBwOB3FxcQD06dOH+fPnY7PZeOyxx1i0aBG5ubnce++9ALjdbiZOnMg999zToPPLbGhCiEAg01EKIYRGZCScEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRBKwEEJoRLME/MILLzBp0iQmT57MHXfcgc1mq7XcsmXLGDNmDGPGjGHZsmVNHKUQQviPTlVVVYsLl5eXExUVBcBbb73FwYMHmT9/vleZ4uJipk2bxpIlS9DpdEydOpWlS5cSGxt7wXNnZ2eTnp7ut9iFEKIxaFYD/in5AlRVVaHT6XzKbN68mYyMDOLi4oiNjSUjI4NNmzY1ZZhCCOE3Ri0v/re//Y2PPvqI6Oho3nrrLZ/XbTYbrVu39mxbrdY6myrOZbfbyc7ObtRYhRDiUtX1i9yvCXjmzJnk5+f77J8zZw6jRo1i7ty5zJ07l1dffZV33nmH+++/v1Gua7FYpAniPIqisu2HMk7k2xmUFkO7FhatQxIi5Pk1AS9evLhB5SZNmsTs2bN9ErDVamXr1q2ebZvNxqBBgxozxJDxtyW5fLmjCIDXV55iwe2d6NslWuOohAhtmrUB5+TkeP68Zs0aOnfu7FNmyJAhbN68mZKSEkpKSti8eTNDhgxpwiiDw5kSB2u+K/Jsu9wqSzae0TAiIQRo2Ab83HPPceTIEXQ6He3atePJJ58EYNeuXbz//vs89dRTxMXF8atf/Yrp06cDcO+99xIXF6dVyAFLUeD8vi4utyadX4QQ59CsG5o/STc0XwveyeGrPSUA6HXw+C0dGZx+4e58Qgj/0rQXhGg6v78xmQ07izhR4OCq7jGktIvQOiQhQp4k4BBhNOgY2T9B6zCEEOeQuSCEEEIjkoCFEEIjkoCFEEIjkoCFEEIjkoCFEEIjkoCFEEIjkoCFEEIjkoCFEEIjkoBDVKXdzf7cSqoditahCBGyZCRcCNq6r5S/vH+UKrtCVLiBx2/uSO/OUfUfKIRoVFIDDkH/+uQEVfaamm95lZtXPz2hcURChCZJwCHGraicKXF47bMVOeooLYTwJ0nAIcag15HRw3sayqG9ZI5lIbQgbcAhaO60JKzxZvbnVtKzUxQzrmmldUhChCRJwCEo3GLgznFttQ5DiJAnTRBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKERScBCCKGRoJwLwm63k52drXUYQggBgNFoJCUlxWe/TlVVVYN4hBAi5EkThBBCaEQSsBBCaEQSsBBCaEQSsBBCaEQSsBBCaEQSsBBCaEQScBDYuHEjY8eOZfTo0SxcuNDn9RMnTnDbbbcxadIkbrnlFvLy8jyvPfvss0ycOJGJEyfy2WefefZ//fXXTJkyhYkTJ/Lb3/4Wl8vVJO9F+F9T3i8lJSXce++9TJo0ienTp/PDDz94jnnzzTeZOHEiEyZMYPHixf57w82ZKgKay+VSR44cqR47dky12+3qpEmT1AMHDniVue+++9SlS5eqqqqqX331lfrQQw+pqqqq69atU2fOnKk6nU61oqJCnTp1qlpWVqa63W512LBh6uHDh1VVVdUXXnhB/eCDD5r2jQm/aOr75S9/+Yv64osvqqqqqgcPHlRvvfVWVVVVdf/+/eqECRPUyspK1el0qrfddpuak5PTJH8HzYnUgANcVlYWycnJJCUlYTabmTBhAmvWrPEqc+jQIQYPHgzA4MGDPa8fPHiQAQMGYDQaiYiIoFu3bmzcuJHi4mJMJhOdOnUCICMjg1WrVjXtGxN+0dT3y7nn6tKlCydOnCA/P59Dhw7Ru3dvwsPDMRqNDBw4MCTvMUnAAc5ms9G6dWvPttVqxWazeZVJS0vz3NyrV6+moqKCoqIi0tLS2LRpE1VVVRQWFvLNN9+Ql5dHfHw8brebXbt2AbBy5Uqvn6EicDX1/XLuubKysjh58iR5eXmkpqayfft2ioqKqKqqYuPGjSF5jwXlXBDC2yOPPMKCBQtYtmwZAwYMwGq1YjAYGDJkCLt27WLGjBkkJCTQt29f9Ho9Op2O559/nqeffhqHw0FGRgZ6vXxXh4rGvF9mz57NU089xeTJk0lNTSU9PR2DwUCXLl2YNWsWd955J+Hh4aSlpYXmPaZ1G4i4PDt27FDvuOMOz/Yrr7yivvLKK3WWLy8vV4cOHVrra7/5zW/U9evX++zftGmTev/9919+sEJzWt4viqKoI0aMUMvKynxee+6559R33nmnIW8hqITgV05w6dWrFzk5OeTm5uJwOFixYgWZmZleZQoLC1EUBYCFCxcybdo0ANxuN0VFRQDs27eP/fv3k5GRAUBBQQEADoeDRYsWMWPGjKZ6S8KPmvp+KS0txeFwAPDhhx8yYMAAoqKivI45efIkq1atYtKkSf58682SNEEEOKPRyLx585g1axZut5tp06aRkpLC3//+d3r27MnIkSPZunUrzz//PDqdjgEDBvDEE08A4HK5uOmmmwCIiori2WefxWisuSVee+011q9fj6Io3HjjjVx11VWavUfReJr6fjl06BC/+93vAEhJSeGpp57yxHLfffdRXFyM0WjkiSeeICYmpin/KpoFmY5SCCE0Ik0QQgihEUnAQgihEUnAQgihEUnAQgihEUnAQgihEUnAQgihEUnAQgihERmIIcR5/vnPf/LJJ5+QkJBAmzZt6NGjB3feeafWYYkgJAlYiHNkZWWxatUqPvnkE5xOJ1OnTqVHjx5ahyWClCRgIc6xY8cORo4cicViwWKxMGLECK1DEkFM2oCFEEIjkoCFOEf//v1Zt24ddrudiooK1q9fr3VIIohJE4QQ5+jduzeZmZlce+21JCYmkpqaSnR0tNZhiSAls6EJcZ6KigoiIyOpqqripptuYsGCBfIgTviF1ICFOM+8efM4ePAgdrudKVOmSPIVfiM1YCGE0Ig8hBNCCI1IAhZCCI1IAhZCCI1IAhZCCI1IAhZCCI38f3y+Fmnzoy09AAAAAElFTkSuQmCC\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "df_epang = df2[df2[\"software\"] == \"epang\"]\n", + "sns.catplot(data=df_epang, x=\"g\", y=\"LL difference\", palette=\"muted\")" + ] + }, + { + "cell_type": "code", + "execution_count": 210, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 210, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "df_pplacer = df2[df2[\"software\"] == \"pplacer\"]\n", + "sns.catplot(data=df_pplacer, x=\"max-strikes\", row=\"max-pitches\", hue=\"strike-box\", y=\"LL difference\", dodge=True, palette=\"muted\")" + ] + }, + { + "cell_type": "code", + "execution_count": 211, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "" + ] + }, + "execution_count": 211, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAFgCAYAAACR2P/oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd2AUZf7/XzM7W7PpPSEBAgFCb1IEBZQiFhS7/qxn/+qp5513XvO88zzP8/SKd/bC3dkrih0QBAHpUkMPEEJ6315mfn8sbLLZDUEhuzthXv/IPs8z6yfJzLyf8imCoigKGhoaGhoaPxAx1gZoaGhoaKgbTUg0NDQ0NE4ITUg0NDQ0NE4ITUg0NDQ0NE4ITUg0NDQ0NE6IHiEku3fvjrUJGhoaGqcsPUJIfD5frE3Q0NDQOGXpEUKioaGhoRE7NCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghNCHR0NDQ0DghpFgb0BnLli3jkUceQZZlLrvsMm699dZYm3TcVDW4+ctbB6lu9DC6OJH7Li1AEIRg/9b9Nv49/xA2l8yAXmZ0Ari8CvUtXqxmicunZDFmQGIMf4IfjnvzIlwr3kR2tIIooktMR+o9HH99OX5bI4rHBY5mQEHfZyQJc36KYDDH2uy45t8fHmLh+gb8fgWDXkQQIC1R4rqZuQzuncCCVXU02XycPSqVoX2tsTb3B+Ovr8D20ePItQcBBcFkRSocjiL78JVtRDBaMJ1+BXJ9OXJrPYah0zAMmIDcUovr2/eR7Y0Yh52Fvv+4WP8opxyCoihKrI3oiN/vZ9asWbzyyitkZ2dz6aWX8uSTT9K/f/+I40tLSykpKYmylZ1z6e+3YHfJwc/jByXx0PV9AWh1+Ljyj9uQu/itP3ZLP4YXqeOlIDfX4FozH1/VHvzl277XtbrsIpJueqqbLFM/f3xtPyu2Nnfan5Yo0dAayH4tCvDozeq5b9qj+H00/fUy8Hu+13UJF/8S5+KXkJtrgm3Wyx9C3/+0k22ixjGIy62tzZs307t3bwoKCjAYDJx33nksXrw41mYdFzsP2kNEBGDjntbgvz9aVdeliAAs+a7xZJvWLSg+L63/+znutR9+bxEB8FfvQ3HZu8GynsG32zsXESAoIgCyAl+ua+huk7oF7+7V31tEAFxrPwoREQD31q9Ollkax0lcbm1VV1eTk5MT/Jydnc3mzZs7He92uyktLY2GaV3SbJfD2vQ6JWif4PYf1/f4XU2UltpOqm3dgaF6J2kttT/4ekXQsWPvPhDj8laMOULXQ0Lwuprj5ln4PkiNNjJ+wHU2RU/HjdEmt0J5HP0O4mm3pLvoEU+v0WiMqz/WhNIyvt3eEvx8+5wCSkrSACgpgSWlOzhU5+70+tREiR/NGUCKVd/ttp4ovlQDrSu7GiUAkZdh5slXUjJk2Mk2q8dw2ZRK3lhSE7Ev0SwypI+Vb0sD91p6ksSPLigmO9UQTRNPEiW07PoC/+Gdxx4m6kAOTMbE5CxyLrwH58p38Gz8NNCWkk3e7JsRk36ILGn8UOJSSLKzs6mqqgp+rq6uJjs7O4YWfT9+d21fth+ws6XMxsyxaaR2EIQXfjqIVdubqWrwMG1kCrsOOWl1+Nhb6aRXppFZY9PRid93LhobpJx+GIbPwLN5IQBCchbG0y5EEEQQRARLEvqi0fgr96DIfvC68LfWoziaMA6fgS41N8Y/QXxz3cxcRvazMn9FHTqdwOBCCy1OP7lpBs4elYZOJ1B6wE6T3ceo/omYDHG5W31cJN3wJJ6d3+LZuxbBkoQuORsprxjFacO1bgG6zN6Yz7gauaEC2daIVDAYQacnYfadmMaej2xvQioYgqCLy9dajyYuD9t9Ph+zZs1i3rx5wcP2J554guLi4ojj4+2w/VTEV1OGYm9CKhymPcgaJx3vgS34q/cg9R6OlN0v1uZodCAun3hJknjwwQe5+eab8fv9XHLJJZ2KiEZ8IGX1jbUJGj0U57JXcX3zxpFPApY592EcelZMbdIIJS5XJN8XNa5IFEVhX6WL5ASJjOT4PwvRiE9aHT6WbWnC7vTT4vQjiQIzx6aRl26MtWknBcXnpelvV4C37UxRTC8g+bZnY2iVRkfickXS02myefn1y/vYV+lCFODyqVlcPzNwViDLCqJKzkeOoihy4ExEI6o02338+Kld1DZ7Q9o//raOf989UKWH7h1RgofrbU3H5/moET20pz8GvLe8ln2VLiDg+//W0hr2Hnbwp9f3c8FvN3Pj46Ws39XSxbfEHsXjxPbBYzT9+UKan/4R3r3rYm3SKcVXGxvDRATA7pJVE4fUFYJkwDj6vJA202kXxcgajc7QViRRxOuTaWz1su+wM6RdUeD1xdWsPOIyXNXg4dE3DvDqLwdjMuhiYepx4VzxJt7SZQDITdXYPvgzKT/+L3JrPa7V76G47BhHzkLfbyyKx4mvsQL/4T2BtClFoxHE+P3Z4pXqRg+iCJnJBo61KW1WqfeW3FwDgoCYlBlsM0+/BalgCP6qvUh9RqDvMyKGFmpEQhOSKGBz+nn0jf1s2B05wNBq0rF+d2tIm90lU17rpjjfEgULfxj+ig4+/x4nvsO7sc//M4ozIIrenaswjJ2DZ8PHIVsUUv/TSLz8oShaq268PplH3zjAqu0tCAJMHZFCSYGFBJMYlkkhP93AWaNTY2TpD0OR/dg/fBxv6XIADEOmYrngvsBkw+sGvw9dTj+kXoNjbKlGJDQhiQKvLqrqVEQAstP07D3sCmkz6kUKs0zdbdoJIRUMwXdwS/CzYLIit9YHRSSAgmfdh2HX+vasxVe5GylX88Y7HpZuamLVkRWrosCS75pY8l0TAFkpes6fkE6KVY9RLzK+JAmjXl0rEm/pN0ERAfBsW4p+0CSkwmG0zvsJcmMlALqsviRe/wSCvmc4E/QUNCGJArsrHMfsDwQshgrJ1BEpcf8yME26AtneiKd0ObrkbMwzbkX4HqlOFG/n0f0aoRyu7/x3VdPkpTDLzPiSpChadHLxHxGK9sgNh/E0VwdFBMBfU4Zn50qMQ6dF0zyNLojvN1UPYdgxUntbzSJXTssiLbHtBWw16bj67PiP5BckAwnn3k3qT98h6eZ/oe89HKlgMIZhZwfHiJl9wJgQfrHRgnP56ziXv47i+/7J+k41JpQkcyxnPpdX3Z5M+gHjob3nn6hDXzwu8mTD6wpv04gpuoceeuihWBtxotTV1ZGZmdn1wBhR0juBxlYf5TUudKJAZrIe25F9bY9PYW+li7/c2p8Ek44hfRL48UX5ZKeqd+luGDgRQ8kZGIaehXnKtej7jsJXuQvF2QqCgGBJBmcrSnM1voNbkG2NGAZMiLXZcU1Gsp6iXDNNdh/pSRJNdh/ykaORrBQ9t1+Qj6RT77xQTEhFyi1GtjchpuZimXUHUv4gxOQsPJsXgT/gnSYkpGKZ9X/a1lacoQUknmRKD9p55qMKDte7OX1IMv83Jz/M8+rep3ezszx0u+uF+wbRK1NdD4fcWo/947/jO7AJXU5/Es67B11m707HK7If55JXcK/+ILRDbyL1/ve62dqeRXmNi0UbGjEZRWaflqaKBJ8/FH9jJZ5NC0Enoe8/Fuey1/CVb0PKH4Rl9o/RpcT/6r2no94pTBzi9ck8/L/97K5wYnfJLFzfyKuLqsPGtd/GAjBIAilW9R1XOT7/N76yDSD78R/eif3Dx4853r3hs3ARgRBXT43joyDLxI3n5HLVtOweLSIAutRczFOvw3zG1QER2bsu4CFYthHHgidibZ4GmpCcVCrq3DTafCFtW8pCvbU+XFnLup1tXk06EW48JxerWX0xFb5DoTUf/DVlgVK6nY0v3xreKBmwzLztZJt2SuLxyRyqdeM/nsppKqW9lyCAr3wbihJeA0gjumhCchLJTTeS2EEQBvZqiwNpbPXy4qeVtD8XnTYylYsmqXNGLvUK3U7UZfVFMHTusizlDQhtEESSbvon+r6jusO8U4r1u1q49tHt3PLkDn70eCn7Kp1dX6RCpNzQe0iXW6yl54kDtL/AScSoF3ngqt7kpBkQBZhQksR1M9vqbVQ3evD5Q2eL9S3hKS7UguWcO5GKRoOoQ5c3kIQLf37M8caxF2AYPgNEKXBoesF96NILomRtz+NgjYtHXtvPT57ezaNv7KfFEZih1DR5+es7B2NrXDdhOe8edPmDANDl9CPh/PtibJEGaIft3Ybfr6DTBfw1FUXh/W9qWbqpiYPVLjy+tl/5nXPyOX/iqVPNTZH9uFa+jXfXKsTUfMxTr9OKW/0APD6ZHz1eSn2LL2K/AMz/wzAMcR6L9H3wN1YiGBMQLUkosl9LsRNHqO+EVyUcFRGAj7+t58VP24KqDJJAnxwTU0ekct6E9GC7zelndWkziRaJMQMSVVMl8fvg+PI5PBs+AcBftRd/9R6SbntO2574nuw65OhURCBQ2PhgjYv+cZxi53iRXTbs7/wBX/k2ECVMk6/EPPmqWJul0Q5NSKLAt6XNIZ89PoXbz8+npHdboF5Vg5ufPLOHpiOH9aP6W3nkR0UIQs8RE+eKN4MichS54TByXfkx3YY1wslJNSKKBGNJOmIxiapzJ+8M95oPAyICIPtwLXsVwWBGTExHXzweQeoJ6fLVjTYNjAL5GaEPtE6EFoePyoa2qN0Fq+qDIgKwcY+NrfvtUbOxu1G8blwr3w7v0BsRE0+drb2TRUaynpvOyUMvBSYaRTkmThuYiChATqqBX17ZO64zR38f/A2Hwtqci17A/sGfaXnl3mN6CmpEB21FEgWumpbNjoMOdlc4MUhgMUk89N/9CAJcdHoGt56fj8sTPrWM1KZWFLcDvB1ToQhYZt2BYIqQQkWjSy4+I5MZY1NptvmDqw+/rPS4LVHDgIl4ty+L2CfXHsCzYznG4TOibJVGe7QVSRRITdTzz7sG8NLPBjHrtPTgykNR4IMVdRyscXHOuDSkducqvTKNjOqfGCuTTyrefetpef52Ajv3bZin3ai9AE6QRLMUsoXV00QEwDD4TCzn3IUufxBiBC8/xaMl/4w12ookiuSlG2m2hyfXq2v2Mro4kX/eWcxX3zWSaJE4Z2yosKgV9/blOD56PKQWiZCQiuXcuzEUj4uhZRpqwjh6NsbRs5Fb6mh58U4UVyDQV7CkYCg5I8bWaWhCEmWmjkhh2eam4Of0JD1D+wS2dvrmmrkp1xwr0046/sZKHPMfo+NKBNmniYjGD0JMyiDxR/8I5N4SRIwjZyImJMfarFMeTUiiQJPNxytfVLLrkIMRRVbuv7yAZVuaSbVKXDYlq0f4+vvKt+P85g0Ulw198Wno+4/DX11GmIgA+gETo2+gRo9A8TiRW+ownnYhokW99Vd6GlpAYhT41Ut72binLefWjDGp3HdpYfCz3eVn1fZmzAaRcYOS0EvqEhZf3SFaX7ormOr7KGJKDnJTVUibVDQG0/iLkBurkYpGoUvJiaapqsTtlXG6/ceVnHH5liY+XV2P2ShyxdRsBhaoP47kKN6DW7G/+wcUlx0kAwnn/wTD4DPxHtiCe/V7KLIf02kXou83NtamnnJoK5JuprrRHSIiAKtL25I2rtnRwqOv78flDej5wAILf72tv2rOR/wNh2l95d4wEQHCREQ/cjaCALY3fhtoEET0Q6ZimXYDYmJ62PUa8Omael789DBOt8yo/lZ+dXWfThN8btpr40+vHwh+XruzlafuKqZPjjq2SxVFwbv9a3zl29H1KsEwZGpIHJVz8UsBEQHweXAsfA6/sxXXl8/CkcSNtrLvSLzx70g5/WLxI5yyqGvqq0LeXBKeRj4vPeBls3mfjd/9pywoIgA7yx2s2t4cdk284l63ALzHlyBQyhuAZ+NnbQ2KjHfrV7TM+0nbC0IjSEOLl6c/PITTHXhJbtxj452va0LGyLKC1xfoX7G1KaTP51f4+Qt7cXnUUT3RtfQ/2D98HPeGT3B89FecS+eF9MutdSGfFXsTri+eDopIoFHGu2tVFKzVaI8mJN1MWVW4a+JFkwIBeJ+uro94zfYD6nmpKr7jdb0U8GxZHPk7Wuvx7P725BnVQyivdePvEEq0vyog2mWVTl75vJIrH9nK3N9t4cF5+1i9oyXsO1odftbsaI2GuSeMe8OnoZ/Xh2ZBMAyeEnpBJ2l1RC13W9TRtra6mY5BhRajyOShKQCYDJEfhAIVpbYwjjwnIBD+I1H5SZnoCwaj2JvxHd4FniOVIAURf6R6JEcQ9J2nnz9VkSPU2Rja18KfXt/P8i2hq9a1OzsXi87us7hDbwJ32yRKMIRuyZnPuhExIQX3xs+RmypDVyJHkPqPwzD4zG43VSMUTUi6EZvTz4Hq0PQNiRZdMKFj35zwl2dKgo4zh6dGxb6TgZQ3gKQb/4F72xJEcxKGETMRzYFASsXtwLNzJY5P/xkSR9IRXd4A9Jo7cBjfbAlfYfx3YU1YKYJjMaRPAmMGqCOw1TzlWhyf/IOAp5+A+cxrQvoFUYdp4qW4N34a8Xrj+EuwnP2j7jdUIwxNSLoRvSRgMoghq5LUI543Hq/Mq4tDz08KMg386ab+qquWKFhTMQyajC67KCS1t2C0BJIxdhARITEd4+jzEa2pCOZE9P3GIui0W7EjyRHKLx+viIzqZ+XCSZmMVVEWaeOIGUj5g/BV7EDKH4QuI3KtGsGSAk3tnh3JiPX/PYo+f2CULNXoiPb0diNGvcjFkzN4/avAAakgwKyxaQA02nzYnKEvWKdHISNZXfW33es/wbHoefD7EJOzsV75MLr0/GC/YEkCyQjtzlJM4y/BNO7CWJirKuZMTGfZ5iYq6o59DpVo0TGyn5UdBx3UNnsZOyCRn19RSKJFfY+3LqMA38GttL7+axSXDTExDfOU6zEMboteN0+9HtvbD4EvkLvNOGKGJiIxRn13mkrwywpVDW4+Xd0QbFMUeOGzw4wvSSLRokMQAm1HqWv2smF3K6OL1bEVobjsOBa/FDwfkZurcS77H9a5DwTHOBb8LURExKy+GE+bE3Vb1UiKVc+z9w5kS5mN6kYP/3g/NAvuiH5WrpyWxch+gfulvMbJ5n12BvdJUKWIAHjLNuL4/F/Bz3JjJfb5jyGm5iDlFgME/qs3BYXEvf4T9ANPR99nRExs1tCEpFtYv6uVv71XHrGMrsMls3J7M6OLE4kUCvrRqjrVCIlsbwoRCQC53ZaD4vfiO7gltL+5Gu/Olfhr9qMvGh1W910jFEknBJN3en0K7y4LrG6njUylMMtE/hFX8q82NvLEOweRj9xT/zcnnwtUWHnTW7YxQquCd9/6oJB4yzaCsyWk37P9a01IYogmJCcZn1/hiXcP0tjaefW6JItEbpqRwiwjB2tCX8SSSvazAXTp+eiy+uKvKQu2GUomB/8t6PSI6QXI9eVtbXoT9vf/BIDrm9exXPBTjMPOip7RKuaCiRlcMDGDd5fV8NJngYqbkk7gt9f04X+LqoIiAvDqoirOn5CuusJoUnYRkTbydBlthc8iBa+KVi2gNZaoxC9QPTTavMcUkcGFFiYODiSZe/SmfuSmtVV30+sCNSbUhPWK32MYNRupcBjm6bdiHH9xSH/C+fciHkmDIqQVoNhDg+bcq9+Pmq09AY9X5vV2Tho+v8Kri6rC3MzdXjniijfe0ZecgX7ELAJV5wEEDKNmox8wPjhGyh+EYfj04GcxozfGsedH11CNELQVyUkmM9kQttIYUWShX56Fkf0TGTsgEUEQ+G5vK2t3tHLZlCySEnTUNXuZUJJMdqq6yoaKielYZt2BZ+sS/DX78ZVtQF80Jtgv5Q8i6Y4XUJytCDo9TX+7KtT/X9TmMt8Hn6zg9oWKht3l5/wJ6by6qE1gzh2Xjqii1e1RBFGH9by7UabfjGfHCnyVu9AXDAkZo8h+TBMuwTjmAvB50PUahNBJcKJGdNCEpBv47TV9ee7jCg7WuBg7MIlbzs0NKXu6cH0DT77btt0zdUQKv7hSvTXLHZ8+hWfzQgDcaz7Acs6dGEefG+wXBBHBEliFGceej3vN/CMdIqYJl0XdXjVjMeqYNiKVxRsbg23njk/nkjOyKMwysaXMxsBeFqaNVE8sUkf8jZW4N3yGe/V7AHg2fIqxYieWmbfhq9qD7Z2HUVrrEMyJJMz5mSYicYCW/TcG3P2vXeyuaMtPJQjw+q+GkGKVVFcqVXE7aHryipBVhpBeQMptz3Z6jXffevw1+5H6jkLKLoqGmT0Kn1/h87X17Kt0Mro4MZgpQe0oiox9/uN4SyOU1dXpSfnZO9he+zW+Q9uCzUJCKuZZd6AvGKrVJYkh2ookBnSsP6ITBVocPh578wDf7bXRJ9vETy4tYEAvFaQAF0TQSUFXTACl4RDOr/+Hecq1ES/RF40J2f7S6Jwmm4+H/ltGWaWT/Awjv72mD7npRs6fkMGWMht7Dzspq3LSVyUZfo+Fd/eayCICIOkDaXYaKkKaFXsjjvf/FAhKvPx3mudWjNDWhN2MxyezaEMDby6p5lBtIF3KlVOzQtLEXzQpg/8trOK7vYF08/urXfz5jQOoYbEoGEwYx80NbVQUXCvexLN7dWyM6kH8/Pk97Cx34PEplFW5+Nlze9i4p5Vfv7SXnz+/l+c+Psxd/9zF0k2NXX9ZnCM313TaZ550JYKoCzl0D8Hnxvn1f7vJMo2u0FYk3czv5pUFBeL1r6r58839GDswiRfuG8iG3TYKs40M7WPluj9vD7mussFDi8NPcoIK/kTuyNmK/RU7obiTB1/juCivDXWGbWj18auX9oW0yQq8taSGqSPUey4CBOuwd0RfPB7ThEuAQJJQFAVf5S7kmv2h1zvVkeW4J6KCt5R62VPhCIoIBALKPlpZx+DeCeSkGTl3fFuW36F9E1jyXZtrbEGWURUiorgduNvXGGmHVDg0ytb0PPQ6Ae9x5Nc6njHxjK98G67lr0XsE4wJKF4XtrceCga4GoZOQ07Jw7drZXCcccTMqNiqEU78v6lUTKRgsPZN7y6rYf6KWnSiwEWTMvB4FTbsbqVvrom750ZOWBefdPg5RR3mqdejLxoNgOL34Vz0Ap5tSxES07FMvwV931ExsFN9XDM9m1e+aKs0KRDIjduRuZPVF8XeHu+BLZE7dBLGMedh//zfIVkSPFuXkHDZg+h7DcJfU4ZUNBrjUC2wNVZoZyTdSL88M2PapTsxSAIXTQoEHK7f1cJLn1VS3+KjpsnL859UkpOmZ9LQZM4bn0HvbHXU5xCMFoxjzmvfQsIFPw1uRUDAJdi9/mMUlw259gC2t36HY+HzYRXvNMK59Mws5k7OICtFz5gBVjJSQpN6jhlg5dGbizhvvLqFRMorDmvTD5lK0o+eAr0R75avwvoVWyNS4VAUlz3gLrx5UTRM1YiAtiLpZh66vi/fbG2irtnLpKHJ5KYFtrO2lIWfK7y3PPBiXbShkZomD1dOy46qrT8U85EVhr/2iEtvTv+Qfu/BDgWtZD/utR/i3bmSpNueQ9Crp5BXtJm/oo4PvgncFzVN4bnbThuYFEza2B6/X6Gh1UtGsl4VaVL0RWMwTboS15r5IAiYxl0UrEfiWvNh+AWiDl3+IFr/+zPwBFzpHYe2I5iTMGi1baKOJiTdjKQTgoegHq/M20urKS13cLiL1OBfrG1QjZB4tizGu2ctuoxCdKl5Yf1S3kB8e9eFtcsttXjLNmIYMCEaZqqSVdubj9n/zZZm6pq9nDUqNegCvKXMxmNvHqS+xUt+uoHfXNOHPipwDzZPuRbTGVcD4D+8E/vHf0cwWpDyw2PETGdeE8jh5nGGtHt3rdSEJAZoQhJF/vZeOUs3NXU9kECNCTXgWvMBzkUvAuAF3Os+wjjuIkwTL0XQBbZhTBMuQW6owLNtadj1gskaRWvVR16Gka37I3vFAWzdb2frfjsfrqzjydv70z/fwt/ebcs8XVHv4d8fVfD4rf07/Y54QhB1+A7vpPXVB4IF0Tzbv8Z4+hWBjAiyH+OY8zBNvAx/zb6w63Vp+WFtGt2PdkYSJTxemWVbjk9E9JLAtTNyutmik4Nn8+KQz4rLhmvZqzgXvxxsE/RGEi68n6Q7X0FMaktKqS8ej17z7Dom15ydTZ8j52VGvcAVU7OYOiKFQQWhKwyvT+HztQ2U1zipbPCE9B3sUO453vFsXRJSVVOxNyE3VQU8VRQZxe0ARUbK7odp4mVwpCqn1HsExtHndfa1Gt1I3K1IHnvsMZYsWYJer6ewsJBHH32UpKSkWJt1wuh0AlazjhZ757XLAa6YmsWciRmkJamjUqKQEDk9h6d0OZaZt4W06ZKzSLrtObxlGxFMVk1EjoPMFANP3zOAijo3qYl6EkyBl+bmfTZ+8cLekLHrd7fwyer6sO84baC6np+jedna493+dfDfns0LkXqVYBw5C/O0GzCOm4vicaBLzY2mmRrtiLsVyaRJk/j4449ZsGABffr04bnnnou1SScFnShw8+y8LpPdTh+dqhoRAQIHoobwVC5iSuTzHUFvxDBggiYi3wNBEOiVaQqKCMCwvgmM6Ne2LSgKUNUQehifZNFx7rh07pijru0e46hzEdttUemywvOx+avaRFRMSNZEJMbE3Ypk8uS2wkgjR47k888/j6E1J5cZY9IY2d9KWaWL/vlmqho8PPXBIfZXuxCFgKtnr0x1uP0eRcofRMqP/4Nr1TsBjxufB8GchOXsm2JtWo+mrsWLJIJRErCYdDTawmvgTBqazI/n9oqBdSeGmJBM0i1P4zuwGcFoQUhIoeXZW0O2u6R2cUieXd/iXDIPxdmCYcQMzFOv1zICR5m4zv57++23M3v2bC688MJjjvvuu+8wGtXpQqooChUNClaTQEpC/LtpHgvB60RqrcGbnAc69ayq1MizC73sqjz2o3v7dIkBeT3jhWqs3Ia19EsEnxtH0ek4+p8JgOhqIfOLRxDaiUzzyEtx9o0fT0A1ZSb/ocRkRXLDDTdQVxcejHbvvfcyfXqg8tkzzzyDTqdjzpw5XX6f0dG216QAACAASURBVGhU9R9rcKwN0FAd+17b3GmfIEB+ugGXLo0BA7LQ6dQ9QQGgpATOujT40d9UjeJswd/swyGHnjtmeutIUPH7QI3EREjmzZt3zP7333+fpUuXMm/ePFUEU2loRItmu495X1Qi6QR87fJr9c4yMqgwgXW7Wqlv8XKozsO8L6pwumVumNWzzg8ci17AveZDQEFML6Bj4hhdbniUvEb3EndnJMuWLePFF1/k1VdfxWyO/yAqDY1osOuQg683N7F6ezMV9aHuvcX5Zu6/vJCkBIkv1jWE9C3b3NSjhMRfs7+twiYg15cj9T8N/+HdKC4bhmFnYRyj1W+PNnEnJA8//DAej4cbb7wRgBEjRvCHP/whxlZpaMSOzfts/PKlvchyeF9ygo5/3jUAAK9PJtGso9XZttWTnWqIlplRQW4Jr1kiGhOw3vMqKDKCqI5A3p5G3AnJwoULY22ChkZc8cnq+ogiAoR4+eklkTvm5POP98txexVSEyVuOrfnrEYApMLhCJYUFEdbcK9+8JmBLXBBE5FYEXdC0lMoPWjng29qkWWYc3oGw4u6TgVSVunE7ZUZWGDpMWdDsq0Bf105Ut4ABIO2VflDMOoj3wsZyXpuPz80t9m0kamMLrayprQFm8tPWaWL/HQjZqN6XrKK7MdfsRPBnIiYmI5sb0KXlofi8+KvKSPh0t/g2fgZsqMZ4/DpGLTiaTFHE5JuoKrBzQMv7MXjCxwArt7RwlN3FQcT523c08pbS2pwe2XmnJ7BlOEp/OmNA6zYGkjQN7CXhT/dXIRFRQ9/JNybvsTx2b9B9oHRgpQ/CKW1AaloNOYp1yJIPWvbpbu45IwsVm5rxu4KLEsmDk7ihlm55KcbwzyyDtW6+e28fVS1S5PyzrIa/nlnMSZD/N9Psr2Z1tceQK47GGgQRFBkxPReyI4WcLaATsJyzl0YR8yIrbEaQTQh6Qa+LW0JigiAz6+wYmszfXLMVDW4eXBeWdDjZsdbB6lr9gZFBGDnIQdfrG1g7uTMsO9WC4rfi3PxiwERAXA78O3bAIC/dj/4vVhm3h47A1VE72wTL/60hG+3N1Pd6MFkDMSGRHLrffnzwyEiAlBe42bF1mbOHp0WFXtPBPfaD9tEBEAJiKdcf6itze/DsfD5QN4tlw39oMnoOsmkoBEdNCHpBjKTw4PxMlMCs+8Nu20hbpsAyyMkc2xoDa89oSYUjwvF1XnWWu/uNaAJyXFR1+zlH++Xs2F3K/KRW+fVRdU8fGPfsFokew87I3wD+Ds5Y4k3ZFt4rrCIeBw4Pv0nAM7lr5N4/V+Rsvp2o2Uax6JnhL3GGRNKkpk4uC1R3qj+VqaOCCQ37JUZHoG/uyL04RdFOHN45GSIakE0JyL1G9t5f7r6UnfEiqc+KGfdrjYRgcAq92jBq/aYDOGPdIpVx6Sh4YkQ4xHDkKnf/yKvC/e6BSfdFo3jR1uRdAM6ncCD1/blQLULWVbom9t2yDy8yMpFkzL4aFUdsgzZqXqqG0NXH9dOz6Y4PzwRotqwXvQLXKvexV9ThpCYhqf0G3DZEFPzsEy/OdbmqYZO65EcyW6kKAprd7ayr9JJXrqBgzWhRdN+fkXvkISP8Yy+7yisl/8O96aFIIgoHgeKsxX9wNPx1ezHt3s1eNWVFv9UQBOSbqRj3fXSg3a27bdzxrAUrpqWjc+v8OHKOt7+OtQ3PlLpVDWguOx4dnwDgKHkDASjBfPU64L9lhm3IbfWoyjgXr8AxevCOHIWUt7AWJmsCkoKE1i/uzWkTSfCRUfO0J77+DAfrmxbnZgNIk5P28H8qP7qup/0/ceh7x9a5dD2zu/x7V4T+QLJgHGMVocklmhCEiU+XFnLswsOBz/fel4ecydnct74dN5bXhOyh+3yqmRDux2ys5XWl+9Bbq4GwLXybRJ/9A/EdhUQBcmAYE6k9dlbURwB5wLPlsUk3vAkUna/mNitBn48txdPvHOQLWV2clL1TChJ5pxx6fTONuFw+8NqkGSl6rl8SjYpVolR/dVfgdJXvS9wptYZOglBr7mWxxJNSKLEW0tCVx1vLqlm7uRMymvdYQehi9Y3qm5V4tm2NCgiAHJTFd7tyzCOPjdknHfHyqCIAOD34dm6RBOSY5CdauAvt/ZHUZRgfJHPr/D64mrW7GzBL4c6b4iCwFmjUmNhavcgH7sYHG4H7o2fYjlb2y6NFZqQRImOD7vD5UeWlYiHo0kqqdd+FNllA58nrF3p8ALwVe3BsfiFsHGiWV0V/GJBq8PHhj2tZKcYyEox8NbXNXy0MvywXThS10bNKB4XKDKeHSvw7l2LLrMPuvxB+Ct2dH5RZ6H/GlFBE5IoUVKYwOodLcHPPhme+7iChesbQ8ZlJuuDe9/xjuLzYP/oCbw7VoBkCFRK9DgAEKzpGAafGTLeuWQeuB0hbWJaPoaR50TLZFWyr9LJL57fi8117Jn5xWdkcOawVAYWqNdRw/n1/3Ctfh/83qAzgXfHCnR9RmGZfRf+hsN4dqxAabf6xWDCOEq7h2KJJiRRoiDLyOoOE6qPv60PcemcODiJB67sjUGvDq9s94ZP8B45XMcX8BQyjr8Y0ZqGYeg0xA61t+WW2rDvsF71R0SLtiI5Fm8vrelSRABSrXpVi4jvUCmuFW9G7PPv34h+zk9Rti8LFRHAMOgMdBmF0TBRoxPU8cbqAZw5PCWkXrtBEuiw24XbK6tGRCCQ0rsjUu4ATOPnIiaEx8HoS84I+azLLUaXrO5tmGhQ2xy+bRiJQ7XurgfFMf6ass47BTHgDuwOd4X2bF6I7/DObrRMoyvU89ZSOcX5Fv54YxGnD0lm6ogU/nRTX8QOGS5ElSVq1Lermw2AZEAqHNrpeEEXugCW7c1h5yga4TRHqMceibED1eWg0RGpzwjoLA28IuPZvBDDkCkghm+kHNOrS6Pb0ba2osio/olBn/4D1a6wFUmzXV1pUQxDpiLbGnBv/ALBbMV85jWI1s69hbx71oZ8Vlpq8Ncd1FJbdEFtc+f3hdWsIyNJzznj0pg8VN3ZEHRp+SRc/EtcK95CtjehdNgK9deUYZp4KeaZt+H8/N8hfVqmhNiiCUmMyErRhwSOATTb/VQ3elRVjMg0/mJM4y8+rrG6jMJQzxtBwH94tyYkXTCqOJHVpS0R++wuP/97oEQVmX2PB8OAiRgGTER2NNP8rxuDZ28AUt+RABhHzcZfvQ/Pd1+AoqAffGaYY4dGdNE99NBDD8XaiBOlrq6OzEx1eDodRS+JZKUaWL2jObgysbtktpbZOXd8emyN6yak3GK8e9ehONteit7da9D3G4uY2DN/5pPByH5WDtW6qagLPwNJS5S47MysHlO/5iiC3oSUNxB/QwWCZMA4fi7GsXMQBAFBEDAUj8M4+lxM4y7EOGIGgqDt0scS7bcfQ6aNTCU9KXT1seewk0aVZ/7tDDExHePw6R1aFbx718XEHrWQYtXz0PV90UV4Wm8/Px+x42FbD0HKLcYwZCqGYWdjGDwFQRCQHc24Vr6Nc8k8FGeLNgGJE7StrRjTN8dEdWObV05qokSSpef+WXSZvY+rTSOc4nwzO8rbMkXnpxs4Q+VZojtD8Xtp+e/9yHUHgCMpd679C/b5jyE3Vgba1n5E0g1PosvqE0NLNUBbkcScW87No8+R5I4pVomfXFIQsWBRT0BRFNzbvg5pMwyfjn7gxBhZpC5+PLeAnLTACjYjWc99l/Xc2Alf2cagiAQaPLS+8dugiATa3Lg3fRl94zTC6LlTX5WQl2HkmXsHUtvsIdWqR+qhIgLg278J77YlIW2eHSvQZffDdNqcGFkV32zaa+Ppjw5R1eBh8tAUnr57ADann7QkPboeuqUFRHTxxdUa1iTow+v7aEQfTUi6GZdHZvM+G5nJevrmmnF7ZeavqGV3hZORRVbOHZ+OKApkJqvHU6sjituB98Bm/PWHkOsPBfa2R52D0CEmIFJkOx4nzoXPIeUWI/UqiZLF6sDtlfnja/uxOQOxNl9910haksRNs/OOed3OcgefranHZBCZc3oGeenqe9lKfUYgJGeHRbELSRkoLYEcY4I1LSwpqEZs0ISkG6moc3P/83tobA0ElF0wIZ0Wh5+vNwdK667Y2kxdi5cbZuXG0swTwv3dlzi+fDbETdOzeSHewzuxXnBfyFh9v7FgMIMnvBys79B2TUg6UF7jCorIUbYf6Lx8MQRK7f702d3BjNILVtXxiysLOXO4yrIBKwpSwRC87YVE1JFwyW9QWutQ3E4MxeMRTAmxs1EjiCYk3cg7X9cERQRgwbfh9ajf+bqGQ7Uufjy3F8kJ4bXe4xn7l8/jWfdhxD7vlsW48gYg5Zcg5QRSxIvWVBKveQzHohfwH9wSMl7K10SkI70yTVjNuhAxGdz72C/OJd81hpQlkBX4y1vlDC9KJMWqnsfdtfp9vFu/amsQRIwTL0PKLkLILY6dYRoR0Q7bu5Fme9epLWQFVmxr4cbHd9BkU4/br+xoxrPuo2OOcX7xDK0v341rzfxgm5TTj6Rr/ozpzGsRTAkI5iTM029GKhjc3SarDpNB5FdX96ZXphFJJ3Dm8GSuPiu70/ELVtWxeGNjWLtfVthx8NgrmXjDd2BzaIMi417xJrbXf42iaCnj4w1NSLqRGWPSjnus0y3z5frwl0C8ItvqAaXLcQCub14Pe/jNk68k5b63SfnJG5jGze0GC3sGo/on8osrC+mVYWTZ5mZ+88q+EHfxo3xb2szTH1XQFCEvlyhCvzz1VBBUZD+yvTlin+/gFnz7N0fs04gdmpB0I6cPSeb31/flrJGpGKSuPWxcHvXMtMSkLOD4vIYUX1ttCY3vz1/eOsj+ahcA2w84+Nf8Q3i8Mlv324LBq+t3RfBoApITJO69uIDMFPU4c3i++wK5Zl+n/YrXFUVrNI4H9WyaqhBFUdi638bqHc0YJBGPr/NMt0a9yPTR6jkQFU1WjOMuxH1020rUkXDxLxEtKXj2rse94o22wTqJpievwDj0LMwzbg3LAqzROS6Pn/Ka0NQopQftXP+XUppsPiSdwB0X5JOfES4Uf72jH4N6Jagu8t1XuavTPjEtD33RmJA2RVHA70WQ1COWPQ3tie5GFm9s5J2v21xeBSF8Yj5+UCKFWSZ6ZRp56D9l1DZ7mTI8hf+bkx/3tUks029BP2Aict1BpL6j0KUGvM+kXoEDdu/edXg2LwxWRXRv+AT0Rixn3xRLs1WFyaCjON/M7oo2TzeDJNJ4ZAvL51d48dPDJCWEulr3yjQyuNAaVVtPFlLhMDybFrY1CCLGUecgJmdhGDELQWpzSvHu34Tj038gN1Uj9R1Fwpz7EROSI3yrRncS328qldPRVVNRYObYVLJT9CRZdEwbmcLPLu/NFdOyefbjw5TXunF5ZL5Y18A7y2piZPX3Q184FOPoc4MichSpaDT+ugPQod6Ie8NngRmkxnEzvMiKQRIQCGQ/cHtDf6dOj0x1Y6ijxuE6Nx98U8u+ynBX67hHVhCTs0CnR0jOJmHuA1jOuRPTxMtCqmkqfi/2+X9Bbgq4CPvKNmJf8ESsrD6l0YSkGykpDHXVFAS49IwsctKMtDj8LPmuiVuf3MH6Xa043aHnI9v2q8vLpj2K10XLs7fiP1Qa3ul1ojgiH6RqhPPFunreW16Lx6egAE02Hw53qBAnmMIfY1mB5z85zF1P7eKrCJ5c8Ypnxwocn/wNubkG/F4UeyOCOXIpZrmlDsXRFNLm27c+ENekEVU0IelGzh6VysVnZGIyiKQlBg49q5s8bNpnC45ptPnYfsCGyRD6p+goQmrCs2MFSmtdxD4hJRfMSfjKt+HdtwHFf3zV/05FnG4/z398uMtxDlfnThqKAm8ure60P97w7FgR2uDzYHvtl3jLNoaNFZOzAiuXDrjXf4JsU4949gS0M5JuRBQFbjk3j1vODaS0aHX4uPOf4QeJ2w84EAXQiQGH2mkjUrl8qoprmXsie9WI6b2wXPBT7G89iO/Ii0HMKCTx2r8gmtVdJrY7WLW9BYe7a08+USQkCLEjPp96thJ1qbmER1MpuNd/ElbaWRB1JFz6G1r/9wB4HO2Gyyh+9cRk9QS0FUkU+XxtQ1jZVFGA3RVOHG45+DK4dkYOxjg/aD8W+pLJYAk98DRNvY7k255DsTcFRQRArjuIR8vgGpFjOVsdrU2SYBK7rKg5d7J6ir4Zx12EGCkt/BGPLMXrxrtvA/76Q4Hm7H4kzL4rZKi+/zh0EVYqGt2HtiKJIi2O8G2cfnmhHjmyHDikV1O53Y4oHhcJ596Dr2oPyH6MI2eiS8kJ9DnDz0dkZ+Qysqc6EwYnU5BlDLr/plgl7rwwn5QEiQG9zJTXesjPMLBht41HXtsfrLQJYJAEZo9LY8LgZEb2U89qTzQnknzzv2l960F8e9cHGg1mTOPn4q8/ROurD6DYA9tWhiHTME+/BcOQKQjWVLy7vkWX3gtDWPE0je5GE5IoMm1kKvNX1OHzB574BLNI/w6unQAWo3pWI4rXhe/gVsTkLHQZhTi+ehn3t+8DCmJ6LxL/36OI1rYIf33/8QjmpLZyuzoJ45BpsTE+zjEZRP7xf8Us29KEx6twxrAUUqwStc0evlzXSGaKgb45Jk4fkswL9w1i6eZGahs9FGabOXtUKkkJ6ny8FUXGMHwmgjkFMTkT05jzEK1p2Bc8GRQRAM+2JXhKl2E598cYh89A33t4DK0+tVHnnaYC6pq9fL25EbNRx7QRKZiNOopyzfz1tv58vraeLWV2KurcfLamIezaTXttjC+Jf194f105ra89gGIPeM7oh54VkmhPrj+E69v3sUy/OdgmWpJIvOFJ3OsWoHhdGEeeo1W4OwZmo45ZY9vKye4st/OLF/bi9gYmI2cMS+ZXV/dh7c4W3ltWi8src9ZIuGBiRqxMPmEcnz7Vtt0piEjZRRgGTYp8gC77cS56AcPgqSHxJRrRRROSbuBwvZt7/rUbmyvg779gZR3/vKsYvSQysMCCxyfz+dpwATmKxaTrtC+ecC79T1BEgNBsrUeQW9riYRRFAdmHLjUXy4xbo2JjT2JnuYMHXtwXFBGA5VuaWbOjmWfbeXct2tDIgF4WVYqJ7GjGs3lRW4Mi41r9PoZBk9DlFOEr2xB2jeKyo7gdCFL8T756KpqQdAPvfF0dFBGA/dUu1u5s5fQhgRu9obULjxKVONn4OqSCj4Rh8BQAPNuX41j0PIq9Cf2gSSScfy+C3tTdJqqejXtaeXdZDT6fQkW9K2I+tv1V4V5yew+rMBARiJS/TTniheUr3x7xCqnPCC2aPcaoZzNeRazfZQtrE9o9H2MHJIWltGjPe9/U4vHFfwJHRTj27WMcf3FgS8LRjH3BEyi2BlBkvKXLca16N0pWqpdDtW4enFfGht02NpfZqW8Jz9U2osjKtFGpYSWaR/ZTZ3oU0ZKEYcTMkDb58C6a/3FNIEix4/i0fBLm/jJa5ml0giYk3UAk76zR/dse7ASTjiduK2b2uDT65YXPyl0eGbcKMgEbhh77kFyXFoif8dcegA5+/b7K3d1mV09hzY6WoGNGJPrlmTh/QjqPvXmQnFQDWSl6slMN3DArh6kj1ZMAtCOW2XeScOlvkQqHBdsUly0wEemAecp1WgxSHKBtbZ1k1u8Kd2UdXGjBaAhPqnf33AJkWeEXL+xla7uUKGMHJJJoif8/jWXq9eB24Nm5AvQmsDfB0bojBjP64vEosh/P1iVh1+rbvSQ0IpMXIaMvBFa3yRYdM0an8uibB5Dltva/3VHMwAJLFK08ufgbDuNe9xGKzxMeVKjIGMddhGf7ckDGeNpFGEom4974Ge5NXyKYEjGfcTVS/qCY2H4qIyg9IINeaWkpJSWxL9XqdPu55tHtIdHIOakG/nxLv2PGhciywsff1rN+dyt9c0xcNiWLBJUcuLfHe3Ar7vWfIEh6jOMuQsouwrVuAc4OuY8Mw2dimX2nlk6+C2RZ4ZHX9rNye+Q4G12EiPaZY9IwGUSSEyTOm5BOsopcgGVHCy3P3dbmGi6IbRMTQDBZSf7xf0LO1jw7VmB//09tX2Iwk3znK9oqJcqo5y5TAQdr3GEpLTKObDccC1EUmHN6BnNOV5+XTXv0hUPRFw4NafNX7AgbJxUMxrP9a8TEDKTewxEEddXLiBaiKPDba/vy5fp65n9TR02zB7uz7f6KlBZl0YaGYGDikk2NPHPPwLDzk3jFu2d1m4gAKDK6/EEojmbExAzM024Ic9Dw7l4d+iUeJ74DmzEMmhQFizWOclxCsm7dOg4cOMAll1xCQ0MDdrudgoKC7rZNdfTONmE167A52w5FLUaRsionNqefpz44REW9m/GDkrjv0kISTCI+v4JeUv9RlSL7QRAQOhzASwVD8Gxb2tYg6nAsfA48Aa8iw5CpJFx4fxQtVR8zx6Qzc0w6H62s45kFFSF9k4cms3J7M4oSWP1WNrSV4T1U62ZLmY1R/dUxOxct4Z5XhoGnY5pwSefXpPcKa9NFaNPoXroUkn/9619s3bqVsrIyLrnkErxeL/fffz9vvvlmNOxTFSaDyG/+Xx+eXVBBRb0bn09hzY5W1uxoxWwQcR45QF+1vYXH3jzAwRoXdc1expck8dPLClW5naXIfhyf/xvP5sUIRgvmqddjHHVOsN8wchb+hsN4Nn2JYElCtKbhK98W7PdsW4pp0pXoMrSJSVecMy6NzftsrNjWjKQTuOzMTK6bmUuz3YeiKLy5pIYPV4ZmXbYY1XNPSUVjkPqODsaKiBm90Q87G8VlRzAl4FrzIa5Vb4MCpgkXY5pwCaaxF+Dbvwnf/u9AJ2E6/Qp0mb1j/JOcenR5RnLhhRcyf/585s6dy/z5gbKqF1xwAQsWLIiKgcdDvJyRtOf2v+/kQHXntaU7Vku8eHImt5yXFwXLTi7u9Z/g+OLpkLaEK/6Aod+YiONt7/4R765VIW2JN/4dKbe422zsaTS2ejHoxbCJR1WDm/ue2ROsnjhxcBIPXts3FiaeEL7y7Sg+N/7Gw7iW/gfF5UAqGBwyAQGwXv0n9H1GAARcgw1m7WwkRnS5p6LX6xEEIbiP7XA4urji5PDyyy8zcOBAGho6jwCPZ5zuzuuzQ3jJ3V2HovN7PdlEqq/t+Oivga2uCOgHTKB90JkubwC6nP7dZV6Pw+7yk5QgRVy95qQZeeGng3jgqkL+fHM/fntNn+gbeBKQCgYjpuXj/PwZFJcdUMJEBAhpE5OzNBGJIV1ubc2ePZsHH3yQlpYW3n77bd577z0uv/zybjWqsrKSFStWkJenvhn6UfIzjdQ0dR7BbjWJ2NoVJBpWpM4AMqlwWGhKC0BxtuDZuRLDoEkhZybe8u04Pn+ao6H7YloeiVc9oh22HweNrV4efeMAW8rspCdJ3D23gHGDwisHJph0TBmu3hgSCMQY2d59mK5SPGhuvvFDlyuSm266iVmzZjFz5kzKysq4++67ufbaa7vVqEcffZT7779ftS8YWVYoPUap3Lx0A71zAt4nggCji61cqdJCVoZhZ6PLLgprd3zwZ1qeux1/UzWyvYnW13+N7X/3g88dHCM3HMbfUBF2rUY4876sYktZ4J6qb/Hx+NsHcXvjP2j1h2D/+G8orfVh7VLvEQgmK4IpAdOU69AXjY6BdRqR6HJFUl5eztixY5k0KeBO53K5OHToEL16dY9nxKJFi8jKymLQIHXPNjydVKVLS5Qo7mXm602BuhyKAlvL7NS3eMlNN0bTxJOCIAhYr/g9tnf+gL9DtLrcUIFr2asgCIHD0EioP4wpKnTMnWVz+qlt8iAIAh6vTN9cMwArtjWzYFUdep3AZVOyGK6yla7f0YxceyBin+/AJrCmYRoxE9PESwFQ/F7cGz7DX7UHqfdwDMPOVu0EVM10KST33HNPiIeWKIrcc889vPfeez/4f3rDDTdQVxde0/vee+/lueee4+WXX/5e3+d2uyktLf3B9nQHZgPY3aFtd58jUZAh8Nyi0AAzj0/hlid2cOUkHWOK1ONlE8KE29BX7yB95YshzbbD+xC8TiIl+HZn9KOq2Q/N8fW3i0cKU33sbVe+PcUCT7+/i437A0LcL1tgxjCR5xb5gxtC3+1p5YGL9KRZ1fFilVoqSVv29LG3SWwNuFa8SW1DE46+E0na8hHm8kABLM+WxVTv2Ya9ZOaxviHqxJsjUHfQpZD4/X4MhraAOoPBgNd7YvWQ582bF7F9586dHDp0iAsvvBCAqqoqLr74Yt555x0yMzsvF2o0GuPuj3XB6ZW8uaQtydyAXmaMSZnUeGBsiYs9VaEJ6PwKfPKdwFWzB6E7Vo3VOEYZNJCWnZ8jHymDCpA86mzkpmrc6z9uGyjp0Q+ZSvL0W8kxqjedRzTp28/P/mf2UFblQgAKsxPYWNa2fbq3WiEx0YxCW8JQnwyN/iwmlagj0NX+0ad4vMeXtThxx5ckln4e1p5UsYHCi+852aZpdEGXQpKWlsbixYs5++yzgcDWU2pq9xzmDRw4kFWr2lxDzzrrLN59913S0tKOcVV8cs3ZORgkkbU7W8jPMLK1zMbjbx8EAmck54xN5Yv1jSE7O812H16fjM6gzlWJIIhYr3wY17JX8TdWYhh4OsZxF4HHieJx4tn5LXgc4PPi3bSQlr0bSP7xvLAgRo1wNu+zU3YkXbwCbC4LP4OLNP/Iz1DPdqniDv+ZdDn98VftiTA48vmQ5rkVG7oUkt///vf87Gc/4+GHH0ZRFHJzc3nssceiYZsq2XPYwcufVVLd6OGMYSk8fmt/Pl9Xz6INbdXdDtd7uPiMTKwWiXeX1QbbJ5QkYVKhiMit9TgWvYC/chdS4XDMM25FNLXbmzdaSLjgPmTbb/CVbQw2K7Z6XKs/wHyMyGWNAGVV4TN1SScEswOLAlx1VjZ6SWR1aQuCALPGpqkqnbxh5Dl4zms1LgAAIABJREFUd6/hqLeWLrcY63V/xbX8Ndwr3+76C3QS5qnXda+RGhHpUkgKCwt5++23sdsDs4WEhIRuN+ooX30VXnEvnvH4ZB6cV0ZjayAg7K2lNVhMOgxS+FTR61O4YWYuKVaJ7/bY6Jdn5gqVem7ZP3w8WOTK07QQxefBetHPw8YpjvDkg3JTVbfb1xMYXZzIvC9Cf1dXn5XF/mo3Hq/M+RPSGdrHytA+Vqoa3Eg6kYxkdZWeNRSPQ/h/j+ItXYaQlIlx9LmIOgnL1OvRFw7Fs3UpmBLx7lgeklLeeNqFSL1KkAqGIFrVt3vRE+hSSDweD1988QUVFRX4fG11Nu66665uNUyNlFW6giJylA27WvnFlYW8tbSGpiMRxylWiSnDU9DpBC45I4tLzlCngAAoPm9YpcRI5VABDGPPx/nJP0LajGPndJttPYlUqx6B0MiKA9VufnlVeDqQnDT1bGd1RN97GPre4SUG9EVj0BcFsiX4RkzH+dXLyE1V6AeejmHwFOTmapCOnRxVo/voUkjuuOMOEhMTGTJkSMihu0Y4eekGDJIQ4vrbJ8dEaqKep+4awMINgVnUjNFppCaqa7bYGYKkR0zLR24XD6LL7BNxrGnETHC24lr7IYJkwDTtRiQtx9Zx0er0hYXnNdvDC6gdpbbZg8kgkmjueQm+pewiEq/6IwDOr/9H68t3AyAYE7Be/YiWbicGdHmXVVdX89JLL0XDFtWTaJG45+ICnllQgc3pZ3hRAleflQ1ARrKeq6Zlx9jC7iHh/Huxz/8LckstYnovLLPu6HSsacIlx8zmqhGZvjlmivPN7K5oOyuZMaZtG0eWFURRwOXx88dXD7B+dyuSTuCKqVlcMz0nFiZ3O7K9Gdeqd4KfFbcd1zdvYL3swRhadWrSpZCMGjWKnTt3MnDgwGjYo3rOGpXK5GHJOFwyKdbIv16vT6a22Ut2qkG1rr7tkXoNJunOl/HX7EewpqNLCE8HrnHiPPKjIj5YUUd1g4fJw5KZODgZl8fP3987xPKtTWQk6RleZGX97lYAfH6F1xZXM3loMn1yzDG2/uSjeBzQIaeb7GyNkTWnNl0Kyfr16/nggw/Iz88P2dqKp+y/8YZBEjFYI7u0btpr47E3D9Bo85GVoue31/ahf566YykUnxf7h3/Bu3MliDqMp83BcvbNsTarx5FokbhuRujq4s0lNXy9uQmAmiYvS75rDLuuvNbdI4VEl5qLLrsf/uq9wTbjiPgKRjxV6FJIXnjhhWjY0aORZYW9lU5SrXr+/n55MM13TZOXZz6q4Inb1b2n69m8KCAiALIf9+oPENMLkBsPIyZlYhw+HRQFb/k2BFGHVDAEQTsYPSnsLA/NGt2xaqLJIDJCZWlSjhdP6fIQEZEKh2McMSOGFp26dCkk+fn5ESskahwfdc1efvXSXspr3WE1SOD/t3ef8VGVaR/Hf2dqegiQQocQeq8WUJBeREDUldVFFNuiIrpr37Ws4rqLq6uyLvhYWHHFBVSQqhikKUjvoRNJKCEQ0mcmM3PO82JgwpCQIJPkzAzX9/PxxTlnJrnAkP+c+9z3dUPGKUf5bwwi7jMZZc7ZlrzH+TlGji1LUc8eB5dn9z55KFp12jeLZNuh0tXskWEGfjcgiRXbzhIVbuSu/knEBNG+7b+GfcN8n2NXxi7UolwMkbV0qujqVemS4mnTpvHhhx/ywQcfAHh3SBSXZ87KLDKyPWFRXn/Ca9qUbQUebMwpPX1PKL4TVdXsdG+IgOehqG3FJzVTXIg6fsbB4p9P065pJMN61iEyzEBSnIVh19Thps5xvPNIS6bc15y2TWpu3VeNK68jgjRs1EWlH1WWL1/u3SERIDExUe5ILsOidadZsyuX42dKylzr0TKa7DwnHZIjuXdwPR2qq1rmZp0J6zcBx7q5noefkbFoOccrfI+an13hdXFpG/bm85dZR7zDWLf3iadT82h+2p3H3FXZLNuQwz8eTqFRQpi+hVYRzWnH/uP/cB3djbFha8J7j0WxhBN23W0UzXvN2y7F2mVIufu+i+pXaZDotUNiMFu07jT/+qb8fTYSapl5cVwzTMbQ+eSkqW5KtixCs51bue4oAqMJ3J5nQUpsIlpels97LG1vrOkyQ8YXP2T5PAv5as1p3GrpHWCBzc3XP2YzaXRorNEp/vbf3s3TXJm7UfOyiRr9DJYW12C8/184D2/CWLcxpuTyt3cW1S8gd0gMdmt35ZU517pRBI0SrIztlxhSIQLgPnEANdc3KEwNWmNp1xdDTDzGRu2wr/mckrTVKIoRS5chspbED7YS3yfqF4aI9zWO0Nn0qmTvjz7Hzn0/omkaiqJgjG+MMb6xTpWJ8yoNkgkTJvDjjz8SGRnp3SHx/CZXonyJtS1wuPTYZFR4aVxTakWFxmr2ixli4j3j1Rd0ZDXUaYy1y1DvcUT/CUT0n6BHeSGncbyV9HOdgM+zmhUcTk+gKAoMu6aOHqVVC2NsIu7sdO+xITbRO0Li/GUntlX/QSvKw9KxP2HX/0Y2ttJBhUHidrsZP348s2bNkvD4FX7bL5Hd6UUcO+3AZFS4d3C9kA0RAEN0HcL7jsO2ahaobgx1GhHe6zd6lxWSjp6y+8zSOu98iAB0So6iQ7PgnvKraSrgGVIPH/QQRV9OQbMXolgjvZ0TVFsBhXNeBqcnVO2rZmGIjMPaebB+hV+lKgwSo9GIwWCgoKCA6Gjp83+5EuMsfPBEKw6ftFE3xhzSIXJe2HW3Y+kwALUoB2NCM9ljpJpMX3iM/GJ3ha/Znxm8zzE11Y3t+//Dse1bFHMY4X3HYenQH1Pz7jj3rAaz1btviSszzRsi5zmPbJUg0UGlQ1sRERGMGDGC66+/noiI0hXYf/rTn6q1sGBnMCg+K9aL7G52HC6kfh0rTRJDYzbNxQxRcRiiLr3pmZqfjev4fkwNWmGIDo5d+wLN0VP2Mufq1bZwIqd0dmByveBdxV6ycwWOTZ6uGZqrhOKl/8J9OhPn7pWec4U5FH3zJqbGHTAmNC0zpGpMaKZD1aLSIBk0aBCDBknbAX8cOFbM8x8eptDu+SQ59qYExg0K/mm/v0bJ7pUULXzLMz3YYCJy5FNY2vTWu6yg06NVDMs2lu7F0bZJBA8Ob8AbX/zCyZwSGidYeXRUQx0r9I/r2N6LzmhltinA7cKddRhzclfCBz2EbeV/wGHD3PJawnqOrLFaRalKg2T06NHY7XaOHz9OcnJyTdQUcmavyPKGCMCcVae45fr4SzZ1DEXFKz4ubbCnurCtnClBcgUeurk+ZqPC1kOFNK8Xzv3D6lM31szHf2xNXpE76H+mzE06ULLtgr3YFQPGxGSfViiYrRjPdUUI63Yz1k6D0VwO3105RY2q9KduxYoV/O1vf8PpdLJixQrS0tJ45513mD59ek3UFxIu3jfCrUKxPfj/0f8a2kVdWcvbLVFULsxiZOLIsncciqKExM+TuW0fwk5n4Ni6FCxhKIqBkh3LPReNJgxx9YnoP8Fnb3bFZEYxhcZzyE8++YQvv/wSgNtuu40BAwZw//3307lzZ7Zu3Ur79u0ZM2YM7777Ljk5Obz55pt07NiR4uJiXn31VQ4cOIDL5eLRRx9lwIAB2Gw2nn32WQ4cOECzZs04deoUL774Ih06dOCll15i586dOBwOBg8ezKRJk6647stqkTJv3jxiYjytPNq0aUNmZuYVf8Or0aDuvtt/tm8aSf26wbuL3ZWwdhzgc2yR5nqiHIqiEN7nd9Sa/DlhXYejnj1RetHtImLII5ibd9evwGq0a9cuvvrqK+bMmcP//vc/5s6dS35+PkePHuXee+9l6dKlHDlyhIULFzJ79myefvpp7wf66dOnc+211zJv3jw+/fRTpk6dSnFxMZ9//jmxsbEsWbKExx9/nN27d3u/3xNPPMFXX33FN998w8aNG9m79+JhxctX6UcYk8lUZsaWzNP+dQZ3r0NUmJGfdufToK6Vkb2uvgfN4YMexlC3Me5jezE1bIulyxC9SxIBxn3mGPaf/odamIOlfT/UvFNlXqPmlz0XKjZv3syAAQO8k5oGDhzIpk2baNiwoXc/qJSUFK677joURaFVq1YcO+bpoLF27VpWrFjBxx9/DIDD4eDEiRNs3ryZcePGAdCyZUuffaWWLl3KnDlzcLlcZGdnc+jQIVq3bn1FtVcaJCkpKSxcuBC32016ejqzZs2iS5cuV/TNrma92teiV/urtyupYjAS1n0EdB+hdykiAGmuEgr++xxa4RkAXEe2EnbDXXDBTvWKNRJzcmjejVTkwn2gDAaD91hRFNzu0mev77777mU/x87IyODjjz9m3rx5xMbG8uyzz+JwXHkn8kqHtv785z9z8OBBLBYLf/jDH4iKiuKFF1644m94NXK5NVbtOMvXa7M5lVu2iWMoUG0FOLYsxbHtWzRH8K5jCBa/ZNlZvjmH42eCfxsCAFfGbm+InOc+lU7UHS9ibtETS7u+RN39BoaI4O+WfSndu3fn+++/x2azUVxczPfff0/37pcXnL179+azzz5DO9difM+ePQB07dqVpUuXAnDw4EH2798PQFFREeHh4URHR3P69GlWr17tV+2XvCN56qmnmDp1KnPmzOGJJ57giSee8OsbXc1enHmYrQc9q5E/XX6SqQ81D/pdES+kFuWR/8njaOc6+trXzSPmvndQrKHzZwwkC9ed5v1zTUENBnjmN024sWNw3+0aYhO58O4DwFArAXNKz7LbFISodu3aceutt3L77bcDnoft559NV2bixIm8/vrr3HLLLaiqSsOGDZkxYwa//e1vefbZZxk2bBjJycmkpKQQHR1N06ZNadu2LUOHDiUpKYmuXbv6VbuiaeXtkgHDhg3jk08+4YEHHmDWrFlc/LJatQLnBzctLY02bdroXUa59mUUM/n9Az7nBnaL48nbQqfRnH3D19i+/9DnXMSwSbLCuBqoqsZvXttNoa10SKNhvJX/e/LKxrYDiW3lp9jXzQVNxRDfhOixUypc4Coq53a7cblcWK1Wjh49yvjx41m2bJnPcFlVuOQdyZ133sn48ePJyMjg1ltv9QkSRVFITU2t0kJClVpOTquh05jVo5w/kOZy4srYjSGuvvwyqEKaBg6n7993YbGLD5cc50y+k5s6x9GzdXAO/4T3HYe16zDU4lyMic1lUk8VsNlsjBs3DpfLhaZpvPTSS1UeIlDBHUlGRgaNGjXipZde4pVXXqnyb1yVAvmORNM0nvrgELvTPf2BzCaFqQ+m0KpR6Az7qIU55H/0GFpRLgBKZBya6gZbPhhMRAx+2KcTsPDPvxZksmh96fOEmAijT/+tl37XlGvbBvcGT64TB9DshZgad0AxBv/6mFB3yf9Djz/+OF999RXp6ek1WE7oURSFKfcl8/mKLE7mOBhzQwItG4ZOiAAYomoTM2EaJbt/AMVAyYENuH/Z7rmouihO/QhL+5tQzKHZY6ymPTyiASkNwtmfaSMuysh/U32nxH6/5WxQB0nhV6/jPLcHiaF2A6J/93fZhz3AXTJIVFVl+vTppKen88knZffXvvfee6u1sFAyfeExb3+kbQcL+fuDKSHXuNEQFYe152g0WwGObd/5XiyxodmLJEiqiNGgMLh7HQZ3h4xT9jJBEswr3J1Hd3lDBEDNOYZj82LCb7yr3NertgKUsCgZBtPZJaf/vvXWWxgMBtxuN0VFRWX+E5dn4758nyZ7+cVu5q0OvUVVrhMHyJ/xIHn/HItWnOtzzdSoHYbo0NloKRBsOVDA7B+yyCtycfO1pX+3dWPN3HZjvI6V+UcrLru76MU/TwCuU0fIm/EQeW/fSf4HD+PKOlzmNaLmXPKjS3JyMg8++CCtWrWiT58+NVlTSPl46fEy5y7uvRUKihf9EzXH82f1+WUQHkP48Mk6VRWa/vv9ST5LLd3aeOItDXj/8Zbk5Dvp0CwKizl494IxJ3f1PGMrOus5oRiwtL+pzOuKF7+LesbTqkk9k0nx4neIue+dmixVXOCSQbJgwQJGjhzJoUOHOHy4bNrL0NblyTxddsHYxn0FvPzpEZ66ozGRYUYdqqpamur22QrVhy2fgg9+j7XnSMJvuleGIPykqhpfrsn2OTd31SlGXNeWZknBuw/JeYolnOh7/oFj4wI0eyGWToMwNWxb5nXuU76/k9xyR1Kp5557jpUrV1KnTh0WLVpU5vrPP//MxIkTadjQ0xR04MCBPProo5f1tS8ZJDabDYDiYlml7I+EWhaOnym7mv3ntHw+T83igeH1daiqaikGI6YmnXCdf8B+MdWFY/2XmBq2wdLyupot7irjdKmYTcF7RwJgrJVIxMAHK3yNuVkXnAc3eo9Nyf4tqAskqqqxcnsu83/MJjvPSXysmVG94unbqRYGw5V/ELv11lu5++67eeaZZy75mu7duzNjxoxf/bUrXEcCXHYiifJNvrURz3x4iPImWe8/FjohHXnLHyj+bjquzDQ0W37p3iMXcJ84CBIkfjEYFMbcEO8ztHV7nwR+ybLz9//9wuETdlo1iuDpOxqHdIfpiGGPU7x8Bq7MPZgatCFi0EN6l1QlVFXjtf+ms/VgIfYSz3qh3EIX732dydpdufzprqZXHCY9evSots7tlwyS1157rcI3yla7l6dDchTTHmvJ0g1nWL45B4ezNFE6JYfORjyG6DpEjfH0YHPnHMe2djbOXSt8XmNq0lGP0kLOXQOSaNMkkn2ZxXRoFkn7plFMmrafwyc82/Duyyjmna8z+NsDKTpXWn0MUXFEjX5W7zKq3MrtuT4hcp7dqbL1QCGrduRyU+fqW+C7bds2brnlFhISEnjmmWdo0aLFZb3vkkHSrl07ALZs2cLBgwcZNmwYAMuWLaN58+ZVUHLoOnTcxuodudSOMTGoW22S64XzyMiG9Oscx4zFx8k6W8IN7WO5o2+C3qX6zZ1znJJdK1DMYVg6DcQQEYuxdn2ibvkDjiYdsa+bA2431mtGY27aSe9yQ0bXFtF0beHZ3kFVNQ4cs/lc359pK+9tIcOde5KSHakoZguWjgNDZp3J/B+zy4TIeXanytdrs6stSNq1a8eKFSuIjIxk1apVPPLII3z33XeVv5EKgmT06NEAzJ49m88//xyTyfPSO++8k7vuKn9Ot4CdRwp57sNDuM/9LPywLZe3f5+Coii0aRLJPyeWTXhN00jdepYdhwtp2SCCoT3rYDQG/kNp95lj5H/yOJR4fmk5ti4h5v73USye9SLWTgOxXrSBlVqYg2PTIlRbAdaO/TE1CP4eUXozGBTaNY30dk8A6NAsUseKqo+murH/NBf72tmgemY/OjYvJuaB90OiSWh2ntOv6/6IiiodIenTpw+vvPIKOTk51K5du4J3eVT6VC4vL4/CwkLvcXFxMXl5Zed6C4/F6894QwQ8wwx7Myp+FvLZ91n8Y24Gyzef5V/fHGPaguDYgdKx4ztviACouVk4D2645Os1l5OCT/+I/af/UbJ1CQWznsaVuacmSg15T93RmK4toogMM3BNmxgev7WR3iVVC9vyD7CvnuUNEQA1P5uS/et0rKrqxMdWvGVwZdf9kZ2d7e2puGPHDlRVJS7u8u5+Kl0C++CDDzJ69GiuueYaNE1j48aNPPbYY/5VHMIs5rJ3EtZK5vUv2+i7D8P3W84y8ZYGAT/7RjGW0/ytgr2zXelbUXNLHxKjunHs+L7c6Z3i10mMszDlvtAectY0Dcf28odayv1ZDEKjesXz3teZ2J1lh7fCzAZG977yxaZPPvkkGzZs4OzZs9x444089thjuFyeQB47dizffvsts2fPxmg0EhYWxltvvXXZ0/UrDZIxY8Zw4403sn27Z2rnH//4R+Ljg3flbHUb3SuetbvysDk8PwjXtY0huV7p/P68IhduVaN2dOkv3KhwIzkFpZ+wIqwGjH5M86sp1i5DcGxbhlboWblvrNcSc/MePq9x52ZhiIhBsYSjWMtOLlDCQmfCgfCfpmk49/2E++RBTE06Ym5Wuhuroigo1gg0l+90eiWyFrb183Ae2Ux4n3EYoiofiglUfTvVYu2uXLYeKPQJkzCzgS4toujjx74zb731VoXX7777bu6+++4r+tqX7P4bTAKt+29OgZP1afnUjTHTrWW0NxT+/c0xFv18GlWF5vXDGdWrLjd2rMWmfQX8dfYvuNwaigKPjmzIsGuCo6WIai/EuX8dmq0QTXVjSmiKKbkbWtFZCue+ivvEfjCHETHgAaxdhlA471Wc+9cDoETXJeaef2CIufr2sK8KZ/KdrNh6FpNRoX+XOGIig7fH1nlFi9+h5IK7jvCBDxHW4xbvsWP7coqXvAuaCigYE5vjzjrovW5s0JqYe/5RkyVXOVXVWLUjl6/Xlq4jGd07nj4d/VtHUp0kSKrBrvRC/jX/GJmnHVzX1jNevT+zmOc/Krv6tnWjCN58OIWcAie704to0SCCBkE2/7/kwAaK5r167h83WLsOQ9OgZOuS0hcZzcQ+9h8MEbE4j+5CsxVgTu4ijRx/Bbdbw61pWEwGTuc5eeS9feQXedbrJNQy8/7jrYK6U4IzI43CWX/0OWeolUTsxI98zrnPnsB9fD+mhm0omP2CtzXPebGTZgX1XUkwCv6PMAHG6VKZ8t9fyC30DFWt2ZlHrSgTjeLL/4W5N6OYHYcL6ZISTd9OwTnO61g3zxsiAI6tyzDWb+X7IrcTNTcLQ0Qs5sbta7jC4PfNT6eZtfwkdqdK/y5xJMaZvSECcCrXydpduQzuHhx3suVxbF5Y5pzqLtuXzhhXD2NcPQAMcfV9gkQJj0EJj66+IkW5ruhpbt++fau4jNBxIqfEGyLnpf1STLeW0ZguMaXX7Q7um0JNK7uK3XxRywolJh5jYnJNlRRSjp6y8++Fxyi0u3G5Nb7dlMPB42XXiQTDc7UKldMNwRBe8W6PEf0mnNvvHbCEEzFkIoqx+mY2ifJd0R1JCIyGVZt6tS3ERZk4e0GYtGkSQf06Vl4e15RPl5/kwDGbt2VKs6QwuqQE9yeosB6jKDr2N8Dzh7K060tY7zvBYMS5dy2G2ETC+4yTne6u0MFjZUPDajZQJ8bMmXzPuoKG8VZ6tQvezawAzK174dy71vdkOR9SLmSMb0zM7/8P9UwmhtgEFEvwN64MRlf0L1s6uF6a2WTg+bua8K8Fx8jMdnB92xjuGeS5De/WMoZuLWM4keNg5fZcosOM9OsaFxSLDytiaXsDhth4nAc3YqzbGHOb3iiKgfBevyG812/0Li/otW8WidGAz/qknq1i+P2IBqzZmYfJqNC7Qyzh1uB9PgJgadMb2/cfohWWToc3N6m8G4JiMGKMb1KdpYlKXPJhe3m7IoLnbmT69Ols2HDphWc1LdAetgtR1X7ances5ScptLsZ1rMOY/sl6l1StXCdOoLt2+m4z2RgbtGTiIEPyV1GFTlx4gRPP/00Z86cQVEU7rjjDu655x6f12iaxpQpU1i1ahVhYWG88cYb3nZZFbnkHUlFuyCOGzfuV5QvhPDX9e1iuT7Ih64uhymhGdG/+5veZehK01RKdq/CsWE+av5pDDF1sfYchaVdHxTlyhcpG41Gnn32Wdq1a0dhYSFjxoyhV69epKSUNvdcvXo16enpfPfdd2zfvp2XX36ZuXPnVvq1LxkkFbWPnzlz5q/7EwghhKiUpqkUfTkF55Ft4PR0c3YX51K8dBrOvT8SOeb5Kw6ThIQEEhI8jWKjoqJITk4mKyvLJ0hSU1MZNWoUiqLQuXNn8vPzOXXqlPd9l3JFFVV3kMyaNYshQ4YwfPhw/v73v1fr9xJCiEBRsnuVT4h4Oe04j2zFuWd1lXyfzMxM0tLS6NTJ9xlUVlYWSUlJ3uOkpCSysrIufnsZATdra/369aSmpvLNN99gsVg4c+ZM5W8SQogQ4Ngwv2yInOe0Y//5ayzt+vr1PYqKipg0aRLPP/+8T8dff1zRHUl1ztqaPXs2Dz74IBaLZ3FenTrBu8BKCCF+DTX/dMXXCyq+Xhmn08mkSZMYMWIEgwYNKnM9MTGRkydPeo9PnjxJYmLlEzsueUfSpUuXcgND0zQcDsfl1v2rpaens2nTJt5++22sVitPP/00HTtWvLOew+EgLS2t2moSQogr9WtmlBpi6uIuzr309egr70unaRovvPACycnJ3HvvveW+pl+/fnz22WcMHz6c7du3Ex0dXenzEaggSLZu3XrFBVdm/PjxnD5dNlknT56M2+0mLy+POXPmsHPnTiZPnkxqamqFd0FWqzVopv/+kmVnx+FCUhqE06ZxaG4+dCHN7cS5fz1aiR1zy2sxSPsKIS7J2nMUxUunlT+8ZQ4j7JrRV/y1N2/ezIIFC2jZsiUjR44EPK3ljx/3tJgZO3Ysffr0YdWqVQwcOJDw8HBef/31y/raAde0ccKECTzwwANce+21AAwYMIA5c+ZUuEtXsKwjWbntLFPnHEU99zd+z6Ak7rwpNNcDAGhuFwWznsJ9fD8ASlRtYsa/Ld1+hbiE8mZtAWAOw9ysi1+ztqpTwFU0YMAAfv75ZwCOHDmC0+m87F26At3nP2R5QwRgzspTuIK8z1ZFnIc3e0MEQCvMwbFtmY4VCRHYFMVA5JgXiBz2GMakFJTIWhiTUogc9ljAhggEYPffMWPG8Pzzz3PzzTdjNpt54403QqYli9PpGxpOt4aqahDkLVIu6aINiIAymxIJIXwpigFLu75+z86qSQEXJBaLhTfffFPvMqrFLdfX5YPFpS2vh/asjaWSbXiDmTmlB4ZaSai552aBWMKxdio7U0QIEdwC7hnJlQiWZyQAm/bls+1QISn1w7kxgHc8qypqcT4lO5ajOYqxdOiPsXZ9vUsSQlQxCRIhQlDaL0V88UMWxSUqw3vWoW/n0HjOKAJTwA1tCSH8k1Pg5LmPDuNwevrO7zpSREykia7Y8zjIAAAST0lEQVQtZOq1qB4SJEIEoV+y7Eybn8mRkza6tYzh0VENiA73/HPeeqDAGyLnrduTJ0Eiqo0EiRBBRtM0XvtvOpnZng4Tq3fkYjUrPHlbYwDq17GWeU9554SoKqE7ZUjoSnM7KdmzGvuGBaj52XqXE1LOFrq8IXLezsOl+we1aRLJyOvrcn4eR5eUKIb2lJ51ovrIHYmoFoVfvIjrlx0A2FbPInrcVEwJzXSuKjTUijSRUMvMqVyn91zLhr67CD48ogF39E3A4VSpV1vuRkT1kjsSUeVcmWneEAGgxIZj4zf6FRRiDAaFZ8c2oVG8JyA6NY/ioZsblHld7WizhIioEXJHIqpeeTPKNbXsOXHF2jSO5IMnW+Nya5hCtTOCCBpyRyKqnLFhG4wNWpeeMFmwdrtZv4JCmISICARyRyKqnKIoRP92CiW7V6EWnsXS5gaMdcoOvQghQoMEiagWijkMa+fBepchhKgBMrQlhBDCLxIkQggh/CJBIkQQcqsaOfnOyl8oRA2QZyRCBJnthwqZOucoZ/KdNEkM4093NaVhvKwXEfqROxIhgoiqarw1zxMi4GneOGPRMZ2rElc7CRIhgkixQ/VpjQJw5KRdp2qE8JAgESKIRIUbadUowudc95bSHl7oS4JEiCDz/G+b0Kt9LEm1LQztWZsHb5bti4W+ZKtdIYQQfpE7EiGEEH6RIBFCCOEXCRIhhBB+kSARQgjhFwkSIYQQfpEgEUII4RcJEiGEEH6Rpo1ChJAtBwpYtP40FpOBMTfG06JBROVvEsJPEiRChIj9mcX8eeZhVNVzvGFvPh/+oTW1Y8z6FiZCngxt1TBN09jzSxF7fikiBJoKAKAW5uA8tAm1OF/vUq5qa3bmekMEwFai8vPe0P1/4kzfTt70hzj7xkgKv/4bWolN75KuWnJHUoNKXCovfHyYXUeKAGjXNJLX70vGYg7ePHfs+oHiRf8E1QUmK1FjnsfcvLveZV2VEmpZyjkXmncjmtNB0Vevo9kLAXCmrcYWU4eI/vfrXNnVKXh/gwWhtTvzvCECsDu9iFU7cnWsyD+apmFL/cgTIgAuB8UrPta3qKvYwG5xtG8a6T3u07EWXVJCszOw+0ymN0TOcx3bq1M1Qu5IalBOQdmtUc8WunSopIqobjSb79CJVhS8wRjswixGpj6UwqHjNqxmhYbxYXqXVG2MdRqihEej2Qq850wN2+pY0dVN7khq0A0damG9YBjLalbo3T5Wx4r8oxhNWNr28Tln6dBPp2rEec3rh4d0iAAoZisRo55GiaoNigFT086E9/6t3mVdteSOpBptPVjA9IXHOJXr5IYOsTwysiH/eDiFhetOo2oat1xXl/p1gnuv7YhhkzDGN8Z1fD+mxh2wdhvuc11z2ile9j4le3/EUCuRiEEPY27SUadqRTBR809TtPgdXBm7MNVrScTwSRhrN/BeL9myDK0wBwBX+jZK9q/D2v4mvcq9qsl+JNXEXuLm7r/uocheOo1mbL9Exg1M0rGqmmf7YSb2dXO9x0p4NLGP/gfFHNwBKqpfwf9ewnVok/fYWL8VMePfAjwzBfPe/Z3P640NWhNzzz9qtEbhIUNb1eTISbtPiAAs+DGbf36ZQU5+2Wclocp5dJfPsWYrwH36qE7ViGDiytjtc+w+vg/Nfe6ZosEEBqPPdcVUdtaaqBkSJNWkSWIY4Vbfv95ih8q3m3J45v8OsS+jWKfKqpc7+xdK9q9Hc5z786kXTSZQDBjrNKr5wkTQMTXwHWUwJqWgGD2j8YaIGKzdR1xw0UTY9XfUZHniAjK0VY027cvn3wuPcfxMSbnXb+xYi+fGNqnhqqrPhcNYSng0Ub99naJ5r6HmZfm8LvbxzzFEBu8kA1Ez3HmnKF74Fq6juzDWb0HkzU9grNvY5zXOo7tQz2RgSu6GMTZBp0qFBEkNeGnmYTbsKyj32tu/T6F148hyrwUTtTCHvPfuAa10OM/cuheaw4bryBbvOSU8hthJs7yfLIW4FHfuSQxh0Shhwf/vI9TJ0FYNuG9ofRLjyh+/zS9213A11UOzFfqECIBWnE9E/wkYYuI9JyzhRAx5RELkCmXnlbBxXz6FttD4mbkUtTif/P/8gfz3J5D77t3YN3ytd0miEvIvugY0SQzjoz+2ZsXWHN6el8n5W8CEWmY6N4/StbaqYoxvjLFBa9wXrC62dBqIMaEpMRM/Qj2dgaFWIoolXMcqg9fSDWeYtiATVYVwi4GX72lGx+TQ+Nm5mH3d3NKfI1cJttSPsbS+AUNMXX0LE5ckQVJDjAaFgd3qkBhn5fstOcREmBjZq25Q99m6WNQdL+PY+A1q7gnMrXphaXUdAIrBiDGhqb7FBTGnS+XjpSe8DRltJSqfLDvB2xNb6FtYNVFzMn1PaCrunGMSJAFMgqSGdUyOCtlPkobwaMJvvEvvMkJOiUujyOE7nJUbzK11KmFucS3OAxu8x0pELUwNWutYkaiMBIkQAS4yzMh1bWP5aXee91z/rnE6VlS9rJ0Ho5XYKNm9EkNUbcL6/E4WsAY4mbUlRBCwl6h8vTabQydsdE2JZmjP2iiKondZQgABeEeSlpbGSy+9hMPhwGg08vLLL9Oxo/RmEle3MIuBsf0S9S5DV1qJHeeRrRii4mSoK8AEXJBMnTqVRx55hD59+rBq1SqmTp3KrFmz9C5LCKEj99kTFHz6FFrRWQAsHfoTOeJJnasS5wXclCFFUSgq8mz+VFBQQEKCrFYV4mrn+Pkrb4gAlOxMxX0qXb+ChI+Ae0Zy6NAhJkyYgKZpqKrKF198QYMGDSp8z7Zt27Ba5WGcEKEqdsNnhB/b5nPuzA2/x1m3uU4VXb6r4fmtLkEyfvx4Tp8+Xeb85MmTWb9+PT169GDw4MEsWbKEOXPmMHPmzAq/njxsFyK0OQ9vofCLF+Hccl5DnYbEPPA+ykUdgIU+Au6OpFu3bmzatAlFUdA0jW7durFly5YK3yNBIkToc6Zv904JtnYfgSGylt4liXMC7mF7QkICGzZs4JprrmH9+vU0bdpU75KEEAHA3LQT5qad9C5DlCPgguTVV1/l9ddfx+VyYbVa+ctf/qJ3SUIIISoQcENbV0KGtoQQQj8BN/1XCCFEcJEgEUII4RcJEiGEEH6RIBFCCOEXCRIhhBB+kSARQgjhFwkSIYQQfpEgEUII4RcJEiGEEH6RIBFCCOEXCZIAsuNwId/8dJrMbLvepVQJ18lD2Dd+g+v4Pr1LESFIzc/GsXkxzoMb0TRV73KuagHXtPFq9dHS48xbnQ2AwQB/vqsp17aN1bmqK+fYuozipe95j8MHPEBYz1E6ViRCiev4Pgo+ew5cDgDMbW8katQzOld19ZI7kgBQZHcz/8fSjb5UFb5YeUrHivxn+/ELn2P7RcdC+MO+/ktviAA496zGnXNMx4qubhIkAUDVNFTVtwmz0xXkt+oup8+h5nISAo2mRaBwu8qeu+hnTtQcCZIAEB1uon/XOJ9zo3rF61RN1bD2HOlzHNbjFhRF0akaEWqs3UeAUvrry9S0E8aEpvoVdJWT/UgChNutsWpHLulZNrq3jKFjcpTeJfmt5MAGXBm7MNVvhaV1L73LESHGdfIgzr0/YohNwNK+H4rZqndJVy0JEiGEEH6RoS0hhBB+kem/QoiAouZn49i6DDQNS6eBGOPq6V2SqIQEiRAiYKhFeeR/PBmtOBcAx5ZFxNz/LwwxwT35JNTJ0JYQImCU7F3jDREAzV5Eye6V+hUkLosEiRAiYCjmsLInyzsnAooEiRAiYFja9MaY0Mx7bKjdAEv7m3SsSFwOeUZSTRxOFYtJkUV4Qvwaqoql6zDcJw561h+17yvrQ4KABEkVyylw8vcvjrL9cCFJtS08MaZRSCwuFKK6aSV28mc+gXomEwDnoY2YW/SQIAkCMrRVxT5acoLthwsBOJlTwt+++AW3O+jXfApR7Ur2/+QNEQCtMIeSHd/rWJG4XBIkVWz/sWKf45wCF2cKpJmcEJVSyzYqlX1GgoMESRXrdNEwVv06FuJjzTpVI0TwsLS6HkOtJO+xEh6DtcMAHSsSl0uekVSx+4bWw+FU2bAvnyYJYfz+lgbywF2Iy6BYI4i+95+U7PoB3E4s7W/CEFVb77LEZZCmjUIIIfwiQ1tCCCH8IkEihBDCLxIkQggh/CJBIoQQwi8SJEIIIfwiQSKEEMIvEiRCCCH8IkEihBDCLxIkQggh/CJBIoQQwi8SJEIIIfwiTRtFtXIe3IhtzX/R7EVYuwwh7NoxepckhKhiEiSi2qgFpyn88jVwuwCwrfgYQ0wClrY36FyZEKIqydCWqDauo7u8IXKe88gWnaoRQlQXCRJRbYwJzcqeS0zWoRIhRHWSIBHVxhjfhPD+94MlHBQDlvY3Ye0yRO+yhBBVTDa2EtVOcznB7USxRuhdihCiGuhyR7J06VKGDx9O69at2blzp8+1GTNmMHDgQAYPHsyaNWv0KE9UMcVklhARIoTpEiQtW7bkvffeo0ePHj7nDx48yOLFi1m8eDEffvghr7zyCm63W48ShRBCXCZdgqR58+YkJ5d96Jqamsrw4cOxWCw0atSIJk2asGPHDh0qFEIIcbkC6mF7VlYWSUlJ3uPExESysrJ0rEgIIURlqm1B4vjx4zl9+nSZ85MnT2bAgAFV+r0cDgdpaWlV+jWFEKIqXA0TgaotSGbOnPmr35OYmMjJkye9x1lZWSQmJlb6PqvVelX8zxJCiEAUUENb/fr1Y/HixZSUlJCRkUF6ejodO3bUuywhhBAV0KXX1vLly3n11VfJycnhoYceok2bNnz00Ue0aNGCoUOHMmzYMIxGIy+++CJGo1GPEoUQQlwmWZAohBDCLwE1tCWEECL4SJAIIYTwiwSJEEIIv0iQCCGE8EtI7JAoCxKFEIHKZDLRokULvcuoViExa0sIIYR+ZGhLCCGEXyRIhBBC+EWCRAghhF8kSIQQQvhFgkQIIYRfJEiEEEL4RYIkBOXn5zNp0iSGDBnC0KFD2bp1q891TdN47bXXGDhwICNGjGD37t06VSqCwcyZMxk+fDg333wzTz75JA6Hw+d6SUkJkydPZuDAgdx+++1kZmbqVKnQiwRJCJoyZQo33HADy5YtY8GCBTRv3tzn+urVq0lPT+e7777j1Vdf5eWXX9anUBHwsrKy+PTTT/nyyy9ZtGgRbrebxYsX+7xm7ty5xMTEsHz5csaPH8+bb76pU7VCLxIkIaagoICNGzdy2223AWCxWIiJifF5TWpqKqNGjUJRFDp37kx+fj6nTp3So1wRBNxuN3a7HZfLhd1uJyEhwef6ihUrGD16NACDBw9m3bp1yDrnq4sESYjJzMykdu3aPPfcc4waNYoXXniB4uJin9dkZWWRlJTkPU5KSiIrK6umSxVBIDExkfvuu4+bbrqJ3r17ExUVRe/evX1ek5WVRb169QBPO5Do6GjOnj2rR7lCJxIkIcblcrFnzx7Gjh3L/PnzCQ8P54MPPtC7LBGk8vLySE1NJTU1lTVr1mCz2ViwYIHeZYkAI0ESYpKSkkhKSqJTp04ADBkyhD179vi8JjExkZMnT3qPT548SWJiYo3WKYLDTz/9RMOGDalduzZms5lBgwaVmbyRmJjIiRMnAM8HmYKCAuLi4vQoV+hEgiTExMfHk5SUxOHDhwFYt25dmYft/fr1Y/78+WiaxrZt24iOji4z7i0EQP369dm+fTs2mw1N0y758/T1118D8O2333LttdeiKIoe5QqdSPffEJSWlsYLL7yA0+mkUaNG/PWvf2XJkiUAjB07Fk3T+Mtf/sKaNWsIDw/n9ddfp0OHDjpXLQLVu+++y5IlSzCZTLRp04YpU6bw73//m/bt29O/f38cDgdPPfUUaWlpxMbG8vbbb9OoUSO9yxY1SIJECCGEX2RoSwghhF8kSIQQQvhFgkQIIYRfJEiEEEL4RYJECCGEXyRIhDgnMzOTm2++We8yhAg6EiRCCCH8IkEiRDkyMjIYNWoUO3bs0LsUIQKeSe8ChAg0hw8f5sknn+SNN96gdevWepcjRMCTIBHiAjk5OUycOJFp06aRkpKidzlCBAUZ2hLiAtHR0dSvX5/NmzfrXYoQQUOCRIgLmM1mpk2bxvz581m4cKHe5QgRFCRIhLhIREQEM2bMYObMmaSmpupdjhABT7r/CiGE8IvckQghhPCLBIkQQgi/SJAIIYTwiwSJEEIIv0iQCCGE8IsEiRBCCL9IkAghhPDL/wMMyiRe0IpBywAAAABJRU5ErkJggg==\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "df_rappas = df2[df2[\"software\"] == \"rappas\"]\n", + "sns.catplot(data=df_rappas, x=\"k\", hue=\"omega\", y=\"LL difference\", dodge=True, palette=\"muted\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Put it all together\n", + "\n", + "Let's make a final plot, combining all the plots." + ] + }, + { + "cell_type": "code", + "execution_count": 260, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3UAAAMQCAYAAACAN0zJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdd3RUdf7/8ef0SS+k0EKHgBCqFKlKd5WiiA2xre2nK+q6urruYl9Xv7uuLuKKK4hdVsUEpKoooGBDEJXeAyQB0tv0+/sjZtYhQYqBMOH1OIej987n3nnfO5O5930/zWQYhoGIiIiIiIiEJXN9ByAiIiIiIiInTkmdiIiIiIhIGFNSJyIiIiIiEsaU1ImIiIiIiIQxJXUiIiIiIiJhTEmdiIiIiIhIGFNSJyKnvblz55Kens7u3btP2XtOmzaN9PR0fD7fEctMnjyZyZMnB5e//PJL0tPTWbVqVZ3EkJ6ezrRp02rEdHiZf/7zn3Xyfr9WbfEdq71795Kens7cuXOD6+677z6GDh0aUm779u1cffXV9OzZk/T0dD766CMA3n33XUaOHEmXLl04++yzT/wgGoCXX36ZMWPGUJ8zFqWnpx/xX/VnBjB06NDg+o4dOzJkyBCmTJnC9u3ba93vc889R3p6OlOmTDlVh3LCioqKuP/+++nbty89evTg+uuvZ+vWrce07cyZM7n55psZMGAA6enpPP/88zXKGIbBmDFjePnll+s6dBEJQ9b6DkBEJFw9+OCDp/T9Jk6cyKBBg07pe9anW2+9lauvvjpk3d/+9jf27t3LM888Q2xsLK1btyYvL4+pU6cyZswY/vrXv+JwOOop4vpXUlLCCy+8wMMPP4zJZKrXWC6++GIuu+yyGutbt24dsjxw4EBuv/12AoEAO3fuZNq0aUyaNIkFCxbQqFGjYDnDMMjMzATgk08+obi4mLi4uJN7ECcoEAhw8803B7+b0dHRzJgxg8mTJzNv3jxSUlJ+cfs5c+YQHx/PsGHDmDNnTq1lTCYTt956Kw899BATJkwgNjb2ZByKiIQJJXUiIieoXbt2p/T9GjduTOPGjU/pe9anFi1a1Fi3fft2zj77bAYPHhxct3nzZvx+P+PHj6+TWjq/349hGFit4XeJfPfdd7HZbIwYMaK+QyElJYXu3bsftVxCQkKwXM+ePUlLSwsmP9ddd12w3Jo1a8jOzmbIkCEsX76chQsXcsUVV5y0+H+NDz/8kHXr1vHGG28Ev5Pdu3dn2LBhzJw5k/vvv/8Xt1+8eDFmsxm3233EpA5gxIgRPPLII7z33nsh50pEzjxqfikiDUZWVhZjx44lIyODvn37cs8993DgwIGQMpWVlTz44IPBJlG33XYb3377bY2mf8fi8OaXtcnOzmbkyJFcfvnlFBcXB9fPmTMnJNY//elPFBUV/eK+fql546uvvsrQoUPp0aMHV111VY1mXoZhMHv2bEaNGkWXLl0YOHAgjzzyCGVlZSHlysrKeOSRRxg4cCBdunRh1KhRzJ49u0ZTvg0bNnDllVeSkZHBoEGDmD59+jE396usrOShhx4Kfga33HILubm5Ncr9vPllddPWffv2kZWVFWyyd9999wU/g2uvvTa4rtqxnOfqJqwvvvgiQ4cOpUuXLmzZsgWAgoICpk6dyqBBg+jSpQujR4+ucZNd3Tx43bp13H333fTs2ZOBAwfy2GOP4Xa7Q8pWVFTw97//neHDh9OlSxcGDBjA7bffzqFDh4JlsrOzufvuu+nXrx9dunRh3LhxfPjhh8d0bt955x1Gjx6NxWIBwOPx0KdPH5544okaZRcuXEh6ejobNmwAYP369Vx33XX07duXrl27MmzYMB566KFjet+6lJGRAVCjufX777+P1Wrl8ccfJyUlJVhrd6J2797NDTfcQLdu3ejfvz9PPfUUb775Junp6bV+H4/HsmXLaNq0achDhri4OIYMGcLHH3981O3N5mO7PbNarYwaNYp33nnnhGMVkYYh/B5DiojUYs6cOUydOpXf/OY33H333Rw4cICnn36a9evXM3fuXKKiogD4y1/+wuLFi7n99tvp0qULq1ev5g9/+MNJiWnDhg3ceOONZGRk8Mwzz+B0OgH4+9//zssvv8zkyZO59957ycvL45lnnmHr1q28/fbbwRvyYzV//nxat27NAw88gNfr5amnnuLWW29l0aJFwdqmf/7zn8yYMYNJkyZx3nnnsX37dp599lk2bdrE66+/jtlsJhAIcNNNN7FhwwamTJlChw4d+PTTT3niiScoKCjg97//PVCV6FxzzTUkJSXx5JNPYrfbeemll8jJyTmmeKdOncqiRYu47bbbyMjI4PPPPz/qZ9C5c2fmzJnD//t//4+MjAxuvfVWABITE+ncuTOPPfYYU6dOpXPnziQmJh73eZ47dy5paWn88Y9/JCIigpSUFMrKyrjiiitwu93cfvvtNG/enJUrV/LQQw/h8XhqJPT33nsvF1xwAc899xxr167lueeeIzY2Ntj/y+PxcP3117Np0yZuvPFGunfvTmlpKZ999hnFxcUkJSWRk5PDpZdeSqNGjbj//vtJTExk4cKF3H777UyfPp1hw4Yd8Rzt27ePHTt2cMcddwTX2e12Ro8ezYIFC7j33ntDjnnevHl06NCBs846i/Lycm644QYyMjJ44okniIqKYt++faxdu/aYPtMjqa1P6tFqQPfu3QsQ0pzQ5XKxePFiBg0aRHJyMmPGjGHmzJns3LmzRnPOY+HxeLjuuuvw+/08/PDDxMfHM2fOHBYtWlSjbCAQIBAIHHWfFosl2OR169attG/fvkaZ9u3bs2DBAlwuV/D34Nfq3bs3b731Fjk5OTRp0qRO9iki4UdJnYiEPb/fz7PPPkufPn1CBg1p3bo1kyZN4r333uPqq69mx44dfPDBB9x9993ceOONAAwYMACXy8Vrr71WpzGtXr2a2267jdGjR/Poo48Gb6b37t3LzJkzue222/jd734XLN+qVSuuvPJKPvnkE4YPH35c72W1WnnhhRew2WzBdXfccQfr16+nZ8+eFBUVMWvWLC666CKmTp0KwKBBg0hISODee+/lk08+YdiwYSxfvpw1a9bwxBNPcPHFFwNV/Z0qKyuZNWsW1157LYmJibzyyivBddU3kf379+e88847aqzVn8Fdd93FTTfdFHyPiooK3n777SNuFx0dTffu3bHZbCQmJoY062vbti1Q1Ry2ev3xnmfDMJg1a1bIjfb06dPZv38/8+fPp1WrVsHjLC0t5bnnnuOKK64ISVAuvPDCYALXv39/1q9fz4IFC4Lr5s2bx9q1a3n++edDkrPRo0cH/3/atGkYhsFrr71GQkICUPVZ5ebm8q9//esXk7rvvvsOgI4dO4asHzduHHPmzGHVqlXBPpkFBQWsXLmSO++8E6j6XIqLi7nnnntCtq/+HpyIF154gRdeeKHG+tWrVwcTb6g69z6fj0AgwK5du5g6dSpms5lRo0YFy3z44YeUlZUxbtw4AC666CJmzpxJZmYmd91113HH9s4777Bv3z7mzp1L586dARgyZAhjxoypUfbee+9l/vz5R93nU089FYyvuLiYNm3a1CgTFxeHYRiUlJTUWVLXqVMnANauXaukTuQMpuaXIhL2du7cSX5+PmPHjg1Zf/bZZ9OsWTO+/vproKp5mWEYITfRQMjNY11YvHgxN954I1dddRV//etfQ2pHVq1aRSAQYOzYsfh8vuC/bt26ERUVFYz1ePTv3z8koevQoQNAsObsu+++w+v11jg/F1xwAVarNfieX3/9NWazmQsvvDCk3NixY/F6vaxbtw6ounns1q1byA1kZGRkjZEqa7N+/XoCgQDnn39+jVjq0vGe50GDBtW4yV65ciXdunWjefPmIfsYOHAgRUVFbNu2LaT8ueeeG7LcoUMH9u/fH1z+/PPPSU5O/sXEbOXKlQwZMoSYmJga77lp06YazWV/rrqp8c8TJoBevXrRokULsrKygusWLFgQPD9QlezGxsby4IMPkpWVdcy1rr9kwoQJvPvuuzX+HT6gxwcffEDnzp3JyMhgzJgxHDhwgGeffTaYbAFkZmYSGxsbPHft27enc+fOzJs374RG+fzuu+9IS0sLeQ+TyVRrX8Q77rij1uM4/N+QIUOOO466UP15H97UXETOLKqpE5GwV91HKjk5ucZrSUlJwderb3p+PqJebcu/1pIlS3A6nVx00UU1XsvPzwc44kAWR+tXV5vDRwC02+0Awf5cRzo/VquV+Pj4YF+/6tEEq7evlpSUFHwd4ODBg7U2LTuW83iqPoPjPc+1jUZYUFDA7t27Q278f2kftX0OHo8npPzRRj0sKCggMzPziP3FCgsLiY6OrvW16s/78M8PqhLzWbNmUVFRQWRkJFlZWfTr14/U1FQAYmJiePXVV3n++ed5+OGHKS8vp3379tx+++0n/NAjOTk52D/ulwwePJgpU6ZgsVhISUkJft+q5eXlsWrVKsaMGYPL5cLlcgFVD2OefvppvvzyS/r163dcsR08eLDW79zh7w3QrFmzY6oB+/nDm7i4OEpKSmqUKS4uxmQy1elIldUPIw7vvykiZxYldSIS9uLj44GqG7XDHTp0KHhTXn1DnZ+fT2RkZLBMdQJQVx599FFmzZrF5MmTefXVV0OaYVXHOmvWrFpv7Kpfr0vV+zx06FBIMubz+SgqKgomI3FxcRQXF+PxeEISg+pBPKrLJScn13rOjuU8nqrPoC7Oc3x8PImJiTzwwAO1vn68fbkSEhKOOk9ZfHw8vXr1CjYPPlx1EnakbaEqcTi81nHcuHE899xzLF26lG7duvH999/z5JNPhpTp1KkT06ZNw+fz8cMPPzBjxgzuvPNOsrKygrW/J0NcXNwvJn9ZWVkEAgGysrJCahurZWZmHndSl5yczLfffltj/c8HrKl2Is0v27VrxzfffFOjzLZt22jevHmdNb2E/z1sqW6uKyJnJiV1IhL2WrduTVJSEgsXLmTixInB9d9++y379u0LDvWdkZGByWQKNo+stnjx4jqNJzo6mpdeeokbb7yRq6++mldeeSXY72vAgAGYzWb279/PgAED6vR9j6Rbt27YbDYWLFjAOeecE1y/cOFCfD4fffr0AaBPnz7MnDmTxYsXhzTVnD9/PjabLdhfrUePHsycOTNkYIaKigqWLVt21Fi6du2K2Wxm0aJFwT51UNUcsC7VxXkeNGgQr7/+Ok2bNq2TmsQBAwawYMECli1bdsSmqoMGDWLt2rW0b9/+uG/8qx8eZGdn10j+WrRoQY8ePZg3bx67du0iMjLyiLWYVquV7t27c8cdd7Bs2TK2b99+UpO6o8nMzCQtLY3HH3+8xmszZsxgyZIlTJ06NeQhwdF069aNrKwsfvzxx+BDH8MwWLp0aY2yd9xxB9dcc81R95mWlhb8/6FDh5KVlcWaNWvo1asXUDWH4Keffvqr+inWJjs7Gzj+hwwi0rAoqRORsLFy5Uo2btwYsi4mJoYBAwYwZcoUpk6dyh/+8AfGjh0bHOmwVatWTJgwAagaUOPCCy/k2WefxTAMOnfuzBdffMEnn3wC1D6M+NKlS2usT05ODt6oHUl1YnfzzTcHE7t27drRokULbrzxRh599FF27txJnz59cDgc5OTk8PnnnzNx4sTjrnU4mvj4eK6//npmzJhBREQEQ4YMYfv27TzzzDP06tUr2Bds8ODB9OrViwcffJCCggLat2/P8uXLeeedd7j55puDfXeuueYa3nzzTa6//npuv/324OiXx5KEtGnThgsvvJB//etfBAIBMjIy+Oyzz1ixYkWdHnNdnOdrr72WhQsXcuWVV3LttdfSunVrKisr2bFjB9988w3//ve/jyumsWPH8s4773D33Xdz00030a1bN8rLy1m5ciXXXHMNbdu2ZcqUKUycOJFJkyZx1VVX0axZM0pKStiyZQvZ2dm1Tk1QrWvXrtjtdr7//vta5+sbN24cjzzyCFu2bGH48OHBEWGhajLvOXPmMHz4cJo3b05lZSWvvfYaUVFR9OjRA6gaXXPEiBHceuutIYPPHMmBAweC/TB/rmnTpkdthlpt/fr1bN++nTvvvJO+ffvWeL2iooLPP/+cpUuXMn78eHw+H507d+aSSy6pNQmsNnHixOBAOnfeeScJCQm8/fbblJeXA4RM3J6WlhaSsB2LkSNH0rVrV/7whz9wzz33BCcft1gs/Pa3vw2W8/v9ZGRkcMkll/DII4+EHPf+/fuDo4du27Yt+PDp3HPPDflbW79+PXa7nW7duh1XjCLSsCipE5Gw8eijj9ZY1759ez744AMuu+wynE4nM2fO5NZbbyUqKorBgwdzzz33hDzBf/TRR4mKiuKll17C6/XSt29fpk6dys0331xrX6XaRtY799xzmTFjxlHjjYqK4sUXX+SWW27h6quvZvbs2XTo0IHf//73tGnThjfffJM333wTk8lE48aNOeecc4KjLNa1u+66i8TERN566y3eeust4uPjGT9+PHfffXcwaTWbzbz44os8/fTTvPTSSxQVFdGsWTPuv//+kJqKxMREZs+ezeOPP84f//hH4uPjufzyy/H7/UyfPv2osTzyyCNERkYya9as4Gfw97//nSuvvLJOj/nXnueYmBjefvttpk+fzn/+8x8OHDhATEwMrVu3ZuTIkccdj81mY+bMmTz33HP897//Zfr06cTHx9OzZ89g08mmTZvy3nvvMW3aNJ5++mkKCwuJj4+nffv2jB8//hf373A4GDZsGMuWLat1Iurf/OY3PP744xw8eDDYTLBay5YtcTqdPP/88xw8eJCoqCgyMjJ4+eWXgxPeG4YRnJj9WMydO7fWuR/vvffekMTml2RmZmI2m2vtnwpVDyKq56wbP348lZWVQO19437Obrfz8ssv88gjjzB16lSioqIYO3Ys3bp145///OcR+y0eK7PZzIwZM3jqqaeCU2D06NGD1157LaQWtfqc+v3+kO1fffXVkCafCxYsCNZmL1++PPiZAMGRXGvrSykiZw6TcSLDRomINCAzZ87k//7v/4ITBouEqy+//JJrrrnmjP0uL1++nN/97nd8/PHHx1wb+HO//e1v2bt3L0uWLDkJ0dW9nJwchg4dyiuvvBJsRi0iZybV1InIGeWTTz5h69atdOzYEbPZzDfffMOsWbM4//zzz8ibYGlY+vbtyznnnMNLL70UnJPwTPLVV18xYcKEY0roZs6cSUxMDC1atKCsrIxFixbx2Wef8dhjj52CSOvGSy+9xDnnnKOETkSU1InImSUqKoqPPvqIF198kcrKSlJTU5k8eXJwgmiRcPfnP/+Zjz/+GMMwQvqGnQnuueeeYy5rs9mYNWsWubm5+P1+2rRpw1//+tdgH9zTnWEYpKamMmnSpPoORUROA2p+KSIiIiIiEsZqDvUmIiIiIiIiYUNJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBiz1ncAIg3J0KFDOXToEBaLhYiICAYPHsxf/vIXbrnlFtatW4fVasVut9O7d2+mTp1KSkoK9913H6mpqdx111217tMwDIYPH47D4WDhwoU1Xl+5ciUvvPACGzZswOFw0K5dO6677jqGDRvG3LlzeeCBB3A6nSHbLF68mNTU1JB4IyMjGTRoEH/5y1+Iioo6KedHRETkWPz8+lTtoosuokuXLsHrmslkIi0tjTvvvJPzzjsvWC47O5sRI0Zw2WWX8fDDD9dH+CKnnGrqROrYCy+8wNq1a3n//ff54Ycf+Pe//w3A1KlTWbt2LUuWLKGkpIQnnnjimPb39ddfU1BQQHZ2NuvXrw95bfHixdxxxx2MHz+eFStWsGrVKqZMmcInn3wSLNO9e3fWrl0b8i81NbVGvJmZmWzYsIEXX3yxDs6CiIjIr1N9far+N3XqVOB/17VvvvmGSy65hDvvvJPi4uLgdllZWcTFxbFo0SI8Hk99hS9ySimpEzlJUlNTGTRoEFu3bg1ZHx8fz6hRo2qsP5L333+foUOHMmTIEDIzM4PrDcPgb3/7G7feeisTJ04kJiYGs9lMnz59eOyxx4473uTkZAYOHMjGjRuPe1sREZFTzWw2M2HCBFwuF3v27AGqro2ZmZnccccdWK1Wli1bVs9RipwaSupETpKcnBxWrFhBp06dQtYXFBSwZMmSGutrU1lZyZIlSxg7dixjxoxhwYIFwaeOO3bsICcnh1GjRtVJvLm5uaxcuZIWLVrUyf5EREROJp/PxzvvvENkZCStWrUCYM2aNeTm5nLBBRdw/vnnhzwMFWnI1KdOpI7ddtttWCwWYmJiGDJkCLfccgs33ngjjz32GE8++SQRERH06dOH+++//6j7Wrp0KXa7nQEDBuD3+/H5fCxfvpwRI0ZQVFQEQEpKyi/u47vvvuPss88OLsfHx/PRRx+FxAtQUVFBv379mDJlyokctoiISJ2qvp5Wu/fee7FarcHrmsVioWXLlkyfPp2YmBigqnXL4MGDiYuL48ILL+Sqq64iPz+fRo0a1ddhiJwSSupE6tj06dPp379/jfV//vOfmThx4nHtKzMzk/PPPx+r1YrVamXkyJG8//77jBgxgvj4eAAOHDhAWlraEffRrVs33nrrraPG+9VXX3H33XdTWFhIbGzsccUpIiJS12q7ns6dO/eI1zWXy8XixYuDXRB69OhBkyZNmD9/Ptdee+2pCFmk3qj5pchpKjc3ly+++IJ58+YxYMAABgwYwJIlS1ixYgUFBQW0adOGJk2asHTp0jp5vz59+nDxxRfz5JNP1sn+RERETqUPP/yQsrIyHn744eB1My8vT00w5YygmjqR00AgEMDtdgeXTSYTWVlZtGrVildffTWk7OWXX86CBQuYPHky9913Hw888EBw8JXIyEi+/fZbsrKyePTRR487jmuuuYahQ4eyadMmOnbs+KuPS0RE5FTJzMxkwoQJIVME5eXlcckll7B582bS09PrMTqRk0tJnchp4MUXXwyZSqBnz54UFhYyadIkkpOTQ8pefvnlvP/++0yePJnRo0cTGRnJCy+8wGOPPYbD4aB9+/b89re/DZZft24dPXr0CNnHK6+8QteuXWvEkZiYyLhx45g+fTrTpk2r46MUERE5drfccktIn7r+/fszbNiwWsvm5eWxevVq3n///ZDrZnJyMoMGDSIzM5M//vGPJz1mkfpiMgzDqO8gRERERERE5MSoT52IiIiIiEgYU1InIiIiIiISxpTUiYiIiIiIhDEldSIiIiIiImFMSZ2IiIiIiEgYaxBTGqxbtw6Hw1HfYYiIyEngdrvp3r17fYcRNnRNFBFpuI50TWwQSZ3D4aBTp071HYaIiJwEGzdurO8QwoquiSIiDdeRrolqfikiIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYey0nKcuJyeHe++9l/z8fEwmE5deeinXXHNNfYclIiIiIiJy2jktkzqLxcJ9991H586dKSsrY8KECQwYMIB27drVd2giIiIiIiKnldOy+WVKSgqdO3cGIDo6mjZt2pCXl1fPUYmIiIiIiJx+Tsuaup/bu3cvGzdupFu3bvUdyilRWu5md24pPn+AtNQYGsVF1HdIpz3PoX14i3KxRidiT07DZDntv9byM9l5peTml9MoLoKWjWOwWE7LZ01hy3NoL97CPKwxidhTWmAyW+o7JBEREaljp/Xdb3l5OVOmTOFPf/oT0dHRRyzndrvZuHHjKYzs5LDanXy4tohFq/cAcFarBH57QRv87tJ6juz01djhp2jukxjuCjCZSbjoD+SZ4jAMo75Dk2PgtcTx6Mvf4vL4MZtN/PnaXsRYSvX51ZHGDj9F7z2B4XGByUzixfeSS4zObwPXUK6JIiJy7E7bpM7r9TJlyhTGjBnDyJEjf7Gsw+GgU6dOpyiyk2fX/mIWrV4fXN6wqxBXwEHXTs3rMarTW+HKd6oSOgAjQMUXmXSY9CBmu7N+A5NjMmv+D7g8fgACAYOslbt48IZ+2KyqTaoL+Z+8XpXQARgByr+eR4crpmK22uo3sOOkBOX4NJRrooiI1HSka+Jp2c7JMAweeOAB2rRpw3XXXVff4ZwyVqsZ62FNz3Rz+8vMjsjQ5cgYMJ+WX2upRXSEPWQ5LtqB2WSqp2gaHosztIWDJTIOk86viIhIg3Na3v2uWbOGrKwsvvjiC8aNG8e4ceNYvnx5fYd10jVuFMVtl3TDZjVjNsHEYe1pkRpT32Gd1iLb9cDetD0AlphEEgZfjtlqP8pWcrro37UJbZrGApCcEMGE89qrT10diuzQG3vjNgBYYpNIGDBBfU5FREQaIJPRADpXbNy4scE0NfH7A+QWVBAIGKQkRuCw6QbsaPyVpfhK8rFExGCNbVTf4chxKil3c6jIRXy0g8Q4NZuta76KEvylBVgiY7HGJNZ3OCekIf3Gnwo6XyIiDdeRfuOVMZxmLBYzzZKPPCiM1GSJiMESoRrNcBUb5SA2ylHfYTRY1shYrJGx9R2GiIiInERq5yQiIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1Ij+/QooAACAASURBVCIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUiciIiIiIhLGlNSJiIiIiIiEMSV1IiIiIiIiYUxJnYiIiIiISBhTUneaKyp1UV7pre8wGoyAx42vtADDCNR3KHKYSreXwhJXfYcRFgIeF77SQn2PRUREBABrfQcgtSuv9LJy3T7eWrqZlIQIbr4og3ZpCfUdVljzHNpL/sev4t6/ldieo4jtNQprtM7p6WDHvmJezPye/QfLuGRoe4aenUZ0pL2+wzoteQ5mk//xq3hytxPTczSxPUdijY6v77BERESkHqmm7jQRCBjk5pezLbuQ0nIPO/YXM/3d7ygocbFpdyH/9/oaikrd9R1m2Ar4PBQsf4vKbWsIuCoo37gK155N9R2WAKUVHp6ds5Yfd+RTWOrmP1k/sHNfcY1ygYDB/oNlbMsupKTcUw+R1r+A10PBp29Quf1b/OXFFK2cg2vvkb/HRsCPJ38f7pwd+CvLTmGkImeGMk8FOwr2sL8kj0BANeciUn9UU3ca2HuglM/W7edgUSUtG8fwzrKt9M9oGlImN7+cSreP+BgHfn+ALXsK+Xx9Dk2To+iVnkpqo8h6ij48BDwuPDk7iGjdlYjW3XDn7cRfUYSnIBd7YuP6Du+MVun2sSe3NLh86bAObNpTyFcb8xjQtSkdWiRgNptYt/Ugf539FW6Pn4y2Sdx5RQ9SEhru9z7g8+Leu5nyTV9gT21FRNvumCw23Lk7Q8r5ig4ccR8VW77mQOYzGH4vER36kDzqRqyxiSc7dJEGq8RVyg8HNrPhwFYyUjuyvWA3mZuWYLPYuKPf9fRp3r2+QxSRM5SSunpWXObm/15fw46faiYiHFYuGdqeuGg7EQ4rlW4fAMP7tCAx1oHL7SM7r5T7nv+cQMAA4Nyezbn90u7YbZZ6O47Tmd/jwmS2ED/4cgLlhRQsew2A8h8/o9HI67EnXlDPEZ7ZEmOcXDigFZkrdtD7rFQ27ynku60HAfjgsx38447BJMZG8OrCDfQ+K5UDBZV8v/0QW7OLGnRS587ZRs4bDwaX4/qNI/G8ScT1uZCCj2ZjsjmIaNMDZ+uMWrf3lRzi4MJ/Y2/cBnNEFK4d6/Ac2KmkTuRX+Gb/el74+nUAlm5fwaSuF2EzW/H6vcz4+nXaJbYiMVLNoUXk1Dttk7oVK1bw+OOPEwgEmDhxIjfddFN9h1TnfP4AO/YVBxM6qKq1CAQMXB4/f/vdQLbsLiQmyk7z5Gje/nAL324+wMi+LYMJHcCKdfuYfH4nUhIb7g3uifCVFlC5ZwPeA7so3/I1UV2G4D24J6RM2Y+fEdNzJGaLrZ6iFKvVzEXntuOsNo2wWcw8PPPL4Gs+v8H+g+VERdjo16UJK9fto1lyNHdd0YMIe8N+iOHO2RGyXPb9p8T1GUNMxhBsyS0JVJZQ/OV8Cj95g4TBl+Ns2g6/uxKz1YbJYsUwDBIGX07ljnX4K0pIPG8SmBv2ORM5mXx+H8t2rApZtzV/J01jG7O7aC/egA+/Bi8SkXpiMgzDOHqxU8vv9zNq1ChefvllUlNTueSSS3j66adp165dreU3btxIp06dTnGUv96GnfmYAZ9hUF7hJS7aTlmFF5vNgtVswmwCm81CYakbp8NKhN1CUZmb6Ag7Pn8AsxkqPX5iI+z4AwEsFjNJsU4S4iLq+9DqnLckH++hbEx2JwGvpyoJs1gIVJRgjowFkwXDUwEGmGz2qpvXgB8MA8wmDI8bk9UOgQCuPT9iBHyUrFlCbO/fkDjo0vo+vDNCUWklBwoq8QcMSio8xERWJdIOm5XSSg82swUDg0q3j5hIOyXlbuJjnAB4vH4Mw8BqNWOzWCgoqSQ60k5cpI1mqbH1eVgnTfnmr8h790lSL38Ak8WO4XNjiUrAMFsxKkswvC7MEdG4c7ZjjojFGhkLZitmZySGpxKTzYm/ogizLQLD8GOyOsFiJaJxq/o+tOMWrr/x9UXn69jsLc7lUEU+ETYnXr8Ps8lEwDAwmUxEWp1UeF2UuEuIdcZgwYzD6mBvaS47CvawZNtyUqOTuL7HZfiNAFazBafZjo8ALq+LhMh4KjyVVHgribZHYjZZSHDGUugqxuP34PZ7SYyIw+3zUOIpIzGiqrzJZCLGEUXj6GQcVkd9nyJpQA6WHaLYXY7b58ZmsVLhqSTC7qTCU0msMwaPz4PL5ybKHkmF10WkzUmkNZKCygIi7BGUucuxW+2YMBFli6Cgsgir2YrNYsNhseM02yjylOIL+LGYLXgDXqJtUfgCPko95UQ7ojAZYDPb8Aa8WMwWil0lRNoisJqsJEcnEh8RV9+nKSwc6Tf+tKypW79+PS1btiQtLQ2ACy64gI8//viISV04WrflAPsOlNEsJZo3l2wmOSECt8fPlz/mYjLB+CHt6Nq2EbPnrmd3Tik2q5nfTezOsq/3sGFXAfdd05vXF21kdL9WzF69gYOFFQzv04JeHVPxBSA5oeEkdu5D2RzM+hee3B1gtpI49CqKVmcSf854ir/IJOB2kXLx7zkwfxpJw6/FV1aCyWyiaNVcLM4o4vtfjOfQPhxN2pD/4cv4ywoxOyJJuvB3WGIb1ffhnRHyiyvZtb8Yl9fPrHk/cqCwkgiHldsu6cbunBJio6suFLMXbMDnD9CmWSxXjOzIjPe/JzuvlEHdm2E2Q48OKcz5cAs79hdjs5r5w6ReBDBIS21YF4LKfVvx5O+jyXV/w71vCwUfvQIBP460TiSedxUH3n2KgLscW2JTki68jcrt33Lo87k0uephCpbOwp27g6j0PpgdkZRv/pK4vmPxlRUS0borfncqFkfD+X0QORHb8ncxf/OHbMnfyci2g9l0aDuxjmiW7/oCgBFtB9Espgkxziie/vw/XJoxhte/ex+3z01yVCNu6HU5TquTL/etZem2FQAMbzOQCm8lLeKbYS/N5cNtK8gpO4DdYuN3fa/lUEU+BZVFvPrdexiGQY8mnWkcncKirZ8Q44jmks6/Yfa37zC+0yhSohoxuFU/bJbT8jZNwkx20X5K3GV8m/M9nZLb8/zX/2Vs+gheXzUXj9/Lzb0n8dq6uZR7K4h3xjKu40je+j6LKf2u52DZIRZu+5SD5fk4rA7uPOe3zPp2DnuK92EzW7mq28VE2iNoFJHAq+vepUeTLmRuXILFbOGW3lfxnzVvBf9uJpz1G9w+N2mxTXjh69fZV5qLzWLj1t5X823Oj4xoN1DNl3+F03L0y7y8PBo3/t/gFampqeTl5dVjRHXL4/Mz99NttG4Wy879JWzcVUDLxrF8+WMuUFW5NP+zHRwqcbE7p2oACa8vwNtLNzOyX0sAdu4rpnXTOD74fCe7ckood/nIWrGDvQfLyC+urLdjOxk8+7dXJXQAAR/FX84jumM/ilbNJfqsgRheF0WfvUd8v/EUffYu9kZNKfhoNoGKErwFOeQvew1n8w5U7voef1lh1W7cFZR+9zGY1BztVCgocZFf4mLjzkIOFFZ9PyvdPhat3sXOnBKS4yPJWrEdn7+q6VJCjJMPVu5gy55CKt0+ln65m6S4CJZ+uYvoqKoaPq8vwOwFG/A3sNZOPncl/tJDFH7yOni9FC1/u6rWGXBnb8RXkIOzVWcAvAX7CbjLKVr1PjE9R1D8zUJc2RswvC7KfliBOSKGgLsCb8F+KjatJlBejLeo4fyWipwIn9/HzsJsVmd/y8AWvcnatJSOSW2DCR3Ah9tXEm2PYPnOL+iQ1IaVu77C7asagfpgeT5FlSUUuUqCCR3ARzs+o3VCC0yYKHWXkVNWNYiRx+/l/Q2LMTB46/ssqhtIrc35kXhnVUuDUncZmw/uoEV8U+Zv/oiCymJyyw6eqlMiDZzb52ZfaS4RVifrcn+kVXxzPt21GrffQ4ek1ny9dx3l3goAilwl7C3JId4Zx3s/LiTGGcPB8nwAfH4v+0vy2FO8DwBvwMeirZ9QWFnM/tJczknryfzNH2Fg0K1xJ5ZuWxHyd7OvJAcD2Feax77Sqnter9/LuxsW0CQ2mb0lOaf+5DQgDeIRkNvtZuPGjfUdxjGLjon7qU+c6YhlTMDhLWMDhhHcwjAMkuIjWLF2X0gZj8dPwPCH1fk4mmY/3dAGGQEwmf73XwDDDyYThgEBT0VI8UBFCQZUZcuH7cdkMhrUuTpdWSJTfvruhn4G/kCg6qPEIPCzzyc5PoJvNoUmHxVuHxazGf/Phg03DAMMGtRnmJIYjyN4KgwObx9vYGAKeR5XdWYtkXG4skPPg7+8CLMz+qfvfvXfir7zDV24XRNPNYfTGfKXdaReKAYmAkYAMNX4S/QFfJhq/HX+b8vDy1ddv4+8RVWZwE9lqkqVlpawcX/R0Q5H5BeZTCYsSVVNec0mM0b1dzH4vTcRqPF9DWCqbo582P4O/24bhoFhGFRdmn/+t1Lz76Z6f7XtA6ruCfTbdeJOy6QuNTWV3Nzc4HJeXh6pqalHLO9wOMKu/8DF51nJKyinddNY2jWPZ++BMnqmp/Dt5qone+ef04rk+AiaJkWx/1A5FrOJy0Z04ONvsgkEDNo0i2PR6l0M653Gki92A1UjZ7ZpHkdyQhTJ8cn1eHR1y5XjwJaUhvdQNpjMxPUZS/FX84k75yJKvl6AyWIjfsAEDi2cQeKwyWCyYImOx19WdTGMbH82/pICIlplUL7lKwIVJZjsTuJ6XwAWG506tannI2z4DhVVUlzhJcJpY+W6/RSUuLBbzVwwoA27cko4UFjBuMFtmb1gA4GAQU5+BaP6teKNxVVzsNmsZhrFOUlvkcDri6t+8C1mE1f/5iysVlPY/f0fTaW3lKizBuL3+0gYOLFqxFYjgL1JO2yJTcnfNQsAa1wyZkcUsX0upOTbJSQMvpz8xS9W7cRkxp6cRvnmL7EnNQezBXNkLLaEVDo1Dq9BlXSRPz7heE081QKHTJzdtCurstcwtuMItuTv5Jy0XqzOXgPAua3P4Zv93zGq/RBmfPU6l2WMZVfRXrx+L4kR8UQ7oom0RTC0zQCW7fgcgCGt+rGrMJuWCc2JdcSQHJnIwYoCrGYrF501Gp/h59IuF/Lmd5kYGHRO7kCJu6o1TpQtkrNS2vPF3m8Zkz6CWEc07Rq3wd5cg3jJr7encB/NYhrzVek6ejTpzKw1cxjbaSTZxfvZkr+Dm3pdycaD23D53MQ4omkR35yVu79mSr/rKKwoJjEinoLKIsxmC81jm9A0OpX9ZXlYTGbO73AekbZIGkXG8/q6uVzQYSjzNn3I+twN3HT2pJC/m7S4Jri8btJim5IalURe+SEsZgsXn3U+e4tzGNV+CI2aJtT36TrtHemaeFoOlOLz+Rg1ahSzZ88ODpTyj3/8g/bt29daPhw7hRuGwZY9hcGO2QWlLhKinVS6vYAJi8WE2WzCbrVQUOIi0mElwmmloMRFdIQNI2BgMpvw+Q1Kyz2Uubw0S46maVIUjRrgQCnuA7vx5u/HHBmL4fdhstowmSwEKoowR8RWPf3xe6vqMOwOCATwFuzHZLFijW+M4a0EmxOzxYKv+BCW6ARMkXE4GzWp70M7YxQWVXCguDL4nY2JsmM2mbBZzZRWeLFZTfgD4PP5iYqw4fL4qXD5KKnwkJoQic1mrupkbav6m4iwW4iJstGmWcO7ABiGQcXWbwADS2wKgYpiDJ8HS3QCARPgqsDwuDBHx1c1ITZV1UibbE4MVxm+0gJsiU0wDAOzxYoRCGC2OzFHJ2KPC79+pOH4G1+fdL6Oze6ifewvycNhsWMygd1ix2/48fq9BAyDKHskdrOVAAYen5cIm5MSdymR9gjcPg+R1gj8hp8KbyUWs5UIqwNvwFc1UEpEPJU+F+XeCqLtUVgwk7lxKd2adCIlKgmP30NyZCKegI8iVzEx9mjyKwuJcUQT74ihSUwqTpsGSpG6k1tygFJPOR5/1SAllT8NlFLuqSDeEYs74KXSW0m0PYpybyVRtggi7U4OlRcRZY+g1F2Gw+qoGijFHsGhikLsFhs2sw2n1YHDbKXQU4o/4MdsMuMN+IixR+H1+yj1lBFjj8JkMmE323EH3FhNVgpdxUTZIrGZrSRFJZAY2fCu5ydDWA2UYrVamTp1KjfccAN+v58JEyYcMaELVyaTifSWVfNFlZR7mPbfdez+aQLmSaM7Bmsoqk0a3ZFvNuZx6fAOfLflEE2So+jaLqlBJnC1caS0xJHSMrh8aOksSr5e8L/Xm7TFnnYWUe174c7ehK1RE4pWZRLbfSgH5/0rWM4cGUtMj+EUz51Lk8mPgJK6UyYhPpL4uAi+2pDHs3PW4fL4mHBee/p0TmXeyh14vH6u/s1ZzP5wMz9szw9u1z+jCTn55ezKKeHJ3w1i74FSCkvcdGufTJtmDbNDtb+yhPylM/EV/69PTVzfMZSu+5iAu4KE8yYR03043kN7Mfwuct98GLMziqTRN+M5lI3JasMwAhQuewOzM4rE8ybhbNawfkNFfq1iVwn/XP2f4HJ6o7bcP/g2Iu1V19UKTyVLt6/gvz98gMNi55beV9G7eTfMphMbjuBcdz+mfTkbt8/NsNYDuLzrWOKcsXj9Xg6WF5AUlUBKVBIm05G7ZoicqMaxKTQ+erEa0uKa1bq+VUJajXVN0T1VfTotkzqAIUOGMGTIkPoO45RwuX3sO1gWXA4EDBw2C25vVV8ys6mqqdnm3YVs2lXAJUPbEem011e4pwV7UvPQ5SZtSRxyORa7E2ezdPLeewpf4X6MgC+kXKCiBLP1p3N3+lVSN2ger58d+4r5xxtrqHRXfS5zPtpCj/QU/nRtb7btLWLV+n20bBwbktQlJ0SyYWcBhlHVZ3To2S3q6xBOGbPNiS25RUhSZ3ZGE/BWdTj35OzA2j8Wa4uzcO3fBkDAVc6BzKexp7am0egb8VeUknzRXdhiG2HWE3+RGopcxSHLu4qyq4Zy/ymp21uSw39/+AADg3JvBc99OZun4h+gSUzKEfe5qzCbQxUFmDDhNwK0iGtG45iq7hC9m3Xj/0Y9gMfnJSW6Ec6fpiywWWw0jU0luziHVXu+IdIeSZuENOKcDXO6FhE5OU7bpO5Mkhjr5MIBrclcUTXC42fr9jHlsu68smADPr/BmEFt+HRNNgDFZe4zPqEDiGjTg5geIyj7YSXOFp2J63MhFnvVnGb4ffhKDlX9fyCAJSoOf3nVxTuibQ/c+7cR2aE3tsMSQzm5ftyRz+Y9hbg8oYm2y+3DZrWQkhDJ9n3F9DmrMb07pbB+ez5nd0zFZjVTVOamcWIETZOj6in6U8tsc5B43lXkuytx5+0gpttQfEUHgqNgRnXqHyxrS2xCTK/RlK5ZDIA9tRVFK/9L5Y51WGIa0fTqRzHHH7lPssiZqnlsU2xmK96fHv6NbDeY+Ij/JVK+gJ9xHUf+1JTMS9bGpXj93iPub1dhNtNWv0zPZhlkbVoKQGJEPHf0+y0Hyg/RNDaV1gktsJprjrq8tziHqcv+TvlPA32N7ziKSzPG1FpWRKQ2p2WfuuPVEPoPFJa42LKniJIKN0lxEazffpB2TeOIjnLwjzfWUFjqxmox8fj/G8BZrcOvT8zJYPh8+CqKsDijMVcndD8p27iaA3P/jsnmJH7gJZjtTgy/D0tMI3yFOZgjYojrObKeIj8zPfnq1xSVuWkU62T5T6O2NkuJ5pGbziEloWrgjuIyN/sOlhHptBJhr3rmtCW7CL8/QPu0BJqlRNdb/PXB764k4CrDnbsLX1EugcpSLFHxOFt3xfGzhxJ+Vzne/P0EvC6Kv5xH5bZvg681vvzPRLbtUR/h15mG8Bt/Kul8HRvDMNhRuIcth3aQEBFPelIbEn6a/PhgeT4PLnuaQxUFALROaMHQ1ucwpFU/nDYnhZXFWM1WYhz/e9D06c7V7CvJZem2FVT6XMH1l2eM5e3v52EymXj4vN/TMbnmnLur9qzhmdUvBZctZgvPnv8QKdFJJ+vwRSRMhVWfujNRQqyTvl2qWjt/t/Ug735c1aQqLTWGS4d3IDbKTvOUaFo2aViTLP8aJqsVW2ztF7zIdj1oes1f8ZUVYktsQun3n1LyxXyqh9RPOv/mUxipALRpFserCzcyuEczrv5NJ2Kj7HRvnxxM6ADioh3ERYc2FUxtdGbUztXG4ojA4ojA7IjE44wk4CrHltQMe6PQPg4WZxSWZu3xFuXh3rs55DVzxJmVCIscK5PJRNvElrRNbFnjtdyyg8GEDmBn4R5a9bwUMPHx9s95+/ssYpzR3NjrSjr9lKTFOmLY6t1JtD0yJKmr7oNnGAZfZK+tNamLtoeOSJsckYjDqmbTInLsTsvJx890bZvFcd2Ys4hyWrGYoX1aAoN7NKdNs3gsZnWgPhZmmxNn83SiO/bDkdKS2O7DiWjbHZPVTnT34US06V7fIZ5xBnZrxuDuTfni+xy2ZheR0S7pjE7YjofFGUVEy85EpfepkdD9nC0+ldQJ92CNT8UcGUvSBbdhT2r4fRBF6lr1SH3VnFYHcY5YdhfvY8Y3r1PsLmVvcQ5Pr/oP+eWFALRNbEFSZCJjO44kxl7123ZOWq/gJOJRtki6Ne7Eupwf2V20F//P5mBtndCCy7qMwW6x0TQmld/1u5Y4Z8wpPGIRCXdqfnmaMgyD/GIXDpuFmCj1oasLfo8r2HzNbNXcP/XB4/FTXO4mJsqO066GAieLr7wYAn6sMYn1HUqdaIi/8SeTztfxCQQC+Aw/dsv/rgtev5d1OT/y6rr3sFvsXNfzUjqndGDN/u956rN/h2z/9OipNI9rEtxXkasEt8+D36ga2n322v+y4eBWbu1zNS9+8yYV3krMJjN/Gnw7XRt3DImjoLIIh9VOjEM17CJSOzW/DDMmk4mk+DNjuoJTxWJ3/m8wFakXdruFZHt4TXwdjqxRaqYtciz2FufwweaP2FuSy/hOI8lI7YTDasdmsdG7eXfSk9thNpmI/qnmrVlsY2IdMcFJw/un9SLpZ3Nrmc1mEiPjKXKVUFhZTLwzlt/3v5EKbyXLdq6iwlsJQMAI8N8f5pOe1DrYzNJsNpMU1TAexIjIqaekTkRERM44FZ4KZnzzBpsPbQfgqc9e4JGhf6BjcttgmdjDasyaxKTw0Hl3saNwD06rg7aJLXHaQh8WZhfv5x+rXmR/SR7JUY24o9/1HCwvwHLY/HZR9ogTnvNORORwSupERETkjFPmqQgmdNXyKwqAtrVv8JPmcU2CzS1rs2LXl+wvyQOqRtFctOUT9pbm0K95L1rENWVP8X7inLFc2mUMFV4X3+9dy/rcTfRqmkGX1HSi1JpBRE6AkjoRERE541jMVvo0785Xe9dVLZvMvzix+LEq9ZSHLFf4XDitDt758QPObX0O1/e8jJSoJJKiEvl4+2fM+OYNAD7dtZq7+99E37TwnoJEROqH6v1FRETkjGEYButyNvDnj5+iRVwzrsgYx/ntz+Ohob+nVXzar97/sDYDgoOuWM1Wejbpwpb8nQSMAF/v+47knxI6v9/Pp7tWh2z75d61v/r9ReTMpJo6EREROWMcKD/E06texOVzM3/Th1x81vl0SGqDPxDA5XMR+SubP7ZLbMUTI+4jp/QAvoCPosoS4p2xRNuiuOHsK0j+aTAUi8VC76bd2HxoR3Dbro01aqmInBgldSIiInLGqPS6cfncAIzvNIqFWz+h2FUCwORuF3NB+rBfNYCJyWQiLa4p2wt28/xXrxLvjKVP8+4kRsTTJiF03sgBLXsTwOCrvesY3LIv3Rt3PvEDE5EzmppfioiIyBkjOSqRXk27/rRkCiZ0AP/94QMKKouAqmaaB8ryyS09GDJR+JH4A37yyg5yqLwAgBJX1bQHRa4Slm5bwcpdX+EL+EK2aRSZwPhOo3h46O8Z3eFc4iNi6+AIReRMpJo6EREROWNE2SO5odflDG3TH7fPE/JarCMam9mKP+Dni71refHrN/AFfFyeMY7hbQcSYat9rtNyTyVLty3nnR8X4LQ6uLXP1XRtfBaOHxfg9le9x8QuFxxxZEvbzyY+FxE5EUrqRERE5IzSKDKBRpEJFFYWM6LtYD7asZI4Rwy39b2GOGcs+0vyeP7LV/D+VLP22nfv0Sm5He0atap1f9nF+3jr+ywAyjw+pn3xMk+NeoDHh/+R/aW5JETE0Sq++ak6PBE5AympExERkTNSQkQc13S/hAvSh+K02EmMTADA4/cEE7pq1f3wanP4ay6fG4/fS4v4prSIb1r3gYuIHEZ96kREROSMZbfaaBqTGkzogP/P3p3HRVnv/R9/DwMMO26Ampa4IZJbaZaQC6m4IYlWlvudv05mR61zLNKTlR21xZOpleldasc6dbuCS1nuW6aZC5akkispIIKKKMM2vz+8nTsCRNFhGHg9Hw8eh/ley/dzzaH5+p7re12X/L1qKezudtbXDavfo7reASXu4y6f2oWWhzcMlb9nTdsUDADF4EwdAADAH3i4uGto6wHqFPig8gvydbfvXarhUa3E9f08a2pCx+d16uLvcnFyUYPq9WRydr2lPnPzcnU0/bh+Tj2su33vUrBfY/m6ceMUADeHUAcAAPAn1dx9VM29+U2v7+9VS/5etcrc39H043p90wzr62vTQh8p8/4AVC1MvwQAAChFxpUL2nhsh+bs+rf2nvlZ2bklX2NXFvEpvxZ6/d1v23Q5J+uO9gGg8uJMHQAAwA1YLBZtOr5TX/28D6JFNgAAIABJREFUUpK06cROTew0Rq1qB99wuwtXLyr58jm5u7iprnfADR9dcI/vXYVeN/drIjfn4h+hAAB/RqgDAAC4gSu5V7Xl5K5Cbb+kHL5hqDuXdV7/2vHfOpZxUgaDQWMf/C91uLutJCn1cppOXTwjk7Or7vG9Sz5u3mru30RDWvXXut+2KsSvqfo26yZnJ6NNjwtA5UGoAwAAuAF3ZzfdX/derT6cYm0r6Zl11x3POK1jGSclST0ad1bSpbP66uBKtQwI1uKfV+vQuSOSpJ5NumhQy37ydfNRZLOu6hL4kNxc3Ah0AG4JoQ4AAOAGnJyc1KNxZ5mMropP+VXdGj2sYL8mN9zGIIMkqVmtxrqQfVHfHN0kSYr79Ts91fJRa6hbm7hZ3Rt31F0+tSVJXiZPGx4JgMqKUAcAAFAKf69aeqJFX0U373nDa+OuC6xeX8F+jdXcr6lWJKy1tucX5Mv8h4eVuzu7ycWJf44BuD3c/RIAAOAm3Uygk6RanjX0tw7PqN1dLdW4RoNCywI8/WQ0OMnL1VN/fXDEbT0KAQAkztQBAADYhI+bt3zcvPWXdoO09JevlXTprPoFR6h17RC91/M1uTg5q5ZnDXuXCaASINQBAADYUH3funq+/XDl5ufKw9VdEtfOAbizCHUAAAA25mJ0louRf3YBsA2uqQMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB8YDUwAAQJVzPOO0jqWflK+btxrXCFQ1d59y6begoECXci7Ly8VDzjy3DsAdwqcJAACoUk5kJOnVDe8qJz9XkhTVrLsGtugro5PRpv2ev5Khb45u1raTu9SqdnNFB/dQbW9/m/YJoGqocKHu7bff1qZNm+Ti4qK7775b06ZNk49P+Xx7VhFcyMxWSvoVebq7qG4tLzk5GexdEgAAlcrpi2esgU6Svjm6SRGNO6mWZw2b9Hf+SoZOXTijM5nJWvnrd5Kkzcd3ysfVS4Na9ZPBwFgP4PZUuGvqQkNDtXr1aq1atUoNGjTQ3Llz7V1SuUk+n6XJn+7S32dt09h/bda+I6n2LgkAgErHx+RV6HU9nzpyd3GzSV+Xc7I098cvtOjAMh09f6LQsoS0ROUV5NmkXwBVS4ULdWFhYXJ2vnYCsXXr1kpOTrZzRbaXnZOn/UdStSchRUdPX5Ak5eQV6OPl8bp42Wzn6gAAqFwa1rhHQ1pFy8fkraBajfRsu8HydPWwSV+pl89rf/IvOpOZoia1Agst69Gks1yMLjbpF0DVUuGmX/7RsmXL1LNnT3uXYXOHT2botXk79WREs0Lt+QUWWWQpsv6lLLN+SkjV5r1Jan9vbT14bx3V8LHNN4wAAFQ23iZP9QnqqtC728ndxSR3F3eb9eXmbJKL0UW5+bnafnK3hrUeIHNejgJr3K2gmoGl7wAAboJdQt3w4cOVlpZWpH3cuHHq2rWrJGnOnDkyGo3q27dvqfszm81KSEi443WWB2dnZ+37NVsFlmshrk4tT51Ny5KTk0HDewUp9cwpnc3Pt65vMBiUdNFVMxcflCTtPZyq7OwctajvpLw8pnAAQFXnyGNiZWR0dtbo+4dqwYHFOpuZKhc5K8ijgXIv5OjkhZP2Lg9AJWGXULdw4cIbLl++fLk2b96shQsX3tTFwyaTScHBwXeouvJ37uoZSce0ZP0R9QkL1N21vdWkfnXV8/eWi3PRGbJrvvip0Ovvf05R306hcjba9q5dAGAPBJRb4+hjYmXUVFJw7cYqsFhU06O6vcsB4MBKGhMr3DV1W7du1SeffKI5c+bI3d120yEqkpCGNTWqf0vVD/CWLFKLRrUUWNe32EAnSe2aBxR63bHNXQQ6AABuUV5Bvs5cStbpi2dkzsuxaV/V3auVGujSrqTrePopXczOtGktACqfCndN3ZtvvqmcnByNGDFCktSqVStNnjzZzlXZlq+XSb06BKrL/fVkcnEu9TEGrZv66cUn79OWfUlqH1Jb7UPqlFOlAABUDnn5edpy4gd9svcrFRQUqG9wd/Vr1l0ef7hhSl5+nn7LOKkjacdU29tfQTUbycfN6wZ7Lbtj6Sc1f+//6FxWunxMXhrXYaTu8qltk74AVD4VLtStW7fO3iXYjbvp5u6A5eNpUpe29dWlbX0bVwQAQOWUnHVOn/z0pfItBZKkuIRv1a5uSzWt1VBXcq7IyWDUqYtJmrTxX7JYrt20bFibx9S7aXih/WRcvaj8gnzV9Khe5JKR3LxcXc03F3mEwp/l5ufp9KVk1fEOUKMaDeTr5q19Z34m1AG4aRUu1AEAANhafkG+NdBdl5Ofq31nftZ/4uPk5mJSZFBXebl4KDMnS5K09uhmdbznAXn/b0jbf/aQ5vz4b13NzdZTLR9VpwYPWp93d+ZSshb/vFpH00+oR+PO6tigvXzdvIut5fdLyfpkz39kzr82BfSeancpMqibrQ4dQCVU4a6pAwAAsLUATz/1aNLZ+vregGbydPHQW9s/0smLSTqc9ps+/vFzhTcMta7TpGYDuTlfC20pl9M0Y+d/K+PqRWXnmTV/7//o9MWzkq6dofufn1fr+9M/6VzWeS06sEyH034rsZaUrHPWQCdJJy/8XmIABIDicKYOAABUOW4uJj0W0kcP1b9f+QX5qudTWycv/G6dailJl3OyFFi9vmq6V1fTWoEa0LyXXIzX/ulkzjPram52oX1eyb1y7X/zsnXk/LFCy5IzU0uspZZHDTkZnFTwv2cO/TxrytPZNg9DB1A5EeoAAECV5G3yVLBfY+trc36OfEzeumS+dvfJhtXvVoh/kN7qHiN3Z3e5Ov/fte9+njUVenc77Tj1oyTJ37Om6npfuwbOx+SliMad9J/4WEnXnjEb5NeoxDoaVKunl8Oe08pfv5Onq4cimnRSPV9uggbg5hHqAAAAJAV4+WlS53H6OfVXmYyuau7ftMRpkO4ubhraKloP39NOOfm5uqdaPfl71ZR0LcR1CXxIdbz8deZyipr7NVGj6veU2K/Ryag2dUMUEtBEFotFJmeTTY4PQOVFqAMAAPhfd1erq7ur1S3SnmnOUtKls9emavrWUTU3H1X3qKbqHtWK3Y+vm4/a129zS327Gl3LVDMAEOoAAABu4Gputhb/vErfJm6RJLUKaK7nHxzOzUwAVBiEOgAAUGVlZl/W75nJcnZyVl3vAHm4uhdZJzUrzRroJOlAyiH9fimZUAegwiDUAQCAKulSdqbm7flCu38/IEmKDu6pfs17yORceBqk0WCU0eBU6Ll2zk7Gcq0VAG6k1OfUpaWlacKECRo5cqQkKTExUUuWLLF5YQAAALaUdCnZGugkaUXCWqVknSuyXoBXLf3XfU/I6GSUwWBQv+AeusuHu1MCqDhKPVMXExOj6Ohoffzxx5KkBg0a6IUXXtBjjz1m8+IAAABsxclgKPTaYDDIqZjvu12MLurSMFTN/ZuqwFIgf89aRc7mAYA9lXqmLiMjQ7169ZKT07VVnZ2drb8DAAA4qrt86qhbo4clXQt0Q1pFK8CrVrHrOjsZdZdPbdX3rUugA1DhlHqmzsPDQxkZGTL877dZ+/fvl7c3FwYDAADH5m3y1OCW0era6GEZDU6q7eUvF6NL6RsCQAVzU9MvR40apVOnTmngwIHKyMjQzJkzy6M2AAAAm3J3dVOga/07si+LxaIruVfl4eJu/TIcAMpDqaEuJCREn3/+uY4fPy6LxaLAwEC5uPAtFgAAwHXnr2Row2/btfP0Xj1U/z490ihMNT2q27ssAFVEqRfHffHFF7py5YqaNGmipk2b6sqVK/riiy/KozYAAACH8MPpvVp66Gv9npmspYe+1g+n99q7JABVSKmhbvHixfLx8bG+9vX15ZEGAAAAf3Ag+VCh1/Epv1p/v5xzRT+nHtHupAM6m5la3qUBqAJKnX5ZUFAgi8VinRuen5+v3NxcmxcGAADgKDoHdtD+PwS7zoEPSbr276j1v23Tf+JjJUm+bj56o8uLcnYy6vvTPynx/HE90jBMzf2byORsskvtABxfqaEuLCxM48aN08CBAyVJX331lR5++GGbFwYAAOAoWtcO1ktho5RwLlHBfo3V3K+xJOn81Qwt+WWNdb2L2ZeUmpWmXUn7tOHYDknS7t8P6PUuL6q5fxO71A7A8ZUa6saPH6+vvvpKX375pSSpQ4cOPHgcAADgDzxcPdT2rpZqe1fLQu2uRlf5mryVdiXd2mZyNml30v5C6526+DuhDkCZlRrqnJyc9NRTT+mpp54qj3oAAAAqDV83bz3ffrhm7vxUF8yXFNGok2p7+en+ui21+cRO63r1ferYsUoAjq7UUPfTTz/pgw8+0JkzZ5SXl2e9vm7Dhg3lUR8AAIBDa+7fRNO6xcicn6Na7tXl4uyi/iE95e9VS0fSflP3xp3UqGYDe5cJwIGVGuomTpyoV155Rffee6+cnEq9WSYAAAD+pIZHtUKvA7z8NCCkl52qAVDZlBrqvL291alTp/KoBQAAAABwi0oNde3bt9fbb7+t7t27y9XV1doeEhJi08IAAAAAAKUrNdQdOHBAkvTzzz9b2wwGg/7973/brioAAAAAwE0pNdQtWrSoPOoAAAAAAJRBqXc+SUtL04QJEzRy5EhJUmJiopYsWWLzwgAAAAAApSs11MXExCgsLEypqamSpAYNGjD1EgAAAAAqiFJDXUZGhnr16mV9nIGzszOPNgAAAACACqLUdObh4aGMjAwZDAZJ0v79++Xt7W3zwgAAAAAApSv1RikxMTEaNWqUTp06pYEDByojI0MzZ84sj9oAAAAAAKW4YagrKCiQ2WzW559/ruPHj8tisSgwMFAuLi7lVR8AAAAA4AZuOP3SyclJkydPlrOzs5o0aaKmTZsS6AAAAACgAin1mrqHHnpI3377rSwWS3nUAwAAAAC4BaVeU/fVV19pwYIFMhqNMplMslgsMhgM2rt3b3nUBwAAAAC4gVJD3b59+8qjDgAAAABAGZQ6/dJisSguLk4ffvihJOns2bOKj4+3eWEAAAAAgNKVGupef/117d+/X6tXr5Z07bl1b7zxhs0Lmz9/voKCgpSenm7zvgAAAADAUZUa6uLj4/Xaa6/JZDJJknx9fZWbm2vTos6ePasdO3aobt26Nu0HAAAAABxdqaHO2dlZ+fn5MhgMkqT09HQ5OZW62W2ZNm2axo8fb+0TAAAAAFC8Um+UMmTIEI0ePVrnz5/XjBkztHbtWo0bN85mBa1fv17+/v5q1qyZzfoAAAAAgMqixFB3+vRp1a9fX3379lVISIh++OEHWSwWffTRR2rUqNFtdTp8+HClpaUVaR83bpzmzp2r+fPn39L+zGazEhISbqsmAAAqA8ZEAKh6DJYSnioeHR2t5cuXa9iwYfrss8/KpZjDhw9r+PDhcnd3lyQlJyfL399fS5YskZ+fX4nbJSQkKDg4uFxqBACULz7jbw3vFwBUXiV9xpd4pq6goEAff/yxTpw4oQULFhRZPmLEiDtboaSgoCDt3LnT+jo8PFxLly5VjRo17nhfAAAAAFAZlHjHk/fee09OTk7Kz89XVlZWkR8AAAAAgP2VeKZu27ZteuaZZ5STk6Pnn3++PGuy2rhxo136BQAAAABHUeKZuuXLl0uSNmzYUG7FAAAAAABuTYln6ho1aqTu3bsrNTVVkZGRRZavWrXKpoUBAAAAAEpXYqh77733dO7cOT399NOaM2dOedYEAAAAALhJN3z4uJ+fn1auXFletQAAAAAAblGJoW7s2LGaOXNmsVMvJaZfAgAAAEBFUGKomzhxoiTp448/LrdiAAAAAAC3psRQ5+/vL0ny9vbWiRMnJEmBgYHy9vYul8IAAAAAAKUrMdTl5ORo0qRJWr9+verVqyeLxaIzZ86oa9eueuONN+Tq6lqedQIAAAAAilHic+o++ugj5ebmavPmzYqNjVVcXJw2bdqk/Px8ffTRR+VZIwAAAACgBCWGuvXr1+uf//ynvLy8rG1eXl567bXXtH79+nIpDgAAAABwYyWGOoPBIHd39yLtnp6eMhgMNi0KAAAAAHBzSrymzmAw6OLFi7JYLMUuAwAAAADYX4mh7vLly4qOjibUAQAAAEAFVmKo27hxY3nWAQAAAAAogxKvqQMAAAAAVHyEOgAAAABwYIQ6AAAAAHBgZQp1nTt3vsNlAAAAAADKokyhrrg7YgIAAAAAyl+ZQh2PNAAAAACAiqHERxosWLCg2HaLxaIrV67YrCAAAAAAwM0rMdRlZWWVuNHQoUNtUgwAAAAA4NaUGOqef/75EjdauHChLWoBAAAAANyiMl1TR6gDAAAAgIqBu18CAAAAgAPj7pcAAAAA4MBKvKauTZs2xYY3i8Uis9ls06IAAAAAADenxFC3b9++8qwDAAAAAFAGZZp+CQAAAACoGAh1AAAAAODACHUAAAAA4MAIdQAAAADgwAh1AAAAAODACHUAAAAA4MAIdQAAAADgwAh1AAAAAODACHUAAAAA4MAIdQAAAADgwCpkqFu0aJF69Oih3r1765133rF3OQAAAABQYTnbu4A/++GHH7RhwwatXLlSrq6uOn/+vL1LAgAAAIAKq8Kdqfvyyy/1zDPPyNXVVZJUs2ZNO1cEAAAAABVXhQt1J06c0J49e/TYY49p8ODBio+Pt3dJAAAAAFBh2WX65fDhw5WWllakfdy4ccrPz9fFixe1ePFiHTx4UOPGjdOGDRtkMBhK3J/ZbFZCQoItSwYAwCEwJgJA1WOXULdw4cISl3355Zfq1q2bDAaDWrZsKScnJ2VkZKhGjRolbmMymRQcHGyDSgEA9kZAuTWMiQBQeZU0Jla46Zddu3bVrl27JEnHjx9Xbm6uqlevbueqAAAAAKBiqnB3v+zfv78mTJigPn36yMXFRW+99dYNp14CAAAAQFVW4UKdq6urpk+fbu8yAAAAAMAhVLjplwAAAACAm0eoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHRqgDAAAAAAdGqAMAAAAAB0aoAwAAAAAHVuFCXUJCgh5//HFFRUUpOjpa8fHx9i4JAAAAACqsChfq3n33XY0ePVpxcXEaO3as3n33XXuXBAAAAAAVVoULdQaDQVlZWZKkzMxM+fv727kiAAAAAKi4DBaLxWLvIv7ot99+09NPPy2LxaKCggJ99dVXuuuuu264zf79+2UymcqpQgBAeQsODrZ3CQ6DMREAKrfixkRnO9Sh4cOHKy0trUj7uHHj9MMPP+iVV15RRESEvv76a02cOFELFy684f5MJhMDPgBUUgkJCfYuwaEwJgJA5VXSmGiXUHejkPbyyy9r4sSJkqSePXvqH//4RzlVBQAAAACOp8JdU+fv76/du3dLkn744Qc1aNDAvgUBAAAAQAVmlzN1N/Lmm29q6tSpysvLk8lk0uTJk+1dEgAAAABUWBUu1LVt21bLly+/7f3k5uYqKSlJ2dnZd6CqisnNzU316tWTi4uLvUsBAAAAYCcVLtTdKUlJSfL29laDBg1kMBjsXc4dZ7FYdP78eSUlJSkwMNDe5QAAAACwkwp3Td2dkp2drZo1a1bKQCdde55fzZo1K/WZSAAAAAClq7ShTlKlDXTXVfbjAwAAAFC6Sh3qbsbChQt19erVEpdPnDhRiYmJkqQ2bdqUuZ/Zs2fr008/LfP2AAAAAFCcKh/q/v3vf5cY6vLz8zVlyhQ1bty4nKsCAAAAgJtTpULdlStX9Mwzz6hv377q06ePPvjgA6WmpmrYsGEaMmSIpGtn49566y317dtX+/bt05AhQ3Tw4MFC+0lPT9cTTzyhzZs3S5I++eQT9e/fX5GRkZo1a1aJ/f/666964okn1L17dy1evFjStRuevP322+rTp48iIyP19ddfS5LWrVunYcOGyWKxKDU1VRERETp37pwN3hUAAAAAjqzS3v2yONu2bZO/v7/mzZsnScrMzNTy5cv12WefqUaNGpKuBb+WLVsqJiam2H2kpaVp1KhRGjdunEJDQ7V9+3adPHlSS5culcVi0ahRo/Tjjz+qXbt2RbY9fPiwFi9erCtXrqhfv37q1KmT9u/fr19//VVxcXHKyMjQgAED1LZtW3Xr1k3ffvutvvjiC23btk1//etf5efnZ7s3BwAAAIBDqlJn6po2barvv/9e7777rvbs2SNvb+8i6xiNRkVERBS7fW5uroYPH67x48crNDRUkrRjxw7t2LFDjz76qPr166djx47pxIkTxW7/yCOPyM3NTTVq1FD79u118OBB/fTTT+rdu7eMRqNq1aqldu3aWc8Mvvrqq5o7d65cXV3Vp0+fO/MmAAAAAKhUqtSZusDAQC1fvlxbtmzR+++/rwcffLDIOiaTSUajsdjtnZ2dFRISou3bt+uBBx6QdG365DPPPKOBAwcWWveLL76wTrG8fmbwVu9WmZycLCcnJ6WlpamgoEBOTlUqgwMAAAC4CVUqJaSkpMjd3V1RUVF6+umndejQIXl6eiorK+umtjcYDJo6daqOHTtmDWphYWFatmyZdR8pKSk6f/68Bg0apLi4OMXFxSkgIECStGHDBpnNZmVkZGj37t1q0aKF2rZtq2+++Ub5+flKT0/Xnj171LJlS+Xl5WnChAn617/+pUaNGmnBggW2eVMAAAAAOLQqdabuyJEjeuedd+Tk5CRnZ2e9/vrr2r9/v0aOHCl/f38tWrSo1H0YjUa99957GjVqlDw9PTVo0CD99ttv1jN1Hh4eevfdd1WzZs0i2wYFBWno0KHKyMjQc889p4CAAHXr1k379u1TVFSUDAaDxo8fLz8/P33wwQdq27at2rZtq2bNmmnAgAHq3LmzGjVqdMffFwAAAACOy2CxWCz2LuJ2JSQkKDg4uNS2yqiqHCeAqovPuVvD+wUAlVdJn/FVavolAAAAAFQ2hDoAAAAAcGCEOgAAAABwYIQ6AAAAAHBghDoAAAAAcGCEOgAAAABwYFXqOXXlzWw2a9CgQcrJyVF+fr4iIiI0ZsyYQuvk5OTopZde0i+//KJq1appxowZqlevnp0qBgAAAOBoCHWSCgos2rovSXFbf1PahWzVquamqI6N1LFNPTk5Gcq8X1dXV3322Wfy9PRUbm6unnrqKXXs2FGtW7e2rrNkyRL5+Pho3bp1WrNmjaZPn67333//ThwWAAAAgCqgyk+/LCiwaNpnu/Xh0gNKTLqoC5fNSky6qA+XHtC0z3aroKDsz2Y3GAzy9PSUJOXl5SkvL08GQ+GQuHHjRvXr10+SFBERoZ07d6oSPA8eAAAAQDmp8qFu674k7T9yTtk5+YXas3Pytf/IOW3d//tt7T8/P19RUVHq0KGDOnTooFatWhVanpKSojp16kiSnJ2d5e3trYyMjNvqEwAAAEDVUeVDXdzW34oEuuuyc/IVtyXxtvZvNBoVFxenLVu2KD4+XkeOHLmt/QEAAADAH1X5UJd2Ifu2lt8sHx8ftW/fXtu2bSvUHhAQoLNnz0q6NkUzMzNT1atXvyN9AgAAAKj8qnyoq1XN7baW30h6erouXbokScrOztb333+vhg0bFlonPDxcK1askCR9++23evDBB4tcdwcAAAAAJanyoS6qYyO5uRqLXebmalRUp8Zl3ndqaqqGDh2qyMhIDRgwQB06dFCXLl00c+ZMbdiwQZI0YMAAXbhwQd26ddOCBQv097//vcz9AQAAAKh6qvwjDTq2qacd8WeK3CzFzdWo1k391LH1XWXed7NmzRQbG1ukfezYsdbfTSaTZs2aVeY+AAAAAFRtVT7UOTkZ9MqwB7R1/++K25L4f8+p69RYHVvfdVvPqQMAAAAAW6vyoU66Fuw631dPne+rZ+9SAAAAAOCWVPlr6gAAAADAkRHqAAAAAMCBEeoAAAAAwIER6gAAAADAgRHqbOiVV17RQw89pD59+ljb3n77bfXo0UORkZEaPXq09eHkf7Z161ZFRESoW7dumjdvXnmVDAAAAMDBEOokWSwFyvx5q5I+Ha8T7/+Xkj4dr8yft8piKbit/UZHR+uTTz4p1BYaGqrVq1dr1apVatCggebOnVtku/z8fE2ePFmffPKJ1qxZo9WrVysxMfG2agEAAABQOVX5UGexFChl6btK+/pj5SQfU0HWReUkH1Pa1x8rZem7txXs2rVrJ19f30JtYWFhcna+9iSJ1q1bKzk5uch28fHxuueee1S/fn25urqqd+/e2rBhQ5nrAAAAAFB5VflQd/mX7bp6/IAsueZC7ZZcs64eP6CsX3bYrO9ly5apY8eORdpTUlJUu3Zt6+uAgAClpKTYrA4AAAAAjssuoe6bb75R79691axZMx08eLDQsrlz56pbt26KiIjQtm3bbF7LxV2rigS66yy5Zl3Ytcom/c6ZM0dGo1F9+/a1yf4BAAAAVA12CXVNmzbV7Nmz1a5du0LtiYmJWrNmjdasWaNPPvlEb7zxhvLz821aS17m+Rsuz89Mu+N9Ll++XJs3b9b06dNlMBiKLA8ICCg0LTMlJUUBAQF3vA4AAAAAjs8uoa5Ro0Zq2LBhkfYNGzaod+/ecnV1Vf369XXPPfcoPj7eprU4e9e84XKjd6072t/WrVvpih4lAAAgAElEQVT1ySefaM6cOXJ3dy92nRYtWujEiRM6ffq0cnJytGbNGoWHh9/ROgAAAABUDhXqmjp7XEvm2z5SBhdTscsMLiZVax9Z5n2/+OKLGjhwoI4fP66OHTtqyZIlevPNN5WVlaURI0YoKipKkyZNknTt2P/f//t/kiRnZ2dNmjRJI0eOVK9evdSzZ081adKkzHUAAAAAqLycbbXj4cOHKy2t6NTFcePGqWvXrne0L7PZrISEhEJtubm5unr1aqnbGhveL9e771XOyZ9lyfu/a+sMzia53n2vnBred1P7Kc6UKVOKtP3xmXXXXb16VT4+Ppo1a5a1rwceeECxsbGF1ilObm5ukWMHAFRdxY2JAIDKzWahbuHChbe8TVmvJTOZTAoODi7UlpCQUOL0xj9zfyJGWb/s0IVdq5SfmSajdy1Vax8pz5BQGQwV6mRmES4uLkWOHQAqEwLKrSluTAQAVA4ljYk2C3VlER4err/97W8aMWKEUlJSdOLECbVs2dLm/RoMTvK692F53fuwzfsCAAAAgDvJLqFu3bp1evPNN5Wenq6//OUvCg4O1qeffqomTZqoZ8+e6tWrl4xGoyZNmiSj0WiPEgEAAADAIdgl1HXr1k3dunUrdtmoUaM0atSoO9KPxWIp9pEBlYXFYrF3CQAAAADsrGJfMHYb3NzcdP78+UobfCwWi86fPy83Nzd7lwIAAADAjirUNXV3Ur169ZSUlKRz587ZuxSbcXNzU7169exdBgAAAAA7qrShzsXFRYGBgfYuAwAAAABsqtJOvwQAAACAqoBQBwAAAAAOjFAHAAAAAA6sUlxTZzabS3y6OgDAsZnNZnuX4FAYEwGg8ippTDRYKus9/wEAAACgCmD6JQAAAAA4MEIdAAAAADgwQh0AAAAAODBCHQAAAAA4MEIdAAAAADgwQh1KdenSJY0ZM0Y9evRQz549tW/fvkLLLRaL/vnPf6pbt26KjIzUL7/8YqdKgbI5e/ashgwZol69eql379767LPPCi2fP3++goKClJ6eXuz2K1asUPfu3dW9e3etWLGiPEoGUE62bt2qiIgIdevWTfPmzSuyPCcnR+PGjVO3bt302GOPKSkpSZKUkZGhIUOGqE2bNpo8eXJ5lw3cEaX9/f/444/q16+fmjdvrrVr19qhQlxXKZ5TB9uaMmWKHn74Yc2aNUs5OTnKzs4utHzr1q06ceKEvvvuOx04cECvv/66lixZYqdqgVtnNBoVExOjkJAQXb58Wf3791doaKgaN26ss2fPaseOHapbt26x2164cEEffPCBli1bJoPBoOjoaIWHh8vX17ecjwLAnZafn6/JkydrwYIFCggI0IABAxQeHq7GjRtb11myZIl8fHy0bt06rVmzRtOnT9f7778vk8mksWPH6ujRozp69KgdjwIom5v5+69Tp46mTZum+fPn27FSSJypQykyMzP1448/asCAAZIkV1dX+fj4FFpnw4YNevTRR2UwGNS6dWtdunRJqamp9igXKBN/f3+FhIRIkry8vNSwYUOlpKRIkqZNm6bx48fLYDAUu+327dsVGhqqatWqydfXV6Ghodq2bVu51Q7AduLj43XPPfeofv36cnV1Ve/evbVhw4ZC62zcuFH9+vWTJEVERGjnzp2yWCzy8PBQ27ZtZTKZ7FE6cNtu5u+/Xr16atasmZyciBT2xv8DuKGkpCTVqFFDr7zyih599FFNnDhRV65cKbROSkqKateubX1du3Zt6z+IAUeTlJSkhIQEtWrVSuvXr5e/v7+aNWtW4vp//vsPCAjg7x+oJG7mv++UlBTVqVNHkuTs7Cxvb29lZGSUa52ALTC+ORZCHW4oLy9Phw4d0pNPPqnY2Fi5u7sXO6caqAyysrI0ZswYTZgwQUajUXPnztXYsWPtXRYAAMANEepwQ7Vr11bt2rXVqlUrSVKPHj106NChQusEBAQoOTnZ+jo5OVkBAQHlWidwu3JzczVmzBhFRkaqe/fuOnXqlJKSkhQVFaXw8HAlJycrOjpa586dK7Tdn//+U1JS+PsHKomb+e87ICBAZ8+elXTti9DMzExVr169XOsEbIHxzbEQ6nBDfn5+ql27to4dOyZJ2rlzpxo1alRonfDwcMXGxspisWj//v3y9vaWv7+/PcoFysRisWjixIlq2LChRowYIUkKCgrSzp07tXHjRm3cuFG1a9fW8uXL5efnV2jbsLAwbd++XRcvXtTFixe1fft2hYWF2eMwANxhLVq00IkTJ3T69Gnl5ORozZo1Cg8PL7ROeHi49a633377rR588MESr8EFHMnN/P2j4jBYLBaLvYtAxZaQkKCJEycqNzdX9evX17Rp0/T1119Lkp588klZLBZNnjxZ27Ztk7u7u6ZOnaoWLVrYuWrg5u3Zs0eDBg1S06ZNrRd7v/jii+rUqZN1nfDwcC1dulQ1atTQwYMH9dVXX2nKlCmSpKVLl2ru3LmSpGeffVb9+/cv/4MAYBNbtmzR1KlTlZ+fr/79+2vUqFGaOXOm7r33Xj3yyCMym80aP368EhIS5OvrqxkzZqh+/fqSrn1uXL58Wbm5ufL29tb8+fML3TkQqOhK+/uPj4/X888/r0uXLslkMqlWrVpas2aNvcuukgh1AAAAAODAmH4JAAAAAA6MUAcAAAAADoxQBwAAAAAOjFAHAAAAAA6MUAcAAAAADoxQBzio9evXKzExscTlX375pWJjYyVJQ4YM0cGDB8urNAAAAJQjZ3sXAKBs1q9fr86dOxf7zKO8vDw9+eSTdqgKAAAA5Y1QB9ympKQkjRw5Uq1bt9a+fft07733qn///po1a5bS09M1ffp0SdKUKVNkNpvl5uamqVOnqmHDhlq4cKEOHz6sadOm6fDhw/rb3/6mJUuWyN3dvVAf06dP18aNG2U0GhUWFqZu3bpp48aN2r17t+bMmaPZs2dr4sSJatasmX766Sf16dNHWVlZ8vDw0NNPP23dT0FBgSZMmKCAgAC98MIL2r59u2bPnq2cnBzrg+U9PT2L9Pfyyy+X63sKAMCNfPjhh1q5cqVq1KihOnXqKCQkpNB4B1Q1hDrgDjh16pRmzpypqVOnasCAAVq1apW+/PJLbdiwQR9//LHeeecdffHFF3J2dtb333+vGTNmaPbs2Ro6dKiGDBmidevWac6cOXrjjTeKBLqMjAytW7dOa9eulcFg0KVLl+Tj46Pw8HB17txZPXr0sK6bm5ur5cuXS5Jmz55daD/5+fn6+9//riZNmmjUqFFKT0/XnDlztGDBAnl4eGjevHlasGCBBg0aVKQ/AAAqivj4eH333XdauXKlcnNzFR0drZCQEHuXBdgVoQ64A+rVq6egoCBJUuPGjfXQQw/JYDAoKChIv//+uzIzM/Xyyy/r5MmTMhgMys3NlSQ5OTnprbfeUt++ffXEE0/o/vvvL7Jvb29vmUwmTZgwQV26dFHnzp1LrKNXr14lLps0aZJ69uypUaNGSZIOHDigxMRE6zTN3NxctW7d+pb6AwCgvO3du1ePPPKITCaTTCaTunTpYu+SALvjRinAHeDq6mr93cnJyfraYDAoPz9fM2fOVPv27bV69WrNmTNHOTk51vVPnDghDw8PpaamWtuefvppRUVFaeLEiXJ2dtbSpUvVo0cPbdq0SSNHjiyxjj+f5fujNm3aaNeuXTKbzZIki8Wi0NBQxcXFKS4uTl9//bWmTp16S/0BAADA/gh1QDnIzMxUQECAJGnFihWF2v/5z3/q888/14ULF7R27VpJ0qeffqq4uDhNmTJFWVlZyszMVKdOnTRhwgQdPnxYkuTp6amsrKybrmHAgAHq1KmTxo4dq7y8PLVu3Vp79+7VyZMnJUlXrlzR8ePHS+wPAICK4L777tOmTZtkNpuVlZWlzZs327skwO6YfgmUg5EjRyomJkZz5sxRp06drO1Tp07VoEGDFBgYqClTpmjo0KFq166datasaV0nKytLzz33nPUMW0xMjKRrUy1fffVVLVq0SLNmzbqpOkaMGKHMzEy99NJLmj59uqZNm6YXX3zReuZw3Lhx8vT0LLY/AAAqgpYtWyo8PFx9+/ZVzZo11bRpU3l7e9u7LMCuDBaLxWLvIgAAAICblZWVJU9PT129elWDBg3Sm2++yc1SUKVxpg4AAAAOZdKkSUpMTJTZbFa/fv0IdKjyOFMHAAAAAA6MG6UAAAAAgAMj1AEAAACAAyPUAQAAAIADI9QBAAAAgAMj1AEAAACAAyPUAQAAAIADI9QBAAAAgAMj1AEAAACAAyPUAQAAAIADI9QBAAAAgAMj1AEAAACAAyPUAQAAAIADI9QBAAAAgAMj1AEAAACAAyPUATYUHh6uli1bqk2bNgoNDVVMTIyysrIKrTN79mwFBQXpwIEDhdqXL1+u4OBgtWnTRvfdd5+ioqK0adMmSdKuXbvUrFkztWnTRm3atFFERISWLVtWaPusrCy1adNGI0eOLFLXnj17NHDgQN1///164IEHNHDgQMXHx9/howcA4P/8cUy8/jN58uQbjnfXnT59Ws2aNdNrr71Waj8xMTEKCgoqNK6dPHlSQUFBhdbbsWOHhgwZojZt2qh9+/aKiorSvHnzZDab78wBA+WIUAfY2Mcff6x9+/YpNjZWhw4d0rx586zLLBaLYmNjVa1aNcXGxhbZtnXr1tq3b5/27NmjAQMGaNy4cbp48aIkyd/fX/v27dPevXs1fvx4vfrqq0pMTLRu+91338nV1VXff/+9zp07Z22/fPmynn32WQ0ePFi7d+/W1q1b9fzzz8vV1dWG7wIAAP83Jl7/mTRpkqQbj3eSFBcXJ19fX33zzTfKyckptZ9q1arp/fffL3H5N998ozFjxigyMlKbNm3Srl27NGPGDKWkpOjs2bO3f6BAOSPUAeXEz89PYWFhSkhIsLbt2bNH586d08SJE/X111+XOFA5OTmpf//+ys7O1qlTpwotMxgM6tq1q3x8fAqFuhUrVmjgwIEKCgrSypUrre3Hjx+XJPXp00dGo1Fubm4KCwtTs2bN7uThAgBwy4ob765/ATp27Fg5Oztr48aNpe7n0Ucf1eHDh7V79+4iyywWi9566y2NHj1ajz/+uKpVqyZJatiwoV599VU1aNDgjh4TUB4IdUA5SU5O1rZt23T33Xdb21asWKEuXbqoZ8+eklRkusl1eXl5WrJkiTw8PIoMNgUFBVq3bp0yMzPVtGlTSdLvv/+u3bt3KzIyUpGRkYXOAgYGBspoNOrll1/Wli1bCn0TCgCAPRU33v30009KTk5W79691bNnz2JntvyZm5ub/vKXv2jGjBlFlh07dkzJycnq3r37nS4fsBtCHWBjo0ePVps2bdSpUyfVqFFDY8aMkSRdvXpVa9euVWRkpFxcXBQREVFkoDpw4IDatm2r0NBQrVmzRh9++KG8vb0lSampqWrbtq0efPBBffDBB3rnnXfUsGFDSdemqQQFBalx48bq3bu3EhMTdejQIUmSl5eX/vOf/8hgMOjVV1/VQw89pGeffVZpaWnl+K4AAKqi0aNHq23bttafxYsXS7rxeLdixQp17NhRvr6+6tOnj7Zt26bz58+X2tfAgQN19uxZbdmypVB7RkaGpGszaK574YUX1LZtW7Vq1eqmQiNQ0RDqABv78MMPtW/fPi1atEjHjh2zDibr1q2Ts7OzOnbsKEmKjIzUtm3blJ6ebt22VatW2rNnj3bt2qXFixerQ4cO1mX+/v7as2ePdu/erbi4OPXu3du6LC4uTpGRkZKkgIAAtWvXTitWrLAub9Sokd566y1t3bpVq1atUmpqqqZOnWrT9wEAgA8//FB79uyx/jz++OOSSh7vsrOzrV+ASlKbNm1Up04drVq1StK1a/Su33Tl+vV517m6uuq5557TzJkzC7VXr15d0rUvR6+bMWOG9uzZo+bNm6ugoMA2Bw/YEKEOKCcPPPCAoqOj9fbbb0uSYmNjdeXKFXXp0kWhoaEaO3ascnNzrQNVWe3du1cnTpzQvHnzFBoaqtDQUMXHx2v16tXKy8srsn6jRo0UHR2to0eP3la/AADcaevWrdPly5f1xhtvWMe0lJQU69m0Z5991nrTlcmTJxfZPjo6WpmZmfruu++sbYGBgQoICNC6devK7TgAW3O2dwFAVTJs2DCFh4dr9+7d2rlzp/77v/+70C2WP/vsM8XFxWnYsGFl7iM2NlahoaHW8Chd+6azb9++2rp1q+655x5t2bJFvXr1Uu3atXX27FmtXr1arVq1uq1jAwDgTouNjVX//v31wgsvWNtSUlI0YMAAHT58uMhjCv7M2dlZf/3rXzVlyhRrm5OTk2JiYvSPf/xDXl5eioiIkI+Pj06ePHlT0zqBiohQB5SjGjVqKCoqSsOGDVNwcLDCwsIKLR8yZIgWLFigI0eOlGn/ZrNZ33zzjd5+++1C1wpIUlRUlGJjYzVx4kQdOHBACxYsUGZmpry9vdWlSxe99NJLZT4uAABuxrPPPiuj0Wh93aFDBz3yyCPFrpuSkqKdO3dqxYoVhcY0Pz8/Pfzww4qNjdXLL79cap99+vTRvHnzdOHCBWtbr1695OXlpXnz5mnatGlydXVVnTp19Pjjj6tHjx63cYSAfRgsFovF3kUAAAAAAMqGa+oAAAAAwIER6gAAAADAgRHqAAAAAMCBEeoAAAAAwIFVirtf7t+/XyaTyd5lAABswGw2q3Xr1vYuw2EwJgJA5VXSmFgpQp3JZFJwcLC9ywAA2EBCQoK9S3AojIkAUHmVNCYy/RIAAAAAHBihDgAAAAAcGKEOAAAAABxYpbimDgCqqtzcXCUlJSk7O9vepdw2Nzc31atXTy4uLvYuBQAAh0KoAwAHlpSUJG9vbzVo0EAGg8He5ZSZxWLR+fPnlZSUpMDAQHuXAwCAQ2H6JQA4sOzsbNWsWdOhA50kGQwG1axZs1KccQQAoLwR6gDAwTl6oLuushwHAADljVAHAAAAAA6sQl5Td/bsWb300ks6f/68DAaDHn/8cQ0bNszeZQEAAABAhVMhQ53RaFRMTIxCQkJ0+fJl9e/fX6GhoWrcuLG9SwMAAACACqVChjp/f3/5+/tLkry8vNSwYUOlpKQQ6gDgNixYsEDLli2TJA0YMEBdu3bVyJEj1bp1a+3bt0/33nuv+vfvr1mzZik9PV3Tp09Xy5YtdeXKFb355ps6evSo8vLy9Pzzz6tr1666evWqYmJidPToUQUGBio1NVWTJk1SixYt9Nprr+ngwYMym82KiIjQmDFj7Hz0AABUXhUy1P1RUlKSEhIS1KpVqxLXMZvNSkhIKMeqAKBiyM3N1dWrV0td79ChQ1q6dKkWLVoki8WiwYMHq1WrVjp16pTeeecdvfrqqxo0aJBiY2M1f/58bd68WR9++KHef/99zZ49W/fff78mTZqkS5cuafDgwbrvvvu0ePFieXp6atmyZUpMTNQTTzwhs9msq1evatSoUfL19VV+fr6eeeYZderUSU2bNr2p4+Hz/PYwJgJA1VOhQ11WVpbGjBmjCRMmyMvLq8T1TCaTgoODy7EyAKgYEhIS5O7uXup6P//8s7p3764aNWpIkiIiInTw4EHVq1dPLVu2lCQ1bdpUYWFh8vDwUIsWLTR37ly5u7tr165d2rZtmxYtWiTpWvDKyMhQfHy8hg4dKnd3d7Vo0UJBQUEymUxyd3dXbGysFi9erLy8PJ07d05JSUk3/HLuOhcXlyKf5wSUW8OYCACVV0ljYoUNdbm5uRozZowiIyPVvXt3e5cDAJWSq6ur9XcnJyfra4PBoPz8fOuyWbNmqWHDhje1z9OnT2v+/PlaunSpfH19FRMTI7PZfGcLBwAAVhXykQYWi0UTJ05Uw4YNNWLECHuXAwAOr23btlq/fr2uXr2qK1euaP369Wrbtu1NbRsWFqbPP/9cFotF0rWpnJJ033336ZtvvpEkJSYm6siRI5KuzbJwd3eXt7e30tLStHXrVhscEQAAuK5Cnqn76aefFBcXp6ZNmyoqKkqS9OKLL6pTp052rgwAHFNISIiio6P12GOPSbp2oxQfH5+b2va5557T1KlT1bdvXxUUFKhevXqaO3eunnrqKcXExKhXr15q2LChGjduLG9vbzVo0EDNmzdXz549Vbt2bd133322PDQAAKo8g+X6V68OLCEhgesHAFRJ9vz8y8/PV15enkwmk06dOqXhw4dr7dq1haZ03qrijofP+FvD+wUAlVdJn/EV8kwdAKDiu3r1qoYOHaq8vDxZLBa99tprtxXoAABA2RDqAABl4uXlpeXLl9u7DAAAqrwKeaMUAAAAAMDNIdQBAAAAgAMj1AEAAACAAyPUAQAAAIADI9QBAG7LK6+8ooceekh9+vQpdvmuXbt0//33KyoqSlFRUfrggw/KuUIAACo37n4JAFWEpaBA57Zu15mVq5STdl6utWqqbt9I+XUMk8Gp7N/xRUdHa/DgwXr55ZdLXKdt27aaO3dumfsAAAAlI9QBQBVgKSjQr2+9qwv7D6jAbJYk5V68qN8++ljnd+5Us5fHlznYtWvXTklJSXeyXAAAcAuYfgkAVcC5rdsLBbrrCsxmXdh3QGnbdti0//3796tv374aOXKkjh49atO+AACoagh1AFAFnFm5qkigu67AbNbvcats1ndISIg2btyolStXasiQIRo9erTN+gIAoCoi1AFAFZCTdr6U5Wk269vLy0uenp6SpE6dOikvL0/p6ek26w8AgKqGUAcAVYBrrZqlLK9ls77PnTsni8UiSYqPj1dBQYGqV69us/4AAKhquFEKAFQBdftG6rePPi52CqaTyaS7oiLLvO8XX3xRu3fvVkZGhjp27Ki//vWvysvLkyQ9+eST+vbbb/Xll1/KaDTKzc1N7733ngwGQ5n7AwAAhRHqAKAK8OsYpvPf7yxysxQnk0nV2rRSrYdDy7zv995774bLBw8erMGDB5d5/wAA4MYIdQBQBRicnNQsZrzStu3Q73GrlJOWJtdatXRXVKRqPRx6W8+pAwAA9kWoA4AqwuDkJL9OD8uv08P2LgUAANxBfDULAAAAAA6MUAcAAAAADoxQBwAAAAAOjFAHAAAAAA6MUAcAuC1nz57VkCFD1KtXr//P3n0HxlHdix7/bu9Fu5JWvVpy7x2DDcY2YDqmpRAISUjITXu5Lzf1vpDGfclNcvNSbm56CBBaCLYpxo5tXCju3ZYlq9dVWWl735n3x8LCuoVie23pfP7yzM7OnlmvZuY355zfj+uvv55HHnnklG1kWeb73/8+y5cv58Ybb+To0aM5aKkgCIIgjE4i+6UgCMIYIUsyR/b3sGNbG35vBKvdwILF1UyZWYpC+f6LgatUKr72ta8xefJkgsEgq1atYtGiRYwbNy6zzbZt22hvb2fDhg0cPHiQhx56iGeeeeZcHJYgCIIgjHkiqBMEQRgDZEnm6Uf20No0RCKeAiAUjPPC3w5z7FAfd947530HdoWFhRQWFgJgNpupqamhv78/K6jbtGkTt9xyCwqFghkzZuD3+xkYGMi8TxAuFbIsE/QNg0KBxebIdXME4aIRCQaIxyOYrA7UahFiXGjiGxcEQRgDjuzvyQro3pKIp2htGuLIgV6mzir9wJ/T3d1NQ0MD06dPz1rf399PUVFRZrmoqIj+/n4R1AmXlGQiTtPBnWx/4a+oVCquvOU+aibNRKlS5bppgpBTA91tbHjqt/iGB5gy70pmX3UjZqs9180aU8ScOkEQhDFgx7a2UwK6tyTiKXZsbf3AnxEKhfjCF77AN77xDcxm8wfenyBcbEYG+9j89z+RiEWJhkNseOo3eD39uW6WIORULBpm65rH8A65kSWJwzs2M9D1wa8pwnsjgjpBEIQxwO+NfKDX/5lEIsEXvvAFbrzxRlasWHHK6y6XC7fbnVl2u924XK4P9JmCcKHFoxGQ5cyylEqSiMdy2CJByL1kPI7fO5S1LhoJ5ag1Y5cI6gRBEMYAq93wgV4/G1mW+eY3v0lNTQ0f//jHT7vN0qVLWb16NbIsc+DAASwWixh6KVxy7PlF5JdUZpYr6qZicxTksEWCkHsmq505V92QWdbo9BSWVuWuQWOUmFMnCIIwBixYXM0Lfzt82iGYGq2KBUtq3ve+9+7dy5o1a6ivr+fmm28G4Mtf/jK9vb0AfOhDH2LJkiVs3bqV5cuXYzAYePjhh9/35wlCrpisdlZ+9HMM9XaiUCrJL65AbxRDjQVh/MxFOF1lREIBHIUlOIvKct2kMUcEdYIgCGPAlJmlHDvUd0qyFI1WRU19PlNmlLzvfc+ZM4fGxsazbqNQKPj2t7/9vj9DEC4W1rx8rHn5uW6GIFxU9AYjZbUTc92MMU0EdYIgCGOAQqngznvncORALzu2tr5dp25JDVNmlHygOnWCIAiCIOSWCOoEQRDGCIVSwdRZpeekdIEgCIIgCBcPkShFEAThEie/IxvfpWy0HIcweqSSSQJeD7HoB8sOKwhjiSRJBLwekQHzAhM9dYIgCJcwvV6Px+PB6XSiUFy6QyhlWcbj8aDX63PdFEEAIOD1sHfLizTs3Y6rvIYrb/4YDpfo5RZGn1Qqid8zSCqVxOosQKt9/+fhcNDP4R2bObD9ZazOQq5edb/IhHmBiKBOEAThElZWVkZ3dzeDg4O5bsoHptfrKSsTGdOEi0NP63GO7HwFgN62JnZuWsPyOx9ArRa3TsKlKxYJk0zGMVnsAEipFI3732DL6keQUimmLryaectuwfA+s7oO9nSwe9MaADx9XWx69o/c+smvojeaztkxCKcnzkyCIAiXMI1GQ3V1da6bIQijQsg/wlBfN7IsoVSpQKHIFBsf7u8hmYiLoE64ZPV3tbJ1zaME/SPMW3YL9dPmEwr42LrmL0ipdFbkw29sYtzUuZRWj896bzwWZaivk0gogD2/COcZeq2jkWDWsvTgfT0AACAASURBVG+on0Q8KoK6C0CcmQRBEARBGPOi4SBbVv+FtoYDABRV1DJ57hKO7toCwPTLlqM3GHPYQkF4/8IBHy//9b8JeD2o1Bp2b1qDo6AYvcmSCejecvIyQOvRvWx85vdAurj4qk9/nfziilO2cxaVodHpScSiAMxcfB0ma955OCLhZBdtULdt2zZ+8IMfIEkSd9xxBw888ECumyQIgiAIwijlGx7MBHQA7s4Wpi5chslqR63WYLY7ctg6Qfhg4rEIAd8wE2dfgSXPScjvJRaN4CyuYNaS69m75QUASqrH4yjMrlvqcXeza+OazHIiFqW3/cRpg7r8onJWfeYbeNzd6A1mCkorUSpFXsYL4aIM6lKpFN/97nf505/+hMvl4vbbb2fp0qWMGzcu1007L2LDwyjVajRW6ymvxYNBkj4fCo0GOZFA63CQCoVQarWn3X60SyZTDA+GUKqU5Bemx3tHIwmi0SRmixaVUsnIcBhZlnE4TVm1t8KhOKFgFJ1Og4yMxWpAKWpzXXAJvx8pHkdpNJIMBFCo1KgtZpJ+PyqjCSkSQanTorFYgHQCjfjwMEqNFimRQG0yohoDyTSikQTRSAKFUoFOq0Jv1J6yTcAXIRxOoNUo0erUGE26zG8+Fk3g80bR6pSo1SrMltH/nQnCexEJB4mGgxhNVqKRELIkoVSqkKS3eynUag1SKkkSJSaLDd/wIMlEHO+gG0lKYXMUotUbUCiVJGIRQIFarSUSCaLR6FCrNSRTcWRJRq3Rvjl8U4NapyMRjaDVm0ilEsiShCTLqFQqdAYjeoMYqiacWyarg1lXXEc8HmH3pjVMX7QC//AAkpSicsI0bI4CLI58rPZ8wkE/8XgUSZJAlpBSEhX1Uziya0tmOHJeYQlD7i50ehMKZFKSRCIWQ0ZCliSKymuRgUQ8xvBAL7IsI0syWr2eRCJBKhFDrdUjyxIms00MzzwHVA899NBDuW7EyQ4ePEhTUxP33HMPKpUKv99PW1sbc+bMOe32Q0NDFBQUXOBWfnDJcISBzVto/OGPGdz+KqaqSvSFhZnX/c0t9K9bT8/f/k4yGMJ/9BgDGzcjRaO0/Pb3mGtr0eXn5/AILqzhoSANh9ysffoQB3d1YbbqUCjh2Uf3s3ndcbQ6NfFYktVPHOCNLa0gyxQWW9BoVPR0jvDso/t4dWMzsWgSjVZF01E3+S4zGu1F+WxjVPIfb+T4D35IuKOThN9PxyOPEe3tQ47H6X3hRcIdnTT//FcM79qNeVwtSp0O98vr8ezchf/wEVp/+3v8jU2Ya6rR2EbvQ42BvgB/f2wfm9c1EgzEGB4Mo9OrsVjTgVkslqTxiJvGo/10tAwz2B9ky8tN+LwRHPkm/L4oLzx9iE0vNjA8GCIUjOMbDpPnNKFSX3pPTC/Vc3yuiO/r7GRZpq/jBLs3rmbb83+lr/0EDlcpDXtfY9yU2fR3taFUKpm5eCXuzhbKxk1isKeNgZ4OlEoFuzatYf/2dbQc2cPIYB8y0Nd+gi1r/kJH40EKSito3Pc62194nCF3N9a8fHzDA+zfto4dG57FNzyAxZ7Phqd/SzwaxmS107j/NdydJ9jy3CO0Hz9IYVk1Jqs911+VMIqoVCosdievv/wME2ctorv1OMf3vUbzoV1EQgHyiyvwjwyxb9tL7Nv6EoWllWx7/nEObH8ZpUqF3mimfvoCuluOce1HPkvTgTfY/sITuDuaKSit4h9P/QaT1c725//KkR2voDeaaDmyF5VKTcPeVwkMD7JlzV+wOQrZ88padm9aSyTkx+504fMMkIjFxG/+XTrTOf6ivLr39/dTVFSUWXa5XPT39+ewRedHuLOTlv/+H5KBANGeXo7/3x8THRwCIBkO4929h97Vawl3dNL73BrUJhOB443EhoYgJdH04/8iPuLN8VFcOAN9AV5+7iihQIxQMM7aJw/i7vHT3TFCZY0Dg0HDi387jGcgRDyW4pWXm+jt8hKLJFi/+ijuHj/JpMTeHZ2MeMIMe8K4u/25PqwxIz4yQtOP/4tIXx/m+jraf/8nwu0dDGzcxPDO3Thmz6b3uTWkIhHC7R00/eznhJpb6PjLY+jy8nCvW08yEMC7Zy+dTzxFKh7P9SGdF8lEildebqSrfYRUUuLw3h5QwMurDxONJAAY7AvgHQ6z+7UOzBYdr7/SwoA7wOuvtHBgVxcNB3tpaRxESsk0HRsgHIqz8YXjDLgDOT46Qcg9r2cAd0czJw7tQkol6W1vpGH3NnxDfRzdvY1pC69m6aqP09NyjKaDb3Dw9Y3EImFsjgL8w0P0tTdl9tXb3oTZZmff1hdJxKL4PANsXfMoOoMJKZWip7WB/q5WoqEgJdX1zLz8WoZ6O+lqPsKUeVdy4NX1DPV2YLTYObLjFZKJOEN9nbzy3CNEI+EcfkvCaGS2OyiurMNkc+Bxd2XWtzccQKc3YrY5Gehqo3bKbHZtXM1Ifw+JeIwDr65HpdGwd+uL3PLJf8M/PETTgR1IqSTuzmYOvf4PZi5Zyc6Nq/EPDxKPRXhj/d+onjSDbWsfw1lUxr5t63C6yuhoPIS7oxlJSnHi4E6G+jqJhoKMDPaSTCZz+O1c+kZFF0UsFqOhoSHXzXjPbEOerOWk349vYIC2oUEKzGZ8h49kvR7u7EJf5CLS04vW4cB//DgjA/0Mu/suZLNzQqfTEQ6pkKS3ixNLkkwyIQFQNS4fvzeK3xfNel8wEKWvr5/hoeyLYzSSxGoz4PUGaWgYOv8HIOBQKIl5PKjNZuKe7N9+sLkZ58IFWeti7n5iXi9qk4mEz5f1Wqillb7OTgKJxHlv94Wm15sZPCn4ioTihAIJ3H0DRGJ+khEjqZSELc/A0EB2prGWxkEqa51Z6/zeKCaLjhGPn0DYfd6PQcitS/WaeKHYTFr8I9nnfa/Hjc3pwt3Zwt6tL1I7ZS6JRPrBkd8zQFnNBIb6unC6SlAoFMhvDkFTKBQolaqsfYUDvqyhZCZrHv7hQY7u2oJGq2Pu1Tcx3N9L4YR01tp4LIYsS9ntGXIz4O4lFB195zght6YsXEEkMJy1TqVWM9DTRuOBHSy5+aP4hgfxj2Rfp6VUEv/wILKcwj+SXULH6xnAYDQTCWVfu5LxOLFoGCmVDtZMVjs+T3YnTdA3gkanR05I9Pb0EAqLhxnv10UZ1LlcLtzut288+vv7cblcZ9xep9MxceLEC9G0cyrS10e33U7Cm+5tc16xiPyaGlwGPbIsk1h8Of6jxzLbWydOoOvJp6n6xH20/e6PFK+8joKaGlwaTY6O4MLqVAxTVGrF3ZPuXSsqtWK16YD0cJqWxgGmzirl0J5uAPQGDcVleRS4zCxYEmXzS8cB0GhV5LvM7Hm9netvn0ph0egdxncxkRIJgtddg/vFdehOGjbgmDuHZDiEymQkFUqf0Etuuh5zbTWylEJfUgxKJUjpG5+Sm2+geJTOsQW4bKnMC08fAkCpUpDnNDF9ro7KqhIUylKGh0L4vX1EI3FKyu0cO/j2g50Z88uxWvW8/kpLeoUCKmocdLZ5KK3Ix5F/6c1bEAHKe3OpXhMvlN72JvIKi7Pmz02ZfxVvrH82vYFCQWl1Pa1H9wAwdcFSGva9ytT5S2nYu525V9/M/m3rAJh95fWEg34MJkvmhnb8zEV0t7z9mzVZ7bz20pMAxKJhdm1awzUfepCRwT5sThf2giICI0NotHoS8fSDyVmLV1JeWYNCJJgQzoNYOMzCa+9g18bVqNRqZi+5nmN7tuPz9NNyZB96k4WJsxZxdPdWALR6A0qliuKqesLBABV1Uziy45V3/P1cSeOBHdROnk3LkfTfjdFsRaFUUjl+GrIsYzBZ6G5pYOYV1zLUl+4lVCpVlNZMIBmPoTOaqaiszM0Xcok50zVRIb/1uOkikkwmueaaa/jzn/+cSZTyk5/8hLq6utNu39DQcMlewMJd3YTa2lDqdG/OkXv7CXt0cAj/0WMEjh/HOmkiCrUalcGAlEyiUCgw19agzRs7aWKTyRS9XV48AyGUSgWuUiv2PAO93T6klMyObW2YzFoKiyyoNSrKq+yUlKe/n3AoRlfbCEF/DJvDiFIJZouewmJLjo9qbImPjBBsaUVKpVCqlIzsO4ChrBRTVRVxjwet3f5mb57lzd+3nXBnF5HePpQaDRG3G32RC0tdHRrr6P2/i0bi9Hb58I1EMFt0qDVKCoutmMy6zDZDg0G8njDxWBJJkulsG6ai2kF1XT5arYrOtmF8IxGcBWYS8SS2PCMFRZfmd3Ypn+NzQXxfZ+fp7+GlR3/B1AVXkYjH0inYtQbi0RDhUACbowBZgpGhXsxWBzqjCQUQCYcAmaG+Lkoq69Do9STjccJBH0azlaBvBIVCgSWvgGg4QDjgxWLPJ5mIs/6JX2c+X6FQcM2HP4vH3U1Z7UQUSiWRgB+lSkU4kN5XUeU4DKZL8+9VuDQMuXuIR0MEvcO8/vIzBH3p3rvC0irmLL0RrU5PJBwmEYug0ekJ+YcpLK0hGg6g0eiQZAmfZwCj2YreaGZkoA9znpNENEIiEcfmLMQ/MoTdWUjI70Wj06d/3xYbUjJJKODFnu9CpdIiySnyi8tFgqB36Uzn+IsyqAPYunUrDz/8MKlUilWrVvHggw+ecVtxARMgPdRysD+IUqGgoMiM0aT7528ShFEqlZToaPXw6qZmbHYDC66swVV8afZKi3P8eyO+r7NLJZOcOLSTrWsfQ5YkrrrtPqRUiq1rHqV83GSSiThdzUcBUKpU3PHZf6egJN2DMDzQh394gHWP/5LUm/N/Js+7ksU3fgT/yBB//a9v8M7bqus/9kWcRWWs/dNP8A6mRyBNnH05sWiE+ctvzRRwTsSiRMNBDGYras2pmW4F4VwLB3zs3fIiBWVVbHnuEZKJOCgUrLjrAeqnp6dDxCJhwgEfBrOFeCLOq2sfp/XYPoxmKys/9gX8w4MERobe7uUGVtz96cz7TzbY08Ez//29TA9fUeU4Vtz1aax5Yyfp37lwpnP8RTn8EmDJkiUsWbIk1824oFLxOMlAEI3NilJ90f7XXLTMFr1I236JkiWJSG8vCZ8fXWEh+oL0CT4+MkJ8eBhNXh46h6gR9V709/l5/Lc738o+TXfnCPd99jJMFvGwQxjbVGo142deRkn1eGRZxmpPj5ApqaonFg3z9C+/k9lWSqXwDvVTUFJJR+Nhhno7iUZDmYAOoOXoXqYuWIrNUciU+Us5vGMTACabA0dhCda8fG66718ZHugBFOhMJszWPCy2t89pGp0ejU5cv4QLx2ixMX7mQtobD7Pi7k8Ti4RRqTUUlFQB4B3q55XnHqGntQGHq5Rr7v4My+78JAHvMFq9AYvNQVF5Lat//59Z+z1xcCf10xcgSRLeITfRcBBrXgFmWx6+4cGskiHujmZSSTFv9FwRkcNFItLXR8fjT+DbfxDn5ZdRtuo29IXvPiV1wp+eZzYWa9e9V8FAFJVKieE0db+E3PAfa+DYd76PFI+jLShg8v/5JgAND/+QaF8f2vx8Jn3r65iqq3LazkuJ3xvlneMwPAMhwqG4COoEgfQQyJN7BzRaHQGvhyU3fZRje7Yz2NsBgMXuxD8yxK6Nq7E68iksq856X2lVPduef5x5V9/M3KtvpGridGKRECaLndSbCSKsjnysDtEbIVxcCsuqseUXEQqMkIzHMVnzMFlsALQ17KentQG9ycK4qXPpaWuklPE4XKUoFOl6qP6RISbOvpyS6nqO7txCKOClcvw0ALpbGnjhkZ8hpZLYnIXccO+XsOQ5sxINFZRWiWHG55AI6i4CsizTv+EfeLa/BkD/yxswlpdTcsPKrO0SgSAqgz6rF09KJvHuO0Dr7/+IQqmk5oFPYJ8+DYUqOxuXkJ5X13h0gC3rGjFatKy8bQrlVaL3J9dSkSjtf3kM6c0SBfHBQfwnmol29xDtSycAiQ8N0fvCS9R9/rO5bOolJc9pRKVWkkqmk8uUV+VhtoqAThBOJ+AdZv2Tv8bd0QwKBZddewe2/EImzrqc/OIKAt4hhtydjJs6l7aGAyxaeRftxw9hz3dhMFto2fw8r69/hpvv/wo6vYF1j/0yXWhco+XWT30VV3lNrg9REE4x0NPOodc2Ikkppl++IhPQAYT8IwDMW3oTb6x/lkQ8ikqt5uZP/BslVXUMubtY+4cfEw76UWu0LLn5HqKRMNUTZxKNhHn1pSczWS99ngF625qYMHsRN33iKxzYvh6bo4Ap869CbzTn5NhHIxHU5VAyFCLU0YlCqSTQeCLrtVBLOnNdIhDAd/gIgeON6IuKSARDOOfPJRnwo9TpUahVHP+/P0JOpbuzGx7+ITN+9mOMZWUX/HguJL8vSkeLh+HBIDXjCygpt6NSKYlFEwwNBEkmJJyFpsxwzI5WD4PuAC89my4TEfBHefrPe/nUly7Hajfk8lAEZJS6k4MNmWQwO1V/MhhElqQxmQ3OOxJmoDeASpVOEPRuhhkXFlv41Jcup6lhAAUwYWqR6J0WhDMY6utMB3QAsszuTWu5+wvfweYsBMCSl8/UBUtpbzxI9cQZ+Ec8TJq7mB0bniXwZnkEBQqUSgUHX9uYnp8EJBNxjuzactqgTpIklGPwfCZcHHzDg6z940+IhtPX2vbjB7nzc9/Gnp/ONj9u6jy6mo/R392aycqaSibZv20dRRW1tB3dTziYHiWWTMRpOriDGz72RVRqDdFIGMXJH6hQoFKpKa+dSFn1eBRKJZIknbyV8AGIoC5H5FSK/g3/oP3Pj6Iy6Cm/+66s8gUFV6bnEw7v3EXzL/47s77q/vvoeW414fZOQm1t1P/rlzIBHYCcSJAMhi7cgeSAJMnserWN1zenA99tG5v5+Ocuo6jUys7tbWx5OV0Ytmqck1s/PJNQKM4zj+xl3uVVWfsJBWLEYqLQZa5E+vro37CJwIkTFK1YBoD/0GF0hQUkR7w4FsxjcOs2pFgMhUZD6S03jcmAzu+L8vSf9uLuSdfqmzKzhOtXTUVnOHMpk1RKorN1mNe3tGB3GJm3qApngXgaKghn8tZwsncuv/N8o1ZrmH3lDQy5u0jG44ybOo9EIoZSkd5GrdGy8No70Gj1aPXZDwp1Jy0HvB5OHNxJe+MhJs6+gqqJ0zGI3grhAgv5RzIBHUA8FiHoH8kEda7yGq750IM07NmW9T6tTo8CUChP/ptRwpuhnN5g5PLrP8QLj/yMVDKBvaCIkqr6zLbhYICWo3toPrSLuunzqZ08G+M7egmF90cEdTkS83jo/OtTQHr42fDuPYz/2v8m4fVhrKzEPK6WVCyGe936rPcFT5wgFYliKC8l0ttLsKUNfVkp0e4eAEw11ejPUtNvNAgGYuzf0ZlZliUZd48fvV7D1g1v93i2N3sY6A8gJSXCwTiyDEaTlnAo/QR14rQirDYxMT0XUvE4HY/9Fc+rrwPgP3yECV//KvYpk0kGQ3Q+8RSmmhqm/vBhEsPD6AryMZSX57jVueEZCGYCOoAj+3u57KpaikrPfAHs7/Xz2G92vJ0kpWOEez69AKNJ9NQJwunkF1dQPm4yXc1HUSpVLLn5nlPm3BlMFsprJ2Wtu/WBr+EfGcJgsmRuhqcuWEpH4yGCvmEseflMnH1FZntZljm2Zzu7N60BoLetkWs/8i+MmzLnPB+hIGQz2xwYzFYib/a26QymrOQ9CoUCR2EJE2ZfQcux/fg9A5gsdmZccS0KpZLqSbM4smsrQa8Hrc7AvKtvQvWO6UFlNRO4+/PfIRIOYnXkY7a+XYKrrWEf29Y+BqTrRiqVKibPG1vJEc8HEdTliFKtRmUyZuYR+Y8cpWzVLeRfd21mG1mWsU2bRrC5JbNOX1yMUqNBZTSgLyxEqddT/6XPE+npBYUSS30d2jz7BT+eC0lvUFNWmceJhoHMujyHAYUSlEoFKent7BAqpQK1Xk1RqZU3trSwctVUwqE4Wo0KV4kNrVb8CeRC0h/Au+9A1rrowABdT/8N+c2sckqtBkNREebqqgvfwIuIVpf9G1WrlWi0Z58z6xuJZCVJ6e/xEw7GRVAnCGdgtuVx9e2fxOPuJJlIoDeaiMciaHVnH55vtuVhtmXXi80vLuf2B79FOODFaLVn3cwm4lFaj+7N2r6ntVEEdcIFZ83L56aP/yvH9mxHllJMmrs4M9z4nZyuUlY98HVC/hEMFlsm8HO6Sln1mW8Q9HowmG3YT3qvQqkkr7CY01VTbjnpb6D12D4R1J0D4o42R7QOB3Vf/AJNP/4pyVAI17KlmKqzM2opFAqKrllOKhbFu+8AzssWYJkwHoVCQeOPf0oqFAYg1N5B3ec/i0o/NnqdtFo1y2+ciE6vpq/Hx4LFNZRU2NHpNVx32xRe+tthUMA1N09Gb9CgN2hYunICyWSKQXeQrRuakCUZhVLB/Z9fREn56A6CL0YamxXnZQsY2Lg5s85UVYlj4Xw8219DX1xE9f33oTKMjd/02eS7zKy4aRKbXzqOSq3kxjunkefMLtAqSzLekfT5wJ5nJM9pRKlSIKXSkV15VR5miwjoBOFsBnvbefEv/y+zfN1HPkftlNnvaR+xaIR4LILJbDsl2APQ6gzUz1jIGy8/k1lXUTf5/TdaED6AgpIKltz0kX+6nclqx2Q99V7JYnNk9e69W/XTF9B14mhmuW7avPe8D+FUF23x8ffiUi60Gh0YRIrF0BUUoNKfPjOdnEoRbO+g88mn8e3bT9XHP0bb7/6Ytc3MX/0cY1nphWjyRSOZTJGIp7KSP6SSEsNDIWLRJC8+exiTWUuBy8Ku19pRqRQsXlbHwb3deAbS8w6vv30qsxdW5uoQxrRo/wCeHTsJtrZReNWVWCdNAEki7hlGZTKitYtg+y2SJOMbCaNQKLA7jFmvJRMpDu/r4eXVR5FlmRU3TWL67DL6ev0c3N2FzW5g8owSnIWX7pydS/kcnwvi+3p/Xnrsl1m9aOXjJnHzJ77yrt/vcXezde2jDPV2MXXh1UxbuAyT9dRh0iH/CG3HD9J94hh10+dTNm4iOr3xNHsUhNEpHPTT2XSEtob91EyaRUX9FFHa4D245IqPjxVv1aKLe72M7N1LpKcX25TJmOvGodSkEyGkolFafvlrQq2tAEixeNY+dAUFqE1j74KgVqtQq7OHoanUSvJdZl569jD9vX6WrpzA5peOA5CUZLZuOMHly8axdX06mYrVLnqCckXvKqT05htPWW8oLclBay5uSqXilN65t3iGQrzwzKHMcMuX/n6Esso8KqodVFSLkh2C8G4VV47LCupKqie86/cm4jFee+kpetvS15YDr66nrHYiB15bTzIeY+LsKygsqyIaCRH0jeAqq2b8jIVotKLMiDD2GM1WiipqsDrysdicIqA7R0RQdxGQZZn+jZvpfPTx9AqlkqkPfw/rxPQFRYrHSXhHMtuP7NlLzac/Rc9zq9HmOaj+1P1o8043anlskiUZ73B6KNrJHdGplITZosNs1aWHbYqhl5eUYEsrfS+uQ04lKb7xeizjxuW6STmXiKey5s8hQyKROuP2giCcXu2UOQR9IzQf3kX1pFmMn7HgXb2vv7uNga5Whgd6M+umzL+Kzc/+kYDXA8Dx/a9z+4PfYu/WF2na/wYAC665nRmXr0CtPnMmW0G4mMSjEbqaj3Jk11ZKq8dTP30BVkf+GbcfHuilYe+rhHwjTF24lKLyWhRKJe7OZtb84Sck4lFMVjs3ffxfcRaN7lJcF4II6i4Cca8Xz46db6+QJALHGzNBnTYvj/IP3U3Lr34NQLizC9vUyeQvWohCo0FtHHu9dGejVClZeNU4WpuGCPiiFJfZ6OtOZw+cOb+cCdOKGD/FhcmsOyWNtXB+RYc8hNva6N+4CY3VRvH112GqenfDX6P9Axz99ndJBgIADO/aw/Sf/AhDSfH5bPJFz5FvYtyEQpqPpxMHVdU5ceSfvldPEIQzs+bls+i6O5m9ZCU6owmV6p/fInmH+ln9+x+hN5ion76AfdteAkBvMGUCOoBELIrPM5AJ6AB2bniW6okzcLrSUydi0TD+4SHUGg02RyFK1dkTIgnChebuamHd478CoOvEEUIBH5dff1fmb8U71I/H3YVaq8OeX8T6J/4Hj7sLgOYju7nzX75NXkExuzatzdS+C/m9tB7dJ4K6c0AEdTkU83gY3Lodz2uvkzdzBuaqKvr/sREA/Uk3qgVXLMJUXUXC60VfXDzm5s+9V5U1Dj7xxcsJBWPMWlDB8FCYUDBGNJLANxwRPXQ5ICUS+I8eo/kXv0JOJADwHjrEtP/4AVrHP+9pjns8mYAOIBUOE/d4xnxQZzRpuemuaXgGQ4RDcRz5RkxmMaRLEN4PpUr1nuplBX3DJGJRErEoI0N9LLz2DnR6AwWl1RzeuZlwIP1AUa3RZqV7P1ko4OO1l56i6cAbKFVqlt3+CeqmzxcPHoWLykBPe9Zy27F9zL3qRowWKz7PAKv/8COC3mEAlt3xqUxAByClUvhHBskrOPWaLXPJp/e4KIigLoc8b+yg45FHAQg2t1Dxkbsx1dZQeOUSLBPGZ7aTJYlwVzfDe/aiK8jHWCkSe/wzSoUCSZLpbvdiseoYGQnzxivpOYk7t7XxyS9dfkrCCeH8inu9RN3uTEAHEHP3E/d631VQp813orHZSPjSN0kqkwlt/pmHfYwVqVQK30iEJ/6wm3gsid6g4Z7PzKe4TDy4EITzzWxzoNUbiEcjtB3bT39nC3d89v9gyXNy431f5sCrG0jEI8xafB1Wh4v6GQtoOrADgPnLb8ukkPe4u2k6kO7Fk1JJXnnuEVwVtdgcBTk7NkE4massO0v7uKlz0BvTI0OGB3ozAR1Ab3sjea5SRvrTdZSVKhU6vZFwwMekuYspq51IV/NRht091E5+b1lmhdMTQV0ODe/ei8poxD59Ggmfj0DjCaZ87yHUpuyhU6G2dg5/o14evQAAIABJREFU/VuZ+l0FS66g9nOfRaUVKcrPxN3r55FfvUEqJQEwfoqLqnFO2ps9hENxgoGYCOouMLXJjNZuR6nVZuoz6kuK3/V8UH1hIZMe+nf6/7EROZWi6JoVGIqLzmeTL2qyLNPVNsLRg70E/THisfT5IRpJsO+NTq6/QwR1gnC+2fNd3PLJf+Pori0oVWomzV2MJc8JpNPFL7vjEyDLKJRKABbf+FGmXbYMlVKNPb8oM59OkrLnwaZSSWRJurAHIwj/hKu8lus/9kWO7dlOSVUdtVPmZIYJa/XZNR2H+rqZv+wW2hsOEI2EqKibQsPe15CkVOYBxvzlt7J01cex2JwX/FhGIxHU5VDxymuJTJ/G0LZX0TodlNx0wykBHUCktzcT0AEMbnuVio98CJXLdSGbe8mIxRJEwnGcBSYG3Onheo1H+1myvJ72Zg92pxGr/ewFZYVzT200YJ08iXGf/yyenbvQ2mwULl+GNu/dBx/mmmrMn/7UeWzlpWOoP8ijv9lBgct8SmZMpVqZo1YJwtgRCQUIeofRG0xcdet9p91GoVDAO4ZQ6o0mioy1p2zndJXhKq+mv6sNgAUrbsOaJ0YiCBcPn2eAeCxCceU4qifOOOV1Z1E5C1asYvfmNegNZhatvIuu5iP0d7Wi1RvY/vzjVE2YQTQSzLxnzyvPUz9j4YU8jFFNBHU5pDKZMsMvQ21txDwepnzvITSW7NSuWkd2WnLzuHGnDf7GokgojkKpQG9IP+0ccAd4+bkjuHv8TJpeTFGZjUN7uikps2F3GLjymnomTivGahOlDHLBWFaKvshF3tw5qA0isP4gAv4oqaSEu8fPjHnldLSke6FNFh0z55fnunmCMKr5PANseOo39He1ojOYuOHeL1JcWfe+92exO7juo1/AO9iHWqvDUVgiEqUIF43ulgZeevQXxGMRymonsnTV/ac8dNAbjMxaspL6GfNRqTSYrHb0BiPH977GyGAfOoOJqQuvZu2ffpJ5j0ZnEL/zc0gEdTmU8PqylsNt7SSDwVOCOlNtDfVf+TLdzzyLoayU8jvvQG2+dAsJnwuppERz4wAbn29Ao1FyzS1TKC63s/GFBtqb0xnH9u3oZNkNE5k8o5grltdTWCTqoFwMlGo1yrMkDBDeHVueAb1BQzSS4JV1jVy2tJayijwc+UZseWJosSC8F+GAj96OE/g8/ZRUjaewrOqs2S972hrp70rP045FQrz24lPceP+/ojtpCNpbySEkScLqKER9lnOf2WrHbBXDpoXcSyUT9He30dvehD2/iJ6W48RjESAd4A10t522J1mpVGLNe3seqLOojFUPfpPAyBCJeIy24wcYN2UOzYd3o9HpuXrV/Vhsop7quSLurHLIUFKMQqPJJI5wXrYwa35RdGCQhN+PLt9JweWLyJs1C6VWI26IgcGBAE//aU+mPtcTf9jNJ7+4iIE+f9Z2CgXc/KEZpxQpF4RLnbPAzD0PLuDYgV7UGhUTJhdRIB5cCMJ7Jssyx/ZsZ8eGZ4H0kMnbPv0NiivPXAczmYhnLcdiEWQ5ew5cKpmgYe9rbHv+cWRZYt7VNzN90XK0OjFKQbi49Xe18vff/ZC3brJmL7kekzWPkD9dMzn5joRn/4zF5iAejbD6dz9CklKU1U5i/vJbqaibQuFJiVeED0ZEBzlkqq5i6sPfY2Tffsy1NehchSg16WGEgRPNHPvuD0j6/Rgqypnw1a+IMgbvEA7Gswoux2NJUimJhUtqWL/mGABKlQKdXs1Qfwh3jxcUkF9gwVVqRaMRQd7FTk6lCHf3kPB60blcGIrG3hxSvzeCZzCE3qBBoYTDe3tQq5VMnlFKYbGF4lIbxaXvPv26IAinioT8HN6xObMsyzL9XS1nDepKayZgMFuJBP0oFAouu/YO9IbsaRFezwBb1z6aSXiy8x/PUVE3BVd5zfk5EEE4R7qaj/HOm6yWo3soq51A4/43MNsduMrfWzCWTMQzyYC6W47R3XKMkurxomTHOSaCuhxSKBRY6tNj8E/84lfE3P2UrroV14rldP71SZL+dK9TpLOLkb37TgnqpESCqLufVCSM72gDwaYTFN9wHdYJE1CMwjHKfm+EUDCO1a7HkW/CYtMT8KWLV5ZW2DFb9UycVgwKBZFQHL1Bw/aNzSxZUYd3JMLObW3o9Gquu3UK46eM3ayJuRQf8RLp6Uap0WIoL0NtfHuYYOBEM+6X1qHQaClaeS1Jv59j3/k+cjKJ2mplyvceeteFykeD4aEQT/xhF56BEHqjhpvunI7BqAEZnv/bQe742GysttM/8Q8GYngGgqhUSvJd5sycU0EQTqXVGSgqr6Hl6N7MOquj8KzvcbpKuf3Bb+IbGsBgtuIoPLX2lpQ8NYNlKvnuezgEIVccruz7zaKKccy44lrGz1yEPd/1npP42ByFVE+aSdux/QCUVNWTly/uw841EdTlWMLnp+lnPyfa0wtA1xNPYamrIxWJZG2XDAazl6NR3C+so+Pxv6JQKCi56QZiQ0Mc+dZDTP/xDzHXjq4nge4eH0/8YTcBX5TCYgu33zOLj356Pp2tw6jVSsqrHZjMOiLhOHtf72B4KIQkpZ8yqdUqtm04AUAsmmT1Ewd44MtXnJIxUDi/4sPDNP7kZ/iPHAWg/MN3U3bbLSg1GiI9vXT85VGMlVWkQiHc69YRauvIZH1N+v2M7N2H2mwChQKdc/SnP3b3+PAMhABYsryeNU8eIBZNolIpWXn7FPze6GmDumAgyponD9JyfBCARUtrWby8Do1WnO4F4XTUGi0LrlmFQqmkv7uNaQuXUVRxaobKk2n1RgrLqjN1uk5mdRYycfYVNOzdDkD5uMnYT1N4WRAuNiXV41lwze0c3fkKxVX1zLnyBuwFRVBU9r72pzeauPLmjzF1/lXIsoyzqAyjRYwyOdfEVT7HUvEYsf6BrHXJcIiia5YTbDqBnEqhNpuxTppI1zN/IxWOkL/oMlAo6Hj0MQBkoGf1Wio+8iGCTSeIDgyMqqAulZJ4dXNzplduoC9Aw2E3Vyyro8CVPYfIYNSy6Opa1j55EACzRYf0znGapAO7RDy7JpBw/oU6uzIBHUDXk0+Tv2ghKpOJmGeYgiuX0PP31ciSROmtN5OKxgg2NmW2lyWJw1/7JqlojHH/8iCOeXNGZY/0W5TK9LAUu8PAgDtALJoOcFMpicN7e1i8vJ5YLIlOl30aH+wPZgI6gNdeaWHqrFIKi60XrvGCcInJKyhm2Z2fIhGLYjCdfW5qKpWks+kIr774JEqVisU3fpiymomZWnRv0ekNLLz2dibMugxJSuEoLMVoPvXvMB6L4nF3Ew0HsOcXk1cgejCE3DJZbMxespLJcxaj0RvOmuDnXe/TasckEgGdVyKoyzFtXh6lt95M9zPpCdoqkxFTZSUD27ZT/qG7QJLQF7noePSvhFpaAHCvW8+Er/9b9o7eClyUSvSFZx82csmRZSLh7CEr0ciZh7AUldq4+oaJJOMpUimZPIcRq12P35sOCidMLcLmEBPVL7STAzClWo2MglBzC/ERL8mAn2QgQMLnp/V/fsfk734b7559JINBDGVlIMvEBocAaPrpz5jxsx9jKB2980yLyuxU1Tnx9IdQa7JvFtVqFVtebmTJNfVU12UPg1GddGOpVCoyAaIgCGemVmsyxcDPxjvo5qXHfpEZWvniX37OXZ//zmmDMaPZetpA7p2aD+9m87N/BEBvNHPrA1/D6Rq95zbh0qBQKDCYRfKtS4kI6nJMqVZTfMP1WCdOIBkOY6yswFhRjmPObBp/9GPinmEq7/1oJqDT2KxYJ09GqdfjvGwhntffAMC5cAEAU77/nVE370ilVnHF1XV0NHuQJBmtTsXk6SVn3N5VnE6E4h0OYzBo2LG9lfmLq1EoFKSSEhabHp1OzDG60EyVFRQuu5qBjZtQqNXUfOZTxPr6aHj4hyBJKPV6Ku/5MG2/+yNyKgUKBdN/+iMSfj9IEoe//u+ZfUnxOKloLIdHc/7Z8wzccc9sRobDyLJMa+MgnsEQJrOWukmFrHvuCD2d3lOCuoIiM3MXVbLnjU5UKiUrbpwohhoLwjkUDQez5solE3FikdD72lco4GXnhr9n7Xugu00EdYIgvGciqLsIaO02tLNnZa2LjwyTv2gRKpMRrcOBvriI0ttuAUBKJPEeOEjZXXfgWn41UiKJvqgIU+XoLThcUePgk1+6nMH+INFIgl2vtrFgSQ1FZ8j8p1BAwBclFk0Qi6b4x9qGzGtXXjv+QjVbeAeN1Ur1J+6j+IbrUKo16AryOfb9/4A3b46kaJRQazv6IhcolOiLXOicTvQuFzGPB0N5GeH2DgDy5sxG7xplPdKnYTBpMZi0ANzz4AJam4YYdAfY/NJxkMFZYKLhUB/5LnNmKLLBqGX+FdWUVtiJx1PkF5rZs6Odsoo8isvsotdOED4gq7MQm6MQ33B66oTTVUbQN0wokI/pLPOEUskEsgxqzdsPFVUqDXqTmVDAm1mn0enPX+MF4TwaGXQz0NOOWq2hsKwKi/3M89+lVIpUMolGp7uALRzdRFB3kVLp9PSufT79b4OB+q98meb/90sSvnTB8vK77yTa10fLL39NMhhEqdMx+Xvfxjp+dAYsSqWCVEriucf3Z9Z1dYxw/+cvw2TOvgCGw3HWPnmQjtZhUMANt0+lu32EcCiOI9/IxKlivkKuqI1GzNXpVMhSMonWln0DpHU6KLv7Doxl5Wgdbxck1TmdTPzGVwm1taNQqTDVVKM2my9o23PNajNQWePAPxKhuMzG1Fml7H2jg9amIfQGDff+y0JcxVZikQTr1x7jxLH0DafRrGXuZVX86Zevc//nFlFSIeY0CMI7JZNJpFTiXdePs9gcXH/vF2k/fpBkPIYkS6x/4tesuPsz1E2bd9r39He1smvTGhLxGPOX3UJJVT0KpTKdQOKWe1n3+C8JB3zUTZ2Hq2z0zIkXxg7/yBBr/vifBL3DAFTWT2PF3Q+ge7PURyIRB1lGo9UxMuTm4Pb1uLtamXH5CqomzkRvMJ5t98K7IIK6i5SptoaSW26i7/kXUZnNxIc8mYAO0vPqKj764UxWTCkWo+fZ1Vi++r9HbfKIgP/t4XauEivT55ax57UOXCU2KqodGM3pHg3/SCQd0AHI8I/nG7j7/rmoNUqsdgMWq3gKmgtSIkGwpZWoux99kQtzbQ1lt99GoLGJ2OAgxppqtA4HLb/8H1AomPK9h7BOnJB5v97lQu8ae7Xq3inPaWLxinpmB2P88eevMeIJA+k5pr2dXlzFVnzeSCagg3RNR4USpJRMX49PBHWC8A5DfV3s2rwGv2eQuVffREXdFDTaf95zYDTbOL7/dbyDfUipdOItj7vrtEGdf2SQ5//8X0TD6ev1mj/+hLs+9xDOovQQy+LKcdz1uYdIxGMYLTa0oqdOuAR5B92ZgA6go+kQAa8HncFEX0czOzb8HVlKMW/ZLTQd2sWxXVsA2PjM77nx41+msn5qjlo+eoig7iKltdmovOcjFF17DQqlAs/O3Vmvqy1m0nkv36Y4B9mJLmYOpxGNVkUikWLG3HLWr3k7k+K1t05m3uXpHiC9UYPBpCESSidTicWSaLVqistF+txc8h09xrGHvpdO6qNQMOnb/07ezOlM/dHDJH1+/Mcbaf3t7zPDMXvXvoBlgihOejoqlfKU2nNvLev0GoxmLeFgPGt7AKtdJAgShLdEQgH+8dRv8PT3ALDusV+y6jPfPGvR8bfojSYmz1nM9hf+mllXPm7K6T8nGMgEdABSKkk46MVJOqgLB/20NuynYfc2KuqnMmnOYqyO91YHTBBy7eTSHlqdAY3OgNczwNo//YRELJ2s7vk//5R5y27N2tY31A8iqPvAlP98EyFXlGo1huIilDodCa8Xx/z0E0CNzUbFRz6EzuVC60wPUVNbLJTedsuo7aUDKCy2cu9nF7L8xon4RsJZr+15rYNoOH0Ta88zcvf9cymtsOMsMPHhT85DpVYw6A6QTIhSBrniXrf+7SytskzfS+sA0DkcmKqriLrdmYAO0g8uREB3eiqVghvvmoYtzwAKmD6nDLvDgGcwhNWu56775uAqtWK161l+00Q624ZZcfMkSivEgw1BeEssEsoEdG8J+Ufe9fvrps/nuo9+jvnLb+W2B76Gq/z0wybNNgf2dxRa1htNWPMKMss9bcfZuvovDPS0s+eV5zmy65VTipYLwsXO4Spl2R2fxGi2YnMWsvKez2FzFBAJ+TMBHUAqmcRgfEdWTYWCgtLRleAvV0Z3184oobXbMY+rRecqxDyuFlmSiLr7Kbz6Kqb96D+IDXnQ5tnHxNC0knI7JeV2ju7vBdoy6yuq89C8o15XeZWDjz4wn0QyxbGDfTz5h93Isszi5XUsWFKDTi+yX15oxqoKhnfszCyfnKXVdfVVDO/cTbSvD11hAcUrr73QTbyoxGNJQsEYBqM20wsXCcdpPNrP7lfbqah2cPcn5oIMDYf6+N1/vYpGq2LVPbOon+Ti3s8sICXJaLVqps0qw2QRk9EF4Z1MFjs1k2fRenQfACq1JhN8JRMJRgb7iEVC2PJdWGyOU95vNFupnTyb2smzz/45VjvX3/MF2o4fIJVIUD1pJjbn24me3B3NWdt3NB5i9pU3oNOLnnXh3JMkCZ+nn1gkjDWvAKPl3NQwVWu0TJi1iPJxk1GqVJl6jxabM5NYSK3RMmvJ9RjMFm6490sM9XVSXFlHYWnVOWnDWCeCuouA/Gb9LYVSgS7/9EMudC4Xjf/506yeDNvUKVjqxp3xPaNZRa2DpSsnsG9HB5W1ThZeVZsZYvYWrU7NsCfEiCeM3qghHIyzdcMJxk0spLQiL0ctH7sKlywh1NLGyN595M2eSeFVS7JeN1ZUMPU/vkd8eASN3Y7OeepNlJRMkvB6UZvNqPSjd97J8FCIQ3u7UKtUaPVqKmuduIqt9HSM8MIzh9Lz47p9RCMJps8rY9s/TgCQiKdY+9RBHvhfV2QNtdRoR28PviC8XxqdnstX3k1l/VSCvhGqJkzHWVQGQNuxfax/6jcgy9jzi7jh3i9hz3//D07zCovJKyw+7WsV9VM5+No/MssTZ18uAjrhvOk6cZSXHvs5qWSSgtJqrv3wZ7A5zl026ZMLjJtteVx/7xfpbDqM1VHI1tV/IRTwolSquP7eL1JaM+EMexLeKxHU5VgqFmNw6zba//hnFGo1tZ/9DM75804ZRqmQ5ayADkBOJt/+tyyDLKNQjo0RtRarnkVLa5k1vxytXo1anf19xaIJ9u/sYvO64+h0apasqKe92UPTsX5SKfkMexXOJ63TQdntt1Gw5AqUej3K06Qx1ubloc07fcAd6ujEd+gwiUAAfWEBlokTMI7S4uPdHSN4BsNv9kjDpOnFLF5Rx9BAiCuW1ZFKSry2uZnWE0NMn1uW9d5EPIUkid+4IJyNd9DN8f2vMzzQy/RFy5kw63JUb85LD/m9bH/xicxwce9QOk37Bwnqzqa4so4b7/tftB3bT1FVHeXjJp2XzxGESCjA1rWPknrz/nGwp43+rrZzGtSdjqOwBEdhCYde35Qp3yFJKV5/+RmKKmrR6UXmy3NBBHU5FunuoeVX/5NZbvrp/2PGz36CsSz7ZlVfXETh1VcxsOkVAPLmz8VQmi7AHWxto3fN86TiccpuuQnL+PoLdwA5pFAoMJpPP6RswB1gw9pjQLow7OaXjnPVteOpHZ9PnlM8Ac2FYNMJjnzr25nl8rvvpPzuO886by46MIjn9TeQUil8Bw7iO3QYSM8rrbzvHvQuF8pRmCBIlmWaGwZYvKIOpUKBUqmgq32YDWuPodOrqZ9UyOyFlegNGuxOA85CE56BdPHjq1dOwCYSogjCGcWiEbas+QvdLen6pW3H9nH7g9/KzIlTKJWoVNnnFeV5nK+u1empHD+NyvHTzttnCAKkry0nz9eU5Qs3f1N10vVao9aKufPn0Oi7G7rEpCLZCT/kRIJUNHrKdmqTiaqP30vh0qtAljGUl6OxWokNDnLsoe9lyh149+5j+k//85SgcKyJRZOnLEejCbasbuLjn78Mi1Xc9F5o/objWcsDm1+heOW1aGynT94hp1L0Pv8CfWtfoPK+j/1/9u47PK7qTPz4d3qv0mjUe5fcbdwxGOxgE0wxEMcBQgsbAmzYZEkgZUnIJsCyIT82CZAsofdeTDc2GNxt3GVJVu9lpFGZXn9/jBlbGJCzYI0tnc/z8KB7597RubI05773nPO+8YAOIDg4SGhwiEggMC6DOqNJw6KlxXy89hBeTxCpTMKFq6cxZVYmJrOGrvZBcguTySmwYrbouOy6OTh6XKjUcmypBiSiwLggfCmfezge0EHsRnewvxd7Vj6hYIABRxdzlq5k/SuPEgoGSM8rEWt+hHFBqzey8NurefvpvxKNRDDbUr+wLqLXPUxncx3O3o7Da97ykCu+fi6CjPxSklOzcHS1olRpmLf80uOuDymMbvzdDZ1i1Gnp6PLycDfGkn5YZs5Akzpyioe3vYP+7TsIuVwkzZmNvrAg/lrAOTCifl3E7yfgcEz4oM5gVGO2ahnojwXNpZNSaW2MZTUb6PeSlZvAxk1QurzcEdvmqVOQ6XRfeCxAcNhF36bNAIQGB5EbDISGh+OvqzPSkWvH55SN9EwTDTW9eD2xshyRcJQ9O9rIyDLz0Xu1ANRW9XDx5dOxJusxWTQYzWo6WgfYt7MNuUJGbkESZuv4/PkIwteh1hnIyC+jvSEW2EkkEkzWWDbKzqZDvPbwf2MwJzHt9GXY0rOxpefgdQ0TCgYwJaUcM4onCKeSnJLJfOem3+L3uDElp6A3HrvkoW7vNj56/cnYhkTChT/4ORl5JV/7e5uT7ay45qcMO/tQafWYk07stM+JRnwyJZgqyUrpbT/D3dSERCpDV5BHyOMl4BxAZUsmEghQ+6f7cB2KZcfqXPMmk++5G4lMylB1DdFwmJwrL6fliaeJhsPItFpUKbZRvuv45vUG6e4c5KxzS/D7woTDYbQ6JS89sQuJBKzJ4kY3EQylpRTe9CM633oHY2UF6cvPQSqXEwkGGa6upWf9h2izMkmaOwd1qh2FQU/SnDl0rnmT7rXryPn+5XS++RZht4esVZeM62nGKo0CjV45Yp89zUB9be+IfdX7uyifGpuG3dM1REtDPy2NTpJTdPh9IabNzkKpFB/zgnA0lVrDGRdcQfWnG3H2djBp7tkkp2UDUL9/BwDDA31s/+A1Zp11Pgd3bqSx6lOkMhlnX3wtRVNmiyljwilLJpeTnJr5pa/7vR72bll3ZEc0Skvt/m8kqAPQ6k1o9aK8zokgevuTgNqegtoee1oxsHcfNf/1R0LDw9jOXETat78dD+gAwl4fwaEhGh78O56W1tj5GekU/vhGhqtrsJ99Fpr09IRcx8mis3WAV5/ZE6/NfvrSIuQKGZNnZDB5VhapGeLDJBEUBj22RaejzcsjNDQEh5P6uBsa2f/r2+NJCYbr6ym66QZkKhXpK76NwmxkaH8VIbcb62mzCDoHUKemovySaZvjRXG5nd3bWnF0u9DoFJRPTUOjVdLWdKSOVn7JkQc4jm43778RG3moAWbOy8Ht8qO0io95Qfg8iy2Vud9aecx+W2YubDuybTBZ2f7BawBEwmHWv/oYqTmFGC0TL+u0MDHIlUrScgpx9nTE9yXZJ/bsr1OF6O1PIoHBQer+/Nf4FLPe9R9hmTEDZXISAUdf7CCplLDXGw/oAHztHahsNlIWnZ6IZp906mt64wEdQPXeLpZfPIn6ml7WPL+HxcvLKKm0o1CINO9jrX/7Tmru+SNEIiiTrJT9+pd4OzuPFCUH+jdtIXjFZchSYg87kk9fyMCuPbQ+9QyRQKzAvGXmV9eFGg+SU/Rccf0cBgd8aHVKLFYtOr0KmUzCwX1d5Bfb6GgZQCaTUFyRSlf74Ijz62t6mb+4MEGtF4RTU07xJGac8W1qdm0iu6gSpWbkFPFoJBLLNn0cBhzdNFTtwj3UT9Hk2bFELNEokUjkmIQRgnCykMnkTD99GaFgkI7GGipOWyTKDpwiTrpPlbvvvpv169ejUCjIzs7mzjvvxGj8ZgojnuyiwSAhl3vEPl93NzmXfQ/Hxk2EhodJXriASCiERKEgGoytt5HI5Sj0+kQ0+aTT3uJEpRr5a52WaWLfp61U7e4E4OUnP+WaHy8gI9v8RW8hnCCBwUGan3gyXpoj0NfP0P4D6AtGLtI2TZmMwmAk7PcT8ftRJSVhKCli6EAsm6nSZkObnTXm7U8EvUGN3nCkHp/RpKFyWjptLQNs+6SRWfNz6ekcxjXkIy1z5MhlUXkKBtP4reUnCCeC3mRlzpILmTJvCSqNFo9rEHtWAd2t9QDMW3YpRnPSqO/j93pY/8pj8XV7B7Z9yIU/uJWDOz+hv6edaQuXkVVYjlyhHOWdBGHsmZNTWbzyagI+Dxqd4bimGwcDfoYG+gj6fKg0GoyWZGTyr59cRTh+J11QN3/+fH76058il8u55557+Nvf/sYtt9yS6GaNCWVSEtmXfZfGv/8jvk04jN/hIOz2INfpaHr0cZQWMyW3/ITmJ56GSIS8a6+KlzeYyIKBEGvXVJNk0zF9TjY1B7rJyrEwa0EuD/2/T0Yc63EHEtTKiSsSDCH93A1MlCi6/DzKfnUb7a+9gS4vl5QzFuFqbGRg7z56131I5sUXkrp8GeapUwn7fGizs9CkpSbmIk4CXm+IA7s6mDE3h6rdHfR2uwConJbOilVT2LW1hbyiZKbMzEIqsmAKwj8lerjeq9YQe5hsMCex/PKbGHB0oVCqsdhSj6serN/rxtHZEt/OLp7Mxreep6OpBoCOxlpW/ssvSMstOjEXIghfQ8Dvo72hmv3bPiQ9t4iiSbMxWr98yvGAo5tP3noWR0cLc8+5mLb6g6jcQ6YrAAAgAElEQVTUWspmLhRTN8fQSRfULViwIP711KlTeeeddxLYmrElkUiwLz4TfX4+IZcLpS2Zmnv+RMoZpzNUVRU/zt/TizI5iUl3/g6QoDCIUTqASDRKwB/k0y0t2Ox6yiankpVjwZqkJb84mYZaBwA6vZIk25dnXRRODFWSlazvXEzdn+8n7PWiy8tDX1iITK3GOmsm5unTcNXVc+A/fkvI5cI0qZKk2bOo/+uDlP/m11imTU30JZwUjCY1FVPTMRhV8YAOYP+uDmbMzeaK6+cik41+0ykIwhHhUJCOplp2f/I+FpuditPOwGJLA0BnMKEzjL6G1z00QFtDNUG/j47mWibPPQuvZ5j9W9aTnJrJtnWvjzh+yOmIB3We4SFaDu2jta6K/PLpZBaUodKIpF5CYvS0NvDm4/ehNZiwZ+bRcmgflpR0QIIpyXZMxsyqHRtoOrib2Usu5MNXHiMY8APQULWL875/Mx3NtXQ0HqJo8izS80pQqsQskhPhpAvqjvbSSy+xbNmyRDdjTMk0GoxlR+Yul/3iZ3jbO7HOm0v/4fTutjMXobanotCLwORoKpWCxctKefof2+ntduFy+WMFmrVKvn3xZNqanQSC4Viglyx+dmNNIpGgLyuj9NZbCAwMoE5LG1HmIOz2UPc/fyHkigUqg/v2Y5oyGSQS3I1NqJKSJsy0y6+iVMk5Y1kJ/Yfr0n1Wk1EqlaDWKkVAJwj/B47OVl57+I8QjdJcA12tjZz3/Zv/qcCqbt92otEIn7z5bHzf5HlnUzx1DpmFFTi622jYvxOIFTg321KPOncbG954CoCaXZtYftlN5FdM/4auThCOFQ6HGOzrIRQMYLTaUB+1frS3swWJRMKsxSvY+NZzhIIB1Fo9Z118DetefoQll/wAjc4AxNaZ9rQ3xb6ORuMBXUyU2j1b2H74gUbNro2suOqnZBdXjtVlTigJCequvPJKHA7HMftvvvlmzj77bAAeeOABZDIZK1asGPX9/H4/Bw8eHPW4U5XEqEd/4QrMi89AIpEQNBqoa20Z9byJSC5XsPoH0/B6QugMcgaGOukfaAdApgGtVoLD6cbhHOWNhBNGolYhz8ygLxiE+vr4fpNcTtjjHXGsTKMm+/LvoTSbaX7uBdQXnY8vIKbOyjDy5otVnHVuGXu2t6LVKZl3Zh5Drh76D7YnunlCgo33PvFECLsdIxI2dTUforuzDbc/fFznazRqWuuqMFhGrrdrrNrFnG+tJBgMUjx1LkZzMp7hQVJzCgkEQxw8eBCtRk31rk0jzqs/sBM0Zvx+P4LwTZPL5fgHOtnw2uNEwmFyS6dSuWA53kDsIaHekkJabjGNVbsIBWN9rs/jouXQfvxuFx0tjQSIrZeTSCSUzVxEW10VUqkMmVxOOBR7n+yiSRzau23E924+tB+/REXo8DHCNychQd2jjz76la+//PLLfPjhhzz66KPHtThTpVJRVlb2DbVOEIREkX//cg796T4AlFYLYY+Xlqefxb50CQqNmqz0dOQaTYJbmXh11T0MD/p597UDnLuyks62ITZ/1MhpC/LIK7Iil4+vzK4iQPnniD7xn9fT1ohEKiV6OJFTZkEZ9vQsVOrj/7wZ7ppM0D/ywVRabhG7P36X/p4OTjv7Amp2bUKtM1C7dytnX3wtZdPnAdBbPp2etsb4edlFleTnj0wiJQjflCGng2eeupdIOPbQoql6N5PmLKasbBIQW1On0WrZvWHkEiiZTI7X68ZssZJ0VK27YMDPRdfdyvCQk+WX3cT+rR+i0miZNGcxcoWC3Z90xY/NyCsmv0isJf06vqxPPOmmX27YsIGHHnqIJ598Eo24eROECSVp7mw0GXfh7+nF3dhE2wsvQTRK97vvUfrLW0VAd1hSih6TVYM9zcieHe0018dKntRW9XD1TfPJzLGM8g6CIBwtOS2bC679GQe2fYQ52U7RlNn/VEAHkF8xna7mOuafu4r6/TuwZeSiVKqp3b0FiE259LqH8bpjZYvUuiPr4Uumz0MildBYtYvS6fPJKhLT04QTJxqJxB9gxPdFj2wrVWoy80uRyWR0ttTh97rRm6wkp2eTZM/EnGwfca5CqSL9qOLkWUWVSCQSJBIJk+eejUKlpvXQASpOO4P03OITe3ET2EkX1P3ud78jEAhw1VVXATBlyhTuuOOOBLdKEISxIFOpiAZDeDu7YgGd8IUsVi2XXTeb4UEfjz+w5cgLUXD2eURQJwj/JKlMRkZeCRlH3Zj+s3QGEwWVsRqaxVPm8O6zD9DREMt2qVCqSc8pIiUjF6eji6kLvkVKRm78XIPJyoxF5zJ1/rdEDTvhhDNYkpm37BI2vB5bx2nPyhsx8vaZtJwiLr3xdtyDTlRaLXKFCqMledRZdNKjMsQarcnMPvsCZp75bWQy8bt9Ip10P933338/0U1IiOCwi7DHg9JqQaoQdT2EiScaDiORyeh4403CPh/JCxfg+DhWisJy2kykKlHP6WhJNj16o4qCUhv11b0AqLUKkkVmV0FIOJ3BxKLzLqNm92akUhkFFdOxZeSy4ppbCPp96IzmETe+nxEBnTAWpFIpZTMWkppdSNDvw2xLRWf44tq9JqsNk9X2tb+nCOhOPPETPgm4m5o49D9/JeL3k3nJxSjMZpRWM8M1tQSdA5inT8NQWJDoZp70Bvo9NB5yMDzko7A0BZtdT1+vm0AgTJJNh06vSnQThS8QCQYZqqnF19lJyOXCMmsGPR+sJ6JQkH3ZapRmM5FwGE2GqHXj7HPTVNeH3x8irzAZe7qRpSvK2ZfRhjVZh9sVoLG+D5lCRkqqIdHNFYQJx+9109F0iJba/WTklZBfPp36AzvpaW9GpdVjtCSjFqUKhJOAQqkaMVosnPpEUJdgYZ+fpseexN/rIPOiCzh035+xTJtKNBJhYNduANpeeInJ99yFLjcnwa09eQUCIdauOUjVnk4APllbx8Xfn8GzD2+HKOQWJnHBd6diNIs1WYkS9nrxdnSCBDRpacg0GsJ+P666BsIeD+0vv4qvI/bvl3v1lbQ+8yzejk6Kfnwj2qxMZOqJV9cmHIrg7PPEirTrlLzx/F6a6mLr59QaBVf/63yGBnxIJBI+XluHs88DwJaPGrj6pvmYreLmURDGUltDNW8/+RcA9m35gNPOvoCqbR/h93konbGAMy64ArlczMYRBOGbJ4K6BAp7vbhb2/B1dJI8by4da96ESAR9cRGtzzwXPy4SCODt6BBB3VdwDfmp2tsZ3w6FInS1D8LhDNVNdX30dA2LoC5BQh4Prc+/SMcrrwGQcfFFpJ27jI5X36Dj9TeQKhRkXnoxPevW4+vopG/LVirv+j2+tna63nmPlMVnYiwtRiIbX1kdv0owEGLnlhbWvnGQKHDG0mIi4SMp133eIB6Xn0+3tGBPM8QDOoj9PQz0e0RQJwhj7NCekenb2+qqsGXk0lZfRe3uzcw+64Jjyh4IgiB8E0SV2gTq27KV6t/fhe2M04EoEknsn8Pf6zhmqpnSIhIffBW1RnHMdDOFYuIEACc7X0dHPKADaH/xZTyNTXS89jpEo0QCAdqefxHbwgUAGAoL6HjtDWruuZeetR+w/1f/gbupOVHNT4g+h4f3Xq8iEokSjURZ/04NJZUjM44pVXLSs0xEIlGUqiPP6GQyqZhuLAgJkFVYPmLblpGDszf2wDE1uxClmHopCMIJIoK6BAkOD9O9/kOSF84nGo2iyc4i+/LVSORyHBs+JmPlhVhmzUSbk0PJz/8dXV7ul75XNBolMsGKOHo9AYaHfEQjsZELrU7JhaunUVKZSkqqgZWXT8ds1SKRxjI0FZamkJJmTGSTJ7ToUUV9P/P539lIMIhEpcQ6+zSss2fRu+7Do16M4OvsYiJxDfniI82fsaUayMlPItmu59IrZ5KUoqd8Shr+QIjlKyvJyrWQkW3me9edRnKKnr5eFzs2NbHpw3q6O4YScyGCMIHklk5h3jmXkJyWxayzzienZApavZHS6fM544Ir/ukyCYIgCMdLTL9MkEgwhKGoiO5330dtTyHlrMXIVCqm/r8/EnK7UdvtsYAvFEKu/fIne572drrefhdXfQMZ55+HaeoU5ON87VFbs5OO1gGq9nSSZNMx+/R8UlIN2NONrLx8GqFQBLVaQSgU5rqfLCQYCGNN1qHVieyJiaJJTyd12Tl0vR0rZJr27eVoc3IwlJYwXB1L+Z2+4tvINFrCHg9d77yPvrAAV1197A0kElQpXz/71qliaNDH/l0dFJenUFvVA0BxeQrpmWa+e+0sIuEIam3s99lk1lBUZmfL+npKKu0Uldux2Q14XH5efXo37S0DAGzU1nHNjxdgTRbZMQXhRNEZzUxftJzKOYtRKFVIJBLSsguRymUi+58gCCeU+IRJEE9zM+0vvgyAa3iYSChM6W23oElNHXmg8ssDkbDPR+PDjzGwYycA1VUHqfz9HZgqK05YuxNt0OmlrdnJe69VAdDS0E+/w8Oqq2eiUiuQy2XI5bFpl3K5DLsYnTspyHU6si9bTcriMwDQZKQj1+ko+dm/421rQ6JQgAT23/qr2AlSKXnXXkX6+SsIOJ1oszJRZ06c7JfhUJgDuzoom5LG4mWxulnpORa0+mM/D3q6hnnq71uJRqLU1fSyc0sLV90wj+EhfzygA/B6gvQ73CKoE4QxoFQdebiqUImp0IIgnHgiqEuQ0NDwiG1fRwfRcGTEPr/DERuhSPriRdWh4WGG9u4beU5P7zfb0JOMx+0n4Bs5ba+5oQ+vJ4hKrSAQCONx+dHoFKhUIsPYyUSh16EoLgIgODiEr6cHpdWKKmkyAJ7WNkzTpmKeNhWl0YhUqWDwwMHY6F4kQukvfk7S7NMSeQljxmTWcOH3puLodhEKRWhrcZJblExDbS8Gkxqb/cj60aEBb3waMoDT4cHtCqDVKdDqlXhcAQAkEtAbxM2lIAiCIIxHIqhLEE1mOlKViojfD4Bt0ekEBpxoM9IJBwI4Pt5I40MPI5FJKfjRv5A0e/Yxmf/kJhNJ8+fSu/6j2A6JBE1G+lhfypgKBiMYTGokEvhsmVZpZSpavYr+PjcfvFlN9b4u8oqSWHZBJUkp+sQ2WDjG8KE6Dv3pf/B1d5O+4lzSz1+B0mxGqlahsiUjlUqp++sDRINBtDk5ZF50AW0vvkzrs89jmjQJuXb8r0np7/Ow7q1q+h0eMnPNTJmRxdP/uw25XMq8xQVAlKRkPS5XAKNZg1whJRSMPRRKzTBiMKnR6pR89+pZfPBWNV5vkLOWlY4IBgVBEARBGD9GDeocDgf33nsvPT09PPTQQ9TV1bFr1y4uueSSsWjfuKUwW8i+bDVhlwupSoWnpRV/dw9UVOBta6fuz3+NRy21f7yPSX/4HXKDAbU9JR7cyZRKsr/7HfQFBXhaWkg+fSG6/LxEXtYJEw7H6nUF/CECgRArL59ObVUPFquG0kmpKJUyqvd2cfBwnbqGGgc7NzezZEU5Eokkwa0XPhMcHubQn/+Kt70dgPaXX8NYWYF1xgxcNbUQjtC55i2iwSAQm6acPH8uSKUorVbczc2orBbUdvtXfZtTXn1ND/2OWImC4vJU3nzpyIj8ujerSU0/jd3bqtm9vZWCEhvfveY0qvd3oderKJuSFl8/mpFj4bvXzCISjqLSiJHrU5HogwVBEITjMWr2y1tvvZUFCxbQ0xNbrJ+bm8vjjz9+whs23inNJsJuN20vvkzz40/St2kzutxcAMIez5FhKCAaCjGwZy+7f/wTejd8QjQcjr+mtttJP+9cCm+4HvOkSqSK8XfjFgqG2bmpmQf/+yOe/cd2hgZ9bN3QSHf7IBKZJD760Ns9ckprV8cQkcixWReFxIn4/QQ+N0U47HIDEBgcQiKVEo2MnIYcjUZRp9qxzJjG/l/8mr0//wWe1tYxa3MiHP17G/nczyMUiuD1Btn8YQNed5D9n3awfWMTpZNSWbikiOTPjU4rlHIR0J3CRB8sCIIgHI9Rgzqn08ny5cuRSmOHyuXy+NfC/51EJiN1+TmU/uLnFN50A5Pu+n28bIEmPR1dQX78WGN5Gd72DiKBAHV/fQBvV3diGp0gjl4377x2gEg4SiQSZfP6BuYtLuDMZaVMn52NVBb7fZwyK5OjB+XmLMpHJhO/qycTiUpN6vJl8W2ZVosmOwsAY1kpA/v2kb16FRz+jFGnp2GeNpWkeXNpevQJiEQIOgcYOpwxc7wqLEnBYIwlWvD5QiMCtdwCKz7vyHWlfb1u9n/azvCQb0zbKZx4og8WBEEQjseo0y+1Wi1OpzM+hW337t0YDGJdxjdBaTJhnTljxD53czP927aTeenFEAoRjUQZrqmhc81bQGzULhqeWDXpwqHwMfW6dDoVlmQtTXV9tDU7KSixkZFt5qqb5tPvcGO2aEjLNCWmwcKX8rW3421vI+eKy4iGQkjkciSH03zr8/Mo+9VttDz9bCywi0QIDg/jdzhw7vg0vv4UQPoVWWHHA1uqgav/dR7Ofg86vYqZc3PobBtEIoH0TBNDg74R60rLJ6dRW9WNVCqmGo83og8WBEEQjseoQd2tt97K9ddfT0tLC6tWrcLpdHLfffeNRdsmHF93Dwf+4w6CA7E05Cq7nYrbf0nnm2/Fj0k7dxnqlPG9nujzrMk6yqekUXV4vVxxhR2rTUfVng7eemk/AFs3NLL62tMoLEshM8eSyOYKXyUapX/LNvq3bIttSyRYZkyLvyzX6hjaX0X/pi1H9ul0pJ6zhOYnniLs9mAoL8NYWjLWLR9zJosWk+VIjcqjSxEYTBou/+Fcmuv70OgUOHpcnHVuKTq9yG453og+WBAEQTgeowZ1FRUVPPnkkzQ2NhKNRsnLy0MxDtdtnQx8PT3xgA7A391NcNhFyc/+HU9rK1KFAm12FjL1xLpx02iVLLuwkhlzcwCwpepRKGTs2jpyXVXNgW4Ky1IS0UThOGkyM0hZfAY96z4EIPOSlahS0+KvK60WMldeSNMjjwEg06iRqdW0PvM8qed8C5lKRdL8ueM+UcpoZHIpOQVWZHIJ2z5uQq1RoNOrcA37qTvYQ111D6WTUikosaHRju9RzfFO9MGCIAjC8Rg1qHvqqac477zzKCqK1ZcaHBxkzZo1fO973zvhjZtolGYzUqWSSCBWV0qm1aIwmVAlWVElWRPcusTSGVSYAyHqa3ppqu+jbFIqBaU2OtsG48dk5ooRupOdwmAg95qrSF32LSRSGZqM9BEPKSQSCSlnLUaXl0twaAilxUr13fcQGhqi/aVX0ORkk7rsW4m7gJOIo9vFEw9sIRSKJVKpq+7hnAsreP25PQBU7elk5eXTqZg6vsucjHeiDxYEQRCOx6irrZ9//nmMRmN822Qy8cILL5zQRk1Umox0yn/za4yTKjFNnUL57b9Ck5aa6GadFPz+IO++doC3XtrPx+8f4uH/2UhpZSrzFxeQkmpgyYpyCoqTv/DcgD9EQ62DDe/XcuhgNz5vcIxbLwD4urrpXruOnnUfIlUq0RcWINMcW3NOYdBjnjIZ28IF6IsLKbzxR+gLC0laOJ+Sn9yM4qjPo4lscMAbD+gABp1eBvu9I46p3tc11s0SvmGiDxYEQRCOx6gjdZFIhGg0Gl+kHQ6HCQbFTfGJIDlciyvne6tQJiWjTrGNeD3ocuOur8Pv6EebnYU+P++YguTjlWvIT+2BHvKLk8kvtqFWy/B5g5RPSWf2wjz0hzMFfpGWhn6efmhbfPuiy6ZROS1jLJotHBZyuaj7y/0M7outgZTpdEy55040GV/87xAYGMBVV0/Y40GXl0fF739L2OtFoRfF5D9jsmionJ5Ozf5ugoEwZqsGjXbktLz8ki9+0CGcOkQfnBjRaBRHRwv9PR1oDUZs6bmotbrRTxQEQUiQUYO6BQsWcPPNN7Nq1SoAnn32WRYuXHjCGzYRhAMBwi43CpMRiUzGcE0tB377O8JuDyq7nfJf34bCYsHd0EjY5yPQ30/DA38HYiURJt35nxhKihN8FWNDo1Ewc34OeoOKQ1Xd2FINrHkxFiDMOzOfhWcXoVKPvKF1u/y0NzvZs6NtxP4dm5spm5wmyh2MIb+jLx7QAYTdbrwdnV8Y1EVCIdpeeoXO19cAYF96Niqbje7312IsKyPrO5egyZi4Uwqj0SjN9X28/0YV0Sicu3IS/Q4X+SU29n/awdwz8mlvGaCgxEZeYVKimyt8TaIPToye9iZe/tsfCIdi2aYXrbiMSXPPSnCrBEEQvtyoQd0tt9zCs88+yzPPPAPAvHnzuOSSS054w8Y7T3sHLc88y9De/aRfcB7G8gpannqasNsDxJKk9G/fgUyjoeHB/8W26HRcdXXx86PhMP07P0WXnzcuC44fLRKO0NPtwmLVsn1jE9NnZ7Pu7SN1yjatb6B8cjrp2WbCoQhtLU4aanoxmDT0dg+TnmXm4N4j09Aysy0ioBtjcr0epdVCoN8Z36ewmL/wWL+jj6433wZAIpejstloeSr2+dPb0wtSKYU3Xo9UPurH17jU1+vm6Ye2EQrGpl6+8fxervvpQsxWLT5PkG0bm5h6Wqz236dbWuPlPuSKiTGqP96IPjgxuprr4gEdwKcb3qZg0iy0ejH9WxCEk9Ood0VSqZTVq1ezevXqsWjPhBANh+lcs4a+jzcCIJFIaX3ueaKh8IjjIoEg3WvXAeB3OFCnpuJt74i/LpFIGK6rw1RWNnaNT4CujiGeeHALBcU2TGbN50vWARA5vLejbYDH7t8cr2s3Z1E+eqOK6XOyqa/pJTvPikarwOsJiKyAY0iVnETpbT+n6dEnCA4NkXP5arTZ2ccc525uITAwgDIpCX9PD3KDnpDbPeKY4eoawl4v0glWq8vR7aKrcxCFXBYP6ADC4QgeVwCb3UBxRSo5hVa2bWhi/Tu1AHyyro7v/2guOfli1O5UJPrgxNAZRz50stjSUIzz+piCIJzaRg3qdu7cyV/+8hc6OjoIhULxuf0ffPDBWLRvXIoEggxX1x61HWBw7z7yrrkKV10dkUAAhcmEZdYMHBs3AzB0oIq8a64i7PPhqm8gae4cggMDDO2vigd10XB4XK6xG+j3Eo1Eqa/t5ZwLKuhqH2LSjAzqqntIyzCRkmYgyRZba9XRMjCiUPnBvZ0YTCpaGvvJK0qmq30QrzfIaQvyEnQ1E5ehuIjy//gF0VAI+ResjXM3t7Dv1l9CNELOFZfT+eZbBPr6MFZU0PHaG/FK20nz5hANR445fzxz9Lh49P5NeFwBTluYhzVZR78jFuyarBrMSUfq2YWCUT49utxHFFobnSKoO0WJPjgx0nIKmTJ/Kfu3ric5LYt5yy9FofzytduCIAiJNmpQ98tf/pLbbruNyspKpFIxZe2bINOoSTvv29Td92cApEol0UiEtpdeJvPii5BqNFimT0ObmUHxv91E35ZtDO7Zh7ulBfs5SzE0NjG4dx+uunpKf3krntZWOt96G19XDxkXrMBYXjaupmRakjRIpRIikSjvvV7FvDPyKZuSRka2mardnRiMaoKBEBqNAqvtqIXsEpg8IwOb3UBBiY2tHzeSYtezeFkJCuX4C35PBTL1sTdFwcEh3I2N+B19ZFywgraXXqHp0cdJXrSQtGXn0L91G3nXXo2vpwddVhZDNbVEI+EvePfxy9E9jMcVK3WyY1MTpy8pQqNRIpFCbmEyJvORLKIqtZycfCt7d7bH99nsIsHMqUr0wYmhM1qYt+wSpi5YilKlQaXRfuXxzp4O+no6UKo02NJz0OjE35wgCGNr1KDOYDCwaNGisWjLhGI9bRblt/8Kb3sHhuIiKn/3GzrffpdIMEDygnloMjLwdnXR8cZbOLdtxzx9GtrsLPzdPfh7egj7fORddy36vFwO3vlfuOsbABjYvYfJ/3UnhqLCxF7gN8ieZuTyH85hx6YmLEk6Js/MpL15gHdeOQBAS2M/AHPPKCAz28K5Kyex9ZNGTl9azN4drezY1My02dn88Kenozeq0erEFJpECgwOEvZ4UFqtSKRS2l9/g/YXXwZAYTaTufJCWp5+luGaQ1hmziDgdOLv6WVw/wFcdfVkrrwIufarb7DGm6OTAEXCUXZubuHamxdgOCrrq2vYh0wmRaVWMHV2FuFwhO6OYaaelkl6likRzRa+AaIPThyZTI7BPPoId39PBy///S587mEATjv7AmaeeZ4IwgVBGFOjBnWzZ8/m7rvvZunSpSiPmk9eUVFxQhs23in0OizTp2GZPo2Qz4+ruhpDWQmaVDuyw0/4HBs+pnfd+vjXCpORwb37yLniMvKKilCajHja2uMBHQCRCP7e3nEV1EllUnIKksgpONK5fvRe7Yhjqvd1MXNeLhqdkhnzciiptPPOqweoO9gLwKb19djTjaSkiUXuieSqq6fmv+8l4Byg4IYfIpXL6Xj19fjrwYEBkEgwTqok6zuX4O/pwblzFzKVkvTzz8PT2YFUpaL1+ZcwT52MqaJ8XE45/rzUDBNnnVvKhvcPoTeqOP87U+MBXcAf4uDeTta9XYNGI2fJilgBcr1BRWqGkS0fN5JfbMNgOrYmoHDyE33wya+vqy0e0AHs/HANZdPnY7CIkiKCIIydUYO6PXv2ALB//5F05BKJhMcff/zEtWqCGT5YTdVv7kBhNhMaHibvmqtIO3cZwzWHRhzn6+6h4ne/RapQED2clUtpMWOoKGf4QBUQyxaoTrWP+TWMhWg0iscVQKWWU3g4fftncgqSGBzwkpwSC4gHB310dwyNOL/f4RrT9gojhTxeGh56GF9nFxkrL6Tp4UdJmjsXhclIoK8/fpw6LQ2zRIK7oZGmRx6Lr6XrXruWzIsuovHv/4tUoUSdlorCYkaXlZWoSxozGq2CeWcUUDktA7lCik6vAmJlO5rr+3jt2djn9PAgvPzUp0ydlcXmDxvobBsEwOcVdc1OVaIPPuiv310AACAASURBVPkp1SMfmOiMVmQKMSNEEISxNWpQ98QTT4xFOyY0V30DeddchbezE1VSEp7OTkIeL6nnLMW5Y2f8uNRvLcXf3U3jQw8THBom54rVWGbOpOiG6+n9ZCP+nh7sZ5+FLicngVdzYriGfeze1saOTU1kZJuZNT+XxctKaGsZIDXDiLPPQyh4ZJ1VW5OT8ilpfLw2VgZCIpWQWyCemiZSJBjA3+sAYutIg84Bej/6iJzvrabt5VcIDgxiP/sspBo1LU8+TfbqVfGADiB5/nxq/3QfRGJJUrwPd1D2q9sSci2JIJFKMFlG3jx2tA7i6Bn5sMLrDmK2HjkuKUWH1SbW95yqRB988rOl5zD99GXs3vgeOoOZsy+5VpQ+EARhzI0a1DkcDu699156enp46KGHqKurY9euXaJOzjdIk5FGzV33xLezL1uNTK3COGkSlb//LZ6WVrTZ2ahSbOy95bbYFDWg5u4/Mvm/78ZQVEj2d8b3v0dro5N1b1UDMDTQRXZ+Eo11DnzeEM31fZRUpo7IAGi2aGg85OCsc0vx+0LY7HrSskQnm0hKk4msVZdQ/5cHIBJBplETdntofupp7GctxlBWRvfaD4gcLmEQ9vtRJicTcBwJBD8L6ADCHg/R6PjNgjk04GVwwItOr8KarPvCY1xDPiLhKDqDCvewH4DCEhsSiYTFy0shGsWeYcRoEln7TlWiDz75afVGZi+5iIrZZyBXqNAZxBpWQRDG3qhB3a233spFF13Egw8+CEBubi7/9m//JjqUryk4NMzgvn14WtoIuUc+aXds+Ji0c5ch12oxVVZiqqwEYinfPwvo4u8zODhmbU4kp2NkrbItHzVw2b/MxtnnQSqVYk/Toz6cTCIYCKHVKSmbnIrHFSAYDJOebUapHD8ZQU9VtgXz0aSnE/J4KPvlbbS/+jqRcBh1WhrNTz1D4Q9/gNxkRKZR0/H6GrIuWYlUrUZpNiNVq1BYzASdsb8BbW4O2szMBF/RidHbPcwz/9jGQJ8XpUrO9647jaxc6zHHpWYYef/1g8w7M59wOIrOoCQ734qj20XAH8aSpMWeLh5mnMpEH3xqkMnlmKwpiW6GIAgT2KhBndPpZPny5fz973+PnSCXi4xO34D+rduo+8v9SJVKMlZeOOI1Y0U5IZcb585PifgDGEqK0WZlokpOxjx1CgO7Y2ss5AY9mvS0RDR/zGXlJyGRHJmNN2t+Lkk2fbw+3dEO7uvi1ad3A6A3qrj8X+ZgSdLR1uykvdmJ0awhPcuEyTKxMiieDGQaDaaKcqLhMEM1tZgmTUKZbEWTlkbFrJmobbEpspPu/D1DB6uRqlVoMzPR5uYQ6Ouj8MbrcTc2I9dqMU+bitJsHuU7npoOVfUw0OcFYolQPnynllVXz0IiiW1r9SoikSjhcJRlF1UwOODFlqonO9+KezjAmhf2xdbRSeC7V8/CKJKknLJEHywIgiAcj1GDOq1Wi9PpRCKRALB7924MBsMJb9h4FgkG6X5/bezrQABfVzcZKy+kb+MmTJMqSV2+jLo/34+rrg6VLZmwx0PZr3/J4P4DGMrLMU2dglQuwzR5MgqzmbDPj0ytSvBVnVjpWSauvHE+bU39WJJ0ZOdbvvA417CPD9ZUx7cjkSh9DjceT4AnHthCJBKLCmefnkdRqY2s/CQUivGfPfFk4O/vZ7imltDgEOr0NA7cfkd8OmXy6QspvOlHRw6WSGh9/sXYSHQkQsnP/53keXPRpKVhnTkzQVcwdqKR6IjtSCTC0KCXdW9V09k2xOzT88jMMfPoXzYRiUSRy6XYUg0oVTI624aOJEaJwrp3asjOT0KlHvXjXjgJiT5YEARBOB7HNf3y+uuvp6WlhVWrVuF0OrnvvvvGom3jllShwDJzOsM1sbT8ves/pPiWn5BxwQpkWi2ellbUdhuG0uJ4HTtfTw+Nf38IAIlCgX3JWahSUqj9030oDAZyrrhsXJUx+DyZTEpWroWsXAs9nUNs3dBIJAIV09JJPWp6mUwqRaWRMzwE+cXJZOVZWfdmNdZkHUvOK+f9N6qIRKLs/7SdSDiKUq0gM+eLA0ThmxMJhXBs2kw0ECQaDBIJh0esj3N8spGcy1cjS4lNXxo8UEXQ6Yy/3vzEUxiKipAb9F9YwHy8KSpPYevHjbiG/cjlUhZ9q5j33zhI7YFuAN599QArVk2JP6SIleswMNDvIRgYWZhdo1UglUrG/BqEb4bogwVBEITj8ZVBXSQSwe/38+STT9LY2Eg0GiUvLw+FQqxN+rpsZywiEgzRv3U7KWcuwlhejsIYC04UBj1StZq2518EoG/jJkpu+SlIpRCJEA0GUdvtVN/5X/H5iNV/uIvJ99yNKnn0QqmnskGnl6f+vo3hIR8Au7a2cM2PF2A5nCRFo1Ny7spJPPfIDvKLbaxdcxAAR48Lt8tPcYWd6n1d2OwGnH1uhgd9CbuWiSQwMEDIOUDb4SLjuVdfOeJ1U2UFcv2R0Qe5duR0QbnBQOuLL+Ht6CTvyivQF+Sf8DYnUkqakav/dT4DTg86vQq1RkFbs3PEMRqNAolUwqIlRXR3DtNQ60ClVmBPN5KRbaa9ZQCdQcVZy8tQKMVo9KlI9MGCIAjC8frKoE4qlXLHHXfw6quvUlRUNFZtmhDUKSlkr15FxoUXHHMDK9NoGNi1e8S+oeoa1Ckp+Lq6AJBqYqMVtjPPQJOWSiQQIOTxoGJ8B3VDA954QAfgcQcYdHriQR3EatZd95OFNNQ6sKcbKZucigQJDYd6Scs04feFKC63s+6tg5x1blkiLmPCiYYj9Kz/ML4dGnZReNOP6H5vLeqMdNLOXT7i78BYXhZfP6qwmEk58wy6P1iHdfo0HJu3oEyyjtv1dJ8xW7WYrbHf62gkytIV5ezc3ExroxOJVILBpOaqG+fx6eZmDu7tBGIJhBYvL2HJinKUShkanRKTWaynO1WJPlgQBEE4XqNOv5w7dy7vvvsuS5cujc/pF74ZEonkmIAOQKbVYpkxHW9be3yfoaQY2+kLGNy/H3WKHXV6GpnfuZiBnbvoXf8hMp0W05TJkD2+CzHrDbFRi8/WDCmUMvTGY6fjma1aUjONFJWlsOG9Q0SiUWbOy6FsUipJNh1d7UNc8aN5pKSJtSljQWk0oE5NI9DXj1SpRKqQ0/C/D2OqKMfT2ETvhxvQ5+UikcVGlNR2O8U/+ymBvn7CXi/927ZjKMin7aVXUJjNmCdVjvug7jNeT4ADuzvYuqGRlHQjq66ZhVarJDXdSDAUpqtjaMTxA31eyqcosSaL2nTjgeiDBUEQhOMxalD37LPP8sgjjyCTyVCpVESjUSQSCZ9++ulYtG9CkkilpJ27HKlKxcDuPdiXLsE8dTIKgwFD8ZGnte6GBlyHYsW1w24PjY88xqTf/Qa5fvzezJmTtKy6ZhZbPmogHI6w4KwiklO++HrlMikb19fHk07s2NjM5BmZlE9Jp3xK+lg2e8KTaTTkXXsVh/78VyJ+P4H+fiI+H86dsc+RaCRC2O9Hrj0y4hro7aXuz/fjbmom/7prqT88dTPgcFD35/uZfM9dKC3jP7Brbx7grZf2A9DX6yYcCnPRZTOQyqSoZFJmzstlzQt748fbUg20NDpFUDdOiD5YEARBOB6jBnW7du0ai3YIn6O2p5B56cWkn3cuIZ+fnnUf4u/rI3n+PHT5efi7e4gEgiPOiXi8RCPjtxiz3x9i3842PnirmvyiZBYtKSI924LPF0QulyKXj1w3FIkem0UwHBq/P5+TnT4/j8o7fkPY68Xb0UHXO+/FX7MvXTIioAv7fDQ98jiuunrg2HqMAaeTSCAwNg1PMGe/Z8R2W9MAfm8Q5eF1cgWlNpauKCcSiWC2atm6oZGCUlEva7wQfbAgCIJwPEYN6qLRKK+//jptbW3ccMMNdHZ20tvby+TJk8eifRNOcNiFTKPG39NL+yuvMlR1kOT58/G0tdG3cRODe/eRuuRsGh9+lOzvfgel1UKg3wkSCblXXhFPtjIe9XQOxUcsDu7twmhS4/UG+WRtHRqdkkXfKsaeduT6rUlaTluQy7ZPmgAoKksh2S5GLxJJYdDH/jMaqLjjdgZ270FfUICxsmLEcRG/H29Hx5Ed0SgKk5HgYGyqYcaF56NMOrYY93iUlmlCIpXEH1DMPSOfYCiMs8+NyaLFZNaQkmZgx6Zmqvd3Uz4ljZw8kdF1vBB9sCAIgnA8JNFoNPpVB9x+++1IpVK2bNnC22+/zeDgIFdffTUvvfTSCW3Yww8/zN13383mzZuxWr/65u3gwYOUlZ3aCS+Cwy4cn2yi84016Arzsc6aRe0f/xTPbpl92Wpan32erEsvpvW5F4iGw0iVStLPPw9dQT5qux1tVibScZwV7dDBbp55aDsSCSw5r5yWxn4c3S4qpqbT1uzENeznih/OQaNTxs/xegL0druIhKMk2/XoDeO7nt940vnOezQ88DcA5EYDZb+6jYCjH7lOiy4vD4Vp/D7AOFo4HKGjdYCGWgepaQa83iBvv3KAaCTKkhXllE1K5Zl/bKez7cho5lU3zSMrd/wEvePhM/7/6v/SB0/kn5cgCMJ492Wf8aOO1O3du5dXXnmFCy64AACTyUQwGBzlrK+ns7OTjRs3kp4+cdY9DdfW0vBg7AbW29mJVKnCMmsGzm07AAj098fWykkkRMOxOlSRQIC2F16i4Ef/QthoQiKVJqz9YyHZpsdgUpOWaaJ6Xxctjf0AfPReLWd/u4yN6+vwB0IjgjqNVkl2npVIOEJrk5N3XtmPJUnHlFmZX7oWTzg5JC+YjyY9lUggBJEwoWEX+qJC1Cm2RDdtTLmGfPi8QbJyLej0Sl740yfx+nRvv7wfe5pxREAHMDQgSnWMF4nogwVBEIRTz6hRgFwuJxwOx7Nu9ff3Iz3BwcOdd97JLbfcMqEyffm7YkWF9cVF5F11BRG/D2NZGfalS+L71Wmp6AoLSD13Wfw8Q1kprkN1HPj17bibWxLS9rFiSdZx+fVzmL0wj872kTexJouG0+bn8skHdRw62IPXPXK9VVfnEI8/uIWqPZ1sXFfHG8/tweeZGGuyTlUKvQ5DcTHOHTs4+Pu7OPi7P3DwP+/E73AkumljZmjAy3MP7+CZh7bz5N+2smVDI7mFI8uWRCIRisqOrKGTyiRYbbrY+YM+OtoGGBL1GE9ZieiDBUEQhFPPqCN1l19+OTfccAN9fX386U9/4p133uHmm28+YQ1au3YtKSkplJaWnrDvcTLSFRaATIZt4Xwa//EoAI4Nn5Bx4Qoq/nAHPR+sR67VUnPnf5F92WrKb/8VrvoG/N09dL+/FgBfRyf6/LwEXsWJl2zTY7FomT47i60fNwGQX5xM1Z7OeK2uXVtbWXXVTIrK7fHzBvo8I5KmtDY5cbkCqLVKhLET8ngJDg6gMBqR63R4WttwNzQg02jQFRSg+tw6OV9P74iEKp7mZjytbaiSk8e66Qnh6HGNKFmwZ3sb510ymYbaWGA7bU4WEglMPS2LGfOy8XqC6A1qgoEw9dW9rHlxL4NOLyaLhlXXzBqx5lQ4NYx1HywIgiCcmr40qGttbSUrK4sVK1ZQUVHBli1biEaj3H///RQUFHytb3rllVfi+IKn7TfffDN/+9vfePjhh/+p9/P7/Rw8ePBrtSnRFDIZJb/9D4Z3jkxT7di4Bf3kyfR+sB6pUkkkEKDp0ccp+++7cGzajKehEQCJTEbA6aRnz176VUpGWSp5yiuqNJFsr2DA6aWozM5jf90cfy0aidLTOYxC48frcwOgVI9MNpGRbWLY1U9vX2tC2j8RWSQSep96lqG9+9CXlpB95RXU/uFuQkNDSJVKcq+5Enl2Fn6lkuHD08ssSJCqVER8R0aa/JHIKf/3fryk0ZF1FOUKKdYUNRddPhmZTIbHFWTtmmoKy1Kwpxv55IM6+nrcGM1qlpxXzqDTC8Cg08v2jY2UTdMTmCBZQ091X6cPHg99oiAIgvDP+dKg7sc//jEvv/wy3//+93nssce+diB3tEcfffQL99fU1NDW1sb5558PQFdXFxdddBEvvPACNtuXr6NRqVTjZlG41Ouj45XX4tvm6VNRGY3kXHEZYY8HhclI37ad6JOSKfnJzXS+/Q7+nl4s06bS8foapGo1lf/5WxSGCbBerDD2P783SFauNb7GDsBoUaPTmAgHNJgsaixJOi7/4Ry2fdyIyaIhv9gGYSn5+emoVKMOWAvfgJbnXmBo7z4AXNU19G/4GIXRQGhoiJzvX0bHq2/g6+zEUFpC0Y9vRJOeTjQaRf7jG6n78/1EAgEyL1lJcmkJqTpdgq9mbPh8Qc5cVsJH79WiUMg479LJZOfakEgluIZ9vP/6QezpRvZsb6Og1EZfT+whxtCAj6o9HaRnmeloHQAg6I98o5/jY2kiBihfpw8eT32iIAiCMNKX9YlfejcbiUR48MEHaWpq4pFHHjnm9auuuuqba91hJSUlbN58ZMRl8eLFvPjii6NmvzyVhXw+fO3tRENh1OlpGMpKKfzXG+h66x2MkypJPWcpzu07aX78ydgJUikVv/k1SrMJpdmEefIk2l56hYaHHoZIBLnRSCQ4sZ7EqzQKvnVBOVs3NNLvcDN5ZiYmi5aH7vuEYCCMQinjiuvnkluQhFwu5dWnd7N1Q2yE89IrZ1A6KS3BVzAx+Do6R253daMwmZBptQzu2YevM/b6cHUN/Vu3k3Hh+UgkEpLnzUVfWEA0FEZlSx7XGV4/T61WMO/MAsqnpCGTSTFbj9Tyk8mkpGebePfVKpJSdAT9oRHnBgNhVJpYLTu5QsqsBblj2XTha0pEHywIgiCcur40qLv33ntZu3Yt4XAYt9s9lm2aMMKBAJ1vrKHlyWcASF44n7wfXIP9rMUkL5iPVKkkNDxM55o3j5wUieA6VId5SqxGkTotLXazfLjoeNaqS1GN4yD4y6Rlmjn3ksl4PUGIRnnnlf0EA7EsocFAmB2bmsjInsq2T2KB32e2fdJEcbkdqUwkHjjR7EvPpnfDx7HfVYkE+9mL8bS1425qIugcGHFsYGDktjpl4hbTlsmkJNmOHXnXaJUYjGoA+nrczJqfy8F9XQQDYWQyKZNnZqLWKJhzej5mqxab3XDMewgnL9EHC4IgCP+MLw3qPv74Y6677joCgQA33njjWLYpbt26dQn5vmPF19VNy9PPxbcdH28kddk5KE0mZKpYPTWpWo0uPx/f4eyYAOrU1PjXupxsJt39ezytbSgMBnR54ztRyldRKGQoTDIaanuRK2QjXvtsimV6ppn9nx4pap2ZYxEB3RgxlpYw5Z678HZ2AdD63AuEfX6S580lZdHp/7+9Ow2Msrz3Pv6dfSYzkz2Z7IQEQsKOgICgCIjYAhUVF7oorT621db62NbT6lGrtrU99tjTU7vY5Tn0aGurrXWp2qKouAAigrIvYQ1k3/dJZuZ+XkSjAYLSktwzye/zinuZ8JswzDX/uTZ279kLkQhWp5PUOeeYnDY2ZOclkZ7pp7qihVdf2MfSKyZitVpISPKQmOjB+17RJ7EnGtpgERGJHf1+mn3iiScAWLNmzaCFGW5OumHDcSdtTid5n76SxKlnYff7ybn8Mvxj+64MGpeTQ+qsmSSMH4fdG8dwF+wMkZ2XhD+h5wNtfKKbidNyACiekMHUmXm4PQ7Gn5XN5LNzzYw6rFhsNnyjCkmaOoWjjz1O++EjBKuqqFn7KoZhMO7e7zDuu3cz8Uc/xD96tNlxY0JCkoerPj+dZSsmMeu8AnZtraTiWDMpaT4VdDFObbCIiJyOfnvqCgsLufDCC6murmbp0qUnXH/mmWcGNNhw4MoIMOJzn+bw73rmy6XNm0tcTs4J98Xl5lJ869cJtbXjTEzAYrOdcI98ICMngY2vHWTcpEzcHgd5Bclk5SYCkJgcx0WXjOe8C4vweB3Y7fpdDjZ7XBxp8+Zy+HePYPN6yb50GaU/eZBwRweJ06Yy6kvXn/CYzpoaumrrcCYn4Q4ETvJTh6/uUJinH9tKJPzBirdFYwPk5ieZmEr+VWqDRUTkdJxyTl1NTQ3XXnstv/jFLwYz07BhczrJXPxJEidPxgj3LJTieG9Vv87qGtoOHMCIRPAWjMSTkYHNrW/eP46k5Dguu/osGuracbsdvRsxA7S3Bak42kR7WzfpGX4CWdq3ywxp58/F7o/HCIU4+ucnCHf0LL3fuOltWg8cwJX2wT50bYcOs/Pue+mqb8Du9zPuO3fgGxWbqzgOhEjY6FPQAYRDYZPSyJmiNlhERE7HKddyT0tL4+mnnx6sLMOSzeU6YcPw7uZm9v33gzRv2w6AJzeHcXff1bsxc3dzM1a3G5tTG2e/r6O9C6vNQqg7QlNjBx6Pg9z8vgvGGBGDzeuPULqnhtR0H+teLmXZiskEshJMSj18uZKTyVi4gGBdPUd+/4c+1yLv7VH3vuqXX6GrvgGAUEsLx556hqL/exMWq+ZCAiSlxDH1nDzeXncEgJz8ROK8em8YCtQGi4jIx3XKfep+8pOfnHTYB2jox0DqrKruLegAOsqO0lFejtXlpPa11zn216eJG5FH/uc+TVxenolJzdfdHWb/7hpeem43njgHU2bmsfqpnQBcde108kZ+UNi1NHditVtxexyU7q5mzPgMmps6VdSZyJWSTP7Kayj96c96jgOBE77kMMJ9e50iob5L9w93TpedwqJ0EhLjwDBobOhgw6sHWbx8AjYtAhSz1AaLiMjp6Leou/322wH45S9/OWhhpIfN48HqdBLpem+/OYsFu9dL675SDvzy1wAEq6oIt7dTcvu3scd5TExrrprKFh5btan3uL6uncnTc9jw6kFeeHonn/3iDFzunn3NHA4r5Uca2bujZyXRt14/RIaGX5qi/UgZHeUVOFOSSZk1g7gReYRaW/FkZ+NOT+u5XlmJKyWFtHnnU/PyWkKtrdg8bnIuuVi9dIBhGAQ7Q7jcdo4ermfdywd6r2VkxxMKhVXUxTC1wSIicjr6LerS39sXyu/3c+jQIQBGjhyJ36+9jgaaJzODom/cwoFfPkSkO0TOFcux2Gy9Q9De17Z/P+H2tmFd1LW1BE84dnt6iriurhBG5IO5Ri3NQaoqmvvc31DXPvAhpY/W/fvZdtudRDo7ASi5/Vsknz2993rLvlK233ZHz5caVivj77mLif/5Q4LVNbhSUvBkZ5kVPWo0NXSw5c0j7Hy3grGTMikck876Vw5gvPdyn3V+IS7X8NmkfShSGywiIqej36Kuq6uLO++8kxdffJGcnBwMw6C8vJwLLriAu+++G6fmcw0Yi82GIyGe5JkzsdrtlP/1KWpfe52C6z6PxW7HeG/4WWDRhTiThvcKd8lpXtweB50dPfOwcvOTqKlqxWKBCxaX4I7reZ22tQR59i/bKJmQyetrSnsebIHC4jSzog9b9Zs29xZ0AEf++BhWt4uq1WtIOWcmLbt3f9BLHYlQ9qfHGXvXv+P50P6Mw92ubRW8+sI+AF59YR9Lkjx84abZVFe2kJjk6V3tVWKX2mARETkd/RZ1P//5z+nu7uaVV17B5/MB0Nrayj333MPPf/5zbr755kELORwFq2uofO753uOuhgYcSUlMuO+7NO3YgTstDX9JybDf3iAlzcc1N8zi8P46nG47GVnxtLV2MWdBIanpH3yj3dHRTdnBBrw+F/M/WUxba5CRo1PJztOH38HmiO/b0+BISuLY038jfswYOsor8BYWkvvpq6j6+z/oqm/AkZw0LIdbRiIGlceaqKtpIzHJQ0Z2PA5nz1v2gb01fe7dva2Cs2bkkZ03vL/kGUrUBouIyOnot6h78cUXefzxx/F4Phja5/P5uOuuu7jyyivVoAwwT05Wn165tPPOxZGYiDs9HX+RNmb+sEBW/EduTeCPdzGqOI3d2yrZs6MKr8/JlLNztU+dCRKnTCZh0kSa3t2KKxAgMG8unZVVHPvrU4RaWwHIvHgJmRcvpau2jsAF84fllxdHDzfwvz9fT+S9IcRXfmEaY8b19FZOnp5L6a4ackcmkRbwM2a89u4batQGi4jI6ei3qLNYLH0ak/d5vV4sFsuAhhLw5ucz4b7vUr/pbVxpaSROmqAtDP4FLreDT1w6ngN7a2lt7mT02ADpmVokxQyejAzG/Ns36K5vwO7z0bR9O0Yk0lvQAVQ9v5qsTy3BV1CANz/fvLAmaW3u5N23ynoLOoB1L+2nsCgNu8NGQVEaV98wi02vH2Lfrmq8ficZWQn4E7SX5VChNlhERE7HKYu6pqYmDMM46TUZWBarFX/R6D69cl1NTbSVHiDU2kLciLxh+WH3fcHObsrLmmhu6iA13U9mdjzWj1jpLynFy9RZ3lPeI4PD4fXi8Pb8WyRMnIARjvS57kpLpauhke7WVtLnnz/4AU1Wuqemd8Gf96UGfL2vcafTxo4t5ezcWgHAay+UkpLmY+LUnEHPKgNDbbCIiJyOfou61tZWLr30UjUoUcIIh6l8/h+UPfonAGweN+Pv+x6+kfmm5jLLvl3VPPHIFgAsVgsrb5hF7sjkj3iURCNnQgKJUyaRdcnFVDzzLK7UFDKXLObQqv9lzK1fNzueKcoO1uP19QwZLt1dQ1ZuIjPOHYnV2vPe290doexgfZ/HVB+3suv7DMPQe3YMUhssIiKno9+i7qWXXhrMHPIRuhoaKH/yqd7jcEcn7YePDMuiLhjsZt3L+3uPjYjBnh2VKupimDMhgRGfWUH6BfPpqqunYdNmxtz6DRLGlZgdzRTjJmfxyENvUjgmjbmLisjOS+wzXNjltjN9Tj7P/nlbzwkLjC7pO6+usaGd7ZvLObC3hqkz8ygsTj+h90+il9pgERE5Hf0WdRJdbG4ProwM2g8e6j3n8PvMC2Qiu91GICueymMf9Ewkpw3P38VQYnU48Obk4M3JIWnSRLPjmCo3P4mrQ2AesAAAIABJREFUb5hF+ZFG0jP85Iw4cVXLsRMz8ce7qa1uJXtEItkf2sYgEjHY+NohNqzt2ZD8UGkdK66dzuixWlBFRERkKFJRFyPsPi+jv3oj+376MzorKsle9im8owrNjmUKm83K7HmFNDd0UHaogQlTsyko0n5zMnQ4nHbyC1PIL0zpPddY386eHZVUlDUzYVo2I0YmUzQuQNG4Ewu1zo5u9myv7HPu2JFGFXUiIiJDlIq6GOLJy2X8vXcTCXbiTEoalsu8vy814OeKL0yjsz2E1+/8WFsTGIZBJBzBpm0MJMaEQxHeeHk/b687DMDWt49yzY2zGFGQctL7PR4H4yZn8fqa0t5zuSO1h52IiMhQ9U8Vdeeffz6vvPLKGY4i/QnW1VP7+hvUvv4GqXNmk3ru7GFd0L3P5XLgcn28OUJ1ta28ve4wRw83MuO8kYwak4bLrflF0S7c1UXzjp1UrX4RX2EBqXNm484Yfr1NbW1d7Hy3vM+5qoqWfos6i9XC1Fkj8MW7OLy/nonTsk86hFNik9pgERE53j9V1J1sNS4ZOPVvbuTQ/1sFQOvefVgdDjI/eZG5oWJId1eYl5/bw853e5Z/P3qogc9+cSYFRakmJ5OP0la6n53fuReAunXraT96lMIbvjTs9myM8zooHJPO9s3Hes+lpZ96e46EJA9nzxnJ2XNGDnQ8GWRqg0VE5Hin3tirH1pOeXDVv/XWccebTEoSXcKhCLXVrdTVtBI+bp+zD+vo6ObA3to+5+pqW/u5WwZLuKODxm3bqV77Gq37D2BETvw37DjWt3eqfsObhJpbBiti1LDbbZy/qIjZ8wvJK0zm8mumkvWhhVFkeFEbLCIix+u3p+5//ud/TnreMAza29sHLJCcKG3ueTRufqfP8XDXFQyxaf1hXnp2NxaLhQs/NZYpM3KxO04clur1Ohk/JYtN781HAgh8aHl4MUfdmxvZ9+P/BsBitzPhvu/iLxrd5x53Vmaf48RpU7HH+wctYzRJTvWyYHEJ4XAE23ubkEciBjWVzTQ3dZKU7CU1oFVghwq1wSIicjr6Lera2tr6fdDVV189IGHk5JKmnkXRN2+h8e3NJE49i8QPLffefvQYLXv2YrHbiR8zGndGholJB09dTSsvPrPrvSODvz+5nbyCZAJZJxZrNruVc+YXkpbho/JYM+OnZJOZkzC4gaWPUFsbR//8195jIxSiYcs7JxR1vsICSu64jYpnn8c3ehTp8+YOu6GXx3u/oAM4cqCORx56k0jEwOmycfWXZ6kHb4hQGywiIqej36LuK1/5Sr8PWrVq1UBkkX44/H5SZ80kLjeXtsOHaTtwkLi8XCLdIXZ85x66anqGFvqKxzD2tm/hSBj6vVChUN+heobRMxzzeIZhUHWsmeqqFlLSfIyblEWczzVYMaUfFoeDuJxsOsrKes+50nrmOAbr68EAV0oyNreb5GlTSZ421ayoUSscCvPai/uIRHrmV3UFw+x4p/yEoq6msoWqimZcLgcZOfH4491mxJXTpDZYREROxz+1UMqqVatYuXLlGY4ip9K8dx87br8TIxwGIP8L1+AKBHoLOoDW3XvorKkeFkVdcqqX0WPT2bezGoBxkzNJSo074b6Ko038z4Pregu+eZ8cw9iJmaRos3JT2ZxOcldcSbC+ntZ9pQQWzCNh3Djq1m9g/y8ewogYFFx/HSmzZmB1aJXSk7JY8Hj79lp64voe11a1sOpn6+ho7wZgxrkjWbCk+GNtASLRS22wiIgcT6tfRjnDMGjdf4DaV9b2FnQANa+8StbSxWC1wnsLTNg8buzeU6+IN1R4fS6WXjGR6soWLFhIz/Cf8IEWoOxgfZ8evM3rj5CRGY8/wY3TqW0azeQdkce4u/6dUFsbjsREgpVV7PnRjzFCIQD2/vgnTM67H29+vrlBo5TNZmXO/FEcPdRAc2Mn6Zl+iif0HX5dU9XaW9ABbHzjEGefm09SyvB4nxiq1AaLiMjx/qlPtVp5a/B0HD3Grnu/R+CCBX3Ou9LTqX97M0W3fI0jj/wBi9NJwf+5dtjMqQPw+d34/KceSuZP6Hs9MTmOY2WNBHLiVdRFAbvXixEK0XmsnEg4hMVqpffjaiRCqL3DzHhRLyM7gS/cNIf21iD+BDfe44YWuz19eznjE9w4nOqli3Vqg0VE5Hj9fqqdMmXKSRsOwzAIBoMDGko+0FVfT3djE5GuLlLOmUX9xrfwFY0mfmwJNa+sJWHiBCb+x31gteLwD89VAU8lNz+ZqTPzeGfTUdIz/JRMzOT1NfuYOjPP7GgCtJcdZff9/0nH4SO4AgEKvnQ9pf/9IADegpF4MofPlxT/rPgEN/EJJ/9yIyM7ngWLi3ntxX34E9x86spJH/lFiEQHtcEiInI6+i3qtmzZMpg55D3hYJCuujqsbjeu5GRc6WnYfT7Kn/4b8WNLyPv0VfiKxxCsrGTMN7+OM0GrOJ6KP8HN5LNzycxLxGqxcPRwAxPOyj6hR0PMUfP6G3QcPgJAsKqKpu3bGXvvdwi3teErKMCZlGRyQvO0tQbpaO/GH+/C5f7n5hV64pycc34h48/KxmG3apGgGKI2WERETofGn0WRYH09ZY8+RtULL+JMSmLUTTdi83gYd+93qH3tDaxOJ0nTp+LNy4NxY82Oa6r62jYa69vxxbtJS/dhsfY/HMkd52D3uxWU7qkhJc3Lsk9PwfqhZeHFPKHmpj7H3c0tJI4bi8U2vIcIVpU38eeHN1NX3caoknQ+ccm405oH19zUSV11K263g9SAl4REzwCmFREREbOpqIsibfsPUrX6BaBn2OWBX/2G+JISAPK/sBKHT4sbAFSWN/PwL9fT0daNzWblM9fPIH9USr/3H9xXS+meGgDqatp446VSLv3sFK0AGAXS582j+sWXiXR1YbHZyLnk4mFf0IVCYdau3kdddc8+ZaW7qtk/Np1p53y8//8NdW386X82UV3RAsCSKyYy5exczcMSEREZwlTURZFwZ99FIbobm3DE+zn216fI+MQiHKNHmZQsuuzdXklHW8+KfuFwhLWr95A9YgYOx8mLgbaWvvNPGhvaCYcjKuqigL9oNJP+8z/orKrGlZpCXF6u2ZFMFw5HaGxo73Ou9UOv4fbWIAb0O4S4qrylt6ADWP3UTgqLUklIOnHLDxERERkaVNRFASMcpu3QYezx8djj4wk1NwOQcdGF1K3bAIDV6SDc0YHNo2FUdnvfoZMOp40Pd0IEg904HXY6O7upKm8mJz8Ju8NKqLtna4O5C4twubT3WbSIy8s9aTFnRCKEWlqx+7zDqvfO5XIwd2ERa1fv7Vmt9UgDRWMDGBGD0j01/OPJHYTDEc6/aAwFo1PwJ/R9T7Da+vbI2e3WUw5PFhERkdinos5EHeUVtB06DBjUrdtAw6ZNZF96Ce5AAAODmpfX0llVRfG3b6Xyub/TsmcvmUsWkzzzbBy+4bt5dtG4AFs2llFX04bH62DuhUXY7TbaWoPsfLeczeuPMLIolZz8ZP78u7fx+V0sWFyC1+ckMTmOjKyhvzl7rAvW1FLx/D+ofe11kmdMJ+tTS3Cnp5sda9Akp3nJyEqgsryJuReOITk1jrraNh5ftYnQe/suPv2nd/n0/5l+QlGXkRXP6JJ09u2qxm63suTyicQn6MsgERGRoUxFnUk6yivYfvuddNXXA5C+YB6e3FyO/P5RUs87l1Ff+TLeESOw2KyU/elxal97A4DSn/6MkqREkqeeZWZ8U6UG/Fx9wyyaGzvwel0kpvQMKzu0v47nn9gBQFVFCxPbuknP8FNd2cI/ntzB5SunkjNi+K6mGEvqNm7k2F+eAKDi2edJGD+OujfW0d3SSso5s/CPKjQ54cDp7grz0nN72LO9EoC/Pb6VhCQPLre9t6Bzue3MvbCIA7trKT/SRPGEDNICPVuaxCd6WPbpyTTWd+Bw2khO1VxcERGRoU5FnUnay8p6CzqA6pfXknvFclr37sMdSMPmcuHNH0FXQwNN23f0eWzHsXIYxkUdgD/ejT++735bFWV9V1IsL2skkBlPdWXP/CJtNh47mnfs6v1z+ry5lP3xMdoOHgJ6irxJP/ohcbk5JqUbWB0d3Rw5UNfnXENdGyUTMsnMSaDiaBMzzh3Jay/uo6O9Z27plo1lrLzxnN796jxxTjxxzkHPLiIiIubQuu4mOX5unCM+nnB7O76i0aSdf36f8+nz539wo8UypHsp/hUFRal9jieclU1tdQsWC8w4dyQZ2Rp2GSvS583t/bMrPb23oAOIdHbSWVllQqrB4fU5mTjtQwWrpWdIpdfvYtmKySxePoHElLjegg6gsa6d5saOk/w0ERERGQ7UdWES78iR5F51BUf//ASO+HhGfe0ruNLSyLnichz+D+bLWWw2Mj95EZ7sLNoPHyZ5+nR8WgXzpHJGJPHZL85g365qsnITGTk6lYnTcgh1h0lI8mi1yxgSP34c4+6+k+bde0iYMJ7qF18iWNOzLQVWK86UZHMDDiCbzcqsuQWkZ/ipqWplzPgAmTmJAKRl+EnL8FNT1YLNbiX83nBMT5wDr089cyIiIsOVxTAMw+wQ/6pdu3ZR8t5+brHECIcJ1tZidThwJg/dD6ki/6rWAwc5+ucn6G5qIveK5SSMHzesVsQ8nmEYlB2s59UX9mF32Dj3gtFk5yWaHWvAxOp7vFn0+xIRGbr6e4+Pyp66hx9+mN///vfYbDbmzp3LrbfeanakAWGx2XAHAmbHEIl6voKRFN3yNQzDwObQdhQWi4W8ghSuujYRiwVsw7jAFRERkSgs6jZs2MCaNWt4+umncTqd1NXVffSDRGTIs9qj7u3KdBpSLCIiIhCFC6U8+uijXH/99TidPfNDUlJSTE4kIiIiIiISvaKuqDt06BCbNm3i8ssv57Of/Sxbt241O5KIiIiIiEjUMmU808qVK6mtrT3h/M0330w4HKapqYnHHnuMbdu2cfPNN7NmzRosFku/Py8YDLJr165+rw9VdrsdH2AArYZBOBw2O5IpLBYLLpcXi2EjbHTS1RU0O5KIiGmGa5soIjKcmVLUrVq1qt9rjz76KAsXLsRisTBx4kSsVisNDQ0kn2J1SJfLNexW+op0d1O34U0OPPQbLFYLBV+6nuSzpw/LeUcVR5t46o/vUFfdxvQ5+ZxzfgG+4zYmF5HYpQLl9AzHNlFEZLjor02MuuGXF1xwAW+++SYABw8epLu7m6SkJJNTRZ+OY+XsfeAnhFpa6G5qZu9//hed5RVmxxp0nR3dPPfENqorWgiHI2xYe4DysiazY4mIiIiIDJqo69a57LLLuO2221iyZAkOh4Mf/OAHpxx6OVyF2tshEuk9NkIhQh3tJiYyR1dXmIbavs+7rU3DL0VERERk+Ii6os7pdPKjH/3I7BhRz5OZgbewgLb9BwDwlxTjycgwOdXg88e7mD2/kBee6emKdjhtZOYkmJxKRERERGTwRF1RJx+PMymJ4m99k9b9B7BYrHgLRuJIGH7FjMViYfLZuQQy42lv7yIt4CeQFW92LDlNke5u2o8epbuhEXdGBp6sTLMjiYiIiMQMFXUxzJ2ejjs93ewYpvPEOSkYk2Z2DPkXNG3dxs7v3geRCHa/n/HfuwdPdhYWm03Dr0VEREQ+goo6ETFVqL2dw4/8oXeOqDd/BC179rL/l78ivriIwIUL8WSq505ERESkPyrqRMRUFqsVmyeu9zhp6lns/9kvAGjZuYtgbT2jvnoDNqfTrIgiIiIiUS3qtjQQ6Y8RMSg7WM9f/7CFNc/upraqxexIcgbY3G7yr/ks9vh4rE5nz8quH9K4ZQuhllaT0omIiIhEP/XUScyoqmzhd79YTyRsALB/TzWf/eJM4rzqwYl1/jFFTH7gP+hubiHU1sbRx/7cey313Dk4ErT4jYiIiEh/VNRJzGisb+8t6AAqjzXT2hJUUTdEuNLScKWlEe7qYtw9d1Gz9jW8hSNJnj4dq11vVSIiIiL90ScliRkJSR6sVguRSE9hl57hw+dTQTfU2JxOEidNJHHSRLOjiIiIiMQEFXUSMwKZ8XzuSzPZ8OoB4hM9TJs1gjify+xYIiIiIiKmUlEnMcNqtTCiMIURhSlmRxERERERiRpa/VJERERERCSGqagTkajU1dBAt7YyEBEREflIGn4pIlEl1NFB3RvrOfzI73EkJFD4peuJLyk2O5aIiIhI1FJPnYhElfbDRyj96c/obmik/dBhdv/wfoJ1dWbHEhEREYlaKupEJKp0Nzf3PW5oJNzeYVIaERERkeinok5EokpcTjb2+Pje45TZs3ClppqYSERERCS6aU6diEQVT1YWE753D62lpVjdbvxFo7F53GbHEhEREYlaKupEJOrE5eUSl5drdgwRERGRmKDhlyIiIiIiIjFMRZ2IiIiIiEgMU1EnIiIiIiISw1TUiYiIiIiIxDAVdSIS1ULt7RiRiNkxRERERKKWVr8UkagUrG+gZu1aal56haTp08hYdCHuQLrZsURERESijoq6KGSEw0TCYWxOp9lRYpIRMQiFIzgcNrOjyL+gccs7HF71MADtR8qwOOyMWHGVyanMEeoOU1XZQktTJ8kpcaRnxn/0g0RERGTYUFEXZVoPHuLYX56gu7GJnMsvI37cWKx2/TN9XHXVrbz5+kEqjjYxc24Bo4vTcbr0+4tFzTt39jlu3PwOOcsvw+ZwmJTIPAdL63j0txvBALvDysobzyErN9HsWCIiIhIlNKcuinQ1NLD7+z+k9rU3aNq2nR3fuZf2I0fMjhUzurpCrH56J5veOMyxw4385X83c6ys0exY8k9KmTWzz3HGhQuHZUHX3RXi1dV7weg5DnVH2LO90txQIiIiElXUhRFFupuaCVZXf3AiEqGrth4KCswLFUM62ro5tL+uz7nG+naT0si/KmHcWMbe9e80vvMO/jHFJIwfZ3YkU1isFvwJ7j7nvH6XSWlEREQkGqmoM1Gwro5gdQ2OhAQ8WZk4k5PxjR5N6759AFhdLtwZAZNTRrfurhB1te1EwhESkj2Mn5LFljfLALBYIC3gNzmh/LNsHg9JZ00h6awpZkcxld1u47wLi6iqaKahtp38whRGFZ98wZimxg727qji8P46JkzNJr8whXDYoL6uDafDRnKaF7tdc01FRESGGhV1JmkvO8rOe79PsKoKm8fN2LvuIL6kmNH/9ybq39xId1MTqXNmE5eXa3bUqBUKhXl7/RFWP7MTDJgwNZtzLxhFRnYCNVUtjJuURWZ2gtkxRf5lGVnxfP4r59DR1o0/wY3bc+IwVMMweOfNI6xd3fOl0M53K7jmxlm8vqaU/btrsFgtXHzlRCZMzcFisQz2UxAREZEBpKLOJA1vbyZYVQVAuKOTw79/lLF33EZcdhZxly7rvS9Y30D74cMYkQjeEXm4UlPNihx1GuraefFvu3rnGm17+xhnzchj+ux8Guraqa5s5vDBejKy/MR5NVxNYpvP78bnd/d7vbOjm22by/ucq69pY//uGqBnVdjnnthBXkEKiclxA5pVREREBpeKOpMYxnGbKYfDYBh9TnW3tXHw17+lbt16APzFYyj65tcJt7VhRCJ4MgLYPJ7Bihw1Oju6aahrJxyOYBx3zTAMGuraeeShDTTU9cynO2deAedfNEbDzmRIc7kdFE8IsO7lA73n7Mdt62EYxvFvMyIiIjIEqKgzSdJZZ1Hx9N/oqm/A6nSS95kV2Nx9v4XvrKjsLegAWnbvoXnHTvb9139DJELmxUvJveJyHD7vYMc3TWtzJ6uf3sn2LeVk5SUwZ8EoXnuhZ7hZ8YQMUgN+yo809hZ0AOvXHmTKjDxS0nxmxRYZcFarhWmz83G5HRzYV8tZM/LIyk0kd2QSZQcbwAKLLh5LYtLw+yJIRERkqFNRZxLviDwm/PA+gtXVOBIT8GRnn3CPzeXCYrdjhEK957pqaiDS08tX8dQzpM4+B8eYokHLbbbqyha2b+kZYlZ+pAm328HKG2dhsVpISfMR53XidPXtnXC77djs2r1DYldDbRs11a24PXbSM+Nxu0++tUNiUhznXjCa2fNHYbX2zJu74ppp1NW24XTaSEn3YbFqPp2IiMhQo6LORO70NNzpaf1fz8xg1Fdv4MAvf4URjpD76RXUb9jQ96ZweIBTRpdIuO/YscP76/Fd5iY59YPeykBmPHMWjGLdy/txue0s+/QUEpM0hyjWdVRUUL9xE8HaWlLnzMY/ehQW69Av1utqWvnfX26gpbETgIWfKmHmuQWnLM6sH7rm9bu0BYKIiMgQp6IuilntdtLmnkd8STFGJIIjIYGuhnpadu8BejZndmdnmZxycKVl+skZkcjRwz2bis//5JgThpN5vE7Ou3A0k6bnYrdbSFBBF/NC7e0c+OWvaXznXQAqn/s7E3/0A3wjR5qcbOBVlTf3FnQALz+/h5KJWRpGKSIiIr1U1EU5i8WCO/DBXnW5V1xO6uxZEI7gycnGER9vYrrBl5Do4YqV06itacPpsJIS8GO1ndhbY7fbSEkbPnMNh7ruhobegg7ACIXorKgcFkWd47jFTjxxTuwnec2LiIjI8KWiLsY4fF4cY8aYHcNUvng3vvj+l3aXocfu9xOXn0/7oUM9JyyWYbO9R2ZOAlNm5LJlYxmeOAcXr5iML17DKUVEROQDUVfU7dq1i7vuuotgMIjNZuM73/kOEydONDuWiJjIER9P0S1f49iTT9FZWUXO8kvxjsw3O9ag8MW7WbRsPDPnFuB02jScWERERE4QdUXd/fffz4033sjcuXNZu3Yt999/Pw8//LDZsUTEZN4ReYz+yg1EIhFsjpOv/jhUtDR3sndHFaW7qxk/JZuCojTSAn6zY4mIiEiUirqJGRaLhba2NgBaWlpIT083OZGIRAuLzTbkCzqA7ZuP8eyft7FnexV/eXgzh/fXmR1JREREoljU9dTddtttXHvttfzwhz8kEonwxz/+0exIIiKDprsrxLbNx/qcK91TTfGEDJMSiYiISLQzpahbuXIltbW1J5y/+eab2bBhA9/+9rdZtGgRzz33HLfffjurVq065c8LBoPs2rVrgNKKiAweh8NB0bg0Ko81957LGRHP7t27MQzjFI8U6aE2UURk+LEYUfYpYerUqWzatAmLxYJhGEydOpXNmzef8jG7du2ipKRkkBKKiAys5sYOdr5bwd4dVUw6O4fRJenEeYfvipd6jz89+n2JiAxd/b3HR93wy/T0dDZu3MiMGTPYsGED+fn5ZkcSERlU8YkeZs4tYMZ5I7FYLGbHERERkSgXdUXdvffey/e//31CoRAul4t77rnH7EgiIqZQQSciIiIfR9QVddOmTeOJJ54wO4aIiIiIiEhMiLotDUREREREROTjU1EnIiIiIiISw1TUiYiIiIiIxDAVdSIiIiIiIjFMRZ2IiIiIiEgMU1EnIiIiIiISw1TUiYiIiIiIxDAVdSYL1jcQrK01O8aQ09HeRWN9O6FQ2OwoIqZpbuygpbnT7BgiIiIywKJu8/HhwohEaHh7M6UP/oJIVxf5n7+GtPPmYHO7zY4W8yrLm3nqj+9QXdHCWTPzOPeCUcQneMyOJTJoQt1htm8p5x9P7cBut7Lk8omMHhvAarWYHU1EREQGgHrqTNJZWcme//hPuhsbCbe3s/9nv6D96DGzY8W8rmCIF5/ZSdWxZoyIwdvrDlN2sMHsWCKDqqaqhaf/9C7BzhBtrV385eHN1Ne2mR1LREREBoiKOpOEOzqJdHX1PdfeblKaoaO7O3zCh9dWDT+TYSbYGepzHApF6AqG+rlbREREYp2KOpO4A+kkTZ/2wXFONp6sLBMTDQ1en4vZC0b1HtvtVvIKUkxMJDL4ktN8pGX4e48Li9NITIkzMZGIiIgMJM2pM4nd56PwS9fTduFCIqEQ3oJ8XKkqPs6EcZOySE7x0toSJC3gI5AVb3YkkUEVn+Dmqi9Mo/JYM1abhYysBOLinGbHEhERkQGios5ErtQUFXIDwO1xMHJ0qtkxREyVlOIlKcVrdgwREREZBCrqRERiWDgcofJYE3U1bSQkecjMTsDp0lu7iIjIcKKWX0Qkhh073MDvfr4ew+g5vvLz0xgzPsPcUCIiIjKotFCKiEgM276lvLegA1j38n5CobB5gURERGTQqagTEYlhKel9580FsuKxWfXWLiIiMpyo5Y8yofZ2QtqvTkQ+pqKxAcZPycJitVAwOoWp54zg6OEGaqpaiIQjZscTERGRQaA5dVHCiERo2r6DQ797GLCQv/JzJIwbi0XfuIvIKSSleFl6xSTmf7KYUHeYPz+ymeryFqw2C1esnEbR2IDZEUVERGSAqWKIEh3l5ey853u0le6nrbSUXfd8j47yCrNjiUgMcDhtJCbHUVneTHV5CwCRsMHf/7qdttagyelERERkoKmoixLdTc0Y3d29x5GuLkLNzSYmEpFY9+EFVERERGToUlEXJdyBAK7AB8Ok3FmZfY5FRD5KVm4i6Rl+AKxWC5+4ZDxen8vkVCIiIjLQNKcuSrhSUxh75200bd0OFgsJE8bjSkk2O5aIxJDkVC+f+eIMGuracXvspKb5zI4kIiIig0BFXRSJy8khLifH7BgiEsP88W788W6zY4iIiMgg0vBLERERERGRGKaiTkREREREJIapqBMREREREYlhKupERERERERimIo6ERERERGRGKaiTkREREREJIapqBMREREREYlhKupERERERERimDYflyGvKxji6OEG9u+pISs3kRGFKfj8LrNjiYiIiIicESrqZMg7eriBRx56s/f4okvGcfackSYmEhERERE5czT8Uoa80t01fY43rz9CsKPbpDQiIiIiImeWijoZ8rJzE/scFxSl4nCpk1pEREREhgYVdTLk5Y9KZtHFY0nL8DPj3JFMn52P1WoxO5aIiIiIyBlhSlH3/PPPs3jZloDeAAAKoUlEQVTxYoqLi9m2bVufaw899BALFy5k0aJFvPbaa2bEkyHG63cz47wCPv+Vc1j4qbEkpXrNjiQiIiIicsaYUtQVFRXx05/+lOnTp/c5X1payrPPPsuzzz7Lb37zG+6++27C4bAZEWUIcnsc6qETERERkSHHlKKusLCQgoKCE86vWbOGxYsX43Q6yc3NZcSIEWzdutWEhCIiIiIiIrEhqubUVVVVkZGR0XscCASoqqoyMZGIiIiIiEh0G7AlAFeuXEltbe0J52+++WYuuOCCM/p3BYNBdu3adUZ/poiISCxSmygiMvwMWFG3atWq035MIBCgsrKy97iqqopAIPCRj3O5XJSUlJz23yciItFPBcrpUZsoIjJ09dcmRtXwy/nz5/Pss8/S1dVFWVkZhw4dYuLEiWbHEhERERERiVqm7MD8wgsvcO+991JfX88Xv/hFSkpK+O1vf8vo0aP5xCc+wSc/+UlsNht33nknNpvNjIgiIiIiIiIxwWIYhmF2iH/Vrl27NNRERGSI0nv86dHvS0Rk6OrvPT6qhl+KiIiIiIjI6VFRJyIiIiIiEsNU1ImIiIiIiMQwFXUiIiIiIiIxTEWdiIiIiIhIDDNlS4MzLRgManNaEZEhKhgMmh0hpqhNFBEZuvprE4fElgYiIiIiIiLDlYZfioiIiIiIxDAVdSIiIiIiIjFMRZ2IiIiIiEgMU1EnIiIiIiISw1TUiYiIiIiIxDAVdfKRmpubuemmm7jooov4xCc+wZYtW/pcNwyD7373uyxcuJClS5eyY8cOk5KKnBmrVq1i8eLFLFmyhFtuueWE5YO7urq4+eabWbhwIZdffjlHjx41KamIDIRXX32VRYsWsXDhQn71q1+dcP3YsWNcc801LF26lM997nNUVlb2Xrv//vtZsmQJS5Ys4bnnnus9v379ei655BKWLFnCv/3bvxEKhQbluYi8bzBf101NTdx4440sXbqU5cuXs3fv3t7H/O53v2PJkiUsXryYVatWDdwTHm4MkY9w6623Go899phhGIYRDAaNpqamPtdfeeUV49prrzUikYixZcsWY/ny5WbEFDkjKisrjXnz5hkdHR2GYRjGTTfdZPzlL3/pc88jjzxi3HHHHYZhGMbf/vY342tf+9qg5xSRgREKhYwFCxYYR44cMYLBoLF06VJj3759fe756le/ajzxxBOGYRjGunXrjG984xuGYRjGyy+/bKxcudLo7u422trajEsvvdRoaWkxwuGwcd555xkHDhwwDMMw/uu//qu3XRUZDIP9uv7BD35g/PSnPzUMwzBKS0uNq6++2jAMw9izZ4+xePFio7293eju7jauueYa49ChQ4PyOxjq1FMnp9TS0sJbb73F8uXLAXA6ncTHx/e5Z82aNSxbtgyLxcLkyZNpbm6murrajLgiZ0Q4HKazs5NQKERnZyfp6el9rr/00ktccsklACxatIj169djaMtPkSFh69atjBgxgtzcXJxOJ4sXL2bNmjV97tm/fz8zZ84EYObMmb3XS0tLmTZtGna7nbi4OMaMGcOrr75KY2MjDoeDkSNHAjB79mxWr149uE9MhrXBfl1/+GcVFhZy7Ngxamtr2b9/PxMnTsTj8WC325k+fbr+L5whKurklI4ePUpycjLf/va3WbZsGbfffjvt7e197qmqqiIjI6P3OCMjg6qqqsGOKnJGBAIBvvCFLzBv3jzmzJmDz+djzpw5fe6pqqoiMzMTALvdjt/vp6GhwYy4InKGHd+mBQKBE9q04uLi3g+iL7zwAm1tbTQ0NFBcXMxrr71GR0cH9fX1vPnmm1RWVpKUlEQ4HGbbtm0A/P3vf+8ztE1koA326/rDP2vr1q2Ul5dTWVlJUVERb7/9Ng0NDXR0dPDqq6/q/8IZoqJOTikUCrFz505WrFjBk08+icfjOek4bJGhoqmpiTVr1rBmzZreRuypp54yO5aIRJFbb72Vt956i2XLlrFx40YCgQA2m405c+Ywd+5crrrqKr7+9a8zefJkrFYrFouFBx54gPvuu4/ly5fj9XqxWvURTKLLmXxdX3/99bS0tHDxxRfz8MMPU1JSgs1mo7CwkOuuu45rr72W6667juLiYv1fOEPsZgeQ6JaRkUFGRgaTJk0C4KKLLjqhqAsEAn2+ZamsrCQQCAxqTpEzZd26deTk5JCcnAzAhRdeyJYtW7j44ot77wkEAlRUVJCRkUEoFKKlpYWkpCSzIovIGXR8m1ZVVXVCmxYIBHjwwQcBaGtrY/Xq1b1TE7785S/z5S9/GYCvf/3rvUPTpkyZwh/+8AcAXn/9dQ4dOjTQT0Wk12C/rn0+H/fddx/Qs6DeggULyM3NBeDyyy/n8ssvB+CBBx7QZ8YzRKWxnFJaWhoZGRkcOHAA6FnlqLCwsM898+fP58knn8QwDN555x38fv8Jc5BEYkVWVhbvvvsuHR0dGIbR72v+r3/9KwD/+Mc/mDlzJhaLxYy4InKGTZgwgUOHDlFWVkZXVxfPPvss8+fP73NPfX09kUgEgF/96ldcdtllQM983PeHYu/evZs9e/Ywe/ZsAOrq6oCe1XN//etfc9VVVw3WUxIZ9Nd1c3MzXV1dADz++ONMmzYNn8/X5zHl5eWsXr2apUuXDuRTHzbUUycf6Y477uAb3/gG3d3d5Obmct999/Hoo48CsGLFCubOncvatWtZuHAhHo+H73//+yYnFvnnTZo0iUWLFnHJJZdgt9spKSnhyiuv5Cc/+Qnjx49nwYIFLF++nG9+85ssXLiQhIQEfvzjH5sdW0TOELvdzp133sl1111HOBzmsssuY/To0X3eAzZu3MgDDzyAxWJh2rRp3HXXXUDPlIXPfOYzQE9Pxf3334/d3vNR6ze/+Q2vvPIKkUiEFStWMGvWLNOeoww/g/263r9/P9/61rcAGD16NN/73vd6s3z1q1+lsbERu93OXXfddcICfPLPsRhask1ERERERCRmafiliIiIiIhIDFNRJyIiIiIiEsNU1ImIiIiIiMQwFXUiIiIiIiIxTEWdiIiIiIhIDFNRJxJDjh49ypIlS8yOISIiIiJRREWdiIiIiIhIDFNRJxKjysrKWLZsGVu3bjU7ioiIyKD62c9+xqJFi1ixYgW33HILv/3tb82OJGIqu9kBROT0HThwgFtuuYUf/OAHFBcXmx1HRERk0GzdupXVq1fz9NNP093dzaWXXsq4cePMjiViKhV1IjGmvr6eG264gQcffJBRo0aZHUdERGRQbd68mQULFuByuXC5XMybN8/sSCKm0/BLkRjj9/vJysri7bffNjuKiIiIiEQBFXUiMcbhcPDggw/y5JNP8swzz5gdR0REZFCdddZZvPzyywSDQdra2njllVfMjiRiOg2/FIlBcXFxPPTQQ3z+858nLi6OBQsWmB1JRERkUEycOJH58+fzqU99ipSUFIqKivD7/WbHEjGVxTAMw+wQIiIiIiIfV1tbG16vl46ODj7zmc9w7733arEUGdbUUyciIiIiMeXOO++ktLSUYDDIJZdcooJOhj311ImIiIiIiMQwLZQiIiIiIiISw1TUiYiIiIiIxDAVdSIiIiIiIjFMRZ2IiIiIiEgMU1EnIiIiIiISw1TUiYiIiIiIxLD/D0jhNrpoAqlgAAAAAElFTkSuQmCC\n", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "num_rows = 2\n", + "num_cols = 2\n", + "scale = 6\n", + "fig, axes = plt.subplots(num_rows, num_cols, sharey=True, figsize=(scale * 1.2 * num_cols, scale * num_rows))\n", + "plt.subplots_adjust(left=0.1, top=0.9)\n", + "fig.suptitle(\"LogLikelihood difference (vs. EPA, g=0.1)\", fontsize=16)\n", + "\n", + "palette = sns.color_palette(\"deep\", 8)\n", + "\n", + "k = 0\n", + "order = [\"pplacer\", \"epa\", \"rappas\", \"epang\"]\n", + "for i in range(num_rows):\n", + " for j in range(num_cols):\n", + " color = sns.color_palette()[j]\n", + " \n", + " if order[k] == \"pplacer\":\n", + " ax = sns.stripplot(data=df_pplacer, x=\"max-strikes\", hue=\"strike-box\", y=\"LL difference\", ax=axes[i, j],\n", + " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[0:2])\n", + " ylabel = 'LL Difference'\n", + " title = f'PPLACER'\n", + " elif order[k] == \"epa\":\n", + " ax = sns.stripplot(data=df_epa, x=\"g\", y=\"LL difference\", ax=axes[i, j], \n", + " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[2:3])\n", + " ylabel = ''\n", + " title = f'EPA'\n", + " elif order[k] == \"rappas\":\n", + " ax = sns.stripplot(data=df_rappas, x=\"k\", hue=\"omega\", y=\"LL difference\", ax=axes[i, j], \n", + " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[3:5])\n", + " ylabel = 'LL Difference'\n", + " title = f'RAPPAS'\n", + " elif order[k] == \"epang\":\n", + " ax = sns.stripplot(data=df_epang, x=\"g\", y=\"LL difference\", ax=axes[i, j], \n", + " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[5:6])\n", + " title = f'EPA-NG'\n", + " \n", + " ax.set(ylabel=ylabel, title=title)\n", + " k += 1\n", + " \n", + "fig.savefig(\"ll_diff.png\")\n", + "fig.savefig(\"ll_diff.svg\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.7.6" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} From 50cc5a4e0c4ee42607efe6dc54bcf3173074d359 Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Sat, 2 May 2020 18:28:58 +0200 Subject: [PATCH 12/43] Minor changes in the LL_difference.ipynb plot --- .../ll_difference.ipynb | 102 +++++++++--------- 1 file changed, 54 insertions(+), 48 deletions(-) diff --git a/examples/6_placement_likelihood/ll_difference.ipynb b/examples/6_placement_likelihood/ll_difference.ipynb index 0a98a90..f828822 100644 --- a/examples/6_placement_likelihood/ll_difference.ipynb +++ b/examples/6_placement_likelihood/ll_difference.ipynb @@ -20,7 +20,7 @@ }, { "cell_type": "code", - "execution_count": 198, + "execution_count": 312, "metadata": {}, "outputs": [ { @@ -167,7 +167,7 @@ "4 NaN " ] }, - "execution_count": 198, + "execution_count": 312, "metadata": {}, "output_type": "execute_result" } @@ -178,7 +178,6 @@ "import seaborn as sns\n", "sns.set_style(\"whitegrid\")\n", "\n", - "\n", "df = pd.read_csv(\"run/likelihood.csv\", sep=\";\")\n", "df.rename(columns={'likelihood': 'LL', 'ms' : 'max-strikes', 'sb' : 'strike-box', 'mp': 'max-pitches', 'o': 'omega'}, inplace=True)\n", "df.head()" @@ -193,7 +192,7 @@ }, { "cell_type": "code", - "execution_count": 199, + "execution_count": 292, "metadata": {}, "outputs": [ { @@ -219,7 +218,7 @@ }, { "cell_type": "code", - "execution_count": 200, + "execution_count": 293, "metadata": {}, "outputs": [ { @@ -366,7 +365,7 @@ "4 NaN " ] }, - "execution_count": 200, + "execution_count": 293, "metadata": {}, "output_type": "execute_result" } @@ -378,7 +377,7 @@ }, { "cell_type": "code", - "execution_count": 201, + "execution_count": 294, "metadata": {}, "outputs": [ { @@ -387,13 +386,13 @@ "Text(0.5, 0.98, 'Log Likelihood of extended trees')" ] }, - "execution_count": 201, + "execution_count": 294, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -418,7 +417,7 @@ }, { "cell_type": "code", - "execution_count": 202, + "execution_count": 295, "metadata": {}, "outputs": [ { @@ -485,7 +484,7 @@ "13 0dbcba127f79ff6ec9aa56486685d702f1da0ca7-4235956 -39793.188953" ] }, - "execution_count": 202, + "execution_count": 295, "metadata": {}, "output_type": "execute_result" } @@ -498,7 +497,7 @@ }, { "cell_type": "code", - "execution_count": 203, + "execution_count": 296, "metadata": {}, "outputs": [ { @@ -651,7 +650,7 @@ "4 NaN -39791.397634 " ] }, - "execution_count": 203, + "execution_count": 296, "metadata": {}, "output_type": "execute_result" } @@ -663,7 +662,7 @@ }, { "cell_type": "code", - "execution_count": 204, + "execution_count": 297, "metadata": {}, "outputs": [ { @@ -822,7 +821,7 @@ "4 NaN -39791.397634 0.000000 " ] }, - "execution_count": 204, + "execution_count": 297, "metadata": {}, "output_type": "execute_result" } @@ -834,7 +833,7 @@ }, { "cell_type": "code", - "execution_count": 205, + "execution_count": 298, "metadata": {}, "outputs": [ { @@ -843,13 +842,13 @@ "Text(0.5, 0.98, 'Log Likelihood difference (vs. EPA, g=0.1)')" ] }, - "execution_count": 205, + "execution_count": 298, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAFkCAYAAAAwtcDIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deVhUZf8/8PcszLCDIAyguCMqipLinhaKmoi5ppYWpvm4ppn1tFpqWv2+uWU+mlmPtmkbrliSK1imkQuoWOKKLIOsss4wy+8PHqdGtgGBA2fer+vyujxn7pnzAYY399znPveRGI1GI4iIqMmTCl0AERHVDQY6EZFIMNCJiESCgU5EJBIMdCIikWCgExGJBAO9kQsLC8OpU6cAABs2bMCSJUtq/Bq3b9+Gv78/dDodAGDmzJnYtWsXACAyMhJTpkypu4Kr4O/vj5s3b1rU9p9fa2pqKoKCgqDX6wEAmZmZeOqppxAUFIT33nsPRqMRr776KoKDgzFhwoR6q78hTJ48GZcuXRK6DKtw5MgRLFq0SOgy6hQDvRohISH49ddf6/UYVQV1VFQU+vTpU6fH27p1K8aOHVunr1mffHx8cPbsWchkMgDAN998g2bNmuHMmTN45ZVX8Mcff+CXX37B8ePH8f333wtcbe0dOXIEDg4O6NKlS70dIzIyEp07d0ZQUJDZP7VaDaDs/R4YGIigoCD0798fr7zyCgoLC81e45VXXkGXLl2QkZFRb3VW5eTJkxgxYgS6d++OadOmISUlpdK269atQ3h4OLp06YINGzaYPRYSEoKkpCRcvny5vktuMAx0anJSU1PRvn17SCQSAEBKSgpatGgBe3v7Gr/WvU8tjcHOnTvx+OOP1/txevTogbNnz5r9U6lUpsc3b96Ms2fPYteuXbhw4QI2bdpkeqyoqAgHDx6Ek5MT9u7dW++13i87Oxvz58/HwoULcfr0aXTt2hUvvPBCpe1bt26NJUuWYPDgwRU+HhYWhm+//ba+ym1wDPQH8O233yI0NBS9e/fG7NmzTb0cADhx4gSGDx+Onj174u2338bUqVPx3Xff1fgYlX1CKC0txeLFi7FgwQJotVoYDAZs2bIFQ4cORZ8+fbBw4ULk5uZW+JrTpk0rV8v777+P4OBghISE4Pjx46b9arUas2fPRu/evREaGmr25tdqtVi5ciUGDhyIgQMHYuXKldBqtabHt27danqsup5zcnIypk6diqCgIEyfPh05OTmmx/45ZPTKK69g9+7d+PTTTxEUFISdO3fijTfewLlz5xAUFIQPP/wQAHD06FE8/vjj6NWrFyZPnmzWCwsJCcGWLVsQHh6OHj16QKfTQa1WY8GCBejbty9CQkLw+eefm9pv2LABCxcuxMsvv4ygoCCEhYUhISHB9HhaWhrmz5+Pvn37ok+fPli+fLnpse+//x6PPfYYgoODMWPGjEp7k1qtFr/99ht69+5t+r4HBgaa/QwvXbqEPn36oLS0FDdv3sTUqVPRs2dP9OnTp16GDlQqFR5++GFcuXLFtC86OhrOzs6YO3cudu/eXaPXu3jxIsaMGYOgoCA8//zzWLRoEdauXVuj1/j555/h5+eHxx57DEqlEgsWLMDly5dx9erVCtuPHTsWgwcPhoODQ4WP9+7dG8eOHatRDY0ZA72WTp48idWrV2PdunU4ceIEWrRogcWLFwMo60U8//zzePHFF3Hq1Cm0bdsWZ8+erbNjl5SUYN68eVAoFFi3bh0UCgW++OILHDp0CF9++SViY2Ph4uJiFixViY+PR9u2bfHbb79h5syZeP3113FvRYjFixfDy8sLsbGx+PDDD7FmzRqcPHkSALBp0yacP38ee/bswd69e5GQkID//Oc/AICYmBh89tln+OyzzxAdHW16TmWWLFmCgIAAnDp1CnPnzjWN8d/vvffeQ3h4OGbMmIGzZ89i8uTJWLZsmanX+fzzz+PSpUt47bXXsHz5cpw6dQqTJk3C3Llzzf7YREVFYcuWLYiLi4NUKsWcOXPg7++PmJgYbN++Hdu3b0dsbKyp/ZEjRxAWFoa4uDiEhIRgxYoVAAC9Xo9//etf8PHxwZEjRxATE4ORI0cCAA4dOoSPP/4YH330EU6ePImePXvixRdfrPDrunnzJqRSKby8vACUhWmPHj0QHR1tarNv3z4MHz4cNjY2WL9+PQYMGIDff/8dMTExmDp1apXf39pIS0tDTEwMOnfubNq3a9cuhIWFISwsDNeuXcOFCxcsei2tVov58+dj7NixOH36NEaNGoVDhw6ZHk9NTUWvXr0q/bdv3z4AwJUrV+Dv7296nr29PVq1aoWkpKRafY3t27dHSkoKCgoKavX8xoaBXkv79u3D+PHjERAQAIVCgcWLF+PcuXO4ffs2YmJi4Ofnh2HDhkEul+Ppp59G8+bN6+S4BQUFmDlzJlq1aoV3333XNK68c+dOvPDCC/Dy8oJCocD8+fNx8OBBi4YUfHx88MQTT0Amk2Hs2LG4c+cOMjMzkZaWhjNnzmDJkiVQKpXo3LkzJk6ciD179pi+B/PmzYO7uzvc3Nwwb94808fwH3/8EePGjUPHjh1hb2+P+fPnV3r81NRUJCQkYOHChVAoFKZPCrX1zTffYNKkSejevbvpa7KxscG5c+dMbaZNmwZvb2/Y2toiISHB9FFeoVDA19cXTzzxBA4cOGBq37NnTwwePBgymQyPP/64qccfHx+PjIwMvPzyy7C3t4dSqUSvXr0AlP1MZs2ahfbt20Mul2P27NlITEyssJd+9+7dcr3I8PBw7N+/HwBgNBpx4MABhIeHAwDkcjlSU1ORkZFhdkxLnD9/3iwshw4davb4vHnz0KtXLzz55JMIDg7G7NmzAZT9nE6dOoXw8HA0b94c/fr1s7iXfv78eeh0Ojz99NOwsbHBsGHD0K1bN9PjPj4+iIuLq/Tfva+7qKgITk5OZq/t6OhYbpzfUve+53fv3q3V8xsbudAFNFUZGRkICAgwbTs4OMDV1RVqtRoZGRmmnhYASCQSs+0Hce8XY/Xq1aYxZKDsl23evHmQSv/+Gy2VSpGVlVXta/7zj42dnR2Asl+c3NxcuLi4wNHR0fS4j4+PqVeWkZEBHx8fs8funSjLyMhA165dTY+1aNGi0uNnZGTA2dnZbAzcx8cHaWlp1dZekdTUVOzevRtffvmlaV9paanZSTxvb2/T/1NSUpCRkWEWinq93mz7n98jW1tbaDQa6HQ6pKWlwcfHB3J5+V+l1NRUrFq1Cu+//75pn9FohFqtLvf9cHFxKRdKw4YNw4oVK5CRkYEbN25AKpWaanrppZewfv16TJgwAS4uLpg+fbrFM3y6d++OHTt2VPr4xo0b0b9//3L79+zZg/bt25t67OHh4Xjvvffw73//GzY2NlUeMyMjAyqVyuw9+8+fgaXs7e3L9aYLCwsrHVKpzr3vubOzc62e39gw0GvJ09PTrKd1LwBVKhU8PDzMxtONRiPS09Pr5LgDBgyAv78/IiIi8MUXX5iCxsvLC6tWrULPnj3LPef27du1Opanpyfy8vJQUFBgCvW0tDTTCTRPT0+kpqbCz8/P9Jinp6fpsX8GcmpqaqXH8fDwwN27d1FUVGQK9dTUVLNf/prw9vbG7NmzMWfOnErb3B8sLVu2NBveqMmx0tLSoNPpyoX6vTpGjx5d7eu0atXKFPb3vr8uLi4YMGAADhw4gGvXrmHkyJGmuj08PPDOO+8AAOLi4jB9+nQEBwejdevWNf4aLLV7926kpaVhwIABAMpOKOfm5uL48ePlevn3u/c7YTQaTV9DWloafH19AZT9vMPCwip9/rJlyzB69Gj4+fmZDccVFRXh1q1b6NChQ62+pqtXr6JFixZmnZamjEMuFigtLYVGozH90+l0GDVqFCIjI5GYmAitVos1a9YgMDAQLVu2xODBg/Hnn3/i0KFD0Ol0+Oqrr5CZmVnlMYxGo9kx/jnee7/nnnsOo0aNQkREBLKzswEAU6ZMwbp160x/ZLKzs83GKGvD29sbQUFBWLNmDTQaDS5fvozvv//eFFBhYWHYtGkTsrOzkZ2djY0bN5o+Go8YMQK7du1CUlISiouL8dFHH1V6nBYtWqBr167YsGEDtFot4uLicPTo0VrXPXHiROzcuRPnz5+H0WhEUVERjh07Vuk4aWBgIBwcHLBlyxaUlJRAr9fjr7/+Qnx8fLXHCgwMhIeHB1avXo2ioiJoNBr88ccfAMrmlG/ZssV0UjE/Px8//vhjha+jUCjQr18/nD592mx/eHg49uzZg4MHD5q+t0DZkNa9ToKLiwskEonZp7O6dvbsWSQnJ+O7777D7t27sXv3buzfvx+jRo0yDcGdOnXKbHz7n3r06AGZTIYvv/wSOp0Ohw4dMjuxfG9qamX/7r3nQkNDceXKFRw8eBAajQYbN26Ev78/2rdvX+Fx7/3uGo1G6HQ6aDQa0/UMAPD7779j0KBBdfVtEhwD3QKzZs1CYGCg6d+GDRvQv39/LFy4EAsWLMDAgQORnJxsOmPv5uaG9evX4//+7//Qp08fJCUloWvXrlV+LN2/f7/ZMarr8cybNw9DhgzB9OnTkZubi6effhohISF49tlnERQUhCeeeMKiQKrOmjVrkJKSgocffhjz58/HggULTB/H586di65du2L06NEYPXo0AgICMHfuXADA4MGD8cwzz+CZZ55BaGgo+vbtW+VxVq9ejfPnz6NPnz7YuHEjxowZU+uau3XrhhUrVmD58uUIDg7GsGHDEBkZWWl7mUyGzZs34/LlyxgyZAj69u2LN954w6ITZfeee/PmTTz66KMYNGiQKbRDQ0Mxc+ZMLF68GA899BBGjRqFmJiYSl9r8uTJpnC8JyQkBDdu3EDz5s3RqVMn0/6EhARMnDgRQUFBmDNnDl5//XVTbzcsLKzKKYX3ZgT9819175Vdu3ZhyJAh8Pf3h4eHh+nfM888g6NHjyI3NxdpaWkICgqq8PkKhQIbNmzA999/j+DgYOzduxePPPIIFApFlce9n5ubGzZs2IC1a9ciODgY8fHxWLNmjenxpUuXYunSpabtN998E4GBgdi/fz82b96MwMBAs+9xVFQUJk+eXKMaGjMJb3BR/wwGAwYNGoQPPvig2mAj6zZ58mQsXbq0Xi8uqi+vv/46RowYgYcfftii9hMnTsTkyZMxfvz4eq6sYkeOHMGePXuwfv16QY5fHxjo9SQ2Nhbdu3eHra0ttm7diq+//hqHDh2Cra2t0KURCeL06dNo27YtmjVrhn379uGtt97CoUOHTOdd6MHxpGg9OXfuHJYsWQKtVosOHTpg48aNDHOyatevX8eiRYtQXFyMli1b4sMPP2SY1zH20ImIRIInRYmIRIKBTkQkEgx0IiKRYKATEYkEA52ISCQY6EREIsFAJyISCQY6EZFIMNCJiESCgU5EJBIMdCIikWCgExGJBAOdiEgkGOhERCLBQCciEglRBvq9m/ISEVkTUQa6TqcTugQiogYnykAnIrJGDHQiIpFgoBMRiQQDnYhIJBjoREQiwUAnIhIJBjoRkUgw0ImIRIKBbmXOXc3Hnl/v4PYdjdClEFEdkwtdADWcLftTsOuXTACATJqKN6e2RZ/OzgJXRUR1hYFuJfKLddh7MtO0rTcA3x5XVxjod/K0iIy9g5x8HUKCmqF3J4Y+UVPAQLcSBgNgNJrvK9UZy7Ur1Rnw8sdXkZ6jBQAcj8/FsmfaMtSJmgCOoVsJFwc5Hu3RzGzfmAEe5dpdvFloCvN7Dp/NqdfaiKhusIduRV4Y74uH/JxwU12CYH8ndG3rWK5NM0ebCvbxbULUFPA31YrIZBKEBDWrsk1rlS1G9nHHgVNZAABPVxuMf7h8T56IGh+J0Xj/yGrTl5iYiM6dOwtdRpN2I70YOQU6dG3jABs5R+aImgL20KlCbbzs0EboIoioRtj1IiISCQY6EZFIMNCJiESCY+hUoeSMEvx6KQ+ergo83M0VcplE6JKIqBqCBnpMTAxWrlwJg8GAiRMnYtasWWaPa7VavPzyy7h48SJcXV2xdu1atGzZUqBqGz+d3oBrqUWQyiRo2dwOxSU6yOUSaHUGSCUSNHNSoFijh1arx90SHZzsbCCTSqCQS2CEBLYKKVIzS3D5VgE++C4F96Y/Rcdl4bUprXG3WAdbGxlcHeUo1RuhkEuRX6xDqdaAIq0erVT2gn791LAMBgOMuemQufnAUFIAbepVSFWtIdNpodcWQaKwh0QigdGgh0FfCmNxEXT5GTDodLBt6Q+pVAGJXAaDRILSa+cg9Q2ArFQLY9Fd6OU2kNrYwsbRBUapFPrsVEjsnGHUlkCv10JaqoGiZWdIpDKhvw2NimDTFvV6PYYPH47//ve/UKlUmDBhAtasWYMOHTqY2nz11Vf4888/sXz5ckRFReHnn3/GunXrqn1ta5y2+ElUCiJPZFbfsJ6tnNEWD3XgMgFiVxT7FTSxXwtdBuQPjYLTiDlCl9FoCDaGHh8fj9atW8PX1xcKhQJhYWE4fPiwWZsjR45g7NixAIDhw4fj5MmTEOG0+QeWU1DaKMIcAJZtvyF0CVTPDHpdowhzANCd2Q99TprQZTQagg25qNVqeHl5mbZVKhXi4+PLtfH29gYAyOVyODk5IScnB25ublW+tkajQWJiYt0X3Ugl3NILXYKJVme0qu+9NZLlpaMxXTt8M+4YNC17CF1Gg6psBEKUJ0WVSqVVDbm0aafHf49dELoMAIC9rdSqvvfWyGjshNyjq8sv3ymQNv2GQ+pYdSfPWgg25KJSqZCenm7aVqvVUKlU5dqkpZV9nNLpdMjPz0ezZlWvRWKN7JQyLB5f+5PFEglgawPU5Ap/mRSwtTGf+WIjAz6c51frOqhpkEgksH/834BM+P6g7fB5DPN/EOwn0q1bN9y4cQPJyclQqVSIiorC6tWrzdqEhIRg165dCAoKwsGDB9G3b19IJJw+V5HQXu4I7eVe4WMvb0lCwvVCs30bFvihg4/ls1LuFuqw86gaNzNK0NvfGeH9mkMq5c/CWim7PAxll4eFLoPuI1igy+VyLF26FDNnzoRer8f48ePh5+eH9evXo2vXrhgyZAgmTJiAl156CaGhoXBxccHatWuFKrdJG9DVxSzQfdwVaOdlV65dbEIujp7LQXNnG0wc7AkPV4XpMWcHOWaNatEg9RJR7XC1RStgNBqx92QmTiTkQdVMgalDVfByU5q1iU3Ixaqvb5q2vdwU2Lq4E2S8oIioyRB+EIzqnUQiweP9PfB4/8rnJhw9Z35XovRsLRKTC9G1TfmbYBBR48S1XAgA4OZkfqciiaT8PiJq3BjoBAB44hFPeDX7e8x83EAP+Lgrq3gGETU2HEMnE53eiMRbhXB3soFPc4Y5UVPDMXQykcsk6FbBjaOJqGngkAsRkUiwh26l1Dla/HoxD27OcgwI4HrnRGLAQLdCV1KK8NLHV6EpNQAAHvLLxspn2wtcFRE9KA65WKE9v2SawhwAzlwpwJWUIgErIqK6wEC3QoYKJjYZDBU0JKImhYFuhRxtzW/b5eooh78vbx9H1NQx0K1Q3F/5Ztt5hTqUaBvPTTKIqHYY6FbITmn+Y7dVSCHjUrhETR4D3QpNC/Uym6b4ZIgKNjW5uwURNUq89F8k8gp1iL9aAEiM6NLaEe7OVS+spc7RIuF6Adr72KFtBWujE1HTw3noInA2KR9vbb+GUl3ZtkQCPDvCGxMGeVb6HFUzBVTNanbrrtiEXOw4ooZOb8SYAR4Y2afiOyQRkTD4OVsEtkenm8IcKLt37/bodNwt1FX+pBq6qS7Beztu4np6CZLvaLBh922cu5pf/ROJqMEw0EWgouDW6Y3IKah9oOcX61Cq+3ty+rmr+TDcNzh35goDnagx4ZCLCAzr5Ybt0elm+9p62aKVZ9kSuPHXCrD1QCru5JVicKArZo70qXTtlsISPd7bcRNxf+XD0U6GWWE+CO3phnbe5cfZK9pHRMJhoIvApEc84e5sg/2/ZaKwxIDu7R3xVIgKEokExRo9ln9xHYUlZb3tPb9mws1JjiceUVX4Wt8eyzDNUy8o1uPDXbcR7O+Ebm0dMeVRT/wQewcGY9kfkUHdXBvsaySi6jHQRUAikSC0pxtCe5Y/yXk1tdgU5vecv1ZQaaBfSys229bpjbiVoYGrow2eHuaNSY+qYDAYYaeUVfh8IhIOx9BFrrXKFgq5+fCKX4vKL/N/yM/JbNvJToaOLf9ur7SRMsyJGikGusg52cux5IlWcHOSQyoBBnR1waRHKp/OOLp/c0x+tOz+ol3bOGDZM21hq+DbhKgp4IVFVsJgMEKnN0JhU3k4GwxGRJ3Kwpkr+WjrbYcJgzxgz944UZPBMXQrIZVKoKhmvZYdR9T48rAaAPBb4l1cTSnCsoh2DVEeEdUBfpYmkyPncsy2T/+Zj/ziurs4iYjqFwOdTNyczNd/cbCVQlnFEA0RNS78bSWTiOHecLAte0vIpMCMx3yg4CqMRE0GT4qSmSKNHpdvFaG1yhbuzjbIzCuFptSAFs2VQpdGRNXgSVEyY6+Umeaif7T7Ng6czoLRCAR1cMTSaZzCSNSYCRLoubm5eOGFF5CSkoIWLVpg3bp1cHFxKdeuc+fO6NixIwDA29sbmzdvbuhSrdaF6wWIOpVl2j6bVICDcVl4vL+HgFURUVUE6W5t2bIF/fr1Q3R0NPr164ctW7ZU2M7W1hZ79uzBnj17GOYNLC1bW35fVvl9RNR4CBLohw8fxpgxYwAAY8aMwaFDh4Qog6rQ08/JbIaLRAL0Dyj/KYqIGg9BhlyysrLg6Vl2+bmHhweysrIqbKfRaDBu3DjI5XLMmjULQ4cObcgyrZqbsw3efqYNdp/IREmpAaP6uiOwnaPQZRFRFeot0CMiIpCZmVlu/6JFi8y2JRIJJJKKr2A8evQoVCoVkpOT8cwzz6Bjx45o1apVtcfWaDRITEysXeEETakRX/+iw4VbRjjYAmOCZXCXpSExMU3o0ogIqHQWX70F+rZt2yp9zN3dHRkZGfD09ERGRgbc3Cq+t6VKVbbEq6+vL3r37o1Lly5ZFOhKpZLTFh/A59FpSLiVAQAoKAF2/mpA2KBOcHXkpCiixkyQMfSQkBDs3r0bALB7924MGTKkXJu8vDxotWUn4bKzs3HmzBl06NChQeu0FunZGqz+7hZe2pKEvb9mIim1/JroN9UlAlVHRJYSpMs1a9YsLFq0CN9//z18fHywbt06AEBCQgJ27tyJlStX4urVq3jrrbcgkUhgNBrx3HPPMdDrgd5gxGufXjPNarlwvRD9A5zN2tgrpfBrydvNETV2vFLUyl1JKcLzH10x2xfY1gEtPZQ4fDYHcpkEYX2aY/oIb4EqJCJL8bI/K+fhYlPuhtFebgokXC+EptSIwhIDvj2egei4bIEqJCJLMdCtnKujDZ4d4W0KdV8PJfp3dUHyHY1Zu2Pncyp6OhE1Ipy2YGXuFuqw9cdUXL5VhC6tHTDjMW+MHeiBR3s0Q3Z+KdqobJFToINUChj+cW/p5i42lb8oETUK7KFbmbU/JOPnP3KQfEeDg3HZWBd5GwDg6ihHO287SKUSuDvb4KkQL9y7wZGnqw2mhKgErJqILMEeupX5/c+7ZtunL9+tsN2TQ1QY2rMZ7uSWwt/Xvtw4OxE1PuyhW5lWnrZm261VtpW0BDxdFQho48AwJ2oiGOhW5vmxLeHxv/FwT1cbLBjTUuCKiKiucB66FdIbjMjMK0VzFxvIpJb3vrPzS2GvlPEmF0SNFMfQrZBMKoGqmcLi9vnFOqz88ibOXyuAnVKKmSN9MLK3ez1WSES1wa4WVevbYxk4f60AAFCsMWDT3hRk55cKXBUR3Y+BTtW6kW6+MJdOb8Tt+y48IiLhcciFzHwenYZ9v2XBViHF1KEqDO/ljmB/Z8T9lW9q4+wgg7+vvYBVElFFGOhkEhOfix1H/7cOerEe6yNvo5OvA0b1dUehRo9j53LQ3MUGEcO9zW5PR0SNAwNdBNKzNfjqsBrqHC0GdXPFqH7Na/U6F28Umm0bjcClm4VorbLFlEdVmPIorxYlaswY6E3c/euZJ1wvhEQKhPWpeah3bm2PvSfN93VqxaEVoqaCn5ubuKupxaYwv+dEQl6tXmtwoCsmDvaAnUIKV0c5FoxpibZevLEFUVPBHnoT5+FiA5kU0P9jZUQvN8vnmP+TRCLBsyN88OwInzqqjogaEnvoTVwzJxtM/8d65i09lFwZkchK8dJ/kcgtKEXWXR3aetlCWoPL+YlIPDjkIhKujjZwdeRNKIisGYdciIhEgj10K6A3GPHd8Qz8ejEPPu5KPD3MCz7uyhq/zuVbhdgWnY7s/FKE9GiGSY94QiLh8A5RY8FAtwI/xGRge3Q6AOBKSjGSUoux5QX/Go21F2v0eHPbdRQU6wEA26PT4Wwvx8g+XHWRqLHgkIsVOHnJ/DZzKZkaJNdwca3LyUWmML8n7q+Kb19HRMJgoFuBFs3Nh1eUNlI0d6nZCVRfD1tI73u3VHX7OiJqeAx0K/D0MC9T+NoqpJgz2gcOtrIavUZzFxvMCW8BO2XZW+YhP0dMGORZ57USUe1xHrqVMBqNSMnUws1ZDntlzcL8nzSlBhRr9JwiSdQI8aSolZBIJGjpUfOZLfdT2ki5dC5RI8XfTCIikWCgExGJBAOdiEgkBAn0H3/8EWFhYejUqRMSEhIqbRcTE4Phw4cjNDQUW7ZsacAKiYiaHkECvWPHjtiwYQOCg4MrbaPX67F8+XJs3boVUVFR2L9/P5KSkhqwSiKipkWQWS7t27evtk18fDxat24NX19fAEBYWBgOHz6MDh061Hd5RERNUqMdQ1er1fDy8jJtq1QqqNVqASsiImrc6q2HHhERgczMzHL7Fy1ahKFDh9bXYQEAGo0GiYmJ9XoMIiKhVHbhZL0F+rZt2x7o+SqVCunp6aZttVoNlcqyW6splUpeKVrHfrmQi6hTWbBXyvDEI57o2NJe6JKI6D6NdsilW7duuHHjBpKTk6HVahEVFYWQkBChy7JK567m452vbuJsUgF+uZiHV7ZeRW6BTuiyiOg+VfbQZ8+eXeWTN9E4vl8AABjPSURBVG/eXKuD/vzzz1ixYgWys7Pxr3/9C507d8ann34KtVqNN954A5988gnkcjmWLl2KmTNnQq/XY/z48fDz86vV8ejB/HIhz2y7WGPAmSv5CAlqJlBFRFSRKhfnOn36NAAgOjoamZmZGD16NAAgKioK7u7ueO211xqmyhri4lx164fYDGw9kGa274N/dUBAGweBKiKiilTZQ+/duzcA4L333kNkZKRpf0hICMaNG1e/lVGjMbK3O35LvIsL1wshkZRtM8yJGh+LTooWFxcjOTnZNCc8OTkZxcXF9VoYNR52Shn+b1YH3MoogZ1CCg9XhdAlEVEFLAr0V199FdOmTYOvry+MRiNSU1OxfPny+q6N6kh2filSMjXo2NL+gZa+beXJOxQRNWYW3+BCq9Xi2rVrAIB27dpBoWi8vTSOof/tx9NZ+M/eFOj0Rjg7yPDO9Hbwa8Eph0RiZFEPvbS0FDt37kRcXByAsrH1SZMmwcaGd61pzDSlBnxyIBU6fdnf7LuFevz3pzSsmlH90gtE1PRY9Pn77bffxsWLFzFlyhRMmTIFFy9exNtvv13PpdGDKizRo1hjMNuXebdUoGqIqL5Z1ENPSEjA3r17Tdv9+vUzTWGkxsvNyQbd2zni/LUC075Hu3PuOJFYWRToMpkMt27dQqtWrQCUzXKRyWp/o2FqOK9PbY1vj2XgRnoJevk7Ibxv8zp53WtpxbhwvRD+vvbw9+WYPFFjYFGgv/zyy3j66afNZrmsWrWqvmujOuBkJ8eMx3zq9DWj47KxLjIZ906nPzfSB+Me9qjTYxBRzXGWC9VYxP9LhDpHa9p2tJXh26UBkEgkAlZFRJzlIlLJGSU4fDYHDrYyDO/lBmeHultYs1RnfqK1VG+E0Qgwz4mExVkuInQjvRgLPvoL3xzLwGc/peGFTVegLTVU/0QLPd7ffBw+vJ87pFKmOZHQOMtFhA7GZUNT+vdIWmqWFnF/5aN/gEudvP4Tj6jQxssOCdcL0LGlPQZ2rZvXJaIHw1kuIlTR5f1Km5r1oC/cKMB/f0pDdr4OIT2a4akhKrNeeO9OzujdyfmBayWiusNZLiIU1scd0XHZyPnfTSgC2jigRwcni59fpNHj7e3XUVhSNkzz9RE1XB3lCO9XN1Meiah+cJaLSBUU63HyUh7sbWXo08kZcpnlPfSzSfl47dNrZvv6dnbGW0+3resyiagOWTz14cKFC0hJSYFer8fly5cBAGPGjKm3wujBONrJENrTrVbPbe1pC5kU0P/jPGpbb7s6qoyI6otFgf7SSy8hOTkZnTp1Mo2dSyQSBrpIuTnbYMGYlth6IA2FGj2C/Z0xYRAvHCJq7CwK9AsXLuDAgQO8cMSKDA92R0hQM2hKjXC04wlwoqbAonnofn5+uHPnTn3XQo2MjVzKMCdqQqrsoc+ePRsAUFhYiLCwMAQGBppdHbp58+b6rY6IiCxWZaA/++yzDVUHERE9oCoDvXfv3g1VBxERPaAqA33KlCnYsWMHgoKCzE6IGo1GSCQSnDlzpt4LJCIiy1h8YVFTwguLiMgaVdlDz83NrfLJrq6udVoMERHVXpWBPm7cOEgkElTUiZdIJDh8+HC9FUZERDVTZaAfOXKkoeogIqIHZNGFRUajEXv27MHGjRsBAKmpqYiPj6/XwoiIqGYsvmPRuXPnsH//fgCAg4MDli1bVq+FERFRzVgU6PHx8XjrrbegVCoBAC4uLigtLa3XwoiIqGYsWpxLLpdDr9eb5qJnZ2dDKrXob0GFfvzxR3z00Ue4evUqvvvuO3Tr1q3CdiEhIXBwcIBUKoVMJkNkZGStj0lEJHYWBfq0adMwb948ZGVlYe3atfjpp5+wcOHCWh+0Y8eO2LBhA956661q227fvh1ubrVb15uIyJpYFOijR49GQEAAfvvtNxiNRvznP/9B+/bta33QB3kuERFVzKJA/+677zBx4kSzIP7ggw+wZMmSeivsnhkzZkAikWDSpEmYNGmSRc/RaDRITEys58qIiIRR2ZXwFgV6dHQ0lEolRo8eDQBYtmwZNBpNlc+JiIhAZmZmuf2LFi3C0KFDLTksduzYAZVKhaysLEyfPh3t2rVDcHBwtc9TKpW89J+IrI5Fgb5hwwbMmTMHUqkUsbGxcHJywqpVq6p8zrZt2x64OJVKBQBwd3dHaGgo4uPjLQp0IiJrVOVUldzcXOTm5qKkpATvvPMOtm7dCgcHB8yfP7/adV4eVFFREQoKCkz//+WXX+Dn51evxyQiasqqXG0xJCTEtJbL/Wu6PMhaLj///DNWrFiB7OxsODs7o3Pnzvj000+hVqvxxhtv4JNPPkFycjLmzZsHANDr9Rg1ahTmzJlj0etztUUiskZcPpeISCSqHEM/efIk+vXrh+jo6AofHzZsWL0URURENVdloP/+++/o168fjh49WuHjDHQiosaj1kMuBw8exPDhw+u6njrBIRciska1XpDl3Xffrcs6iIjoAdU60EV4LpWIqEmrdaDfW3mRiIgahypPioaHh1f6WEWX9RMRkXCqDPTNmzc3VB1ERPSAqgz0Fi1aNFQdRET0gCxanCsoKKjcmLmTkxO6du2KV155Bb6+vvVSHBERWc6iQH/mmWfg5eWFUaNGAQCioqJw69YtBAQE4LXXXsMXX3xRr0USEVH1LJrlcuTIEUyePBmOjo5wdHTEpEmTcOLECYwcORJ5eXn1XSMREVnAokC3s7PDgQMHYDAYYDAYcODAASiVSgCcvkhE1FhYdOl/cnIyVq5cibNnzwIoG1N/9dVXoVKpcOHCBfTq1aveC60JXvpPRNaIy+cSEYmERUMu6enpmDdvHvr164d+/fphwYIFSE9Pr+/aiIioBiwK9FdffRUhISGIjY1FbGwsHn30Ubz66qv1XRsREdWARYGenZ2N8ePHQy6XQy6XY9y4ccjOzq7v2oiIqAYsCnRXV1fs2bMHer0eer0ee/bsgaura33XRkRENWDRSdGUlBSsWLEC586dg0QiQVBQEN588014e3s3RI01xpOiRGSNaj3LZdu2bYiIiKjjcuoGA73m7hbqkJKlQXsfOyjktV5VmYgEVOvf3G3bttVhGSSkQ2eyMe29S1i8KQkR7yciKbVI6JKIqBZ4xyIrp9UZ8PG+VGh1ZT/PnAIdPvsxrdL2pToDcgt0DVUeEdWARYtzVYSX/ItDUYkeBSV6s30ZOdoK2x45m4NN+1JQUKxHt7YOeP2pNnBxqPVbiIjqWJW/jRUtmwuU9c41Gk29FUUNx9XRBoHtHBB/rdC0b1Bg+RlM+cU6fLgrGZrSsp58wvVCfH1YjTmjuWY+UWNRZaDfW7uFxO31J9vg6yNq3FCXoFdHJ4wd6FGuTWqm1hTm91xPL26oEonIAvy8THB2kGN2eNU97XbetnBzkiM7/+/x814dneq7NCKqAc5PE4kSrQEXbxQiv6h+TljayKVYFtEWQR0c4eOuwKRHPDH+Yc96ORYR1Q5XWxSBizcK8fbn11FQrIfSRoLFE1pVOA5OROLGHroIbD2QioLispkqmlIjNu9Lgd4gur/TRFQNQQL9/fffx4gRIxAeHo558+bh7t27FbaLiYnB8OHDERoaii1btjRwlU3HnTzzaYa5hTqU6gwCVUNEQhEk0AcMGID9+/dj3759aNOmDT7++ONybfR6PZYvX46tW7ciKioK+/fvR1JSkgDVNn6DA5uZbffp5AxbhUygaohIKILMchk4cKDp/z169MBPP/1Urk18fDxat24NX19fAEBYWBgOHz6MDh06NFidTcX0Ed5wcZTjXFI+OrSww6RHVEKXREQCEHza4g8//IDHHnus3H61Wg0vLy/TtkqlQnx8fEOW1mTIZRI8MdgTTwzmrBMia1ZvgR4REYHMzMxy+xctWoShQ4cCADZt2gSZTIbRo0fX6bE1Gg0SExPr9DWJiBqLymbx1VugV7caY2RkJI4dO4Zt27ZVuLyASqUyu2+pWq2GSmXZUIJSqbSqaYtERIBAJ0VjYmKwdetWbNq0CXZ2dhW26datG27cuIHk5GRotVpERUUhJCSkgSslImo6BLmwKDQ0FFqt1nQbu+7du2P58uVQq9V444038MknnwAAjh8/jlWrVkGv12P8+PGYM2eORa9vbRcWEREBvFKUiEg0eKUoEZFIMNCJiESCgU5EJBIMdCIikWCgExGJBAOdiEgkGOhERCLBQCciEgkGOhGRSDDQiYhEQvD10Kl+/HIhF5En7gAAJgzyRL8uLuXa3FSX4LOf0pCWrcGAABdMHeIFmaz8ypdE1DQw0EUoKaUIq76+iXv3iX7nyxv4aEFHtPX+e2VLvd6IpduuISO3FACwMyMDChsppjzKux0RNVUcchGh3//MN4U5ABiMwO9/5Zu1uaEuMYX5PXF/VnyzbgAoLNFjzfe3MPXdi3jzv9eQmqmp05qJ6MEx0EXI11NZbl8rD/N9Xm4KKG3Mf/ytVbaVvuaWqFT8/EcOsu7qEPdXPlZ+faNOaiWiusNAF6H+XVwQ2rMZpBJAKgFGBLuhdydnszYOtjI8P7YlHG1lAAB/X3tMHepV0csBAM4lmffwr6WVIK9QV/fFE1GtcQxdhKRSCRZPaIVnR3gDkMDV0fzHrNUZsD06Hb9dykOnVvaYONgTge0czdqUaA347KdUxP2ZjzZetmjRXGk2ROPpagMnO1lDfDlEZCEGuoi5OtpUuP/rw2pExpbNgEnN0iI1S4NPFneCVPr3DJfPfkzFvt+yAABp2Vq08lTCv6U9/rxdBFUzBZZM9DVrT0TCY6CLyLmr+YhNyIOqmQJhfdzhYFtxDzruvhOkZaGuRUsPJTSlBtzKKMHp+06Q3srQ4KvXukAhl8JeKWWYEzVCDHSROHkpD8u/uPH39sU8rJ3rV2HbNipbXE0tNm1LJYDSRoILNwqw4osbuFukh+y+syvNnORwsZdznjpRI8aToiLx4+kss+3LyUW4llZcYdvHgt3Mtg1GIOpUFj7el4q7RXoAgN4AyP8X3s2c5Hhxgi/DnKiRYw9dJOyV5YdX7JUV/70uKTWU25eapUF6ttZsn8FgxOf/7gw3JxuGOVETwB66SDzxiCccbP/+cT7W2w1ebuXnowNAQBsHODuY/wHo38UFA7uZLw/Qr4sLPFwVDHOiJkJiNBqN1TdrWhITE9G5c2ehy2hw+UU6/PFXPlTNFOjc2qHKttfTivHVETVy8ksx5CE3jOztDk2pATuPqpFwvRD+Le3x5BBVpSdWiajxYaATEYkEh1yIiESCgU5EJBIMdCIikWCgExGJBAOdiEgkGOhERCLBQCciEglBLv1///33cfToUdjY2KBVq1Z499134ezsXK5dSEgIHBwcIJVKIZPJEBkZKUC1RERNgyA99AEDBmD//v3Yt28f2rRpg48//rjSttu3b8eePXsY5kRE1RAk0AcOHAi5vOzDQY8ePZCeni5EGUREoiL4GPoPP/yAQYMGVfr4jBkzMG7cOHzzzTcNWBURUdNTb2PoERERyMzMLLd/0aJFGDp0KABg06ZNkMlkGD16dIWvsWPHDqhUKmRlZWH69Olo164dgoODqz22RqNBYmLig30BRESNVGVrVQm2OFdkZCS++eYbbNu2DXZ2dtW237BhA+zt7TFjxoxq23Jxrto7f7UA6dka9PJ3hrtzxfckJaLGSZBZLjExMdi6dSu+/PLLSsO8qKgIBoMBjo6OKCoqwi+//IK5c+c2cKXWZd0PyTgYlw0AUNpI8e7MdujcqupleImo8RBkDH3FihUoLCzE9OnT8fjjj2Pp0qUAALVajeeeew4AkJWVhSeffBKjR4/GxIkTMXjw4CrH2unBqHO0iP4j27StKTXg+5g7AlZERDXF9dAJAJCcUYJZa/802/eQnyNWPtteoIqIqKYEn+VCjYOvpy16tHc0bUskwKi+zQWsiIhqijeJJpO3nm6L6LhsqHO06B/ggoA2HD8nakoY6GRiq5BidH/2yomaKg65WKnzVwvwQ2wGklKLhC6FiOoIe+hW6POf07HjiBpA2Vj5ixN8MeQhN4GrIqIHxR66ldHqDIiMzTBtG43AN8cyqngGETUVDHRrYwQM901U1d+/g4iaJAa6lVHYSDGqj7vZvnEPewhUDRHVJY6hW6HnwnwQ0NYRV1OL0aO9IwLbOVb/JCJq9BjoVkgikWBAgAsGBLgIXQoR1SEOuRARiQQDnYhIJBjoREQiwUAnIhIJBjoRkUgw0ImIRIKBTkQkEgx0IiKRYKATEYkEA52ISCQY6EREIsFAJyISCQY6EZFIMNCJiESCgU5EJBJcD93K3b5TgsgTd1BUYsBjvd3RvT1vdkHUVDHQrVhBsR4vbk7C3SI9ACD2Qi4+mNUBnVs7CFwZEdUGh1ysWNyfd01hDgAGA3DsfK6AFRHRg2CgW7FmTjYV7OOHNqKmioFuxQLbOeDhbn/fV7SNly3C+rgLWBERPQh2x6yYRCLBa0+2wbW0YhSV6NGltQOkUonQZRFRLQkW6OvWrcPhw4chlUrh7u6Od999FyqVqly7Xbt2YdOmTQCAOXPmYOzYsQ1dqui187YTugQiqgMSo9FoFOLABQUFcHQsmyL3+eefIykpCcuXLzdrk5ubi/Hjx+OHH36ARCLBuHHjEBkZCRcXl4pe0iQxMRGdO3eut9qJiBojwcbQ74U5ABQXF0MiKf9R/8SJExgwYABcXV3h4uKCAQMGIDY2tiHLbDLOXMnH9zEZSEotEroUIhKIoGPoa9euxe7du+Hk5ITPP/+83ONqtRpeXl6mbZVKBbVaXe3rajQaJCYm1mmtjdn+P3Q4ctEAAJAAeHKgDD3byYQtiojqTWUjEPUa6BEREcjMzCy3f9GiRRg6dCheeOEFvPDCC/j444/x5Zdf4vnnn6+T4yqVSqsZcinRGhD79QXTthHA8T/lmBrWSbiiiEgQ9Rro27Zts6hdeHg4Zs2aVS7QVSoVTp8+bdpWq9Xo3bt3XZYoAkbcfxbk9h0Nln9xHa9Mbg2FDWemElkLwX7bb9y4Yfr/4cOH0a5du3JtBg4ciBMnTiAvLw95eXk4ceIEBg4c2IBVNn62ChlG9i4/d/zkpbuIjssWoCIiEopgY+irV6/G9evXIZFI0KJFCyxbtgwAkJCQgJ07d2LlypVwdXXF3LlzMWHCBADAvHnz4OrqKlTJjdbscB8YYcS+k1lm+29llAhUEREJQbBpi/XJGqctZuaV4tkPElGq+/vH+c70tujZ0VnAqoioIfFKUZFo7mKDFRHtsOOIGkUaPUb1bc4wJ7IyDHQR6d7ekeuZE1kxToEgIhIJBjoRkUgw0ImIRIKBTkQkEgx0IiKRYKATEYkEA52ISCQY6EREIsFAJyISCQY6EZFIiPLSf2u7YxERWRe5XA4/P79y+0W52iIRkTXikAsRkUgw0ImIRIKBTkQkEgx0IiKRYKATEYkEA90KxMTEYPjw4QgNDcWWLVvKPa7VarFo0SKEhoZi4sSJuH37NgAgJycH06ZNQ1BQEJYvX97QZZNIVPf++/333zF27Fh06dIFP/30kwAVigcDXeT0ej2WL1+OrVu3IioqCvv370dSUpJZm++++w7Ozs74+eefERERgQ8++AAAoFQqsXDhQrz88stClE4iYMn7z9vbG++++y5GjRolUJXiwUAXufj4eLRu3Rq+vr5QKBQICwvD4cOHzdocOXIEY8eOBQAMHz4cJ0+ehNFohL29PXr16gWlUilE6SQClrz/WrZsiU6dOkEqZRw9KH4HRU6tVsPLy8u0rVKpoFary7Xx9vYGUHYFmpOTE3Jychq0ThInS95/VHcY6EREIsFAFzmVSoX09HTTtlqthkqlKtcmLS0NAKDT6ZCfn49mzZo1aJ0kTpa8/6juMNBFrlu3brhx4waSk5Oh1WoRFRWFkJAQszYhISHYtWsXAODgwYPo27cvJBKJEOWSyFjy/qO6w8W5rMDx48exatUq6PV6jB8/HnPmzMH69evRtWtXDBkyBBqNBi+99BISExPh4uKCtWvXwtfXF0BZ2BcUFKC0tBROTk747LPP0KFDB4G/ImpKqnv/xcfHY/78+bh79y6USiWaN2+OqKgooctukhjoREQiwSEXIiKRYKATEYkEA52ISCQY6EREIsFAJyISCQY6EZFIMNCJiERCLnQBRE3Vxo0bsXfvXri5ucHb2xsBAQGYMWOG0GWRFWOgE9VCfHw8oqOjsXfvXpSWlmLcuHEICAgQuiyycgx0olo4c+YMhgwZAqVSCaVSiUcffVTokog4hk5EJBYMdKJaeOihh3D06FFoNBoUFhbi2LFjQpdExCEXotoIDAxESEgIRo8eDXd3d3Ts2BFOTk5Cl0VWjqstEtVSYWEhHBwcUFxcjKeeegorVqzgiVESFHvoRLW0dOlSJCUlQaPRYOzYsQxzEhx76EREIsGTokREIsFAJyISCQY6EZFIMNCJiESCgU5EJBIMdCIikfj/sQ85JrtKAyUAAAAASUVORK5CYII=\n", "text/plain": [ "
" ] @@ -876,7 +875,7 @@ }, { "cell_type": "code", - "execution_count": 206, + "execution_count": 299, "metadata": {}, "outputs": [ { @@ -1035,7 +1034,7 @@ "4 NaN NaN NaN -39705.898393 -2.030944 " ] }, - "execution_count": 206, + "execution_count": 299, "metadata": {}, "output_type": "execute_result" } @@ -1048,22 +1047,22 @@ }, { "cell_type": "code", - "execution_count": 207, + "execution_count": 300, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 207, + "execution_count": 300, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1079,22 +1078,22 @@ }, { "cell_type": "code", - "execution_count": 208, + "execution_count": 301, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 208, + "execution_count": 301, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1110,24 +1109,24 @@ }, { "cell_type": "code", - "execution_count": 210, + "execution_count": 309, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 210, + "execution_count": 309, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -1135,28 +1134,31 @@ } ], "source": [ - "df_pplacer = df2[df2[\"software\"] == \"pplacer\"]\n", + "df_pplacer = df2[df2[\"software\"] == \"pplacer\"].copy()\n", + "df_pplacer[\"max-strikes\"] = df_pplacer[\"max-strikes\"].astype(int)\n", + "df_pplacer[\"max-pitches\"] = df_pplacer[\"max-pitches\"].astype(int)\n", + "df_pplacer[\"strike-box\"] = df_pplacer[\"strike-box\"].astype(int)\n", "sns.catplot(data=df_pplacer, x=\"max-strikes\", row=\"max-pitches\", hue=\"strike-box\", y=\"LL difference\", dodge=True, palette=\"muted\")" ] }, { "cell_type": "code", - "execution_count": 211, + "execution_count": 310, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 211, + "execution_count": 310, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1166,7 +1168,8 @@ } ], "source": [ - "df_rappas = df2[df2[\"software\"] == \"rappas\"]\n", + "df_rappas = df2[df2[\"software\"] == \"rappas\"].copy()\n", + "df_rappas[\"k\"] = df_rappas[\"k\"].astype(int)\n", "sns.catplot(data=df_rappas, x=\"k\", hue=\"omega\", y=\"LL difference\", dodge=True, palette=\"muted\")" ] }, @@ -1181,14 +1184,14 @@ }, { "cell_type": "code", - "execution_count": 260, + "execution_count": 338, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -1198,9 +1201,9 @@ "source": [ "num_rows = 2\n", "num_cols = 2\n", - "scale = 6\n", - "fig, axes = plt.subplots(num_rows, num_cols, sharey=True, figsize=(scale * 1.2 * num_cols, scale * num_rows))\n", - "plt.subplots_adjust(left=0.1, top=0.9)\n", + "scale = 7\n", + "fig, axes = plt.subplots(num_rows, num_cols, sharey=True, figsize=(scale * num_cols, scale * 0.9 * num_rows))\n", + "plt.subplots_adjust(left=0.1, top=0.93)\n", "fig.suptitle(\"LogLikelihood difference (vs. EPA, g=0.1)\", fontsize=16)\n", "\n", "palette = sns.color_palette(\"deep\", 8)\n", @@ -1215,7 +1218,8 @@ " ax = sns.stripplot(data=df_pplacer, x=\"max-strikes\", hue=\"strike-box\", y=\"LL difference\", ax=axes[i, j],\n", " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[0:2])\n", " ylabel = 'LL Difference'\n", - " title = f'PPLACER'\n", + " title = f'PPLACER | max-pitches=40'\n", + "\n", " elif order[k] == \"epa\":\n", " ax = sns.stripplot(data=df_epa, x=\"g\", y=\"LL difference\", ax=axes[i, j], \n", " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[2:3])\n", @@ -1229,13 +1233,15 @@ " elif order[k] == \"epang\":\n", " ax = sns.stripplot(data=df_epang, x=\"g\", y=\"LL difference\", ax=axes[i, j], \n", " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[5:6])\n", - " title = f'EPA-NG'\n", + " ylabel = ''\n", + " title = f'EPA-NG | default heuristic'\n", " \n", " ax.set(ylabel=ylabel, title=title)\n", " k += 1\n", " \n", "fig.savefig(\"ll_diff.png\")\n", - "fig.savefig(\"ll_diff.svg\")" + "fig.savefig(\"ll_diff.svg\")\n", + "fig.savefig(\"ll_diff.pdf\")" ] }, { From 06b4b4c463df24c135aca8bad212cbc6950ee717 Mon Sep 17 00:00:00 2001 From: Nikolai Romashchenko Date: Sat, 2 May 2020 19:07:57 +0200 Subject: [PATCH 13/43] LL_difference.ipynb: changed EPA's parameter name --- .../ll_difference.ipynb | 412 +++++------------- 1 file changed, 121 insertions(+), 291 deletions(-) diff --git a/examples/6_placement_likelihood/ll_difference.ipynb b/examples/6_placement_likelihood/ll_difference.ipynb index f828822..42a4f9b 100644 --- a/examples/6_placement_likelihood/ll_difference.ipynb +++ b/examples/6_placement_likelihood/ll_difference.ipynb @@ -20,7 +20,7 @@ }, { "cell_type": "code", - "execution_count": 312, + "execution_count": 351, "metadata": {}, "outputs": [ { @@ -167,7 +167,7 @@ "4 NaN " ] }, - "execution_count": 312, + "execution_count": 351, "metadata": {}, "output_type": "execute_result" } @@ -192,7 +192,7 @@ }, { "cell_type": "code", - "execution_count": 292, + "execution_count": 349, "metadata": {}, "outputs": [ { @@ -218,7 +218,7 @@ }, { "cell_type": "code", - "execution_count": 293, + "execution_count": 361, "metadata": {}, "outputs": [ { @@ -242,142 +242,83 @@ " \n", " \n", " \n", - " pruning\n", - " g\n", + " software\n", + " G\n", " query\n", - " length\n", " LL\n", - " software\n", - " max-strikes\n", - " strike-box\n", - " max-pitches\n", - " red\n", - " ar\n", - " k\n", - " omega\n", " \n", " \n", " \n", " \n", " 0\n", - " 0\n", + " epa\n", " 0.01\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39708.145904\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " \n", " \n", " 1\n", - " 0\n", + " epa\n", " 0.10\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39705.898393\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " \n", " \n", " 2\n", - " 0\n", + " epa\n", " 0.01\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39779.279980\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " \n", " \n", " 3\n", - " 0\n", + " epa\n", " 0.10\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39778.547272\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " \n", " \n", " 4\n", - " 0\n", + " epa\n", " 0.01\n", " 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282\n", - " 0\n", " -39791.397634\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " \n", " \n", "\n", "" ], "text/plain": [ - " pruning g query length \\\n", - "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", - "\n", - " LL software max-strikes strike-box max-pitches red ar k \\\n", - "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", - "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", - "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", - "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", - "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + " software G query \\\n", + "0 epa 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "1 epa 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "2 epa 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "3 epa 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "4 epa 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 \n", "\n", - " omega \n", - "0 NaN \n", - "1 NaN \n", - "2 NaN \n", - "3 NaN \n", - "4 NaN " + " LL \n", + "0 -39708.145904 \n", + "1 -39705.898393 \n", + "2 -39779.279980 \n", + "3 -39778.547272 \n", + "4 -39791.397634 " ] }, - "execution_count": 293, + "execution_count": 361, "metadata": {}, "output_type": "execute_result" } ], "source": [ - "df_epa = df[df[\"software\"] == \"epa\"]\n", + "epa_columns = [\"software\", \"g\", \"query\", \"LL\"]\n", + "df_epa = df[df[\"software\"] == \"epa\"][epa_columns].copy()\n", + "df_epa.rename(columns={'g': 'G'}, inplace=True)\n", "df_epa.head()" ] }, { "cell_type": "code", - "execution_count": 294, + "execution_count": 355, "metadata": {}, "outputs": [ { @@ -386,13 +327,13 @@ "Text(0.5, 0.98, 'Log Likelihood of extended trees')" ] }, - "execution_count": 294, + "execution_count": 355, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -402,8 +343,8 @@ } ], "source": [ - "ax = sns.catplot(data=df_epa, x=\"g\", y=\"LL\", palette=\"muted\")\n", - "ax.set_axis_labels(\"g\", \"LogLikelihood\")\n", + "ax = sns.catplot(data=df_epa, x=\"G\", y=\"LL\", palette=\"muted\")\n", + "ax.set_axis_labels(\"G\", \"LogLikelihood\")\n", "plt.subplots_adjust(left=0.1, top=0.9)\n", "ax.fig.suptitle(\"Log Likelihood of extended trees\")" ] @@ -417,7 +358,7 @@ }, { "cell_type": "code", - "execution_count": 295, + "execution_count": 356, "metadata": {}, "outputs": [ { @@ -484,20 +425,20 @@ "13 0dbcba127f79ff6ec9aa56486685d702f1da0ca7-4235956 -39793.188953" ] }, - "execution_count": 295, + "execution_count": 356, "metadata": {}, "output_type": "execute_result" } ], "source": [ - "baseline = df_epa[df_epa[\"g\"] == 0.1][[\"query\", \"LL\"]].copy()\n", + "baseline = df_epa[df_epa[\"G\"] == 0.1][[\"query\", \"LL\"]].copy()\n", "baseline.sort_values(\"query\", inplace=True)\n", "baseline.head()" ] }, { "cell_type": "code", - "execution_count": 296, + "execution_count": 357, "metadata": {}, "outputs": [ { @@ -521,106 +462,52 @@ " \n", " \n", " \n", - " pruning\n", - " g\n", + " software\n", + " G\n", " query\n", - " length\n", " LL\n", - " software\n", - " max-strikes\n", - " strike-box\n", - " max-pitches\n", - " red\n", - " ar\n", - " k\n", - " omega\n", " LL_baseline\n", " \n", " \n", " \n", " \n", " 0\n", - " 0\n", + " epa\n", " 0.01\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39708.145904\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39705.898393\n", " \n", " \n", " 1\n", - " 0\n", + " epa\n", " 0.10\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39705.898393\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39705.898393\n", " \n", " \n", " 2\n", - " 0\n", + " epa\n", " 0.01\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39779.279980\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39778.547272\n", " \n", " \n", " 3\n", - " 0\n", + " epa\n", " 0.10\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39778.547272\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39778.547272\n", " \n", " \n", " 4\n", - " 0\n", + " epa\n", " 0.01\n", " 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282\n", - " 0\n", " -39791.397634\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39791.397634\n", " \n", " \n", @@ -628,29 +515,22 @@ "" ], "text/plain": [ - " pruning g query length \\\n", - "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", - "\n", - " LL software max-strikes strike-box max-pitches red ar k \\\n", - "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", - "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", - "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", - "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", - "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", + " software G query \\\n", + "0 epa 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "1 epa 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "2 epa 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "3 epa 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "4 epa 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 \n", "\n", - " omega LL_baseline \n", - "0 NaN -39705.898393 \n", - "1 NaN -39705.898393 \n", - "2 NaN -39778.547272 \n", - "3 NaN -39778.547272 \n", - "4 NaN -39791.397634 " + " LL LL_baseline \n", + "0 -39708.145904 -39705.898393 \n", + "1 -39705.898393 -39705.898393 \n", + "2 -39779.279980 -39778.547272 \n", + "3 -39778.547272 -39778.547272 \n", + "4 -39791.397634 -39791.397634 " ] }, - "execution_count": 296, + "execution_count": 357, "metadata": {}, "output_type": "execute_result" } @@ -662,7 +542,7 @@ }, { "cell_type": "code", - "execution_count": 297, + "execution_count": 358, "metadata": {}, "outputs": [ { @@ -686,19 +566,10 @@ " \n", " \n", " \n", - " pruning\n", - " g\n", + " software\n", + " G\n", " query\n", - " length\n", " LL\n", - " software\n", - " max-strikes\n", - " strike-box\n", - " max-pitches\n", - " red\n", - " ar\n", - " k\n", - " omega\n", " LL_baseline\n", " LL difference\n", " \n", @@ -706,91 +577,46 @@ " \n", " \n", " 0\n", - " 0\n", + " epa\n", " 0.01\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39708.145904\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39705.898393\n", " -2.247511\n", " \n", " \n", " 1\n", - " 0\n", + " epa\n", " 0.10\n", " 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050\n", - " 0\n", " -39705.898393\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39705.898393\n", " 0.000000\n", " \n", " \n", " 2\n", - " 0\n", + " epa\n", " 0.01\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39779.279980\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39778.547272\n", " -0.732708\n", " \n", " \n", " 3\n", - " 0\n", + " epa\n", " 0.10\n", " ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769\n", - " 0\n", " -39778.547272\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39778.547272\n", " 0.000000\n", " \n", " \n", " 4\n", - " 0\n", + " epa\n", " 0.01\n", " 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282\n", - " 0\n", " -39791.397634\n", - " epa\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", - " NaN\n", " -39791.397634\n", " 0.000000\n", " \n", @@ -799,29 +625,22 @@ "" ], "text/plain": [ - " pruning g query length \\\n", - "0 0 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "1 0 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 0 \n", - "2 0 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "3 0 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 0 \n", - "4 0 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 0 \n", + " software G query \\\n", + "0 epa 0.01 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "1 epa 0.10 8f43808640b0e835dfd588c7c30a955c78cf2252-17084050 \n", + "2 epa 0.01 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "3 epa 0.10 ea4a29b8b59fb031b08cf1a7575658013dd45baa-9896769 \n", + "4 epa 0.01 64463bb954837427bbaa8bff6de57ba4bbaccb8b-9717282 \n", "\n", - " LL software max-strikes strike-box max-pitches red ar k \\\n", - "0 -39708.145904 epa NaN NaN NaN NaN NaN NaN \n", - "1 -39705.898393 epa NaN NaN NaN NaN NaN NaN \n", - "2 -39779.279980 epa NaN NaN NaN NaN NaN NaN \n", - "3 -39778.547272 epa NaN NaN NaN NaN NaN NaN \n", - "4 -39791.397634 epa NaN NaN NaN NaN NaN NaN \n", - "\n", - " omega LL_baseline LL difference \n", - "0 NaN -39705.898393 -2.247511 \n", - "1 NaN -39705.898393 0.000000 \n", - "2 NaN -39778.547272 -0.732708 \n", - "3 NaN -39778.547272 0.000000 \n", - "4 NaN -39791.397634 0.000000 " + " LL LL_baseline LL difference \n", + "0 -39708.145904 -39705.898393 -2.247511 \n", + "1 -39705.898393 -39705.898393 0.000000 \n", + "2 -39779.279980 -39778.547272 -0.732708 \n", + "3 -39778.547272 -39778.547272 0.000000 \n", + "4 -39791.397634 -39791.397634 0.000000 " ] }, - "execution_count": 297, + "execution_count": 358, "metadata": {}, "output_type": "execute_result" } @@ -833,22 +652,22 @@ }, { "cell_type": "code", - "execution_count": 298, + "execution_count": 360, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "Text(0.5, 0.98, 'Log Likelihood difference (vs. EPA, g=0.1)')" + "Text(0.5, 0.98, 'Log Likelihood difference (vs. EPA, G=0.1)')" ] }, - "execution_count": 298, + "execution_count": 360, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -858,10 +677,10 @@ } ], "source": [ - "ax = sns.catplot(data=df_epa, x=\"g\", y=\"LL difference\", palette=\"muted\")\n", - "ax.set_axis_labels(\"g\", \"LogLikelihood\")\n", + "ax = sns.catplot(data=df_epa, x=\"G\", y=\"LL difference\", palette=\"muted\")\n", + "ax.set_axis_labels(\"G\", \"LogLikelihood\")\n", "plt.subplots_adjust(left=0.1, top=0.9)\n", - "ax.fig.suptitle(\"Log Likelihood difference (vs. EPA, g=0.1)\")" + "ax.fig.suptitle(\"Log Likelihood difference (vs. EPA, G=0.1)\")" ] }, { @@ -875,7 +694,7 @@ }, { "cell_type": "code", - "execution_count": 299, + "execution_count": 365, "metadata": {}, "outputs": [ { @@ -1034,7 +853,7 @@ "4 NaN NaN NaN -39705.898393 -2.030944 " ] }, - "execution_count": 299, + "execution_count": 365, "metadata": {}, "output_type": "execute_result" } @@ -1047,22 +866,22 @@ }, { "cell_type": "code", - "execution_count": 300, + "execution_count": 377, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 300, + "execution_count": 377, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAFgCAYAAACFYaNMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deXhU5d0+8Hv2bJPJQjJJyAIhBMMiUAFREDAQVtlB1FYNiyhgNbxv8e1baf1Bq9TLBXy1VRErihWrhqWQKqUBSUEsgkgQRoRAICyZkH2f/fdHYGAMYUKSmSdz5v5cl1c9J2fOfGtm7pz5znOeR+ZwOBwgIiKvk4sugIjIXzGAiYgEYQATEQnCACYiEoQBTEQkiCQD+OTJk6JLICJyS5IBbLVaRZdAROSWJAOYiMgXMICJiARhABMRCcIAJiIShAFMRCQIA5iISBAGMBGRIAxgIiJBGMBERIIwgImIBFGKLoA6p9oGG/Ydq4RGKcddfXTQqPi3mqijMYCpmfIaC5564yTKqi0AgOTYAKxe1BNqhjBRh+I7SuJqG2zYuu8yNu42wlhhbtVjdhwsd4YvAJy+1IivDdWeKpHIb/EKWMJMFjuy/nwSF0pNAIDP9pTg/55MRdcumps+zmptvk6r1ca1W4k6Gq+AJezAD9XO8AWAepMdOw6WuX1cxh3hCA649tLQh6sxtHeoR2ok8me8ApYwpULWfJ+8+b6fionQ4E9P9ULutxVQq2TIuCMCQRqFJ0ok8msMYAkblKpFSlwgTl1sAADogpUYPySyVY/Vh6vx0Gi9J8sj8nsyh8MhueaewWBAWlqa6DI6BZPFjn3fV6HBbMPwvmHQBfNvLlFnwXejxGlUcqQPDBddBhHdAL+EIyIShAFMRCQIA5iISBAGMBGRIPwSzo/U1Fvx1fEqBGkUGJoWCpWSf3+JRBIawHl5eXj++edht9sxe/ZsLFy40OXnZrMZzzzzDI4dO4awsDCsXr0a8fHxgqr1bSWVZmT96SQqaq0AgLTEILz0eAoUrbgxg4g8Q1gA22w2rFy5Eu+99x70ej1mzZqF9PR0pKSkOI/59NNPERoaip07dyInJwcvv/wy1qxZI6rkTiP/dC1WfHAG9SY7lHIgKECBBpMdVpsD8isXtTb7zc9hOFeP+57Nv+HPFHJAKQfMVsABQAZAqZTBdmU+iOTYADw/vwdCg/gBypc5HHZUvfU4HBUXxRQgVyJ4zgqouw8Q8/ydgLDPoPn5+UhKSkJCQgLUajUmTZqE3Nxcl2N27dqF6dOnAwDGjRuH/fv3Q4L3jdwSm92B37xbgHpTU8Ja7UB1vQ0WmwMONAWvu/B1/xyA6Ur4Ak3/a7E6YHcAdgdw6mIjfvPu6fY9CQlXl/ueuPAFALsVdRuXw95QI64GwYQFsNFoRExMjHNbr9fDaDQ2OyY2NhYAoFQqodVqUVFR4dU6O5uCi/XtDtiOUFjcILoEaifLdztElwDAAWvRMdFFCCPJz5AmkwkGg0F0GR7RaOkcnwCCNZDsf2N/odPGILCsQGgNDgCFVRbYJP5aamlqBGEBrNfrUVxc7Nw2Go3Q6/XNjrl06RJiYmJgtVpRU1OD8HD3t9VqNBpJzwVx/8WL+GTPZWHPr1IAy3+RjLRkrbAaqP3s3Z5H1ZoHAYFtvYAh05E6+B5hzy+asADu168fCgsLUVRUBL1ej5ycHLzyyisux6Snp2Pz5s0YOHAgduzYgaFDh0Im47f2c8fH4YFRehw/V4e0xCB8sLMYW79yned35j1dMGFwJMpqLGgwWSFXyFFZbUFFrRX9e4Sg0WKH4UwNSmqsGNA9BDsPV+LQiVpEhSkxd3ws+nQLQZBGjhNFDYgKU0IbqESjyY6aRitS4oL4e5AAeaAW4f+7HY0Fh2A+dRAqfQ84NEGwlpyC7UIBoA2HJjYFii6JgMUEuTYMluKzsBtPQt5nDGQlpyBL7AvrwRzYdXoog0KhTkyDTBkI64UfoIq/Dday87CWFEIerAOUAZApFDAXHYMiMhGBvYZCpg4Q/Z9BKKGzoe3ZswcvvPACbDYbZs6ciUWLFuG1115D3759MXr0aJhMJixbtgwGgwE6nQ6rV69GQkKC2/P622xonx8ow/9tPu+yb8Wj3THkNk6iTtSZcTpKCTBZ7Ph/75/BdwW1AIAR/XT4nweSIL/FMb7fnKjGXz6/hPJaC8b8LALzxsdynDCRBzGAJaSwuAFKhRzxUTdf8+1GquuseOTF4zBd9yXfE5PjMPXuqI4skYiuw3tRJaRbTGCbwhcATpyvdwlfADh6uq4jyiKiFjCACQDQIzaw2RpyqQlBgqoh8g8MYAIARISq8F+zEhCuVUIhB0YPDMe0YV1El0UkaewBkwuHwwG7HVDcYEVlIupYkrwTjtpOJpNBwRXoibyCLQgiIkEYwEREgrAFIQGHfqzG2zkXUVplwb39w/HE5DiudkHkAxjAPq6u0YY//PUsGs1Nc1T+40AZosNVmDNK7+aRRCQaL5N8XMHFBmf4XvX9Gd5AQeQLGMA+rntsANQq1yFjbb0bjoi8iwHs47SBSgxKdZ2X9+CP1X6/dBORL2AAS8DFUrPL9vnLZhgrzC0cTUSdBQNYAhKjXVsOIYEKRGhVgqohotZiAEvAvAlx6B7TtLKANlCBrBnxUKv4qyXq7DgXhISUVJoRFqKEupVjgOsabVDIZQhQux7vcDi45BCRF3AcsIREh6lvuP/7M7XY9V0FwoKVmHJ3FwQHKLA6uwh7jlRCpZRjzr3RePBePQxn67BmUxGKLptwR08t/nt2IsJC+BIh8hS+uyTuSEEtfvNuAexXPufkHa3C1Lsjsfu7SgBNyxl98M9i3NEzBKs2nsXlKgsA4OCPNVj3j4v41f2Jokonkjw2CiVux8EyZ/gCwIVSEw79WNvsuKNn6pzhe9UPRfWeLo/IrzGAJeLkhXqs+8dFbPr3ZdQ12pz7QwKbzy15e49gl22lQoZhfXWICXdtYfTt1nTcuZJGfH6gDAUXGzxQOZH/YgtCAvJP1+J/3y2A/codyV8eqcBrS3pCJpNhxvAo7Pu+CuU1VgDAyNvDMGN4NBRyGXL+U4YgtQIPjdYjJlyDZ3+ehDe2XMDZkkYM7qXF/ImxyP22HK98VoSrX9VyoU6ijsNREBLwx41nsSe/0mXfq0+kIC2p6Qq20WzDoZO1CA9RondS8I1O0aK5LxlQXH7tpg5toAJ/+20fjpIg6gC8ApYAjbp5J+n6fQFqBYb10bXp3CaL60Q/ZqsddgfAFYuI2o89YAmYMTwKIQHXer3D+uiQHBvYIeeePNR1Yc6JQyKhkDN9iToCWxASUVlrxYEfqhEZqsTAFC3kHRiS+76vxNEzdUiND8Ko/mEdem4if8YWhESEhSgxdlCER849uFcoii6bsCe/EmXVFkwfHgUlexBE7cYAJrfe2HoeOw9VAAAO/FCNkkozlkyNF1wVke9jD5huyuFwOO+au2rX4QpB1RBJCwOYbkomkyFC6/pBKTKUU10SdQQGMLm1cFIc1Mqmnq9GJceCiXGCKyKSBvaAya1hfcPwQfcQnL7UgJS4QGiD+LIh6gh8JxFKqyz4LK8El6ssGNU/DPf0C2t2jC64aXgbEXUcBrCfs9kd+PW6AlwoNQEAvjpWhf95wIFR/cMFV0YkfewB+7kfz9c7w/eq3RzlQOQVQq6AKysrsXTpUly4cAFdu3bFmjVroNM1n6sgLS0NqampAIDY2Fi89dZb3i5V8iK0KshkwPX3Q0ZwlAORVwi5Al67di3uuusu/POf/8Rdd92FtWvX3vC4gIAAbN26FVu3bmX4eog+XI1ZI65NLxkdpsKcUdECKyLyH0KugHNzc7FhwwYAwLRp0/Dwww9j2bJlIkrxaY1mG0wWB3TB7fs1zhsfh/GDIlFabUFaYhBUrVzUk4jaR0gAl5WVITq66SorKioKZWVlNzzOZDJhxowZUCqVWLhwIcaMGdOq85tMJhgMhg6rtzPa9b0N/8y3wWIF+iTI8It7lM6xum2lAnDqZMfUR0TXtDQ5mMcCODMzE6Wlpc32Z2VluWzLZLIWJ/fevXs39Ho9ioqK8OijjyI1NRWJie4XidRoNJKeDe1cSSO2f3vCuf19kQMnK7pg1gi2Doh8iccCeP369S3+LDIyEiUlJYiOjkZJSQkiIm48i5derwcAJCQkYMiQITh+/HirAljqzhobW7WvJcYKM7Z/XQqzxYHxgyPQvYPmDiaiWyOk2Zeeno4tW7YAALZs2YLRo0c3O6aqqgpmc9NSOOXl5fj222+RkpLi1To7q37dg6FRuX5qGNSrdTdJ1DRYkfXnk/gs7zL+vr8UWX8+iXMlrQ9vIuo4QgJ44cKF2LdvH8aOHYuvvvoKCxcuBAAcPXoUzz77LACgoKAAM2fOxJQpU/Doo4/iscceYwBfERaiwsrMZPTtFowkfQAWTorDyNubbpyoa7Rh1+EKfHOiGjZ787n2vz5ejcpaq3PbbHVwdjMiQbgihoSUVJqR9eeTqLiyAvLAlBA8Py/Zpce+/3gVVm4odHnc/Amx7B8TCcDxRhKybX+pM3wB4PCpWuSfrnM5ZnCvUPTtfm1l5K5dNBjnoZU0iOjmOBeEhDSa7S3uq2u0oeBiA5L0AXhxQQ8cOV0Lk8WOn/XUQs1xv0RCMIAlZNygCOw4WA6LtamrFBepxsCeIThSUIsVG86gwWSHSinDr2YnYsTtzWc8IyLvYgBLSErXIKxZ3BO5hysQEqjAhMGRUCvleOcfF9FgaroStlgdeHv7BdzTT9fi+Gsi8g4GsMQkxwYi+SfjesuqLS7blbVWWGyOdt85R0Ttw+afHxjV37XdMLyvjn1fok6AV8B+YMGEOESGqnCkoBY944Nw/8go9w8iIo/jOGCJsNkdUMjZUiDyJbwC9nGVtVa88uk5HDpZg/guGjw9IwF9ugW7fyARCcdGoI979/OLOPhjDRwOoOiyCas2noXNJrkPNUSSxAD2cT8U1btsl1VbUPqTUQ9E1DmxBeHj+nYLxvnL1xbVjAlXI0p3a2u6HfihGt+cqEZidADGDY7gCAkiL2EA+7j5E+LQYLLjwA/VSNQH4MlpXSG/hS/jvvimDK9tOu/cPlJQi+W/6OaBSonopxjAPsrhcOBciQkRWiV+/WBSm8+T87XrclBfHa9CRY0F4VqujEzkaQxgH3S5yozfvncGZ42NUCllmDc+FtOGtW1sb6DGtd2gkMugVrEFQeQNfKf5oI9yjc4liCxWB979/BIqatr2xdtDo/VQXXdL8qwRUQgOUHRInUR0c7wC9kGXys0u21abAyWVbWsbDOihxV9+lYbDp2qQpA9AanxQR5VJRG7wCtgHDeujc9mOCVcjpWvbF9bsolMh444Ihi+Rl/EK2AfdNzQSdrsDeUcroQ9X4xejY3gbMpEP4lwQRESCsAVBRCQIA5iISBD2gP1co9mGXd9VorLGihG36xAfFdDsmFMX67H7u0rogpWYMCQC2kC+bIg6At9Jfsxud+DX607jxJUJff72pREvP5GCnl2vjYYwnKvDM2sLYL0yw9quwxX40y9ToVDwSz+i9mILwo8dP1vnDF8AMFsd2La/1OWYzw+UOcMXAM4aG5F/ptZrNRJJGQPYj91o0p6fDmcLUDd/iQTwVmWiDsF3kh9LSwxC/+QQ53agRo6pd3dxOWba3VEIDb52a/LgXlqkJXHFDaKOwHHAfs5itWP/8WpU1Fpwdx8donTqZsfUNFhxwFANXbASP+upvaXpLomoZQxgIiJB2IIgIhKEAUxEJAgDmIhIEAYwEZEgDGAiIkGEBPDnn3+OSZMm4bbbbsPRo0dbPC4vLw/jxo1DRkYG1q5d68UKiYg8T0gAp6am4vXXX8fgwYNbPMZms2HlypVYt24dcnJysH37dpw6dcqLVRIReZaQyXh69Ojh9pj8/HwkJSUhISEBADBp0iTk5uYiJSXF0+UREXlFp50NzWg0IiYmxrmt1+uRn5/fqseaTCYYDAZPlUZEdEtaujHMYwGcmZmJ0tLSZvuzsrIwZswYTz0tAECj0fBOuHaorLVALpMhNLjT/n0mkgSPvcPWr1/frsfr9XoUFxc7t41GI/R6fTuropux2Rx49bMi7D5SAZkMmHRnJBZN7gqZjHM/EHlCpx2G1q9fPxQWFqKoqAhmsxk5OTlIT08XXZak7cmvxK7vKuBwAHY7sG1/Gb49ybl/iTxFSADv3LkTI0aMwOHDh/H4449j/vz5AJquch977DEAgFKpxO9+9zssWLAAEydOxIQJE9CzZ08R5fqNosuNrdpHRB2Ds6GRk+FsHf777VO4+opQyIG3snrdcJ04Imo/fstCTmlJwfifBxKxZW8plAoZZo+MZvgSeVCrroAPHjyIs2fPYubMmSgvL0ddXZ1zfG5nxCtgIvIFbnvAb7zxBtatW+e8FdhisWDZsmUeL4y8w2K1w26XXBeKyCe4bUHs3LkTW7ZswfTp0wE0DQ+rq6vzeGHkWWaLHWs2FWFPfiVCg5RYMCEWo38WIbosIr/i9gpYpVJBJpM5x4LW19e7eQT5gi1flWL3d5Ww24HKWitWZxehtMoiuiwiv+L2CnjChAn43e9+h+rqanzyySfIzs7G/fff743ayINOnnf9Q2qzA6cvNaCLTiWoIiL/4zaA58+fj3379iE4OBhnzpzBU089hWHDhnmjNvKg25NDsPf7Kue2RiXDbQlBAisi8j9uA7ioqAiDBg1yhm5jYyPOnz+P+Ph4jxdHnjPxzkiUVJqR+20FwkKUmDc+lnM/EHmZ22FoM2bMwMcffwy1Wg0AMJvNePDBB5Gdne2VAtuCw9BuTWWtBR/sLEZhcSPuSNVizig9lArO/0DkaW4veWw2mzN8AUCtVsNi4Zc1UvL7Dwtx/GxTT9hwrh4mix3zxscJropI+tyOgoiIiEBubq5z+1//+hfCw8M9WhR5T2WtxRm+V+27rjdMRJ7j9gp4xYoV+NWvfoXf//73cDgciI2NxYsvvuiN2sgLggMU0AYqUNNgc+6Li9Q4/72y1oLNe0tRVm3ByP5hGNwrVESZRJLU6sl4rt58ERwc7NGCOgJ7wLdmT34F1mSfR6PZjiidCisyu6N7TCBsdgcWvXYCRSUm57HPPdwNQ3vrBFZLJB1ur4DNZjN27NiBCxcuwGq1Ovc/+eSTHi2MvGfk7eEYlBoKY4UZSdEBUFz5Au5EUb1L+ALAPw+VM4CJOojbAF60aBG0Wi369Onj8mUciXf+ciM+3l2CylorxvwsHKMGtL03HxygQHJsoMu+0KDmLw8dh6oRdRi37yaj0Yh3333XG7XQLTBZ7HhmbQEqaps+lRw6WQOVSo5hfTru6jQ+SoP7hkZi+9dlAIAIrRKzR0R32PmJ/J3bAB44cCBOnDiBXr16eaMeaqWjp2ud4XtVXn5lhwYwACyZGo8JQyJRWmVB/x4h0Kg67SpWRD7HbQAfOnQImzdvRteuXV1aENu2bfNoYXRz0eHN20H6cM/M45AcG9isPUFE7ec2gN955x1v1EG3KDE6ALNGRGHTvy/D7gC6xwRgxvAo0WUR0S3gihg+rqTSjOo6K3rEBXL5eCIfwxUxfFx0mBopXYMYvkQ+yG0A79y5E2+++SYCA5t6gFwRg4ioY3BFDCIiQbgiBhGRIK36Em7fvn3Yu3cvAGD48OGdfkUMf/oSjoh8102vgG02GzIzM7Fhw4ZOH7pERL7mpj1ghUIBuVyOmpoab9VDROQ33PaAg4KCMHnyZNx9990ICrq2aOPy5cs9WhgRkdS5DeCxY8di7Nix3qiFiMivuA3g6dOno7GxERcvXkRycrI3aiIi8gtuxwHv2rULU6dOxYIFCwA0jTB44oknPF4YEZHUtepW5M8++wyhoU1rgaWlpeH8+fMeL4yISOrcBrBSqYRWq3XZx3kHiIjaz20POCUlBdu2bYPNZkNhYSE2bNiAgQMHeqM2IiJJc3snXENDA9566y2XO+EWL14MjUZzs4fd1Oeff4433ngDBQUF+PTTT9GvX78bHpeeno7g4GDI5XIoFAps2rSpVefnnXBE5AtavAJetmwZXnrpJXzyySdYunQpli5d2mFPmpqaitdffx3PPfec22Pff/99REREdNhzExF1Fi0G8LFjx2A0GpGdnY1p06bhpxfKYWFhbX7SHj16tPmxRERS0WIAP/DAA8jMzERRURFmzJjhEsAymQy5ubleKXD+/PmQyWSYM2cO5syZ06rHmEwmGAwGD1dGRNQ6LbVEWwzge++9F4888giee+45rFix4pafMDMzE6Wlpc32Z2VlYcyYMa06x8aNG6HX61FWVoa5c+ciOTkZgwcPdvs4jUbDHjARdXotBvDTTz+NTZs2obCwsE0nXr9+fRtLukav1wMAIiMjkZGRgfz8/FYFMBGRL2gxgO12O9566y0UFhbivffea/bzuXPnerSw+vp62O12hISEoL6+Hvv27cPixYs9+pxERN7U4o0Yr776KuRyOWw2G+rq6pr90x47d+7EiBEjcPjwYTz++OOYP38+AMBoNOKxxx4DAJSVleGhhx7ClClTMHv2bIwcORIjRoxo1/MSEXUmbscB79mzByNHjvRWPR2C44CJyBe02ILYunUrpk6dioKCApw+fbrZzz3dgiAikroWA7ihoQEAV0EmIvKUVi3K6WvYgiAiX9DiFfAf/vCHmz6QSxIREbVPi6Mg+vTpgz59+sBkMuHYsWNISkpCUlISDAYDzGazN2skIpKkFq+Ap0+fDqDpbrSPPvoISmXToQ888AB+/vOfe6c6IiIJczshe1VVFWpra53b9fX1qKqq8mhRRET+wO2E7AsXLsT06dNx5513wuFw4JtvvsEvf/lLb9RGRCRprRoFcfnyZRw5cgQA0L9/f0RFRXm8sPbgKAgi8gUchkZEJIjbHjAREXkGA5iISJA2BfCoUaM6uAwiIv/TpgCWYNuYiMjr2hTAMpmso+sgIvI7LY4DvtEqGEDT1S9nSCMiar8WA/hmq1488sgjHimGiMiftBjATz75ZIsP6ogFN4mI/F2besAMYCKi9uMoCCIiQTgKgohIkBZ7wAMHDrxh0DocDphMJo8WRUTkD1oM4MOHD3uzDiIiv+N2PmDyDVabAx/lFuNrQzXiozSYOz4WsREa0WUR0U0wgCXio11GbNxdAgA4U9yIQmMj3s7qxX49USfG2dAk4oDBdZmoohITLpZx8VSizowBLBF1JrvLtlolQ2QoP+AQdWYMYAmoqbfCWO56tRsRokSAWiGoIiJqDQawFMgA+U9+k7oQlZhaiKjVGMASoA1UYtKdXZzbchkwe2S0wIqIqDXYJJSIJybHYXAvLQqNjRiUqkW3mMBbPkej2Y7C4gYk6gMQpGH7gsjTGMASIZPJMKhXKAb1Cm3T4/NP1+L3HxaitsGGQI0cv34gCUNua9u5iKh12IIgAMDb2y+gtsEGAGgw2fHnv18QXBGR9DGACQBgrHAdRXG5ygybnbPeEXmSkAB+8cUXMX78eEyePBlLlixBdXX1DY/Ly8vDuHHjkJGRgbVr13q5Sv9yT78wl+1hfXRQyHkXHZEnyRwCJvfdu3cvhg4dCqVSiZdeegkAsGzZMpdjbDYbxo0bh/feew96vR6zZs3Cq6++ipSUFLfnNxgMSEtL80jtUmW22PHxlyX4/kwteiUE4aF0PQL5RRyRRwn5Em748OHOfx8wYAC++OKLZsfk5+cjKSkJCQkJAIBJkyYhNze3VQFMt06tkuORjBjRZRD5FeGjILKzszFhwoRm+41GI2JirgWCXq9Hfn5+q85pMplgMBg6rEYiovZo6RO5xwI4MzMTpaWlzfZnZWVhzJgxAIA333wTCoUCU6ZM6dDn1mg0bEEQUafnsQB2t3Dnpk2b8OWXX2L9+vU3nDJRr9ejuLjYuW00GqHX6zu6TCIiYYSMgsjLy8O6devw5ptvIjDwxnds9evXD4WFhSgqKoLZbEZOTg7S09O9XCkRkecIGQWRkZEBs9mMsLCmoU/9+/fHypUrYTQasXz5crzzzjsAgD179uCFF16AzWbDzJkzsWjRoladn6MgiMgXCAlgT2MAE5Ev4J1wRESCMICJiARhABMRCcIAJiIShAFMRCQIA5iISBAGMBGRIAxgIiJBGMBERIIwgImIBGEAExEJInxCdmofk8WOdf+4iK8N1YjvosHj98WhW4zrDHMniuqxNucCLpWZMayvDo9NjINaxb+9RKLxXejjPthZjO1fl6G0yoLvCmqxYkMh7NetZmyx2rFiwxkcP1uPilortn9dho27jQIrJqKrGMA+7vDJGpft4nIzLpVfW2K+6LIJFTVWl2O27S/F92dqvVIfEbWMAezjesS5thtCgxSI0qmc23GRagRpXH/NdY12PPuX07hcZQYRicMA9nFzx8eib/dgAEBkqAr/PTvRpb8boFZg2ZxEBKhdf9VmqwOHfnS9eiYi7+KXcD4uQqvCSwtTUNtgQ6BGDoW8+fp6Q9N0mDc+Bn/++0WX/XGRmlY9x7HCOnzxTRnsDgfGDYrE7ckhHVI7kb9jAEtESKDipj8fNygSB0/U4MCJGshlwMQ7Wxek/zhQhtc3n3du7zpciccmxmHGPVHtrpnI3zGA/YRaJceKzGQUl5ugVsoREapy+xiL1Y7svJJm+z/NK2EAE3UABrAPO1JQiwM/VCM+SoMxPwuHSum+pR8Tca3tYLbY8cU35Si63Ight4VicK9QAMD3Z2qxJrsIF8rM0KiatzSIqGMwgH3UrsMVeOmTc87tQydrsPzn3W7pHH/8+Cz2H68GAGz/ugxZM+MxZmAE/vjxOZRVWwAAJkvzNVvvHxnd9sKJyIkB7KO27S912d73fRXKqy2tai0AQHm1xRm+V+V8XYaBKVpn+F4VHabCgB4hsDuAsYMi0K87v4Qj6ggMYB+l+cmtxAo5oFS0vl2gUsmgVMhgtV27wg0KUKBLqAoxEWoUX3czxx2pWjw1PaH9RRORC44D9lFz7o12Cdwpd3VBaHDr/55qA5WYed0XaWqlDA/eq4dcLsNvf9ENvZOCEBKgwL0DwjB/QlyH1k5ETWQOh6N5k8/HGQwGpKWliTC1iXEAAAm+SURBVC7D44wVZhw6WYOEKE2b2wIniupxrqQRA1O06KJrXfuCiDoGA5iISBC2IIiIBGEAExEJwgAmIhKEAUxEJAgDmIhIEAYwEZEgDGAiIkEYwEREggiZC+LFF1/E7t27oVKpkJiYiFWrViE0NLTZcenp6QgODoZcLodCocCmTZsEVEtE5BlCroCHDRuG7du3Y9u2bejWrRvefvvtFo99//33sXXrVoYvEUmOkAAePnw4lMqmi+8BAwaguLhYRBlEREIJn44yOzsbEyZMaPHn8+fPh0wmw5w5czBnzpxWndNkMsFgMHRUiURE7dLS3DQeC+DMzEyUlpY225+VlYUxY8YAAN58800oFApMmTLlhufYuHEj9Ho9ysrKMHfuXCQnJ2Pw4MFun1uj0XAyHiLq9DwWwOvXr7/pzzdt2oQvv/wS69evh0x244nE9Xo9ACAyMhIZGRnIz89vVQDTzdntDpTXWBGhVUJ+g2Xsicg7hLQg8vLysG7dOnz44YcIDAy84TH19fWw2+0ICQlBfX099u3bh8WLF3u5Uuk5UVSPFz4qREmlBTERajz78ySkxAWJLovILwmZDzgjIwNmsxlhYWEAgP79+2PlypUwGo1Yvnw53nnnHRQVFWHJkiUAAJvNhvvuuw+LFi1q1fk5H3DLFr12AoXFjc7t1PhAvLYkVWBFRP6LE7L7mUm/OQL7db9xjUqGLStvF1cQkR/jnXB+5o5Urcv2oNTmN8AQkXcwgP3Mf81KRPqAcMRFqpFxRzienhEvuiQiv8UWBBGRILwCJiISRPidcORZZqsduw5X4GKZCXf11iEtMVh0SUR0BQNY4p7/61kc+KEaAPBZ3mX89hfdcFdvneCqiAhgC0LSLpWbnOELAA4H8Pevmt8eTkRiMIAlTKmQ4ad3eauUvPWYqLNgAEtYlE6NjDsinNsqpQyzRkQLrIiIrscesMRlzYjHiH5huFRmwuDbQqEPV4suiYiuYABLnEwmu3L3m9btsUTkXWxBEBEJwgAmIhKEAUxEJAgDmIhIEAYwEZEgDGAiIkEYwEREgjCAiYgEYQATEQnCACYiEoQBTEQkCAOYiEgQBjARkSAMYCIiQRjARESCMICJiARhABMRCcIVMSTOYrXjr7lG/MdQjYRoDeaNj0VMhEZ0WUQEBrDk/TXXiL99WQIAKDQ24lyJCW9l9RJcFREBbEFI3n9+qHbZPmtsRHG5SVA1RHQ9BrDEJUa5thtCAhQI16oEVUNE12MAS9y8CXHopg8AAIQEKvDUjHhoVPy1E3UGMofD4RBdREczGAxIS0sTXUanYqwwIzxECTXDl6jTEPZuXLNmDSZPnoypU6di3rx5MBqNNzxu8+bNGDt2LMaOHYvNmzd7uUrp0IerGb5EnYywK+Da2lqEhIQAAD744AOcOnUKK1eudDmmsrISM2fORHZ2NmQyGWbMmIFNmzZBp9Pd9Ny8AiYiXyDskuhq+AJAQ0MDZDJZs2P27t2LYcOGISwsDDqdDsOGDcO///1vb5ZJROQxQscBr169Glu2bIFWq8UHH3zQ7OdGoxExMTHObb1e32Krwl+ZLXa8t+PSlRstAvDYxFjERwWILouIWsGjAZyZmYnS0tJm+7OysjBmzBgsXboUS5cuxdtvv40PP/wQTz31VIc8r8lkgsFg6JBzdXZ/P2jFl8ftAIBL5WacuViNX09V3fATBRGJ0VJL1KMBvH79+lYdN3nyZCxcuLBZAOv1ehw4cMC5bTQaMWTIELfn02g0ku8BF5U0otDYiILLxQCu3VhxuRoIj+mBWN5uTNTpCesBFxYWOv89NzcXycnJzY4ZPnw49u7di6qqKlRVVWHv3r0YPny4F6vsnD7LK8HC1SfwwkdncaHU9a42baACkbzRgsgnCOsBv/LKKzhz5gxkMhm6du2KFStWAACOHj2Kjz/+GM8//zzCwsKwePFizJo1CwCwZMkShIWFiSq5U2g02/Hhv671we0OIEAtR6PZjrAQJZ6eHs/hZkQ+gjdi+Jiaeivm/OEYrv+tJURr8OKCHtAGKaFUsPdL5Ct4qeRjtEFK3N3bdRz0+EGRCNeqGL5EPobTUfqgZXMSkfZ1KQqLG3FHqhaj+oeLLomI2oAB7IM0Kjlm3hMtugwiaie2IIiIBGEAExEJwgAmIhKEAUxEJAgDmIhIEAYwEZEgDGAiIkEYwEREgjCAiYgEYQATEQnCACYiEkSSc0H405JERNT5KZVK9OzZs9l+Sc4HTETkC9iCICIShAFMRCQIA5iISBAGMBGRIAxgIiJBGMBERIIwgCUmLy8P48aNQ0ZGBtauXdvs52azGVlZWcjIyMDs2bNx/vx5AEBFRQUefvhhDBw4ECtXrvR22dTJuXtdffPNN5g+fTp69+6NL774QkCFvokBLCE2mw0rV67EunXrkJOTg+3bt+PUqVMux3z66acIDQ3Fzp07kZmZiZdffhkAoNFo8PTTT+OZZ54RUTp1Yq15XcXGxmLVqlW47777BFXpmxjAEpKfn4+kpCQkJCRArVZj0qRJyM3NdTlm165dmD59OgBg3Lhx2L9/PxwOB4KCgjBo0CBoNBoRpVMn1prXVXx8PG677TbI5YyUW8H/WhJiNBoRExPj3Nbr9TAajc2OiY2NBdB0e6RWq0VFRYVX6yTf0prXFbUNA5iISBAGsITo9XoUFxc7t41GI/R6fbNjLl26BACwWq2oqalBeHi4V+sk39Ka1xW1DQNYQvr164fCwkIUFRXBbDYjJycH6enpLsekp6dj8+bNAIAdO3Zg6NChkMlkIsolH9Ga1xW1DWdDk5g9e/bghRdegM1mw8yZM7Fo0SK89tpr6Nu3L0aPHg2TyYRly5bBYDBAp9Nh9erVSEhIANAUzrW1tbBYLNBqtfjLX/6ClJQUwf+PqDNw97rKz8/Hk08+ierqamg0GnTp0gU5OTmiy+70GMBERIKwBUFEJAgDmIhIEAYwEZEgDGAiIkEYwEREgkhyVWSi9igtLcWqVavw3XffQafTQaVSYcGCBcjIyBBdGkkMA5joOg6HA0uWLMG0adPwyiuvAAAuXLiAXbt2Ca6MpIjjgImus3//fvzpT3/Chx9+KLoU8gPsARNd5+TJk+jdu7foMshPsAVBdBMrVqzAoUOHoFKpkJ2dLbockhheARNdp2fPnjh+/Lhz+7nnnsP69es5ZzJ5BAOY6DpDhw6FyWTCRx995NzX2NgosCKSMn4JR/QTJSUlWLVqFY4cOYKIiAgEBgbiwQcfxMSJE0WXRhLDACYiEoQtCCIiQRjARESCMICJiARhABMRCcIAJiIShAFMRCQIA5iISJD/D1qqYQ+8zfieAAAAAElFTkSuQmCC\n", "text/plain": [ "
" ] @@ -1072,28 +891,31 @@ } ], "source": [ - "df_epa = df2[df2[\"software\"] == \"epa\"]\n", - "sns.catplot(data=df_epa, x=\"g\", y=\"LL difference\", palette=\"muted\")" + "epa_columns = [\"software\", \"g\", \"query\", \"LL\", \"LL difference\"]\n", + "df_epa = df2[df2[\"software\"] == \"epa\"][epa_columns].copy()\n", + "df_epa.rename(columns={'g': 'G'}, inplace=True)\n", + "\n", + "sns.catplot(data=df_epa, x=\"G\", y=\"LL difference\", palette=\"muted\")" ] }, { "cell_type": "code", - "execution_count": 301, + "execution_count": 378, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 301, + "execution_count": 378, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1103,28 +925,30 @@ } ], "source": [ - "df_epang = df2[df2[\"software\"] == \"epang\"]\n", + "epang_columns = [\"software\", \"g\", \"query\", \"LL\", \"LL difference\"]\n", + "df_epang = df2[df2[\"software\"] == \"epang\"][epang_columns]\n", + "\n", "sns.catplot(data=df_epang, x=\"g\", y=\"LL difference\", palette=\"muted\")" ] }, { "cell_type": "code", - "execution_count": 309, + "execution_count": 379, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 309, + "execution_count": 379, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAAFgCAYAAAB38TSsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3de1xUZf4H8M+ZGWaAAVFUBjXzhhfyhpukpomieAc0bU1N05Uou6i5laalrXnJcrvprunPUtPW1itsmqZpapquaSheRssLCioDCMIwwFzP7w/WSQIE4cCZgc/79fL18lzmnC8w8JnznOc8jyCKoggiIiIJKeQugIiIah6GCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCtcLGjRsRFxcHANi2bRsMBsN999+2bRvmz59fHaWV6YUXXsCwYcOcy3fu3MGkSZMwYMAATJo0CdnZ2TJWR1QyhgvVCmPGjMHw4cMBANu3b0daWprMFZXPnj17oNVqi6xbtWoVevTogT179qBHjx5YtWqVTNURlY7hQtUiJSUFgwYNwqxZszBw4ED89a9/xU8//YSnn34aAwYMQGJiIgAgMTERo0ePxvDhw/H000/jypUrAIC1a9fizTffBABcvHgRw4YNQ35+frHzhIeH4/3330dkZCRGjRqFa9euAQCWLVuGzz//HLt378bZs2fx2muvITo6GgUFBUhMTMTTTz+NqKgojBo1Crm5uQCAtLQ0TJ48GQMGDMD777/vPMfhw4cxevRojBgxAlOnToXJZAIALF26FEOGDEFkZCSWLFlS6e+ZyWTCmjVrMGXKlCLr9+3b5wzK4cOH4/vvv6/0uYgkJxJVg+TkZDE4OFi8cOGCaLfbxREjRoizZs0SHQ6HuHfvXnHKlCmiKIqi0WgUrVarKIqieOTIEfHll18WRVEU7Xa7OHbsWHHPnj3iiBEjxBMnTpR4nr59+4r//Oc/RVEUxe3bt4uxsbGiKIrip59+Kq5evVoURVF85plnxMTERFEURdFsNovh4eHi6dOni5x/69atYnh4uJiTkyMWFBSIffr0EW/evCnevn1bHDt2rGgymURRFMWVK1eKy5YtEzMzM8UBAwaIDodDFEVRzM7OLlbb0aNHxaioqGL/Ro8eXeLXsnDhQnHPnj1icnKyOHToUOf6Rx991Pl/h8NRZJnIVajkDjeqPR566CG0bdsWABAUFIQePXpAEAS0bdsWN27cAAAYjUbMnDkT165dgyAIsFqtAACFQoH33nsPUVFRGD16NB599NFSz3P3/sTQoUOxePHi+9Z09epVNGzYEJ06dQIA+Pj4OLf16NEDvr6+AIBWrVrhxo0bMBqNuHTpEsaMGQMAsFqtCAkJga+vLzQaDWbPno2+ffuiT58+xc7VvXt3xMfHl+dbBb1ej+vXr2P27NlISUkpdT9BECAIQrmOSVSdGC5UbdRqtfP/CoXCuSwIAux2OwDgk08+Qbdu3fCPf/wDKSkpmDBhgvM1SUlJ8Pb2LnK/ZPLkycjIyECHDh2wcOHCKqtXqVTCbrdDFEX07NkTH374YbH9t2zZgqNHj2L37t3YsGEDvvzyyyLbjx07VmLYeXl54euvvy6yLiEhAWfPnkV4eDhsNhsyMzMxfvx4rF+/HvXr10daWhoCAgKQlpYGf39/ib5iIukwXMilGI1G6HQ6AIU33u9dv2DBAmzYsAHvvvsudu/ejUGDBuHzzz8vdoxdu3YhNjYW3377Lbp06VJsu1ardd4nadGiBdLT05GYmIhOnTohNzcXnp6epdYXEhKC+fPn49q1a2jWrBny8vJgMBgQEBCAgoIChIWF4U9/+hP69+9f7LUPcuUyduxYjB07FkDh/aoXXngB69evB1B4XykuLg6xsbGIi4tDv379ynVMourEcCGXEhMTg1mzZmHFihUICwtzrl+0aBHGjRuHFi1aYOHChZgwYQJCQ0NRv379YsfIzs5GZGQk1Gp1iVcYI0aMwLx58+Dp6Yl///vf+Oijj7BgwQIUFBTA09MTa9asKbU+f39/LF68GDNmzIDFYgEATJ8+HVqtFi+++CLMZjMAYNasWZX9VpQqNjYW06dPx5YtW9C4cWN8/PHHVXYuoooSRJFD7lPNER4eji1btrCpiEhm7IpMRESS45ULERFJjlcuREQkOYYLERFJrkaGy2+//SZ3CUREtVqNDBebzSZ3CUREtVqNDBciIpIXw4WIiCTHcCEiIskxXIiISHIMFyIikhzDhYiIJCdbuNy6dQvjx4/HkCFDMHToUKxbt67YPqIoYsGCBYiIiEBkZCTOnTsnQ6VERPSgZBtyX6lUYtasWWjfvj1yc3MxcuRI9OzZE0FBQc59Dh06hKSkJOzZswenT5/GO++8g82bN8tVMhERlZNsVy4BAQFo3749gMKpZVu2bAmDwVBkn3379mH48OEQBAEhISHIyckpMgshERG5JpeYLCwlJQV6vR6dO3cust5gMCAwMNC5HBgY6Jz1737MZjP0en2V1EpEVB7BwcFylyAr2cPFZDJh6tSpmD17Nnx8fCQ5pkajqfU/WCIiOckaLlarFVOnTkVkZCQGDBhQbLtOp0NqaqpzOTU11Tm/ujuxO0Qcv5CD29lWdHukDhr6qeUuqco4cjNhvXgUgrYuPFp3g6CU/fNLrXI1NR+nL+eiVWMvdGwhzYe16mRPS4I16TSUga3g8XAHucuhSpDtN18URcyZMwctW7bEpEmTStwnPDwcGzZswNChQ3H69Gn4+vqW2STmihZsSMIxfQ4A4PPdt/DB860Q1Nhb5qqkZ0+/hpwvXwPMeQAAVYsu8B2zQOaqao8Dp7Pw/r+v4+70f+P66fBM/8D7v8iFWM4dgCl+KYDCL8DziXHwemKsvEVRhcl2Q//kyZOIj4/HsWPHEB0djejoaBw8eBAbN27Exo0bAQBhYWFo2rQpIiIi8Pbbb2PevHlylVthV2/lO4MFAAosDsQdzpCxoqpT8PN/nMECALarCbDduCBjRbXLv39Iw73zym45lA6L1SFfQQ8o/8gm3A0WACg4tgWizSpfQVQpsl25dO3aFRcvXrzvPoIguGWg3MtqLz6LtNXmPr/wD8RRfKoD0c7pD6rLH99rdocIhztNYu74Q5A47IBYQ39XagE+oV/F2jzkjfbNtc5llVJAZI8GMlZUdTR/GgIoPZzLykatoWr6iIwV1S7DexZ9Xw1+zB+eavf5FdeERhdd7jIYgodGpmqosgRRFN3ps0256PV6l+otVmBxYH9CFjKyLXiiU120CPSSu6QqY0+/Dov+EARtXWg69oOgrrlfqytKuGTEqUuFN/R7dfCDQiHIXdIDsV49BVvSKSh1reAR3BOC4D7hSEUxXIiISHL8WEBERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSkzVc3nzzTfTo0QPDhg0rcft///tfPProo4iOjkZ0dDSWL19ezRUSEVFFqOQ8+ZNPPolnnnkGM2fOLHWfrl27YuXKldVYFRERVZasVy6hoaHw8/OTswQiIqoCsl65lMepU6cQFRWFgIAAzJw5E61bty7zNWazGXq9vhqqIyIqWXBwsNwlyMqlw6V9+/bYv38/tFotDh48iJdeegl79uwp83UajabW/2CJiOTk0r3FfHx8oNVqAQBhYWGw2WzIzMyUuSoiIiqLS4dLeno6RFEEACQmJsLhcKBevXoyV0VERGWRtVlsxowZOH78OLKystC7d2+88sorsNlsAIAxY8bgu+++w8aNG6FUKuHp6YkPP/wQgiDIWTIREZWDIN69NKhB9Ho977kQEcnIpZvFiIjIPTFciIhIcgwXIiKSHMOFiIgkx3AhIiLJMVyIiEhyDBciIpIcw4WIiCTHcCEiIskxXIiISHIMFyIikhzDhYiIJMdwISIiyTFciIhIcgwXIiKSHMOFiIgkx3AhIiLJMVyIiEhyDBciIpIcw4WIiCTHcCEiIskxXIiISHIMFyIikhzDhYiIJCdruLz55pvo0aMHhg0bVuJ2URSxYMECREREIDIyEufOnavmComIqCJkDZcnn3wSq1evLnX7oUOHkJSUhD179uDdd9/FO++8U33FERFRhckaLqGhofDz8yt1+759+zB8+HAIgoCQkBDk5OQgLS2tGiskIqKKUMldwP0YDAYEBgY6lwMDA2EwGBAQEHDf15nNZuj1+qouj4ioVMHBwXKXICuXDpeK0mg0tf4HS0QkJ5fuLabT6ZCamupcTk1NhU6nk7EiIiIqD5cOl/DwcMTFxUEURZw6dQq+vr5lNokREZH8ZG0WmzFjBo4fP46srCz07t0br7zyCmw2GwBgzJgxCAsLw8GDBxEREQEvLy8sWrRIznKJiKicBFEURbmLkJper+c9FyIiGbl0sxgREbknhgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkZA2XQ4cOYeDAgYiIiMCqVauKbd+2bRu6d++O6OhoREdHY/PmzTJUSURED0ol14ntdjvmz5+PNWvWQKfTYdSoUQgPD0dQUFCR/YYMGYK5c+fKVCUREVWEbFcuiYmJaNasGZo2bQq1Wo2hQ4di3759cpVDREQSku3KxWAwIDAw0Lms0+mQmJhYbL89e/bg559/RosWLfDmm2+iUaNGZR7bbDZDr9dLWi8R0YMIDg6WuwRZyRYu5dG3b18MGzYMarUaX3/9NWbOnIkvv/yyzNdpNJpa/4MlIpKTbM1iOp0OqampzmWDwQCdTldkn3r16kGtVgMAnnrqKZw7d65aayQiooqRLVw6duyIpKQkJCcnw2KxYOfOnQgPDy+yT1pamvP/+/fvR6tWraq7TCIiqgDZmsVUKhXmzp2LmJgY2O12jBw5Eq1bt8Ynn3yCDh06oF+/fli/fj32798PpVIJPz8/LF68WK5yiYjoAQiiKIpyFyE1vV7Pey5ERDIqV7PYiRMnsHXrVgBAZmYmkpOTq7QoIiJyb2WGy/Lly7F69WrnE/RWqxWvv/56lRdGRETuq8xw2bt3L1asWAEvLy8Ahb28TCZTlRdGRETuq8xw8fDwgCAIEAQBAJCXl1flRRERkXsrs7fY4MGDMXfuXOTk5GDTpk3YunUr/vznP1dHbURE5KbK1VvsyJEjOHz4MACgV69e6NmzZ5UXVhnsLUZEJK8ywyU5ORkBAQHQaDQAgIKCAmRkZOChhx6qlgIrguFCRCSvMu+5TJs2zXm/BQAUCgWmTZtWpUUREZF7KzNc7Ha7c3wvAFCr1bBarVVaVG12LsmEhEtG2B017tlWiOY8WH47DnvGdblLcSsp6QU4ps+GqcAudynVRrSaYb30M2xpV+UuhSqozBv6/v7+2LdvH/r16wcA+P7771GvXr0qL6y2sdtFvL32ChIu5QIAWgR64oPng6D1VMpcmTRsaVeR+9VsiPk5AADPx0fDq88Ematyff/aZ8D67wsHePXxVGJRTEu0buItc1VVy555E8YNMyHmZgIANI8Og/fAKTJXRQ+qzCuXv/3tb1i5ciX69OmDsLAwrF69GvPnz6+O2mqVny/mOIMFAK6mFmDPiUwZK5JWweGNzmABgIKjm+HIzZKxIteXY7Jh4w8G53JugR1f7TPc5xU1Q8GxLc5gAQDzyR2wZ96QsSKqiDKvXB5++GFs2rTJ+eCkVqut8qJqo2yTrVzr3JWYl/OHFQ6IBUbAh1fBpTGZ7bDZizaPZufWnPdEacS87BLW5QD+TWSohiqqzHCxWCz47rvvcOPGDdhsv7+xX3755SotrLbpFuwHX69bMOYXtqt7qAT06VxX5qqko+7UD7brZ5zLyibtoGzwsIwVub5G/hp0aKHF2au/j4gR8ai/jBVVD3XH/rD+esy5rKjfFMombWWsiCqizHCZMmUKfH190b59+yI39kladX1UWPpCEL7YdRMFFgfG9Q9E80AvucuqFFvyedizDfBo+SdoOkVAUHvDcuEIlPUaQfPYcLnLcwvvTGiB7YfTcSPDjA7NtdB4CLh0Iw9BNfi+i7Lhw9A8NhyOzFtQNGwKz9DhEATZpp6iCiozXAwGAz7//PPqqKVWsztELI9LwZn/fUq9tek6Pn6xNer5eshcWcWYdi2HJWFX4YLGG77j3oO6XU+o27n2A7iuRuupxDP9A3HqshHz1l6FxVbYTDYmXIcJEYEyVyc9c+L3yNvxMQARgADvdj2gYNOpWyrz40CXLl1w8eLF6qil1riRYcZbX1zB2EXnsHTTdaRnWzB37RVnsABA2h0rvj1+W8YqK85+xwBLwu7fV5jzCm/gZ6chd9PfcOeTccjdvgSOEtrWqWT/2mdwBgsAbDmUVmrX5EyjFQs2JGHswnN4d0MSMnPc59GB/IPrURgsACD+b7l0oigi/9AGZC+fiJwvpsF6NaHKa6TyKfPK5eTJk9i+fTuaNGlSpFnsm2++qdLCarKFXyXhamoBAGBfQhYuJJtwI8NSbL88s6O6S5OGJR+//4EoJJrzkBv/Aewp5wEAVv0h5Dns8Bk5W4YC3c8f3wtWmwiL1VFiV/WPtiTjxK9GAMBP57JRYLFj4V/cY4pw0VJ0YFzRfP+Bci2nvkPB4Y2FCznpyN38LvxeXgOFt19VlUjlVGa4/N///V911FFr3Mm1OYPlrluZxYPFQwlE/Mk9mwOUAc2hfOgRZ5AAAtSdIpAX916R/WzXTlV/cW5qSLf6WLY9xbn8eHu/UptMT13Ove+yK9N0GQzzsa1Flu/HmnS66AqbGbYUPdRtuldFefQAygyXJk2a4MSJE7h27RpGjhyJzMxMzudSCXW8lQio64G0O783VdTVqpBp/L0nnqdagaXPt3LrG/q+o/8Gc8IuOO4Y4NGuJzyad0bB4X/Bcc/T+Uqde3yadgVDHquP+r4eOH4xB80CPDEotPReY60ae+Fi8u+f+Fs1cp/3kVffSVA2bA5bynmoHgqGukP4ffdXBbaCVX/o9xWCAkpdyyquksqjzIErly9fjrNnz+Lq1av47rvvYDAYMG3aNHz99dfVVeMDc/WBK88m5eLvm5ORmmlBu6beiB3WGCt33MTF5DwE1PXAjFEPo3MrH7nLlJzt1m8wxS+FIzMFSl0raIfPhLI+n12QWlJqPpZ8fR1JhgI003li5tMPo4Ubf1C5H9FmgWnHR7DqD0PQeMOr7yRougySuyxCOcIlOjoacXFxGDFiBOLi4gAAkZGRLn3PxVXC5U6uDf8+YEBKuhndH6mDIY/Vdw4CeuG6CRv2piLP4sDAUH8M7Fofufl2eGsUUCiEMo7svkRRhDlhN6wXDsNhzAAUSngEhcLriXEQVOzqXh5mqwMbvk/FqUu5aNXYCxMHBqKujwfOXM3Ff37KgEopYESvhmjSQOO2wweJBSYYty2E/dZvUNQJgDbqNah0LUrf31IAqDwgKNzz6127di1Gjx7tnPH3j+bMmYNJkyYhKCgIXbp0QUJCxTouLFu2DN7e3pg8eXJlyi2XMpvFOBNlxb2z7iouphR+v078aoTFKmJEr4Yw5tsw54srzpu0+ut58PVS4fH2Nf8mpOXMPuTvXl5knTn9GmA1w3vACzJV5V5Wf3sTO44V9iS8dDMfqZkWPB/ZGLM/v+J8ov/o+RysmtHWbcMl58vXnE2ojvQkGL98DXVf/RqCquT7TILaszrLk9yXX36JqKioEsPFbrdj4cKFMlRVOWV2Rf7jTJSTJk3iTJTlcPO22Rksd/1wqnAsrdOXc4v1/jl6vnZ0y7WcO1jy+nueyKb7++N75fSVXPyQkFVkqBiz1YFj53P++FK34Mg3Frk3BwCwFsB2s2Y8EpGXl4fY2FhERUVh2LBhWL58OdLS0vDss89i/PjxAAofAXnvvfcQFRWFhIQEjB8/HmfOnClynMzMTIwePRoHDhwAAKxevRojR45EZGQkPv3001LPf+HCBYwePRoDBgzApk2bABS2KCxZsgTDhg1DZGQkvv32WwDA3r178eyzz0IURaSlpWHgwIFIT08v19dZ5pXL5MmTceTIEWi1Wly9ehVTp051+ZkoXYGfVgW1SijybELDuoXNPk0aaIrt/1DDwnU/ncvGF7tvIdtkw4BH/fGXwY2grCHNZPb0a7AbLpe4Tclxo8qtjrcKt3N+7wBSv44HAv2LNyk28HOfB3BFmxV5ez6DRX8Igm8DQKEEHEWf41HWayRTddL68ccfERAQgFWrVgEAjEYjtm3bhnXr1sHfv7CjRl5eHjp16oRZs2aVeIyMjAxMmTIF06dPR8+ePXH48GFcu3YNW7ZsgSiKmDJlCn7++WeEhoYWe+3FixexadMm5OXlYcSIEQgLC8OpU6dw4cIFxMfHIysrC6NGjULXrl0RERGB7777Dl999RV+/PFHvPLKK2jYsGG5vs77hovdbsfEiROxfv36KgmUQ4cOYeHChXA4HHjqqacQGxtbZLvFYsEbb7yBc+fOoW7duvjoo49cegbMe2k9lfjLoEb4bMdN57oL13Iw6YOzsJjtUAjA3Slb1B7AyV/v4PTlbCRcynfuv+1wOhr5qzGsR4PqLv+B5Z8/goLvlgM2G+BVB/DSQtD6Q8xMAUQBcFgAY0bJL/bwhGfPp6u3YDdizLPi7S+uwGS2Id/iwO2cole9GhVwMdmEOl4CcvIL31QCgI+3JuFggi/eGNMCSqVrf0ApOLYFllOFD96K5hLm+xEUyP7qLcCUBai9AJsFcFgBuwOoGwDBtz68Q6OgDir+x9TVtGnTBkuWLMEHH3yAvn37omvXrsX2USqVGDhwYImvt1qtmDhxIubOnYvHHnsMQOFU9EeOHMHw4YXDKuXl5SEpKanEcOnXrx88PT3h6emJbt264cyZMzh58iSGDh0KpVKJBg0aIDQ0FGfOnEG/fv3w9ttvY9iwYQgJCcGwYcPK/XXeN1yUSiUUCgWMRiN8fX3LfdDysNvtmD9/PtasWQOdTodRo0YhPDwcQUFBzn02b96MOnXqYO/evdi5cyeWLl2Kjz/+WNI6qorN7igSLACQmSsCKP5UtcUKnLlaUGw9APzjPzfQJ6QefLxct+086/2RgO2e+q15QA4gouSrlGKsBcj9ahY04X+Bd/eRVVOkm1r97Q1s/bGUUP6fm5lW3My8U2SdCMCYDxw6Z8ShtxLx0ZQgtHvYdUc0LzhRRgch0QFk/i90zMai2zKuQcy4BtPVX2DybYB6r6yrmiIl0qJFC2zbtg0HDx7Exx9/jO7diz+To9FooFSW/DuvUqnQvn17HD582BkuoigiNjYWTz9d9EPaV1995Wz6unuldO/MwuWRmpoKhUKBjIwMOBwOKBTlG+etzL28vb0RGRmJ2bNnY8GCBc5/lZWYmIhmzZqhadOmUKvVGDp0KPbt21dkn/3792PEiBEAgIEDB+Lo0aMoo3Oby1iwIUmyY638xnXnsig4vbdosFSC+Ye1khynJikrWMrrrTXlDHoZOGwWQKqhgIwZMP92XJpjVRGDwQAvLy9ER0dj8uTJOH/+PLRabbmfHxQEAYsWLcKVK1ecgdGrVy9s3brVeQyDwYDbt29j3LhxiI+PR3x8PHQ6HQBg3759MJvNyMrKwvHjx9GxY0d07doVu3btgt1uR2ZmJk6cOIFOnTrBZrNh9uzZ+Pvf/45WrVphzZo15f46y7znMmDAAAwYMKDcBywvg8GAwMDfB97T6XRITEwstk+jRoXtrCqVCr6+vsjKynK2S5bGbDZDr9dLXvODOJ9U/Kn7itJfy4Je75oPrvqe+QlSfR4WRQf0584WtreTpPIKRNl/J0qjyM9GgITHMyT+BJNN2paWiijtcYhff/0V77//PhQKBVQqFd555x2cOnUKMTExCAgIwPr19x9PDShsVfrwww8xZcoUaLVajBs3DpcvX3ZeuXh7e+ODDz5A/fr1i722bdu2mDBhArKysvDiiy9Cp9MhIiICCQkJiI6OhiAIeP3119GwYUMsX74cXbt2RdeuXdGuXTuMGjUKffr0QatWZT8AXeZzLgBQUFCAmzdvomVL6Z583b17N3788UdnF7u4uDgkJiZi7ty5zn2GDRuG1atXO0Oof//+2LRpU5nh4grPucT/lIbPvrklybHeHtcMj3dwzbld7Jm3kPNZjCTHEnwbou4rayU5Vk0x/O3TMEswP9hjbX3wt4muOyJC1nvRgEOaidD8ZvwbCs+a9xCyuymzWWz//v2Ijo5GTEzhHxC9Xo8XXqj88wg6nQ6pqanOZYPB4Lxsu3efW7cK/0DbbDYYjUbUq+ce421FPx6AwHplXhiWKapHfZcNFgBQ+jeCR9izlT6OUDcQdWL/IUFFNcu/3+oAj0q+jR5q4IF3nnXtIVGUbaS5Ee/ZP5bB4iLKfNsuX74cW7Zscfa/Dg4ORkpKShmvKlvHjh2RlJSE5ORk6HQ67Ny5E3//+9+L7BMeHo7t27ejS5cu+O6779C9e/cHvhklpzVvtMc1QwGyTTa0b6bFl3tTselgWrlf/1RYQ/xlUOMqrFAaPj3/DFuzDjCun1l447UUvjHLoQpoAXt2GnI+iwXsd8dXE+ATOQMKjevecJaLRqPEf97tDACIeus0rPf0B1GrAIVCgQJL8e95PR8VVr/WDt4a92hiFLNSy97pHop6jeA3ZTVEqxmWC0cgFuRCHdwLCp+aP1OnuyjzyuXuvY57SfEHXqVSYe7cuYiJicGQIUMwePBgtG7dGp988onzxv6oUaNw584dREREYM2aNXjttdcqfd7q1kzniU4tfaBUCsXmQy/L979kVVFV0rOc/eG+wQIAsBWGie3yyXuCBQBEPkRZDn27FP3DGd7FHw5Hye+prFwbTv5qLHGbKxI8HmzsM9Fu+9/rNNB0DIdnaBSDxbDblY4AABuVSURBVMWUeeUSFBSEb775Bna7HUlJSVi/fj26dOkiycnDwsIQFhZWZN20adOc/9doNPd90tTVZRmt2H4kHel3rOjTuR4Ghvpj1/HbyP/fJ83G9dVIu2OBreQ5n1DPp/LNatVF0N6/uVLwqQ/zif9A7NgPCv/iV2N8iLJsL0U3QSN/NfTXTHikuRZDutXHbyl5uHyr5N567vT+8Wj56D1TNJRN6d8EjrwcKLzrVGFVVBll3tDPz8/HZ599hsOHDwMo7PL24osvQqMp/pS5q3CFG/oOh4gXPrmI5DSzc91bzzRHi0BP/HAqC94aJSIe9cfR89n4539uoMDigL+vCtkmG+wOQOMh4O1nmuPRNu7xy+MwZcO44Q04bhc2mXq06wWvXmNgPncAloTdEAvufooW4DPmXVgvHoU5YRcgOuDRpju0w2eVOm4UlWzBV0k4crbkLrxPdPTD7LHNq7egShDNecjdPB+262cAhQrqzgNg/fUoRFPh1bu6+0jAZoXlxH+cr1E2ao06k9zjubfaqNRwef311/HBBx9g3bp1ePbZyt+wrU6uEC766ybMWHGpyLruwXUwb0LxkV3zzHZk5ljRpIEGt3NsuHIrH+2aeqOO1n0+eQKA6LDDdv0sBE8tVIGFD8Nar/yC3K/fLrKfukNfaKNegyM3E7DboPCTsiNq7WAqsOOp+Wdx729vgzoq/PXPD8PHU4mgJt7yFVcJ9qxbEDTeUHj7QbSaYUs+C0WdACgbNIXxq9mwXSs6OVidmH9AGdBcnmLpvkr963Xu3DkYDAZs3boVw4cPL/bwYt26rtuDyRX4eRf/1l5MycPWH9MwomfDIsPqe2uU8G5YeOO1gZ+HW40JBRQONFjw479gN1yBqkUXePYY5dwmlDDd7N11bCOvOLVKgLdGAVPB7/e5BEHAv/YZ8HCAJ8bV8Sh1pkpXpqzXCJZfj8H8y7cQPDzh+fhTUDZoCqCE95KggOAl//Ms7shsNmPcuHGwWCyw2+0YOHAgpk6dKuk5Sg2Xp59+GhMnTkRycjKefPLJIuEiCEKxp+mpqMYNNIjq0QD/Ofr7E9ZZRhtWf3sL+WYHnukfeJ9XuxfT9vdgSyqcstiWfBai2QTvfoXzRagCW8GjdTdYf/tv4c4qNdSdIuQq1e3tPZmJr38wwGYX8afWvjhyNhsOEVAqgPRsK9KzrThz1YTLN/Px0Yut5S73gVmvn4VpywIUDmADWC8eAfwCoe0zAV49R8N27TTE/z3N79l9JBS+xR8SrEkcDhEHTt9B3JF0pGdb0dDPA8N7NkSfznUrNe+TWq3GunXroNVqYbVaMXbsWPTu3RshISGS1V5quPTt2xcTJkzAvHnz8Le//U2yE9YmU6Ka4E9BWryz/lqR9TuOZdSYcHEU5DqD5S7rhcPA/8JFFEXYb9/Tdd1mQe6GmagzeRmbwx7QpZt5+HBLsnM57U42OrXUIvGKCfY/dNS7kJyH9GwLGvq51wRs1os/4W6wOGWnwhT/Aeo8vxJ+L34B2/UzUNTVQdngYVlqrC4Oh4gFXyUh4VKus7v5nVwblm1PweGzd/DWuOYVDhhBEKDVFnb9t9lssNlskj/mUWpX5Lu9tpKSkiQ9YW1zK8tabF1JzyW4K0HtVay5QuH3+8OwYu5tODKLjo0mFuQi/8evqqW+muTMleJDACWWsA4AvDUK+Hq51z07APf5wCHCev0MBLUnPIJCa3ywAMCB03eKBMtdBVYHEn7LxcHEO6W8snzsdjuio6Px+OOP4/HHH0fnzp0rdbw/KvXd53A48NlnnyEpKanEwcomTZokaSE1lc1WvL9EgJt9mrwfQaGE94ApMO38CLCaIWjrwqvf71OoCtp6gHddIK/oL4Lt+lmIDrvbTksrh9ZNyvcsiFol4Jn+gVi/NxUZ2Rb0CamHHo+4xyynmpBBsP56rLDX2B/Uthv3cUfSS/0gWmB1YPvhdPQNqfiIJUqlEvHx8cjJycFLL72EX3/9FW3atKnw8f6o1CuXDz/8EAqFAna7HSaTqdg/Kp/racWfQXgh0vWfun8Q6keegN8rX8J34kfwe2ktVI1+b+sXFEpoR8wsNhil404q8vd/Ud2lurUOLXwwYUAgvNQKaDwEDA71h4eqaFPGq08+hHUzg/Ht8dvYdjgdh85kY/76JBw5W7lPudVFUHvC95n3oGr7eLFt9lu/yVCRfNKzi7d6PMj28qpTpw66deuGH3/8UZLj3VXqlUvLli0RGxuLtm3bFnvQkcrvmL7oVLMKBdA5qOb1cFF4+kDRuORPPepmnaCIXQHjZ8/j3vZ0y5n98O7/XDVVWDOM6avDn3sHQASgUgroE1IPmw+mwWoTEfl4A/Rs74fz10xISTcXed2+hCz0dOEx6v7Iq/czMF78qcg6y5n98OwaKVNF1a+hnwfu5JY+mGfDSvQqzczMhEqlQp06dVBQUICffvoJzz0n7e9iqeESHx+P6OhoXL58GVeuXCm2nc1i5VPXVwVj/u+P4NfzUdWYaYsfhMq/CQSfehBzM53rBK37/LFzJffOKtmppQ86tSw6UGPdEp6PqutGT+sD/+umrvQoMkyQwqd2vV+G92yIZdtTUGAt3jTm6aHAiF7lm264JGlpaZg1axbsdjtEUcSgQYPQt2/fypRbTKnvuPz8wul28/LyJD1hbRMzuDEWfpUEi02ESilg5BMBSLhkRPtmWqg9yjejm7uzZ96EI9sArz4Tkfftp4VDq6s08A7/i9yl1Rh3B0h9pJkWjRtoMLxnA8QdKewG38DPA0+FuUfPPNFmgS35PBR+DeHVexzyf1gHQITg5QvPJ8bJXV616tO5Lg6fvYOE33KLBIynhwJdWvsgrFPFw7Zdu3aIi4uTosxSlWs+F3fjCk/o3yvbZMOvKXn45Tcj4n/KgCgC9XxVWBLTCk0DPOUur0rlH9qAgsNfAxAhaOtBO2IWYC2AslEbjgslkU+2JWP3z4VXhI3rq/F+bBDq1/HA9bQCZGRb0aGFFmqV63+QsWfegHHDmxBzbwMAPB//M9SdB8CRdROqhx6BoH6wwS1rAodDxMHEO9h++PfnXEb0aoiwTpV7zqU6lBouZU1l/NZbb1VJQVJwtXABgMwcK8YvOQ/HPVe4fUPq4o3RzeQrqoo5cjORvezZIqMlq9v3gTb6dRmrqlku38zHy8t+LbJu5BMNETPE/TqNmL75EJYz9zycLSjg99IXUNSpePMPyafUjzPt27dH+/btYTabce7cOTRr1gzNmjWDXq+HxSLdFL61xR2TrUiwAECmUZqZ91yVIy+72DD8jlz3mUbAHWQZi/cYynLT95XD9If3huiAw+QevdyouFLvuYwYMQIAsHHjRvzrX/+CSlW469NPP41x42pX26cUWgR6omUjT1y5Z3j0fl3cY1bNilI2bA5lQAvY064616k7SnvTsLbr1NIHDfw8kHFPt9TwLu5541vdIRy2K784lxUNmkEZ6LpTM9P9ldmFJDs7G7m5uc6BKvPy8pCdXfIw31Q6QRCw8C8tseVQOlKzLHiiox/COtXscBEEAT5jFqDg2FY47higDu4F9SO95S6rRlF7KLD0+SBs/TEN2bk29H/U322mafgjTYe+EJQqWM7/CEXdAHh2exKC4Pr3iqhkZd7Q37p1K5YvX45u3bpBFEX8/PPPeOWVV5xXNq7IFe+5EBHVJuXqLZaeno7TpwvnUejcuTMaNnTtG2wMFyKi0r355ps4cOAA6tevjx07dgAAlixZgh9++AEeHh54+OGHsXjxYtSpU/GrYHZFJiJyUaLogOXcQZiPx8GRkwFFnQbQPDYc6vZhlWoy/Pnnn+Ht7Y2ZM2c6w+Xw4cPo3r07VCoVPvjgAwCFk0ZWFBs0iYhckCg6YNq6EHm7lsOeegli3h3YUy8hb9dymLYugihWfHT10NBQ+PkVHcy0V69ezo5bISEhSE1NrVT9DBciIhdkOXcQ1qunAOsfBr+1FsB6NQHW84eq7Nxbt25F796V63xToXDp06dPpU5KRET3Zz4eVzxY7rIWoOC/26vkvCtWrIBSqURUVFSljlOh0exq4G0aIiKX4sjJuP924/23V8S2bdtw4MABrF27ttIzU1YoXKSeDpOIiIpS1GkAe17pIxQofBtIer5Dhw5h9erV2LBhA7y8Kj+OW6nhUtLsk0DhVUtlR0q+c+cOXn31Vdy4cQNNmjTBxx9/XOzmEgAEBwc7Z0Zr1KgRPvvss0qdl4jIXWgeG468XctLbhrz8IRnt4o/azhjxgwcP34cWVlZ6N27N1555RWsWrUKFovFOZ1K586dMX/+/Aqfo9SuyMuXL7/vC19++eUKn/T9999H3bp1ERsbi1WrViE7O7vELm9dunRBQkLCAx+fXZGJyN3d7S1W7Ka+hyc8WnSBduRslx7BoNQrl/uFx9q1ayt10n379mH9+vUAgOHDh2P8+PGV6k9NRFTTCIIC2pFzYD1/CAX/3Q6HMQMK3wbw7DYCHo/0dulgASr4EGWfPn1w4MCBCp+0a9euOHHiBIDCZrbQ0FDn8r0eeeQRtGvXDiqVCrGxsejfv3+5jn/q1CloNJoK10dEVFm1vfWkynqLTZw4ERkZxXszTJ8+vciyIAildhD44YcfoNPpkJycjGeffRZt2rTBww8/XOa5NRpNrf/BEhHJqcp6i92v6ax+/fpIS0tDQEAA0tLS4O/vX+J+Op0OANC0aVM89thjOH/+fLnChYiI5FVquHTp0qXEEBFFEWazuVInDQ8PR1xcHGJjYxEXF4d+/foV2yc7OxteXl5Qq9XIzMzEL7/8gpiYmEqdl4iIqocsA1dmZWVh+vTpuHXrFho3boyPP/4YdevWxZkzZ/D1119j4cKF+OWXXzBv3jwIggBRFDFhwgQ89dRT5To+e4sREcmLoyITEZHkXLsvGxERuSWGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeRkCZddu3Zh6NChaNeuHc6cOVPqfocOHcLAgQMRERGBVatWVWOFRERUGbKES5s2bbBs2TKEhoaWuo/dbsf8+fOxevVq7Ny5Ezt27MClS5eqsUoiIqoolRwnbdWqVZn7JCYmolmzZmjatCkAYOjQodi3bx+CgoKqujwiIqokWcKlPAwGAwIDA53LOp0OiYmJ5Xqt2WyGXq+vqtKIiMoUHBwsdwmyqrJwmThxIjIyMoqtnz59Ovr3719VpwUAaDSaWv+DJSKSU5WFy9q1ayv1ep1Oh9TUVOeywWCATqerZFVERFQdXLYrcseOHZGUlITk5GRYLBbs3LkT4eHhcpdFRETlIEu47N27F71790ZCQgKef/55TJ48GUDh1clzzz0HAFCpVJg7dy5iYmIwZMgQDB48GK1bt5ajXCIiekCCKIqi3EVITa/X854LEZGMXLZZjIiI3BfDhYiIJMdwISIiyTFciIhIcgwXIiKSHMOFiIgkx3AhIiLJMVyIiEhyDBciIpIcw4WIiCTHcCEiIskxXIiISHIMFyIikhzDhYiIJMdwISIiyTFciIhIcgwXIiKSHMOFiIgkx3AhIiLJMVyIiEhyDBciIpIcw4WIiCTHcCEiIskxXIiISHIqOU66a9cuLF++HJcvX8bmzZvRsWPHEvcLDw+HVquFQqGAUqnEtm3bqrlSIiKqCFnCpU2bNli2bBnmzZtX5r7r1q2Dv79/NVRFRERSkSVcWrVqJcdpiYiomsgSLg9i8uTJEAQBo0ePxujRo8v1GrPZDL1eX8WVERGVLjg4WO4SZFVl4TJx4kRkZGQUWz99+nT079+/XMfYuHEjdDodbt++jUmTJqFly5YIDQ0t83UajabW/2CJiORUZeGydu3aSh9Dp9MBAOrXr4+IiAgkJiaWK1yIiEheLtsVOS8vD7m5uc7/HzlyBK1bt5a5KiIiKg9ZwmXv3r3o3bs3EhIS8Pzzz2Py5MkAAIPBgOeeew4AcPv2bYwdOxZRUVF46qmnEBYWht69e8tRLhERPSBBFEVR7iKkptfrec+FiEhGLtssRkRE7ovhQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCRESSY7gQEZHkGC5ERCQ5hgsREUmO4UJERJJTyV1AbVFgceDA6Sxkm2x4omNdNK6vkbskIqIqw3CpBg6HiFmrL+Nich4AYOP+NPz9hSC0auwlc2VERFVDlmaxJUuWYNCgQYiMjMRLL72EnJycEvc7dOgQBg4ciIiICKxataqaq5TO2SSTM1gAwGx1YMexDBkrIiKqWrKES8+ePbFjxw588803aN68OVauXFlsH7vdjvnz52P16tXYuXMnduzYgUuXLslQbeUphJLWlbCSiKiGkCVcevXqBZWqsEUuJCQEqampxfZJTExEs2bN0LRpU6jVagwdOhT79u2r7lIl0b65Fh2aa53LXhoFInvUl7EiIqKqJfs9l61bt2Lw4MHF1hsMBgQGBjqXdTodEhMTy3VMs9kMvV4vWY1SaOJng14A7CLQsqEIY8ZV6LN49UJUUwUHB8tdgqyqLFwmTpyIjIzi9xWmT5+O/v37AwBWrFgBpVKJqKgoSc+t0Whc6gd79VY+vjv9q3P5XIqIJGMAIns0kLEqIqKqU2Xhsnbt2vtu37ZtGw4cOIC1a9dCKOH+g06nK9JcZjAYoNPppC6zWlxNLSi27rcbefjPTxnIyrUirFNdNA9kzzEiqjlkuedy6NAhrF69GitWrICXV8l/VDt27IikpCQkJyfDYrFg586dCA8Pr+ZKpdGppQ9UyqIBevaqCSu+uYGvf0jDK8t/g/66SabqiIikJ0u4vPvuuzCZTJg0aRKio6Mxd+5cAIVXJ8899xwAQKVSYe7cuYiJicGQIUMwePBgtG7dWo5yK62BnwfemdAc7Zp6o2mABtGP18etTItzu80uYuex2zJWSEQkLUEURVHuIqSm1+td6p7LH11MzsP0f/5WZN2gUH9Me7KpTBUREUmLY4vJoG1Tb3Rt4+tc9tYoMLxnQxkrIiKSluxdkWurdya0wPGLOcgy2tD9kTrw9/WQuyQiIskwXGSiVAro8Yif3GUQEVUJNosREZHkGC5ERCQ5hgsREUmO4UJERJJjuBARkeQYLkREJDmGCxERSY7hQkREkmO4EBGR5BguREQkuRo5/IsrTnNMRLWLSqVy22lCpFAjh9wnIiJ5sVmMiIgkx3AhIiLJMVyIiEhyDBciIpIcw4WIiCTHcCEiIskxXGqZnJwcTJ06FYMGDcLgwYORkJAgd0nk4m7duoXx48djyJAhGDp0KNatW1dk+xdffIG2bdsiMzNTpgrJFdXIhyipdAsXLsQTTzyBTz/9FBaLBQUFBXKXRC5OqVRi1qxZaN++PXJzczFy5Ej07NkTQUFBuHXrFo4cOYLGjRvLXSa5GF651CJGoxE///wzRo0aBQBQq9WoU6eOzFWRqwsICED79u0BAD4+PmjZsiUMBgMAYPHixXj99dchCIKcJZILYrjUIikpKfD398ebb76J4cOHY86cOcjLy5O7LHIjKSkp0Ov16Ny5M77//nsEBASgXbt2cpdFLojhUovYbDacP38eY8aMQVxcHLy8vLBq1Sq5yyI3YTKZMHXqVMyePRtKpRIrV67EtGnT5C6LXBTDpRYJDAxEYGAgOnfuDAAYNGgQzp8/L3NV5A6sViumTp2KyMhIDBgwANevX0dKSgqio6MRHh6O1NRUPPnkk0hPT5e7VHIRvKFfizRs2BCBgYG4cuUKWrZsiaNHj6JVq1Zyl0UuThRFzJkzBy1btsSkSZMAAG3btsXRo0ed+4SHh2PLli3w9/eXq0xyMQyXWubtt9/Ga6+9BqvViqZNm2Lx4sVyl0Qu7uTJk4iPj0ebNm0QHR0NAJgxYwbCwsJkroxcGYfcJyIiyfGeCxERSY7hQkREkmO4EBGR5BguREQkOYYLERFJjuFCdB/ff/89Ll26VOr2jRs3Ii4uDgAwfvx4nDlzprpKI3JpfM6F6D6+//579OnTB0FBQcW22Ww2jBkzRoaqiFwfw4XcQkpKCmJiYhASEoKEhAR06NABI0eOxKefforMzEwsXboUQOGUAmazGZ6enli0aBFatmyJtWvX4uLFi1i8eDEuXryIv/71r9i8eTO8vLyKnGPp0qXYv38/lEolevXqhYiICOzfvx/Hjx/HihUrsGzZMsyZMwft2rXDyZMnMWzYMJhMJnh7e2Py5MnO4zgcDsyePRs6nQ6vvvoqDh8+jGXLlsFisTgfXNVqtcXON3PmzGr9nhJVKZHIDSQnJ4vBwcHihQsXRLvdLo4YMUKcNWuW6HA4xL1794pTpkwRjUajaLVaRVEUxSNHjogvv/yyKIqiaLfbxbFjx4p79uwRR4wYIZ44caLY8TMzM8UBAwaIDodDFEVRzM7OFkVRFGfOnCnu2rXLud8zzzwjzps3z7n86aefiqtXr3ZuS0hIEF999VXxn//8pyiKonj79m1x7NixoslkEkVRFFeuXCkuW7as1PMR1RS8ciG38dBDD6Ft27YAgKCgIPTo0QOCIKBt27a4ceMGjEYjZs6ciWvXrkEQBFitVgCAQqHAe++9h6ioKIwePRqPPvposWP7+vpCo9Fg9uzZ6Nu3L/r06VNqHUOGDCl129y5czF48GBMmTIFAHD69GlcunTJ2XxmtVoREhLyQOcjcke8oU9uQ61WO/+vUCicy4IgwG6345NPPkG3bt2wY8cOrFixAhaLxbl/UlISvL29kZaW5lw3efJkREdHY86cOVCpVNiyZQsGDRqEH374ATExMaXW8cfmtHt16dIF//3vf2E2mwEUDvrYs2dPxMfHIz4+Ht9++y0WLVr0QOcjckcMF6oxjEYjdDodAGD79u1F1i9YsAAbNmzAnTt3sHv3bgDA559/jvj4eCxcuBAmkwlGoxFhYWGYPXs2Ll68CADQarUwmUzlrmHUqFEICwvDtGnTYLPZEBISgl9++QXXrl0DAOTl5eHq1aulno+opmCzGNUYMTExmDVrFlasWFFkxN5FixZh3LhxaNGiBRYuXIgJEyYgNDQU9evXd+5jMpnw4osvOq84Zs2aBaCwCeztt9/G+vXr8emnn5arjkmTJsFoNOKNN97A0qVLsXjxYsyYMcN5JTV9+nRotdoSz0dUU3BUZCIikhybxYiISHIMFyIikhzDhYiIJMdwISIiyTFciIhIcgwXIiKSHMOFiIgk9/9pi0pqVJ6I6wAAAABJRU5ErkJggg==\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1134,31 +958,34 @@ } ], "source": [ - "df_pplacer = df2[df2[\"software\"] == \"pplacer\"].copy()\n", + "pplacer_columns = [\"software\", \"max-strikes\", \"max-pitches\", \"strike-box\", \"query\", \"LL\", \"LL difference\"]\n", + "df_pplacer = df2[df2[\"software\"] == \"pplacer\"][pplacer_columns].copy()\n", + "\n", "df_pplacer[\"max-strikes\"] = df_pplacer[\"max-strikes\"].astype(int)\n", "df_pplacer[\"max-pitches\"] = df_pplacer[\"max-pitches\"].astype(int)\n", "df_pplacer[\"strike-box\"] = df_pplacer[\"strike-box\"].astype(int)\n", + "\n", "sns.catplot(data=df_pplacer, x=\"max-strikes\", row=\"max-pitches\", hue=\"strike-box\", y=\"LL difference\", dodge=True, palette=\"muted\")" ] }, { "cell_type": "code", - "execution_count": 310, + "execution_count": 381, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 310, + "execution_count": 381, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1168,8 +995,11 @@ } ], "source": [ - "df_rappas = df2[df2[\"software\"] == \"rappas\"].copy()\n", + "rappas_columns = [\"software\", \"k\", \"omega\", \"red\", \"query\", \"LL\", \"LL difference\"]\n", + "df_rappas = df2[df2[\"software\"] == \"rappas\"][rappas_columns].copy()\n", + "\n", "df_rappas[\"k\"] = df_rappas[\"k\"].astype(int)\n", + "\n", "sns.catplot(data=df_rappas, x=\"k\", hue=\"omega\", y=\"LL difference\", dodge=True, palette=\"muted\")" ] }, @@ -1184,12 +1014,12 @@ }, { "cell_type": "code", - "execution_count": 338, + "execution_count": 383, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -1221,7 +1051,7 @@ " title = f'PPLACER | max-pitches=40'\n", "\n", " elif order[k] == \"epa\":\n", - " ax = sns.stripplot(data=df_epa, x=\"g\", y=\"LL difference\", ax=axes[i, j], \n", + " ax = sns.stripplot(data=df_epa, x=\"G\", y=\"LL difference\", ax=axes[i, j], \n", " linewidth=0.5, edgecolor='w', dodge=True, palette=palette[2:3])\n", " ylabel = ''\n", " title = f'EPA'\n", From 1c2f7989392a92136facce9aa0f7e55e94d481d6 Mon Sep 17 00:00:00 2001 From: Benjamin Linard Date: Mon, 4 May 2020 16:02:21 +0200 Subject: [PATCH 14/43] * minor parameter changes in examples to reduce analysis time now that D652 is used --- examples/5_CPU_RAM_requirements_evaluation/config.yaml | 8 ++++---- examples/6_placement_likelihood/config.yaml | 8 ++++---- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/examples/5_CPU_RAM_requirements_evaluation/config.yaml b/examples/5_CPU_RAM_requirements_evaluation/config.yaml index ae94bcb..dd15588 100755 --- a/examples/5_CPU_RAM_requirements_evaluation/config.yaml +++ b/examples/5_CPU_RAM_requirements_evaluation/config.yaml @@ -81,8 +81,8 @@ config_pplacer: #it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic #(Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [6,24] - strike-box: [3,12] + max-strikes: [12] + strike-box: [6] max-pitches: [40] #pre-masking, 1=yes, 0=no @@ -128,12 +128,12 @@ config_rappas: #panel of k that is tested #integer in [3,16] (~8-10 recommended, >12 often produces too long computations) - k: [7,8] + k: [6,8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1.5,2.0] recommended. For amino acids, values in [5,15] recommended. - omega: [1.5,2.0] + omega: [2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment diff --git a/examples/6_placement_likelihood/config.yaml b/examples/6_placement_likelihood/config.yaml index 856eec4..9e22283 100755 --- a/examples/6_placement_likelihood/config.yaml +++ b/examples/6_placement_likelihood/config.yaml @@ -54,9 +54,9 @@ config_pplacer: # it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic # (Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [6, 24] - strike-box: [3, 12] - max-pitches: [40] + max-strikes: [6] + strike-box: [3] + max-pitches: [20] #pre-masking, 1=yes, 0=no premask: 1 @@ -106,7 +106,7 @@ config_rappas: #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1.5,2.0] recommended. For amino acids, values in [5,15] recommended. - omega: [1.5, 2.0] + omega: [1.75, 2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment From 4ab65643e077c0bf201896d1170ee89e0ec13a25 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Mon, 11 May 2020 17:45:59 +0200 Subject: [PATCH 15/43] Update README.md --- README.md | 32 +++++++++++++++++++++++++++++--- 1 file changed, 29 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index f3a37a1..4a46698 100755 --- a/README.md +++ b/README.md @@ -156,9 +156,9 @@ Identify the Snakefile corresponding to your needs. Procedure | Snakefile | Description --- | --- | --- -Accuracy (ND + eND) | `eval_accuracy.smk` | Given a reference tree/alignment, compute both the "Node Distance" and "expected Node Distance" for a set of software and a set of conditions. This procedure is based on a pruning approach and an important parameter is the number of prunings that is run (see documentation). -Ressources | `eval_ressources.smk` | Given a reference tree/alignment and a set of query reads, measures CPU/RAM consumptions for a set of software and a set of conditions. An important parameter is the number of repeats from which mean consumptions will be deduced (see documentation). -Likelihood Improvement | `eval_likelihood.smk` | Given a reference tree and alignment, compute tree likelihoods induced by placements under a set of conditions, with a lower likelihood reflecting better placements. +Pruning-based Accuracy evaluation (PAC) | `eval_accuracy.smk` | Given a reference tree/alignment, compute both the "Node Distance" and "expected Node Distance" for a set of software and a set of conditions. This procedure is based on a pruning approach and an important parameter is the number of prunings that is run (see documentation). +Ressources evaluation (RES) | `eval_ressources.smk` | Given a reference tree/alignment and a set of query reads, measures CPU/RAM consumptions for a set of software and a set of conditions. An important parameter is the number of repeats from which mean consumptions will be deduced (see documentation). +Likelihood-based Accuracy evaluation (LAC) | `eval_likelihood.smk` | Given a reference tree and alignment, compute tree likelihoods induced by placements under a set of conditions, with a lower likelihood reflecting better placements. @@ -175,6 +175,32 @@ Mostly related to the formatting of the jplace outputs. Note that these options * *Evolutionary model* A very basic definition of the evolutionary model used in the procedures. Currently, only GTR+G (for nucleotides), JTT+G, WAG+G and LG+G (for amino acids) are supported. +**Notes on memory usage for alignment-free methods:** +Alignment-free methods such as RAPPAS require to build/load a database in memory prior to the placement. +This step can be memory consuming for large datasets. Correctly managing the memory is done as follows: + +1. Do a test with a single RAPPAS run on the tree/alignment you use as input dataset.Write down the peek memory. To measure peek memory you can use: +``` +/usr/bin/time -v command +#Make sure to use the full path, it's not the default linux 'time' command. +``` +And search for the line "Maximum resident set size". + +2. For instance, You conclude RAPPAS requires 8Gb per analysis. Then force RAPPAS to use 8Gb of RAM via the snakemake config file and the field: +```yaml +config_rappas: + memory: 8 +``` + +3. Finally choose the maximum amount of RAM that can be used by snakemake during its launch. For instance, if you have a machine with 32Gb of RAM : +``` +snakemake -p --cores 8 --resources mem_mb=32000 \ +--snakefile eval_accuracy.smk \ +--config workdir=`pwd`/examples/2_placement_accuracy_for_a_bacterial_taxonomy/run \ +--configfile examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml +``` +With this setup, snakemake will wait for 8gb to be free before each RAPPAS launch and will run with a maximum of four RAPPAS launches in parallel (32/8=4). + **4. Test your workflow:** It is strongly recommended to test if your configuration is valid and matches the analyses you intended. From 5526adc466acb74bc8bcb16b85b9e7032bd42ec5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Fr=C3=A9d=C3=A9ric=20Mah=C3=A9?= Date: Thu, 23 Jul 2020 10:20:04 +0200 Subject: [PATCH 16/43] Typo --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 4a46698..db2095d 100755 --- a/README.md +++ b/README.md @@ -238,7 +238,7 @@ Refer to the snakemake [documentation](https://snakemake.readthedocs.io/en/stabl ## Contacts *B Linard, N Romashchenko, F Pardi, E Rivals* -MAB team (Methods and Algorithms in Bioifnormatics), LIRMM, Montpellier, France. +MAB team (Methods and Algorithms in Bioinformatics), LIRMM, Montpellier, France. ## Licence From 0c28a428262c806fa013392ddc026f15b97b6e1e Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Thu, 23 Jul 2020 10:53:35 +0200 Subject: [PATCH 17/43] Update README.md --- README.md | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/README.md b/README.md index db2095d..0313b4d 100755 --- a/README.md +++ b/README.md @@ -2,6 +2,28 @@ [![Build Status](https://travis-ci.com/phylo42/PEWO.svg?branch=master)](https://travis-ci.com/phylo42/PEWO) +- **Documents:** [**Usage**](https://github.com/phylo42/PEWO/wiki), +[**Tutorial**](https://github.com/phylo42/PEWO/wiki), +[![license](https://img.shields.io/github/license/shenwei356/seqkit.svg?maxAge=2592000)](https://github.com/phylo42/RAPPAS/LICENSE) +[![Cross-platform](https://img.shields.io/badge/platform-any-ec2eb4.svg?style=flat)](https://github.com/phylo42/RAPPAS) + + +- **Please cite:** [![doi](https://img.shields.io/static/v1?label=doi&message=10.1093/bioinformatics/btaa657&color=blue)](https://doi.org/10.1093/bioinformatics/btaa657) + + + + + +**PEWO: a collection of workflows to benchmark phylogenetic placement** + +*Benjamin Linard, Nikolai Romashchenko, Fabio Pardi, Eric Rivals* + +*Bioinformatics, btaa657, 22 July 2020* + +*doi.org/10.1093/bioinformatics/btaa657* + +
+ **Benchmark existing placement software and compare placement accuracy on different reference trees.** ## Overview From d2fc346825ccb5f4e2ccb53da4136c8128403d2c Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Thu, 23 Jul 2020 10:54:14 +0200 Subject: [PATCH 18/43] Update README.md --- README.md | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/README.md b/README.md index 0313b4d..efce349 100755 --- a/README.md +++ b/README.md @@ -1,11 +1,9 @@ # PEWO: "Placement Evaluation WOrkflows" -[![Build Status](https://travis-ci.com/phylo42/PEWO.svg?branch=master)](https://travis-ci.com/phylo42/PEWO) +[![Build Status](https://travis-ci.com/phylo42/PEWO.svg?branch=master)](https://travis-ci.com/phylo42/PEWO) [![license](https://img.shields.io/github/license/shenwei356/seqkit.svg?maxAge=2592000)](https://github.com/phylo42/RAPPAS/LICENSE) [![Cross-platform](https://img.shields.io/badge/platform-any-ec2eb4.svg?style=flat)](https://github.com/phylo42/RAPPAS) - **Documents:** [**Usage**](https://github.com/phylo42/PEWO/wiki), [**Tutorial**](https://github.com/phylo42/PEWO/wiki), -[![license](https://img.shields.io/github/license/shenwei356/seqkit.svg?maxAge=2592000)](https://github.com/phylo42/RAPPAS/LICENSE) -[![Cross-platform](https://img.shields.io/badge/platform-any-ec2eb4.svg?style=flat)](https://github.com/phylo42/RAPPAS) - **Please cite:** [![doi](https://img.shields.io/static/v1?label=doi&message=10.1093/bioinformatics/btaa657&color=blue)](https://doi.org/10.1093/bioinformatics/btaa657) From 18899a200f8b3fd01eda9c7497820513b53a5af6 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Thu, 23 Jul 2020 10:55:13 +0200 Subject: [PATCH 19/43] Update README.md --- README.md | 15 +++------------ 1 file changed, 3 insertions(+), 12 deletions(-) diff --git a/README.md b/README.md index efce349..ad834fa 100755 --- a/README.md +++ b/README.md @@ -4,22 +4,13 @@ - **Documents:** [**Usage**](https://github.com/phylo42/PEWO/wiki), [**Tutorial**](https://github.com/phylo42/PEWO/wiki), - - **Please cite:** [![doi](https://img.shields.io/static/v1?label=doi&message=10.1093/bioinformatics/btaa657&color=blue)](https://doi.org/10.1093/bioinformatics/btaa657) - - - - -**PEWO: a collection of workflows to benchmark phylogenetic placement** - -*Benjamin Linard, Nikolai Romashchenko, Fabio Pardi, Eric Rivals* - -*Bioinformatics, btaa657, 22 July 2020* - +**PEWO: a collection of workflows to benchmark phylogenetic placement**
+*Benjamin Linard, Nikolai Romashchenko, Fabio Pardi, Eric Rivals*
+*Bioinformatics, btaa657, 22 July 2020*
*doi.org/10.1093/bioinformatics/btaa657* -
**Benchmark existing placement software and compare placement accuracy on different reference trees.** From b18b7e6302b9ba6487aa97fbf267bdfa57108a3d Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Thu, 23 Jul 2020 10:56:01 +0200 Subject: [PATCH 20/43] Update README.md --- README.md | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index ad834fa..f8e01dd 100755 --- a/README.md +++ b/README.md @@ -1,20 +1,19 @@ # PEWO: "Placement Evaluation WOrkflows" +**Benchmark existing placement software and compare placement accuracy on different reference trees.** + [![Build Status](https://travis-ci.com/phylo42/PEWO.svg?branch=master)](https://travis-ci.com/phylo42/PEWO) [![license](https://img.shields.io/github/license/shenwei356/seqkit.svg?maxAge=2592000)](https://github.com/phylo42/RAPPAS/LICENSE) [![Cross-platform](https://img.shields.io/badge/platform-any-ec2eb4.svg?style=flat)](https://github.com/phylo42/RAPPAS) - **Documents:** [**Usage**](https://github.com/phylo42/PEWO/wiki), -[**Tutorial**](https://github.com/phylo42/PEWO/wiki), +[**Tutorial**](https://github.com/phylo42/PEWO/wiki) - **Please cite:** [![doi](https://img.shields.io/static/v1?label=doi&message=10.1093/bioinformatics/btaa657&color=blue)](https://doi.org/10.1093/bioinformatics/btaa657) **PEWO: a collection of workflows to benchmark phylogenetic placement**
*Benjamin Linard, Nikolai Romashchenko, Fabio Pardi, Eric Rivals*
*Bioinformatics, btaa657, 22 July 2020*
-*doi.org/10.1093/bioinformatics/btaa657*
-**Benchmark existing placement software and compare placement accuracy on different reference trees.** - ## Overview PEWO compiles a set of workflows dedicated to the evaluation of phylogenetic placement algorithms and their software implementation. It focuses on reporting placement accuracy under different conditions and associated computational costs. From df75dfa746680f86db62d9e422cdc57b031d6c87 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Tue, 20 Oct 2020 11:52:50 +0200 Subject: [PATCH 21/43] Update README.md --- README.md | 18 ------------------ 1 file changed, 18 deletions(-) diff --git a/README.md b/README.md index f8e01dd..28e5774 100755 --- a/README.md +++ b/README.md @@ -106,24 +106,6 @@ Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/wiki/IV.-T ### PEWO procedures -* *Node Distance (ND)* : -This standard procedure was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. The reference tree is pruned randomly. For each pruning, the pruned leaves are placed and accuracy is evaluated as the number of nodes separating expected and observed placements. - - -This distance measure between two placements was introduced with EPA and reused in PPlacer and RAPPAS original manuscripts. It is computed as follows: The reference tree is pruned randomly, which removes one sequence of the tree. The pruned sequence serves as query for placement against the pruned tree, but one knows the true solution (ie. the position of that sequence in the original tree). The resulting placement is compared to the true solution by evaluating the accuracy as the number of nodes separating the true and observed placements. This procedure is repeated with numerous prunings. One drawback: the running time. - -* *Expected Node Distance (eND)* : -An improved version of ND, which takes into account placement weights (e.g. Likelihood Weight Ratios, see documentation). - - - -* *Likelihood Improvement (LI)* : -Rapid evaluation of phylogenetic placements designed for developers and rapid evaluation of changes in the code and algorithms. Following placement, a re-optimization simply highlights better or worse results, in terms of likelihood changes. - -* *Ressources (RESS)* : -CPU and peek RAM consumptions are measured for every step required to operate phylogenetic placement (including alignment in alignment-based methods and ancestral state reconstruction + database build in alignment-free methods). This procedure mostly intends to evaluate the scalability of the methods, as punctual analyses or routine placement of large sequence volumes do not induce the same constraints. - - **Software currently supported by PEWO.** * **EPA**(RAxML) (Berger et al, 2010) From 612f462b2b7c467b589f3a47a3104c32d3699ccf Mon Sep 17 00:00:00 2001 From: erivals Date: Thu, 5 Nov 2020 18:18:15 +0100 Subject: [PATCH 22/43] Update README.md update snakemake documenation URL in wiki --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 28e5774..0b76087 100755 --- a/README.md +++ b/README.md @@ -223,7 +223,7 @@ snakemake --snakefile [snakefile].smk -p --core [#cores] Note that the workflow can be launched on a grid environment such as [SunGridEngine](https://en.wikipedia.org/wiki/Oracle_Grid_Engine) or [SLURM](https://en.wikipedia.org/wiki/Slurm_Workload_Manager) (i.e., with `qsub` command). -Refer to the snakemake [documentation](https://snakemake.readthedocs.io/en/stable/executing/cluster-cloud.html#cluster-execution) to learn how to configure a workflow for such environments. +Refer to the snakemake [documentation](https://snakemake.readthedocs.io/en/stable/executing/cluster.html) to learn how to configure a workflow for such environments. From fc74aeae22bf59a80c03bf4b5dcee8d32734dc2f Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Fri, 17 Apr 2020 13:26:34 +0200 Subject: [PATCH 23/43] Basic changes to PEWO to add own implementation of Alignmant-free Phylogenetic Placement with filtered SPAced word Matches. Added Snakemake module for APPSPAM. Configured config for minimal running example. --- config.yaml | 42 +++++++++++++++--------- eval_accuracy.smk | 2 ++ pewo/software.py | 1 + pewo/templates.py | 9 +++++- rules/op/operate_likelihood.smk | 24 ++++++++++++++ rules/op/operate_nodedistance.smk | 3 +- rules/placement/appspam.smk | 53 +++++++++++++++++++++++++++++++ rules/utils/workflow.smk | 3 ++ 8 files changed, 120 insertions(+), 17 deletions(-) create mode 100755 rules/placement/appspam.smk diff --git a/config.yaml b/config.yaml index 36df7d2..87c065e 100755 --- a/config.yaml +++ b/config.yaml @@ -10,20 +10,20 @@ dataset_align: examples/1_fast_test_of_accuracy_procedure/alignment_150.fasta dataset_tree: examples/1_fast_test_of_accuracy_procedure/RAxML_bipartitions.150.BEST.WITH #working directory -workdir: /home/nikolai/dev/pewo/run_accuracy1 +workdir: /home/matthias/Documents/test_PEWO #states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 #which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [epa, epang, rappas, pplacer] +test_soft: [pplacer, appspam] #epa, epang, rappas, # READ GENERATION # Read lengths to generate -read_length: [150,300] +read_length: [150] # Number of random prunings to compute -pruning_count: 50 +pruning_count: 1 ## IF "ACCURACY" IS EVALUATED @@ -68,7 +68,7 @@ config_epa: #proportion of top scoring branch for which full optimization is computed #float in ]0,1] - G: [0.01,0.1] + G: [0.1] ### PPLACER ############################### @@ -79,9 +79,9 @@ config_pplacer: #it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic #(Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [6,12] - strike-box: [3,6] - max-pitches: [40,80] + max-strikes: [12] + strike-box: [6] + max-pitches: [80] #pre-masking, 1=yes, 0=no premask: 1 @@ -98,13 +98,13 @@ config_epang: # h3: heuristic equivalent to pplacer defaults, fast # h4: no heuristic, very very slow but should produce the best accuracy #(Barbera et al, 2019 ; doi: 10.1093/sysbio/syy054) - heuristics: ["h1", "h2", "h3", "h4"] + heuristics: ["h1"] #heuristic-specific parameters can be setup in following lines h1: - g: [0.999,0.99999] + g: [0.99999] h2: - G: [0.01,0.1] + G: [0.1] h3: options: none #reserved if any option appears in future versions h4: @@ -125,12 +125,12 @@ config_rappas: #panel of k that is tested #integer in [2,16] (8~10 recommended, >12 often produces too long computations) - k: [7,8] + k: [8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1,2] recommended. For amino acids, values in [5,15] recommended. - omega: [1.5,2.0] + omega: [2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment @@ -181,7 +181,7 @@ config_apples: # BE : k=1 (Beyer et al., 1974) #methods: ["OLS","FM","BE"] #!warning, be sure to set methods VALUES as UPPER CASE - methods: [OLS,BE] + methods: [BE] #List of placement criterion to test. #Possible values are: @@ -190,8 +190,20 @@ config_apples: # HYBRID : MLSE then ME #criteria: ["MLSE","ME","HYBRID"] #!warning, be sure to set criteria VALUES as UPPER CASE - criteria: [MLSE,ME] + criteria: [ME] +### APP-SPAM +############################### + +config_appspam: + + #appspam placements are based on distance computations between the query and the reference tree + + #List of matching modes to test. + mode: [all] + + #List of number of dont care positions to test. + d: [20] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/eval_accuracy.smk b/eval_accuracy.smk index 2a8eb77..55d56c8 100755 --- a/eval_accuracy.smk +++ b/eval_accuracy.smk @@ -47,6 +47,8 @@ include: #distance-based placements, e.g.: apples include: "rules/placement/apples.smk" +include: + "rules/placement/appspam.smk" #results evaluation and plots include: "rules/op/operate_nodedistance.smk" diff --git a/pewo/software.py b/pewo/software.py index 30c7c87..862d15b 100644 --- a/pewo/software.py +++ b/pewo/software.py @@ -17,6 +17,7 @@ class PlacementSoftware(Enum): APPLES = "apples" RAPPAS = "rappas" RAPPAS2 = "rappas2" + APPSPAM = "appspam" @staticmethod def get_by_value(value): diff --git a/pewo/templates.py b/pewo/templates.py index 2a71c92..e04fc3b 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -67,7 +67,9 @@ def get_experiment_dir_template(config: Dict, software: PlacementSoftware, **kwa elif software == PlacementSoftware.APPLES: return os.path.join(software_dir, input_set_dir_template, "meth{meth}_crit{crit}") elif software == PlacementSoftware.RAPPAS: - return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}", "k{k}_o{o}") + return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}") + elif software == PlacementSoftware.APPSPAM: + return os.path.join(software_dir, input_set_dir_template, "mode{mode}_d{d}") def get_experiment_log_dir_template(config: Dict, software: Software) -> str: @@ -158,6 +160,8 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) return get_common_queryname_template(config) + "_meth{meth}_crit{crit}" elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" + elif software == PlacementSoftware.APPSPAM: + return get_common_queryname_template(config) + "_mode{mode}_d{d}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: @@ -206,6 +210,9 @@ def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs template_args["o"] = config["config_rappas"]["omega"] template_args["red"] = config["config_rappas"]["reduction"] template_args["ar"] = config["config_rappas"]["arsoft"] + elif software == PlacementSoftware.APPSPAM: + template_args["d"] = config["config_appspam"]["d"] + template_args["mode"] = config["config_appspam"]["mode"] else: raise RuntimeError("Unsupported software: " + software.value) return template_args diff --git a/rules/op/operate_likelihood.smk b/rules/op/operate_likelihood.smk index 5765d02..641c006 100644 --- a/rules/op/operate_likelihood.smk +++ b/rules/op/operate_likelihood.smk @@ -179,6 +179,15 @@ rule extend_trees_apples: run: make_extended_tree(input.tree, output.ext_tree, input.jplace) +rule extend_trees_appspam: + input: + jplace=get_output_template(config, PlacementSoftware.APPSPAM, "jplace"), + tree=config["dataset_tree"], + output: + ext_tree=get_output_template(config, PlacementSoftware.APPSPAM, "tree") + run: + make_extended_tree(input.tree, output.ext_tree, input.jplace) + rule calculate_likelihood_epa: """ Calculates likelihood values for the placements produced by EPA. @@ -307,6 +316,21 @@ rule calculate_likelihood_apples: run: _calculate_likelihood(input, output, params, wildcards) +rule calculate_likelihood_apples: + """ + Calculates likelihood values for the placements produced by APPSPAM. + """ + input: + alignment=_get_aligned_query_template(config), + tree=get_output_template(config, PlacementSoftware.APPSPAM, "tree") + output: + csv=get_output_template(config, PlacementSoftware.APPSPAM, "csv") + params: + workdir=cfg.get_work_dir(config), + software=PlacementSoftware.APPSPAM.value, + model="GTR+G" + run: + _calculate_likelihood(input, output, params, wildcards) def _get_csv_output(config: Dict) -> List[str]: """ diff --git a/rules/op/operate_nodedistance.smk b/rules/op/operate_nodedistance.smk index ee11ec6..f9d42dd 100755 --- a/rules/op/operate_nodedistance.smk +++ b/rules/op/operate_nodedistance.smk @@ -25,7 +25,8 @@ rule compute_nodedistance: compute_pplacer=1 if "pplacer" in config["test_soft"] else 0, compute_rappas=1 if "rappas" in config["test_soft"] else 0, compute_apples=1 if "apples" in config["test_soft"] else 0, + compute_appspam=1 if "appspam" in config["test_soft"] else 0, jar=config["pewo_jar"] shell: "java -cp {params.jar} DistanceGenerator_LITE2 {params.workdir} " - "{params.compute_epa} {params.compute_epang} {params.compute_pplacer} {params.compute_rappas} {params.compute_apples} &> {log}" + "{params.compute_epa} {params.compute_epang} {params.compute_pplacer} {params.compute_rappas} {params.compute_apples} {params.compute_appspam} &> {log}" diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk new file mode 100755 index 0000000..f86faaa --- /dev/null +++ b/rules/placement/appspam.smk @@ -0,0 +1,53 @@ +""" +Module to operate placements with APPSPAM +""" + + +__author__ = "Benjamin Linard, Nikolai Romashchenko" +__license__ = "MIT" + +import os +import pewo.config as cfg +from pewo.software import PlacementSoftware, AlignmentSoftware +from pewo.templates import get_output_template, get_log_template, get_software_dir, \ + get_common_queryname_template, get_benchmark_template, get_output_template_args + + +_working_dir = cfg.get_work_dir(config) +_alignment_dir = get_software_dir(config, AlignmentSoftware.HMMER) + +_appspam_place_benchmark_template = get_benchmark_template(config, PlacementSoftware.APPSPAM, + p="pruning", length="length", mode="mode", d="d", + rule_name="placement") if cfg.get_mode(config) == cfg.Mode.RESOURCES else "" + +appspam_benchmark_templates = [_appspam_place_benchmark_template] +appspam_benchmark_template_args = [get_output_template_args(config, PlacementSoftware.APPSPAM),] + + +# FIXME: These are the same methods as in the epang.smk +def _get_appspam_input_reads(config) -> str: + return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_refs") + + +def _get_appspam_input_queries(config) -> str: + return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_queries") + +def _get_appspam_input_tree() -> str: + return os.path.join(_working_dir, "T", "{pruning}.tree") + +rule placement_appspam: + input: + r=_get_appspam_input_reads(config), + q=_get_appspam_input_queries(config), + t=_get_appspam_input_tree(), + output: + jplace=get_output_template(config, PlacementSoftware.APPSPAM, "jplace") + log: + get_log_template(config, PlacementSoftware.APPSPAM) + benchmark: + repeat(_appspam_place_benchmark_template, config["repeats"]) + version: "1.0" + shell: + """ + fswm -i {input.r} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} + """ \ No newline at end of file diff --git a/rules/utils/workflow.smk b/rules/utils/workflow.smk index aee7861..889b5e5 100755 --- a/rules/utils/workflow.smk +++ b/rules/utils/workflow.smk @@ -222,6 +222,9 @@ def _get_resources_tsv(config: Dict, software: PlacementSoftware, **kwargs) -> L elif software == PlacementSoftware.APPLES: software_templates = apples_benchmark_templates + hmmer_benchmark_templates software_template_args = apples_benchmark_template_args + hmmer_benchmark_template_args + elif software == PlacementSoftware.APPSPAM: + software_templates = pplacer_benchmark_templates + hmmer_benchmark_templates + software_template_args = pplacer_benchmark_template_args + hmmer_benchmark_template_args else: raise RuntimeError("Unsupported software: " + software.value) From 128d800bc9778bd62bfd2bdfd978f4a151ff4602 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Fri, 17 Apr 2020 18:39:32 +0200 Subject: [PATCH 24/43] Fixed bug in appspam-snakemake module. --- rules/placement/appspam.smk | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index f86faaa..b64dbad 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -49,5 +49,5 @@ rule placement_appspam: version: "1.0" shell: """ - fswm -i {input.r} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} + fswm -r {input.r} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} """ \ No newline at end of file From cc1a39bbb01828eaf48660d96c8089b58d65ce16 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Tue, 21 Apr 2020 10:19:38 +0200 Subject: [PATCH 25/43] Fixed small bugs to fully integrate APPSPAM to PEWO. --- config.yaml | 2 +- rules/op/operate_likelihood.smk | 2 +- rules/utils/workflow.smk | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/config.yaml b/config.yaml index 87c065e..7af7b55 100755 --- a/config.yaml +++ b/config.yaml @@ -200,7 +200,7 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree #List of matching modes to test. - mode: [all] + mode: [ALL] #List of number of dont care positions to test. d: [20] diff --git a/rules/op/operate_likelihood.smk b/rules/op/operate_likelihood.smk index 641c006..80295c7 100644 --- a/rules/op/operate_likelihood.smk +++ b/rules/op/operate_likelihood.smk @@ -316,7 +316,7 @@ rule calculate_likelihood_apples: run: _calculate_likelihood(input, output, params, wildcards) -rule calculate_likelihood_apples: +rule calculate_likelihood_appspam: """ Calculates likelihood values for the placements produced by APPSPAM. """ diff --git a/rules/utils/workflow.smk b/rules/utils/workflow.smk index 889b5e5..1358ec1 100755 --- a/rules/utils/workflow.smk +++ b/rules/utils/workflow.smk @@ -223,8 +223,8 @@ def _get_resources_tsv(config: Dict, software: PlacementSoftware, **kwargs) -> L software_templates = apples_benchmark_templates + hmmer_benchmark_templates software_template_args = apples_benchmark_template_args + hmmer_benchmark_template_args elif software == PlacementSoftware.APPSPAM: - software_templates = pplacer_benchmark_templates + hmmer_benchmark_templates - software_template_args = pplacer_benchmark_template_args + hmmer_benchmark_template_args + software_templates = appspam_benchmark_templates + hmmer_benchmark_templates + software_template_args = appspam_benchmark_template_args + hmmer_benchmark_template_args else: raise RuntimeError("Unsupported software: " + software.value) From 48021cccffaed89d2e5bfd5a0a5984782446574e Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Wed, 22 Apr 2020 16:58:40 +0200 Subject: [PATCH 26/43] Added plotting support for appspam. Changed input query files for appspam to unaligned reads. --- rules/placement/apples.smk | 1 - rules/placement/appspam.smk | 10 ++++------ scripts/R/eval_accuracy_plots.R | 3 ++- 3 files changed, 6 insertions(+), 8 deletions(-) diff --git a/rules/placement/apples.smk b/rules/placement/apples.smk index 0d28a4b..101cd65 100755 --- a/rules/placement/apples.smk +++ b/rules/placement/apples.smk @@ -28,7 +28,6 @@ apples_benchmark_template_args = [get_output_template_args(config, PlacementSoft def _get_apples_input_reads(config) -> str: return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_refs") - def _get_apples_input_queries(config) -> str: return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_queries") diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index b64dbad..85f190b 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -2,7 +2,6 @@ Module to operate placements with APPSPAM """ - __author__ = "Benjamin Linard, Nikolai Romashchenko" __license__ = "MIT" @@ -25,19 +24,18 @@ appspam_benchmark_template_args = [get_output_template_args(config, PlacementSof # FIXME: These are the same methods as in the epang.smk -def _get_appspam_input_reads(config) -> str: +def _get_appspam_input_sequences(config) -> str: return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_refs") - def _get_appspam_input_queries(config) -> str: - return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_queries") + return os.path.join(_working_dir, "R", "{pruning}_r{length}.fasta") def _get_appspam_input_tree() -> str: return os.path.join(_working_dir, "T", "{pruning}.tree") rule placement_appspam: input: - r=_get_appspam_input_reads(config), + s=_get_appspam_input_sequences(config), q=_get_appspam_input_queries(config), t=_get_appspam_input_tree(), output: @@ -49,5 +47,5 @@ rule placement_appspam: version: "1.0" shell: """ - fswm -r {input.r} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} + fswm -s {input.s} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} >& {log} """ \ No newline at end of file diff --git a/scripts/R/eval_accuracy_plots.R b/scripts/R/eval_accuracy_plots.R index d0fcd52..f259908 100755 --- a/scripts/R/eval_accuracy_plots.R +++ b/scripts/R/eval_accuracy_plots.R @@ -26,8 +26,9 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") +appspam<-c("mode","d") -soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples) +soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) #load data From 5414d5ff7d969cc9a80bdbbf9297094d2bcf12ee Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 23 Apr 2020 09:25:35 +0200 Subject: [PATCH 27/43] Added more parameters for appspam to test accuracy. Adapted config-files for first and second example. --- config.yaml | 12 ++++--- .../config.yaml | 36 ++++++++++++++----- pewo/templates.py | 8 +++-- rules/placement/appspam.smk | 2 +- scripts/R/eval_accuracy_plots.R | 2 +- 5 files changed, 41 insertions(+), 19 deletions(-) diff --git a/config.yaml b/config.yaml index 7af7b55..23673ee 100755 --- a/config.yaml +++ b/config.yaml @@ -17,13 +17,13 @@ workdir: /home/matthias/Documents/test_PEWO states: 0 #which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [pplacer, appspam] #epa, epang, rappas, +test_soft: [pplacer, appspam, apples] #epa, epang, rappas, # READ GENERATION # Read lengths to generate read_length: [150] # Number of random prunings to compute -pruning_count: 1 +pruning_count: 3 ## IF "ACCURACY" IS EVALUATED @@ -190,7 +190,7 @@ config_apples: # HYBRID : MLSE then ME #criteria: ["MLSE","ME","HYBRID"] #!warning, be sure to set criteria VALUES as UPPER CASE - criteria: [ME] + criteria: [MLSE] ### APP-SPAM ############################### @@ -200,10 +200,12 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree #List of matching modes to test. - mode: [ALL] + matchingmode: [ALL, FIRST] + + assignmentmode: [LCA, BEST, RANDOM] #List of number of dont care positions to test. - d: [20] + d: [16, 32] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml index a6e999f..f7933c9 100755 --- a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml +++ b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml @@ -10,16 +10,16 @@ dataset_tree: examples/2_placement_accuracy_for_a_bacterial_taxonomy/RAxML_bipar # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /path/to/examples/2_placement_accuracy_for_a_bacterial_taxonomy/run +workdir: /home/matthias/Documents/test_PEWO # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas,epang,pplacer] +test_soft: [appspam,pplacer] # read lengths to test -read_length: [300] +read_length: [150] ## IF "ACCURACY" IS EVALUATED @@ -28,7 +28,7 @@ read_length: [300] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 5 +pruning_count: 1 ## IF "RESOURCES" ARE EVALUATED @@ -81,8 +81,8 @@ config_pplacer: #it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic #(Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [1,3,6] - strike-box: [1,3,6] + max-strikes: [6] + strike-box: [3] max-pitches: [40] #pre-masking, 1=yes, 0=no @@ -128,12 +128,12 @@ config_rappas: #panel of k that is tested #integer in [3,16] (~8-10 recommended, >12 often produces too long computations) - k: [6,7,8] + k: [8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1.5,2.0] recommended. For amino acids, values in [5,15] recommended. - omega: [1.75,2.0] + omega: [2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment @@ -159,7 +159,6 @@ config_rappas: #currently, only raxml-ng can use multiple threads # #!!! warning, be sure to set arsoft VALUES as UPPER CASE !!! - arsoft: [RAXMLNG] arthreads: 1 #maximum amount of memory available to rappas process @@ -198,6 +197,25 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE,ME,HYBRID] +### APP-SPAM +############################### + +config_appspam: + + #appspam placements are based on distance computations between the query and the reference tree + + #List of matching modes to test. + matchingmode: [ALL, BEST] + + #List of assignment modes to test + assignmentmode: [LCA] + + #Filtering threshold. Number of dont care positions will be multiplied by this constant. + #All words below the threshold will not be considered + filteringthreshold: [0, 10, 20, 30, 40] + + #List of number of dont care positions to test. + d: [32] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/pewo/templates.py b/pewo/templates.py index e04fc3b..6474174 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -69,7 +69,7 @@ def get_experiment_dir_template(config: Dict, software: PlacementSoftware, **kwa elif software == PlacementSoftware.RAPPAS: return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}") elif software == PlacementSoftware.APPSPAM: - return os.path.join(software_dir, input_set_dir_template, "mode{mode}_d{d}") + return os.path.join(software_dir, input_set_dir_template, "matchingmode{matchingmode}_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}") def get_experiment_log_dir_template(config: Dict, software: Software) -> str: @@ -161,7 +161,7 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" elif software == PlacementSoftware.APPSPAM: - return get_common_queryname_template(config) + "_mode{mode}_d{d}" + return get_common_queryname_template(config) + "_matchingmode{matchingmode}_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: @@ -212,7 +212,9 @@ def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs template_args["ar"] = config["config_rappas"]["arsoft"] elif software == PlacementSoftware.APPSPAM: template_args["d"] = config["config_appspam"]["d"] - template_args["mode"] = config["config_appspam"]["mode"] + template_args["matchingmode"] = config["config_appspam"]["matchingmode"] + template_args["assignmentmode"] = config["config_appspam"]["assignmentmode"] + template_args["filteringthreshold"] = config["config_appspam"]["filteringthreshold"] else: raise RuntimeError("Unsupported software: " + software.value) return template_args diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index 85f190b..45850b5 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -47,5 +47,5 @@ rule placement_appspam: version: "1.0" shell: """ - fswm -s {input.s} -q {input.q} -t {input.t} -m {wildcards.mode} -d {wildcards.d} -o {output.jplace} >& {log} + fswm -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -d {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} """ \ No newline at end of file diff --git a/scripts/R/eval_accuracy_plots.R b/scripts/R/eval_accuracy_plots.R index f259908..c93b77f 100755 --- a/scripts/R/eval_accuracy_plots.R +++ b/scripts/R/eval_accuracy_plots.R @@ -26,7 +26,7 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") -appspam<-c("mode","d") +appspam<-c("matchingmode","assignmentmode","filteringthreshold","d") soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) From 5553ba8b5d0daca26afa8609b61ae3b58b267f16 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Mon, 27 Apr 2020 07:45:55 +0200 Subject: [PATCH 28/43] Changed name of program. --- rules/placement/appspam.smk | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index 45850b5..ce995b2 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -47,5 +47,5 @@ rule placement_appspam: version: "1.0" shell: """ - fswm -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -d {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} + appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -d {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} """ \ No newline at end of file From c22b287025b857ae6cb16f43ed8aacf9dd91351d Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Fri, 8 May 2020 09:12:45 +0200 Subject: [PATCH 29/43] Adjustments for evaluation of appspam. --- config.yaml | 13 ++++---- .../config.yaml | 20 ++++++------- .../config.yaml | 30 +++++++++++++++---- .../config.yaml | 6 ++-- rules/placement/appspam.smk | 2 +- scripts/java/PEWO_java | 2 +- 6 files changed, 47 insertions(+), 26 deletions(-) diff --git a/config.yaml b/config.yaml index 23673ee..f11301b 100755 --- a/config.yaml +++ b/config.yaml @@ -10,7 +10,7 @@ dataset_align: examples/1_fast_test_of_accuracy_procedure/alignment_150.fasta dataset_tree: examples/1_fast_test_of_accuracy_procedure/RAxML_bipartitions.150.BEST.WITH #working directory -workdir: /home/matthias/Documents/test_PEWO +workdir: /home/matthias/Documents/test_PEWO_xmp1 #states used in analysis, either '0' for nucleotides or '1' for amino acids @@ -23,7 +23,7 @@ test_soft: [pplacer, appspam, apples] #epa, epang, rappas, # Read lengths to generate read_length: [150] # Number of random prunings to compute -pruning_count: 3 +pruning_count: 10 ## IF "ACCURACY" IS EVALUATED @@ -200,12 +200,15 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree #List of matching modes to test. - matchingmode: [ALL, FIRST] + matchingmode: [ALL, BEST] + + assignmentmode: [BEST, LCA, RANDOM, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] + + filteringthreshold: [0, 40] - assignmentmode: [LCA, BEST, RANDOM] #List of number of dont care positions to test. - d: [16, 32] + d: [8, 12] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml index f7933c9..45c6dfd 100755 --- a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml +++ b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/2_placement_accuracy_for_a_bacterial_taxonomy/RAxML_bipar # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /home/matthias/Documents/test_PEWO +workdir: /home/matthias/Documents/test_PEWO_xmp2 # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [appspam,pplacer] +test_soft: [pplacer, appspam, apples] # read lengths to test read_length: [150] @@ -28,7 +28,7 @@ read_length: [150] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 1 +pruning_count: 5 ## IF "RESOURCES" ARE EVALUATED @@ -38,7 +38,7 @@ pruning_count: 1 # number of identical runs to launch when evaluating RAM/CPU consumption # final measurements are reported as mean of the the runs. -repeats: 1 +repeats: 3 # defines queries source, one of the following: # user: query sequences are loaded from a file target by parameter "query_set" # simulate: queries are simulated from the input alignment (reserved for future upgrades, currently not implemented) @@ -186,7 +186,7 @@ config_apples: # BE : k=1 (Beyer et al., 1974) #methods: ["OLS","FM","BE"] #!warning, be sure to set methods VALUES as UPPER CASE - methods: [OLS,FM,BE] + methods: [FM] #List of placement criterion to test. #Possible values are: @@ -195,7 +195,7 @@ config_apples: # HYBRID : MLSE then ME #criteria: ["MLSE","ME","HYBRID"] #!warning, be sure to set criteria VALUES as UPPER CASE - criteria: [MLSE,ME,HYBRID] + criteria: [HYBRID] ### APP-SPAM ############################### @@ -205,17 +205,17 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree #List of matching modes to test. - matchingmode: [ALL, BEST] + matchingmode: [ALL] #List of assignment modes to test - assignmentmode: [LCA] + assignmentmode: [BEST, LCA, RANDOM, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] #Filtering threshold. Number of dont care positions will be multiplied by this constant. #All words below the threshold will not be considered - filteringthreshold: [0, 10, 20, 30, 40] + filteringthreshold: [0, 40] #List of number of dont care positions to test. - d: [32] + d: [8, 12] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml index a65afe7..cb08883 100755 --- a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml +++ b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/3_placement_accuracy_for_HIV_genomes/compendium-norec-exp # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /path/to/examples/3_placement_accuracy_for_HIV_genomes/run +workdir: /home/matthias/Documents/test_PEWO_xmp3 # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas,epang] +test_soft: [epang, appspam, pplacer] # read lengths to test read_length: [500] @@ -28,7 +28,7 @@ read_length: [500] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 5 +pruning_count: 50 ## IF "RESOURCES" ARE EVALUATED @@ -100,14 +100,14 @@ config_epang: # h3: heuristic equivalent to pplacer defaults, fast # h4: no heuristic, very very slow but should produce the best accuracy #(Barbera et al, 2019 ; doi: 10.1093/sysbio/syy054) - heuristics: ["h1","h2"] + heuristics: ["h2"] h1: #value in ]0,1], value close to 1 recommended by authors g: [0.9, 0.999, 0.99999] h2: #values in ]0,1], higher values increase accuracy but analysis is slower - G: [0.01, 0.1, 0.5] + G: [0.1] h3: options: none #reserved if any option appears in future versions h4: @@ -128,7 +128,7 @@ config_rappas: #panel of k that is tested #integer in [3,16] (~8-10 recommended, >12 often produces too long computations) - k: [7,8] + k: [8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids @@ -198,6 +198,24 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam placements are based on distance computations between the query and the reference tree + + #List of matching modes to test. + matchingmode: [ALL, BEST] + + assignmentmode: [BEST, LCA, APPLES, WEIGHTED, KMER, KMERLCA] + + filteringthreshold: [0, 40] + + + #List of number of dont care positions to test. + d: [8, 12] + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/5_CPU_RAM_requirements_evaluation/config.yaml b/examples/5_CPU_RAM_requirements_evaluation/config.yaml index dd15588..28646e7 100755 --- a/examples/5_CPU_RAM_requirements_evaluation/config.yaml +++ b/examples/5_CPU_RAM_requirements_evaluation/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/5_CPU_RAM_requirements_evaluation/tree_652.nwk # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /path/to/examples/5_CPU_RAM_requirements_evaluation/run +workdir: /home/matthias/Documents/test_PEWO_xmp5 # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas,epang,pplacer] +test_soft: [rappas,epang,pplacer,appspam] # read lengths to test read_length: [150] @@ -45,7 +45,7 @@ repeats: 2 query_type: user # queries used in resource evaluation, >10000 sequences recommended. # MUST be an absolute path (some software do not accept relative path). -query_user: /path/to/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_10000.fas +query_user: /home/matthias/Documents/PEWO/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_10000.fas diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index ce995b2..12ceb66 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -47,5 +47,5 @@ rule placement_appspam: version: "1.0" shell: """ - appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -d {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} + appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -k {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} --histogram >& {log} """ \ No newline at end of file diff --git a/scripts/java/PEWO_java b/scripts/java/PEWO_java index 1b3bfd7..183676c 160000 --- a/scripts/java/PEWO_java +++ b/scripts/java/PEWO_java @@ -1 +1 @@ -Subproject commit 1b3bfd7fa14c4167e1392e061b12b3bd25f5c90e +Subproject commit 183676ccfefd75e41498d3e9adb3c4d9772d94ce From 758d131abf586f1ede9e46bc1dab339210f7a53e Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Mon, 11 May 2020 12:08:03 +0200 Subject: [PATCH 30/43] Bug fix for RAPPAS templates. Added appspam to other plotting functions. --- pewo/templates.py | 7 +++---- rules/placement/appspam.smk | 2 +- scripts/R/eval_accuracy_plots.R | 2 +- scripts/R/eval_likelihood_plots.R | 3 ++- scripts/R/eval_resources_plots.R | 5 ++++- 5 files changed, 11 insertions(+), 8 deletions(-) diff --git a/pewo/templates.py b/pewo/templates.py index 6474174..23843bb 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -67,9 +67,9 @@ def get_experiment_dir_template(config: Dict, software: PlacementSoftware, **kwa elif software == PlacementSoftware.APPLES: return os.path.join(software_dir, input_set_dir_template, "meth{meth}_crit{crit}") elif software == PlacementSoftware.RAPPAS: - return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}") + return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}", "k{k}_o{o}") elif software == PlacementSoftware.APPSPAM: - return os.path.join(software_dir, input_set_dir_template, "matchingmode{matchingmode}_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}") + return os.path.join(software_dir, input_set_dir_template, "assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}") def get_experiment_log_dir_template(config: Dict, software: Software) -> str: @@ -161,7 +161,7 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" elif software == PlacementSoftware.APPSPAM: - return get_common_queryname_template(config) + "_matchingmode{matchingmode}_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}" + return get_common_queryname_template(config) + "_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: @@ -212,7 +212,6 @@ def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs template_args["ar"] = config["config_rappas"]["arsoft"] elif software == PlacementSoftware.APPSPAM: template_args["d"] = config["config_appspam"]["d"] - template_args["matchingmode"] = config["config_appspam"]["matchingmode"] template_args["assignmentmode"] = config["config_appspam"]["assignmentmode"] template_args["filteringthreshold"] = config["config_appspam"]["filteringthreshold"] else: diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index 12ceb66..ab955c7 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -47,5 +47,5 @@ rule placement_appspam: version: "1.0" shell: """ - appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.matchingmode} -g {wildcards.assignmentmode} -k {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} --histogram >& {log} + appspam -s {input.s} -q {input.q} -t {input.t} -g {wildcards.assignmentmode} -k {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} """ \ No newline at end of file diff --git a/scripts/R/eval_accuracy_plots.R b/scripts/R/eval_accuracy_plots.R index c93b77f..b36a0f3 100755 --- a/scripts/R/eval_accuracy_plots.R +++ b/scripts/R/eval_accuracy_plots.R @@ -26,7 +26,7 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") -appspam<-c("matchingmode","assignmentmode","filteringthreshold","d") +appspam<-c("assignmentmode","filteringthreshold","d") soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) diff --git a/scripts/R/eval_likelihood_plots.R b/scripts/R/eval_likelihood_plots.R index 4aedd45..d6aad62 100755 --- a/scripts/R/eval_likelihood_plots.R +++ b/scripts/R/eval_likelihood_plots.R @@ -26,8 +26,9 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") +appspam<-c("assignmentmode","filteringthreshold","d") -soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples) +soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) #load data diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index f2af394..c108979 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -29,6 +29,8 @@ rappasplacement<-c("k","o","red","ar") apples<-c("meth","crit") hmmbuild<-NULL ansrec<-c("red","ar") +appspam<-c("assignmentmode","filteringthreshold","d") + soft_params<-list( "epa-placement"=epa, @@ -41,7 +43,8 @@ soft_params<-list( "rappas-placement"=rappasplacement, "apples-placement"=apples, "hmm-align"=hmmbuild, - "ansrec"=ansrec + "ansrec"=ansrec, + "appspam"=appspam ) From 22147a2cef9b00a3fe1e89379f32a0833e2c28a4 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Mon, 11 May 2020 12:08:31 +0200 Subject: [PATCH 31/43] Changes to configs to reduce runtime while testing around. --- config.yaml | 3 --- .../config.yaml | 23 ++++++++++++++++--- .../config.yaml | 6 +---- .../config.yaml | 3 --- .../config.yaml | 2 +- 5 files changed, 22 insertions(+), 15 deletions(-) diff --git a/config.yaml b/config.yaml index f11301b..14992f2 100755 --- a/config.yaml +++ b/config.yaml @@ -199,9 +199,6 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree - #List of matching modes to test. - matchingmode: [ALL, BEST] - assignmentmode: [BEST, LCA, RANDOM, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] filteringthreshold: [0, 40] diff --git a/examples/1_fast_test_of_accuracy_procedure/config.yaml b/examples/1_fast_test_of_accuracy_procedure/config.yaml index e6b045c..ab56af8 100755 --- a/examples/1_fast_test_of_accuracy_procedure/config.yaml +++ b/examples/1_fast_test_of_accuracy_procedure/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/1_fast_test_of_accuracy_procedure/RAxML_bipartitions.150. # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /path/to/examples/1_fast_test_of_accuracy_procedure/run +workdir: /home/matthias/Documents/test_PEWO_xmp1 # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas,epang,pplacer] +test_soft: [appspam,epang,pplacer,rappas] # read lengths to test read_length: [150] @@ -28,7 +28,7 @@ read_length: [150] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 2 +pruning_count: 14 ## IF "RESOURCES" ARE EVALUATED @@ -198,6 +198,23 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam placements are based on distance computations between the query and the reference tree + + #List of assignment modes to test + assignmentmode: [BEST, LCA, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] + + #Filtering threshold. Number of dont care positions will be multiplied by this constant. + #All words below the threshold will not be considered + filteringthreshold: [0, 40] + + #List of number of dont care positions to test. + d: [8, 12] + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml index 45c6dfd..7123b56 100755 --- a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml +++ b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml @@ -199,16 +199,12 @@ config_apples: ### APP-SPAM ############################### - config_appspam: #appspam placements are based on distance computations between the query and the reference tree - #List of matching modes to test. - matchingmode: [ALL] - #List of assignment modes to test - assignmentmode: [BEST, LCA, RANDOM, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] + assignmentmode: [BEST, LCA, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] #Filtering threshold. Number of dont care positions will be multiplied by this constant. #All words below the threshold will not be considered diff --git a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml index cb08883..6997394 100755 --- a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml +++ b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml @@ -205,9 +205,6 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree - #List of matching modes to test. - matchingmode: [ALL, BEST] - assignmentmode: [BEST, LCA, APPLES, WEIGHTED, KMER, KMERLCA] filteringthreshold: [0, 40] diff --git a/examples/5_CPU_RAM_requirements_evaluation/config.yaml b/examples/5_CPU_RAM_requirements_evaluation/config.yaml index 28646e7..cd60642 100755 --- a/examples/5_CPU_RAM_requirements_evaluation/config.yaml +++ b/examples/5_CPU_RAM_requirements_evaluation/config.yaml @@ -16,7 +16,7 @@ workdir: /home/matthias/Documents/test_PEWO_xmp5 states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas,epang,pplacer,appspam] +test_soft: [epang,pplacer,appspam] # read lengths to test read_length: [150] From ed6ba580a550d887ebfd061f716ae1e73a5557be Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Mon, 11 May 2020 19:24:12 +0200 Subject: [PATCH 32/43] Added code for resource evaluation of appspam. --- eval_resources.smk | 2 ++ .../config.yaml | 28 ++++++++++++++----- rules/placement/appspam.smk | 4 +-- rules/utils/workflow.smk | 5 ++-- scripts/R/eval_resources_plots.R | 5 ++-- 5 files changed, 29 insertions(+), 15 deletions(-) diff --git a/eval_resources.smk b/eval_resources.smk index 7c72d78..1e528e1 100755 --- a/eval_resources.smk +++ b/eval_resources.smk @@ -46,6 +46,8 @@ include: #distance-based placements, e.g.: apples include: "rules/placement/apples.smk" +include: + "rules/placement/appspam.smk" #results and plots include: "rules/op/operate_plots.smk" diff --git a/examples/5_CPU_RAM_requirements_evaluation/config.yaml b/examples/5_CPU_RAM_requirements_evaluation/config.yaml index cd60642..d997637 100755 --- a/examples/5_CPU_RAM_requirements_evaluation/config.yaml +++ b/examples/5_CPU_RAM_requirements_evaluation/config.yaml @@ -15,8 +15,8 @@ workdir: /home/matthias/Documents/test_PEWO_xmp5 # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 -# which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [epang,pplacer,appspam] +# which software to test, at least one of : epa, epang, pplacer, rappas, apples, appspam +test_soft: [epang, apples, appspam] # read lengths to test read_length: [150] @@ -38,17 +38,14 @@ pruning_count: 0 # number of identical runs to launch when evaluating RAM/CPU consumption # final measurements are reported as mean of the the runs. -repeats: 2 +repeats: 1 # defines queries source, one of the following: # user: query sequences are loaded from a file target by parameter "query_set" # simulate: queries are simulated from the input alignment (reserved for future upgrades, currently not implemented) query_type: user # queries used in resource evaluation, >10000 sequences recommended. # MUST be an absolute path (some software do not accept relative path). -query_user: /home/matthias/Documents/PEWO/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_10000.fas - - - +query_user: /home/matthias/Documents/PEWO/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_100000.fas ######################################################################################################################## # PER SOFTWARE CONFIGURATION @@ -198,6 +195,23 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam placements are based on distance computations between the query and the reference tree + + #List of assignment modes to test + assignmentmode: [BESTSCORE] + + #Filtering threshold. Number of dont care positions will be multiplied by this constant. + #All words below the threshold will not be considered + filteringthreshold: [0] + + #List of number of dont care positions to test. + d: [8] + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index ab955c7..a06c500 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -16,14 +16,12 @@ _working_dir = cfg.get_work_dir(config) _alignment_dir = get_software_dir(config, AlignmentSoftware.HMMER) _appspam_place_benchmark_template = get_benchmark_template(config, PlacementSoftware.APPSPAM, - p="pruning", length="length", mode="mode", d="d", + p="pruning", length="length", mode="assignmentmode", d="d", threshold="filteringthreshold", rule_name="placement") if cfg.get_mode(config) == cfg.Mode.RESOURCES else "" appspam_benchmark_templates = [_appspam_place_benchmark_template] appspam_benchmark_template_args = [get_output_template_args(config, PlacementSoftware.APPSPAM),] - -# FIXME: These are the same methods as in the epang.smk def _get_appspam_input_sequences(config) -> str: return os.path.join(_alignment_dir, "{pruning}", get_common_queryname_template(config) + ".fasta_refs") diff --git a/rules/utils/workflow.smk b/rules/utils/workflow.smk index 1358ec1..7e50639 100755 --- a/rules/utils/workflow.smk +++ b/rules/utils/workflow.smk @@ -14,7 +14,6 @@ from pewo.software import AlignmentSoftware, PlacementSoftware from pewo.templates import get_software_dir, get_common_queryname_template, get_common_template_args, \ get_output_template, get_output_template_args - def get_accuracy_nd_plots() -> List[str]: """ Creates a list of plot files that will be computed in the accuracy mode. @@ -223,8 +222,8 @@ def _get_resources_tsv(config: Dict, software: PlacementSoftware, **kwargs) -> L software_templates = apples_benchmark_templates + hmmer_benchmark_templates software_template_args = apples_benchmark_template_args + hmmer_benchmark_template_args elif software == PlacementSoftware.APPSPAM: - software_templates = appspam_benchmark_templates + hmmer_benchmark_templates - software_template_args = appspam_benchmark_template_args + hmmer_benchmark_template_args + software_templates = appspam_benchmark_templates + software_template_args = appspam_benchmark_template_args else: raise RuntimeError("Unsupported software: " + software.value) diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index c108979..44335eb 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -42,9 +42,9 @@ soft_params<-list( "rappas-dbbuild"=rappasdbbuild, "rappas-placement"=rappasplacement, "apples-placement"=apples, + "appspam-placement"=appspam, "hmm-align"=hmmbuild, - "ansrec"=ansrec, - "appspam"=appspam + "ansrec"=ansrec ) @@ -287,6 +287,7 @@ analyses["epang_h3"]<-c("hmmer-align", "epang-h3-placement") analyses["epang_h4"]<-c("hmmer-align", "epang-h4-placement") analyses["pplacer"]<-c("hmmer-align", "pplacer-placement") analyses["apples"]<-c("hmmer-align", "apples-placement") +analyses["appspam"]<-c("appspam-placement") analyses["rappas"]<-c("ansrec", "rappas-dbbuild","rappas-placement") results<-list() From 788c601df7bff85591f08e04ba38112299faa037 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 28 May 2020 12:00:04 +0200 Subject: [PATCH 33/43] Small fixes. --- scripts/R/eval_resources_plots.R | 2 +- scripts/java/PEWO_java | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index 44335eb..a4a12dc 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -29,7 +29,7 @@ rappasplacement<-c("k","o","red","ar") apples<-c("meth","crit") hmmbuild<-NULL ansrec<-c("red","ar") -appspam<-c("assignmentmode","filteringthreshold","d") +appspam<-c("mode","threshold","d") soft_params<-list( diff --git a/scripts/java/PEWO_java b/scripts/java/PEWO_java index 183676c..1b3bfd7 160000 --- a/scripts/java/PEWO_java +++ b/scripts/java/PEWO_java @@ -1 +1 @@ -Subproject commit 183676ccfefd75e41498d3e9adb3c4d9772d94ce +Subproject commit 1b3bfd7fa14c4167e1392e061b12b3bd25f5c90e From 2fed0d1a2911f4f449a8b574a5e08e6583612af8 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Mon, 13 Jul 2020 09:18:15 +0200 Subject: [PATCH 34/43] Changed parameters for appspam. --- config.yaml | 7 ++----- pewo/templates.py | 9 ++++----- rules/placement/appspam.smk | 5 ++--- scripts/R/eval_accuracy_plots.R | 2 +- scripts/R/eval_likelihood_plots.R | 2 +- scripts/R/eval_resources_plots.R | 2 +- 6 files changed, 11 insertions(+), 16 deletions(-) diff --git a/config.yaml b/config.yaml index 14992f2..4100bb0 100755 --- a/config.yaml +++ b/config.yaml @@ -199,13 +199,10 @@ config_appspam: #appspam placements are based on distance computations between the query and the reference tree - assignmentmode: [BEST, LCA, RANDOM, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] - - filteringthreshold: [0, 40] - + mode: [LCACOUNT] #List of number of dont care positions to test. - d: [8, 12] + w: [8, 12] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/pewo/templates.py b/pewo/templates.py index 23843bb..6f490a1 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -69,7 +69,7 @@ def get_experiment_dir_template(config: Dict, software: PlacementSoftware, **kwa elif software == PlacementSoftware.RAPPAS: return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}", "k{k}_o{o}") elif software == PlacementSoftware.APPSPAM: - return os.path.join(software_dir, input_set_dir_template, "assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}") + return os.path.join(software_dir, input_set_dir_template, "mode{mode}_w{w}") def get_experiment_log_dir_template(config: Dict, software: Software) -> str: @@ -161,7 +161,7 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" elif software == PlacementSoftware.APPSPAM: - return get_common_queryname_template(config) + "_assignmentmode{assignmentmode}_filteringthreshold{filteringthreshold}_d{d}" + return get_common_queryname_template(config) + "mode{mode}_w{w}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: @@ -211,9 +211,8 @@ def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs template_args["red"] = config["config_rappas"]["reduction"] template_args["ar"] = config["config_rappas"]["arsoft"] elif software == PlacementSoftware.APPSPAM: - template_args["d"] = config["config_appspam"]["d"] - template_args["assignmentmode"] = config["config_appspam"]["assignmentmode"] - template_args["filteringthreshold"] = config["config_appspam"]["filteringthreshold"] + template_args["w"] = config["config_appspam"]["w"] + template_args["mode"] = config["config_appspam"]["mode"] else: raise RuntimeError("Unsupported software: " + software.value) return template_args diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index a06c500..c92243b 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -11,12 +11,11 @@ from pewo.software import PlacementSoftware, AlignmentSoftware from pewo.templates import get_output_template, get_log_template, get_software_dir, \ get_common_queryname_template, get_benchmark_template, get_output_template_args - _working_dir = cfg.get_work_dir(config) _alignment_dir = get_software_dir(config, AlignmentSoftware.HMMER) _appspam_place_benchmark_template = get_benchmark_template(config, PlacementSoftware.APPSPAM, - p="pruning", length="length", mode="assignmentmode", d="d", threshold="filteringthreshold", + p="pruning", length="length", mode="assignmentmode", w="w", rule_name="placement") if cfg.get_mode(config) == cfg.Mode.RESOURCES else "" appspam_benchmark_templates = [_appspam_place_benchmark_template] @@ -45,5 +44,5 @@ rule placement_appspam: version: "1.0" shell: """ - appspam -s {input.s} -q {input.q} -t {input.t} -g {wildcards.assignmentmode} -k {wildcards.d} -o {output.jplace} --threshold {wildcards.filteringthreshold} >& {log} + appspam -s {input.s} -q {input.q} -t {input.t} -g {wildcards.mode} -w {wildcards.w} -o {output.jplace} >& {log} """ \ No newline at end of file diff --git a/scripts/R/eval_accuracy_plots.R b/scripts/R/eval_accuracy_plots.R index b36a0f3..d3b59bf 100755 --- a/scripts/R/eval_accuracy_plots.R +++ b/scripts/R/eval_accuracy_plots.R @@ -26,7 +26,7 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") -appspam<-c("assignmentmode","filteringthreshold","d") +appspam<-c("mode","w") soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) diff --git a/scripts/R/eval_likelihood_plots.R b/scripts/R/eval_likelihood_plots.R index d6aad62..a3c6e40 100755 --- a/scripts/R/eval_likelihood_plots.R +++ b/scripts/R/eval_likelihood_plots.R @@ -26,7 +26,7 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") -appspam<-c("assignmentmode","filteringthreshold","d") +appspam<-c("mode","w") soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index a4a12dc..06bb076 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -29,7 +29,7 @@ rappasplacement<-c("k","o","red","ar") apples<-c("meth","crit") hmmbuild<-NULL ansrec<-c("red","ar") -appspam<-c("mode","threshold","d") +appspam<-c("mode","w") soft_params<-list( From f7e9977880c0a35f9327cb3188a1b4166ede0c08 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 16 Jul 2020 11:49:07 +0200 Subject: [PATCH 35/43] Update templates.py Fixed missing underscore in query_template for appspam. --- pewo/templates.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pewo/templates.py b/pewo/templates.py index 6f490a1..2bd36b0 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -161,7 +161,7 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" elif software == PlacementSoftware.APPSPAM: - return get_common_queryname_template(config) + "mode{mode}_w{w}" + return get_common_queryname_template(config) + "_mode{mode}_w{w}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: From 3c2b14099f2e0e78525af6ea8c00e6ed4b262b12 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 12 Nov 2020 10:12:47 +0100 Subject: [PATCH 36/43] Added number of patterns to list of parameters. --- pewo/templates.py | 5 +++-- rules/placement/appspam.smk | 4 ++-- scripts/R/eval_accuracy_plots.R | 2 +- scripts/R/eval_resources_plots.R | 2 +- 4 files changed, 7 insertions(+), 6 deletions(-) diff --git a/pewo/templates.py b/pewo/templates.py index 2bd36b0..5c92352 100644 --- a/pewo/templates.py +++ b/pewo/templates.py @@ -69,7 +69,7 @@ def get_experiment_dir_template(config: Dict, software: PlacementSoftware, **kwa elif software == PlacementSoftware.RAPPAS: return os.path.join(software_dir, input_set_dir_template, "red{red}_ar{ar}", "k{k}_o{o}") elif software == PlacementSoftware.APPSPAM: - return os.path.join(software_dir, input_set_dir_template, "mode{mode}_w{w}") + return os.path.join(software_dir, input_set_dir_template, "mode{mode}_w{w}_pattern{pattern}") def get_experiment_log_dir_template(config: Dict, software: Software) -> str: @@ -161,7 +161,7 @@ def get_queryname_template(config: Dict, software: PlacementSoftware, **kwargs) elif software == PlacementSoftware.RAPPAS: return get_common_queryname_template(config) + "_k{k}_o{o}_red{red}_ar{ar}" elif software == PlacementSoftware.APPSPAM: - return get_common_queryname_template(config) + "_mode{mode}_w{w}" + return get_common_queryname_template(config) + "_mode{mode}_w{w}_pattern{pattern}" def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs) -> Dict[str, Any]: @@ -213,6 +213,7 @@ def get_output_template_args(config: Dict, software: PlacementSoftware, **kwargs elif software == PlacementSoftware.APPSPAM: template_args["w"] = config["config_appspam"]["w"] template_args["mode"] = config["config_appspam"]["mode"] + template_args["pattern"] = config["config_appspam"]["pattern"] else: raise RuntimeError("Unsupported software: " + software.value) return template_args diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index c92243b..b1cfb2f 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -15,7 +15,7 @@ _working_dir = cfg.get_work_dir(config) _alignment_dir = get_software_dir(config, AlignmentSoftware.HMMER) _appspam_place_benchmark_template = get_benchmark_template(config, PlacementSoftware.APPSPAM, - p="pruning", length="length", mode="assignmentmode", w="w", + p="pruning", length="length", mode="mode", w="w", pattern="pattern", rule_name="placement") if cfg.get_mode(config) == cfg.Mode.RESOURCES else "" appspam_benchmark_templates = [_appspam_place_benchmark_template] @@ -44,5 +44,5 @@ rule placement_appspam: version: "1.0" shell: """ - appspam -s {input.s} -q {input.q} -t {input.t} -g {wildcards.mode} -w {wildcards.w} -o {output.jplace} >& {log} + appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.mode} -w {wildcards.w} -p {wildcards.pattern} -o {output.jplace} >& {log} """ \ No newline at end of file diff --git a/scripts/R/eval_accuracy_plots.R b/scripts/R/eval_accuracy_plots.R index d3b59bf..04828b3 100755 --- a/scripts/R/eval_accuracy_plots.R +++ b/scripts/R/eval_accuracy_plots.R @@ -26,7 +26,7 @@ epang_h4<-c() pplacer<-c("ms","sb","mp") rappas<-c("k","o","red","ar") apples<-c("meth","crit") -appspam<-c("mode","w") +appspam<-c("mode","w","pattern") soft_params<-list(epa=epa,epang_h1=epang_h1,epang_h2=epang_h2,epang_h3=epang_h3,epang_h4=epang_h4,pplacer=pplacer,rappas=rappas,apples=apples,appspam=appspam) diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index 06bb076..632c0fc 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -29,7 +29,7 @@ rappasplacement<-c("k","o","red","ar") apples<-c("meth","crit") hmmbuild<-NULL ansrec<-c("red","ar") -appspam<-c("mode","w") +appspam<-c("mode","w","pattern") soft_params<-list( From 83c7f746d8bccf93d34199b0d5701d774f2588c4 Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 12 Nov 2020 11:06:30 +0100 Subject: [PATCH 37/43] Updated config files in examples with appspam. --- config.yaml | 49 ++++++++++++------- envs/environment.yaml | 1 + .../config.yaml | 36 ++++++++------ .../config.yaml | 48 +++++++++++------- .../config.yaml | 40 +++++++++------ .../coleoptera_12S/config_12S.yaml | 26 ++++++++++ .../coleoptera_16S/config_16S.yaml | 26 ++++++++++ .../coleoptera_cob/config_cob.yaml | 25 ++++++++++ .../coleoptera_cox1/config_cox1.yaml | 25 ++++++++++ .../config.yaml | 42 ++++++++++------ examples/6_placement_likelihood/config.yaml | 25 ++++++++++ 11 files changed, 264 insertions(+), 79 deletions(-) diff --git a/config.yaml b/config.yaml index 4100bb0..1eddc7e 100755 --- a/config.yaml +++ b/config.yaml @@ -10,20 +10,20 @@ dataset_align: examples/1_fast_test_of_accuracy_procedure/alignment_150.fasta dataset_tree: examples/1_fast_test_of_accuracy_procedure/RAxML_bipartitions.150.BEST.WITH #working directory -workdir: /home/matthias/Documents/test_PEWO_xmp1 +workdir: /home/nikolai/dev/pewo/run_accuracy1 #states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 #which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [pplacer, appspam, apples] #epa, epang, rappas, +test_soft: [epa, epang, rappas, pplacer] # READ GENERATION # Read lengths to generate -read_length: [150] +read_length: [150,300] # Number of random prunings to compute -pruning_count: 10 +pruning_count: 50 ## IF "ACCURACY" IS EVALUATED @@ -68,7 +68,7 @@ config_epa: #proportion of top scoring branch for which full optimization is computed #float in ]0,1] - G: [0.1] + G: [0.01,0.1] ### PPLACER ############################### @@ -79,9 +79,9 @@ config_pplacer: #it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic #(Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [12] - strike-box: [6] - max-pitches: [80] + max-strikes: [6,12] + strike-box: [3,6] + max-pitches: [40,80] #pre-masking, 1=yes, 0=no premask: 1 @@ -98,13 +98,13 @@ config_epang: # h3: heuristic equivalent to pplacer defaults, fast # h4: no heuristic, very very slow but should produce the best accuracy #(Barbera et al, 2019 ; doi: 10.1093/sysbio/syy054) - heuristics: ["h1"] + heuristics: ["h1", "h2", "h3", "h4"] #heuristic-specific parameters can be setup in following lines h1: - g: [0.99999] + g: [0.999,0.99999] h2: - G: [0.1] + G: [0.01,0.1] h3: options: none #reserved if any option appears in future versions h4: @@ -125,12 +125,12 @@ config_rappas: #panel of k that is tested #integer in [2,16] (8~10 recommended, >12 often produces too long computations) - k: [8] + k: [7,8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1,2] recommended. For amino acids, values in [5,15] recommended. - omega: [2.0] + omega: [1.5,2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment @@ -181,7 +181,7 @@ config_apples: # BE : k=1 (Beyer et al., 1974) #methods: ["OLS","FM","BE"] #!warning, be sure to set methods VALUES as UPPER CASE - methods: [BE] + methods: [OLS,BE] #List of placement criterion to test. #Possible values are: @@ -190,20 +190,35 @@ config_apples: # HYBRID : MLSE then ME #criteria: ["MLSE","ME","HYBRID"] #!warning, be sure to set criteria VALUES as UPPER CASE - criteria: [MLSE] + criteria: [MLSE,ME] ### APP-SPAM ############################### config_appspam: - #appspam placements are based on distance computations between the query and the reference tree + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. mode: [LCACOUNT] - #List of number of dont care positions to test. + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. w: [8, 12] + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] + + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE ######################################################################################################################## diff --git a/envs/environment.yaml b/envs/environment.yaml index 9bca367..397d0f9 100644 --- a/envs/environment.yaml +++ b/envs/environment.yaml @@ -7,6 +7,7 @@ dependencies: - rappas - epa-ng - pplacer + - appspam - dendropy - _libgcc_mutex=0.1=conda_forge - _openmp_mutex=4.5=0_gnu diff --git a/examples/1_fast_test_of_accuracy_procedure/config.yaml b/examples/1_fast_test_of_accuracy_procedure/config.yaml index ab56af8..5965d9d 100755 --- a/examples/1_fast_test_of_accuracy_procedure/config.yaml +++ b/examples/1_fast_test_of_accuracy_procedure/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/1_fast_test_of_accuracy_procedure/RAxML_bipartitions.150. # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /home/matthias/Documents/test_PEWO_xmp1 +workdir: /path/to/examples/1_fast_test_of_accuracy_procedure/run # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [appspam,epang,pplacer,rappas] +test_soft: [rappas,epang,pplacer] # read lengths to test read_length: [150] @@ -28,7 +28,7 @@ read_length: [150] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 14 +pruning_count: 2 ## IF "RESOURCES" ARE EVALUATED @@ -203,18 +203,26 @@ config_apples: config_appspam: - #appspam placements are based on distance computations between the query and the reference tree - - #List of assignment modes to test - assignmentmode: [BEST, LCA, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] - - #Filtering threshold. Number of dont care positions will be multiplied by this constant. - #All words below the threshold will not be considered - filteringthreshold: [0, 40] - - #List of number of dont care positions to test. - d: [8, 12] + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml index 7123b56..896d24a 100755 --- a/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml +++ b/examples/2_placement_accuracy_for_a_bacterial_taxonomy/config.yaml @@ -10,16 +10,16 @@ dataset_tree: examples/2_placement_accuracy_for_a_bacterial_taxonomy/RAxML_bipar # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /home/matthias/Documents/test_PEWO_xmp2 +workdir: /path/to/examples/2_placement_accuracy_for_a_bacterial_taxonomy/run # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [pplacer, appspam, apples] +test_soft: [rappas,epang,pplacer] # read lengths to test -read_length: [150] +read_length: [300] ## IF "ACCURACY" IS EVALUATED @@ -38,7 +38,7 @@ pruning_count: 5 # number of identical runs to launch when evaluating RAM/CPU consumption # final measurements are reported as mean of the the runs. -repeats: 3 +repeats: 1 # defines queries source, one of the following: # user: query sequences are loaded from a file target by parameter "query_set" # simulate: queries are simulated from the input alignment (reserved for future upgrades, currently not implemented) @@ -81,8 +81,8 @@ config_pplacer: #it uses a 2-step heuristic similar to EPA but called the "baseball" heuristic #(Matsen et al, 2012 ; doi: 10.1186/1471-2105-11-538) - max-strikes: [6] - strike-box: [3] + max-strikes: [1,3,6] + strike-box: [1,3,6] max-pitches: [40] #pre-masking, 1=yes, 0=no @@ -128,12 +128,12 @@ config_rappas: #panel of k that is tested #integer in [3,16] (~8-10 recommended, >12 often produces too long computations) - k: [8] + k: [6,7,8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids #For DNA, values in [1.5,2.0] recommended. For amino acids, values in [5,15] recommended. - omega: [2.0] + omega: [1.75,2.0] #reduction setup, e.g. gap/non-gap ratio #above which a site of the input alignment @@ -159,6 +159,7 @@ config_rappas: #currently, only raxml-ng can use multiple threads # #!!! warning, be sure to set arsoft VALUES as UPPER CASE !!! + arsoft: [RAXMLNG] arthreads: 1 #maximum amount of memory available to rappas process @@ -186,7 +187,7 @@ config_apples: # BE : k=1 (Beyer et al., 1974) #methods: ["OLS","FM","BE"] #!warning, be sure to set methods VALUES as UPPER CASE - methods: [FM] + methods: [OLS,FM,BE] #List of placement criterion to test. #Possible values are: @@ -195,23 +196,34 @@ config_apples: # HYBRID : MLSE then ME #criteria: ["MLSE","ME","HYBRID"] #!warning, be sure to set criteria VALUES as UPPER CASE - criteria: [HYBRID] + criteria: [MLSE,ME,HYBRID] ### APP-SPAM ############################### + config_appspam: - #appspam placements are based on distance computations between the query and the reference tree + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] - #List of assignment modes to test - assignmentmode: [BEST, LCA, PHYLOKMERS, APPLES, WEIGHTED, KMER, KMERLCA] + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] - #Filtering threshold. Number of dont care positions will be multiplied by this constant. - #All words below the threshold will not be considered - filteringthreshold: [0, 40] + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] - #List of number of dont care positions to test. - d: [8, 12] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml index 6997394..de4b59a 100755 --- a/examples/3_placement_accuracy_for_HIV_genomes/config.yaml +++ b/examples/3_placement_accuracy_for_HIV_genomes/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/3_placement_accuracy_for_HIV_genomes/compendium-norec-exp # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /home/matthias/Documents/test_PEWO_xmp3 +workdir: /path/to/examples/3_placement_accuracy_for_HIV_genomes/run # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [epang, appspam, pplacer] +test_soft: [rappas,epang] # read lengths to test read_length: [500] @@ -28,7 +28,7 @@ read_length: [500] ### this section matters only when you run the "eval_accuracy.smk" workflow # number of random prunings to compute -pruning_count: 50 +pruning_count: 5 ## IF "RESOURCES" ARE EVALUATED @@ -100,14 +100,14 @@ config_epang: # h3: heuristic equivalent to pplacer defaults, fast # h4: no heuristic, very very slow but should produce the best accuracy #(Barbera et al, 2019 ; doi: 10.1093/sysbio/syy054) - heuristics: ["h2"] + heuristics: ["h1","h2"] h1: #value in ]0,1], value close to 1 recommended by authors g: [0.9, 0.999, 0.99999] h2: #values in ]0,1], higher values increase accuracy but analysis is slower - G: [0.1] + G: [0.01, 0.1, 0.5] h3: options: none #reserved if any option appears in future versions h4: @@ -128,7 +128,7 @@ config_rappas: #panel of k that is tested #integer in [3,16] (~8-10 recommended, >12 often produces too long computations) - k: [8] + k: [7,8] #panel of omega that is tested, rappas probability threshold is Thr=(omega/#states)^k #integer in ]0,#states] with #states=4 for nucleotides and 20 for amino acids @@ -203,16 +203,26 @@ config_apples: config_appspam: - #appspam placements are based on distance computations between the query and the reference tree - - assignmentmode: [BEST, LCA, APPLES, WEIGHTED, KMER, KMERLCA] - - filteringthreshold: [0, 40] - - - #List of number of dont care positions to test. - d: [8, 12] + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_12S/config_12S.yaml b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_12S/config_12S.yaml index 3c9bac5..884e685 100644 --- a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_12S/config_12S.yaml +++ b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_12S/config_12S.yaml @@ -198,6 +198,32 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_16S/config_16S.yaml b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_16S/config_16S.yaml index 22b891d..6457345 100644 --- a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_16S/config_16S.yaml +++ b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_16S/config_16S.yaml @@ -198,7 +198,33 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] + + ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE ######################################################################################################################## diff --git a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cob/config_cob.yaml b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cob/config_cob.yaml index b453c59..7876b03 100644 --- a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cob/config_cob.yaml +++ b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cob/config_cob.yaml @@ -198,6 +198,31 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cox1/config_cox1.yaml b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cox1/config_cox1.yaml index af7a78b..413c126 100644 --- a/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cox1/config_cox1.yaml +++ b/examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cox1/config_cox1.yaml @@ -198,6 +198,31 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/examples/5_CPU_RAM_requirements_evaluation/config.yaml b/examples/5_CPU_RAM_requirements_evaluation/config.yaml index d997637..32fc1f0 100755 --- a/examples/5_CPU_RAM_requirements_evaluation/config.yaml +++ b/examples/5_CPU_RAM_requirements_evaluation/config.yaml @@ -10,13 +10,13 @@ dataset_tree: examples/5_CPU_RAM_requirements_evaluation/tree_652.nwk # working directory #MUST be an absolute path (some software do not accept relative path) -workdir: /home/matthias/Documents/test_PEWO_xmp5 +workdir: /path/to/examples/5_CPU_RAM_requirements_evaluation/run # states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 -# which software to test, at least one of : epa, epang, pplacer, rappas, apples, appspam -test_soft: [epang, apples, appspam] +# which software to test, at least one of : epa, epang, pplacer, rappas, apples +test_soft: [rappas,epang,pplacer] # read lengths to test read_length: [150] @@ -38,14 +38,17 @@ pruning_count: 0 # number of identical runs to launch when evaluating RAM/CPU consumption # final measurements are reported as mean of the the runs. -repeats: 1 +repeats: 2 # defines queries source, one of the following: # user: query sequences are loaded from a file target by parameter "query_set" # simulate: queries are simulated from the input alignment (reserved for future upgrades, currently not implemented) query_type: user # queries used in resource evaluation, >10000 sequences recommended. # MUST be an absolute path (some software do not accept relative path). -query_user: /home/matthias/Documents/PEWO/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_100000.fas +query_user: /path/to/examples/5_CPU_RAM_requirements_evaluation/EMP_92_studies_10000.fas + + + ######################################################################################################################## # PER SOFTWARE CONFIGURATION @@ -200,17 +203,26 @@ config_apples: config_appspam: - #appspam placements are based on distance computations between the query and the reference tree + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) - #List of assignment modes to test - assignmentmode: [BESTSCORE] - - #Filtering threshold. Number of dont care positions will be multiplied by this constant. - #All words below the threshold will not be considered - filteringthreshold: [0] - - #List of number of dont care positions to test. - d: [8] + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## diff --git a/examples/6_placement_likelihood/config.yaml b/examples/6_placement_likelihood/config.yaml index 9e22283..2be1421 100755 --- a/examples/6_placement_likelihood/config.yaml +++ b/examples/6_placement_likelihood/config.yaml @@ -172,6 +172,31 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [8, 12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE From 8a4af9dcf5616e0a4cbf0640c8ac87864acc600d Mon Sep 17 00:00:00 2001 From: Matthias Blanke Date: Thu, 19 Nov 2020 15:04:30 +0100 Subject: [PATCH 38/43] Added appspam to TravisCI tests. --- .../tests/1_travis_accuracy_test/config.yaml | 27 ++++++++++++++++++- .../2_travis_likelihood_test/config.yaml | 27 ++++++++++++++++++- 2 files changed, 52 insertions(+), 2 deletions(-) diff --git a/travis/tests/1_travis_accuracy_test/config.yaml b/travis/tests/1_travis_accuracy_test/config.yaml index 86c1960..32f89ed 100755 --- a/travis/tests/1_travis_accuracy_test/config.yaml +++ b/travis/tests/1_travis_accuracy_test/config.yaml @@ -16,7 +16,7 @@ workdir: travis/tests/1_travis_accuracy_test/run states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas, epa, epang, pplacer, apples] +test_soft: [rappas, epa, epang, pplacer, apples, appspam] # read lengths to test read_length: [150] @@ -196,6 +196,31 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE diff --git a/travis/tests/2_travis_likelihood_test/config.yaml b/travis/tests/2_travis_likelihood_test/config.yaml index ca94c11..282a4ed 100755 --- a/travis/tests/2_travis_likelihood_test/config.yaml +++ b/travis/tests/2_travis_likelihood_test/config.yaml @@ -16,7 +16,7 @@ workdir: travis/tests/2_travis_likelihood_test/run states: 0 # which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [rappas, epa, epang, pplacer] +test_soft: [rappas, epa, epang, pplacer, appspam] # read lengths to test read_length: [150] @@ -198,6 +198,31 @@ config_apples: #!warning, be sure to set criteria VALUES as UPPER CASE criteria: [MLSE] +### APP-SPAM +############################### + +config_appspam: + + #appspam calculates phylogenetic distances between all query and reference distances based on + #filtered spaced word matches. The placement position is determined with different heuristics (mode). + #(Blanke, Morgenstern, 2020 ; https://doi.org/10.1101/2020.10.19.344986) + + #List of placement heuristics to test. + #Possible values are: + # MINDIST : Above reference with smallest phylogenetic distance. + # SPAMCOUNT : Above reference with most filtered spaced word matches. + # LCADIST : LCA of two leaves with smallest phylogenetic distances. + # LCACOUNT : LCA of two leaves with most filtered spaced word matches. + # APPLES : Our calculated distances are used as input matrix for APPLES. + mode: [LCACOUNT] + + #List of weights for the pattern to be tested (number of match positions). + #Largest values tend result in shorter running times. w between [8, 16] recommended, use 12 as default. + w: [12] + + #Number of pattern from which spaced words are generated. + #At the moment 1 is heavily recommended. + pattern: [1] ######################################################################################################################## # OPTIONS COMMON TO ALL SOFTWARE From 8229d11a1b7977c79db6ea4676e2a85d64b4a379 Mon Sep 17 00:00:00 2001 From: Benjamin Linard <22132778+blinard-BIOINFO@users.noreply.github.com> Date: Fri, 20 Nov 2020 10:45:52 +0100 Subject: [PATCH 39/43] Update appspam.smk Change __author__ to Mathias Blanke --- rules/placement/appspam.smk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/rules/placement/appspam.smk b/rules/placement/appspam.smk index b1cfb2f..84cc798 100755 --- a/rules/placement/appspam.smk +++ b/rules/placement/appspam.smk @@ -2,7 +2,7 @@ Module to operate placements with APPSPAM """ -__author__ = "Benjamin Linard, Nikolai Romashchenko" +__author__ = "Matthias Blanke" __license__ = "MIT" import os @@ -45,4 +45,4 @@ rule placement_appspam: shell: """ appspam -s {input.s} -q {input.q} -t {input.t} -m {wildcards.mode} -w {wildcards.w} -p {wildcards.pattern} -o {output.jplace} >& {log} - """ \ No newline at end of file + """ From f8a2c05de756a6f8115f44dd0190ed5927d1e777 Mon Sep 17 00:00:00 2001 From: Benjamin Linard Date: Fri, 20 Nov 2020 12:06:53 +0100 Subject: [PATCH 40/43] * added app-spam in readme * corrected example directories in .gitignore --- .gitignore | 13 ++++++++++--- README.md | 5 +++-- 2 files changed, 13 insertions(+), 5 deletions(-) diff --git a/.gitignore b/.gitignore index ce8280d..5c67671 100755 --- a/.gitignore +++ b/.gitignore @@ -11,6 +11,13 @@ run *.csv *.tsv #directories built by demos -demos/HIV_accuracy_test/run -demos/16SrRNA_accuracy_test/run -demos/16SrRNA_resource_test/run +examples/1_fast_test_of_accuracy_procedure/run +examples/2_placement_accuracy_for_a_bacterial_taxonomy/run +examples/3_placement_accuracy_for_HIV_genomes/run +examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_12S/run +examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_16S/run +examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cob/run +examples/4_search_for_most_accurate_taxonomic_marker/coleoptera_cox1/run +examples/5_CPU_RAM_requirements_evaluation/run +examples/6_placement_likelihood/run + diff --git a/README.md b/README.md index 0b76087..114c68b 100755 --- a/README.md +++ b/README.md @@ -113,11 +113,12 @@ Please read the [dedicated wiki page](https://github.com/phylo42/PEWO/wiki/IV.-T * **EPA-ng** (Barbera et al, 2019) * **RAPPAS** (Linard et al, 2019) * **APPLES** (Balaban et al, 2019) +* **App-SpaM** (Blanke et al, [preprint](https://doi.org/10.1101/2020.10.19.344986)) + PEWO can easily be extended to integrate new tools for phylogenetic placement, and new tools are welcome. -As of March 2020, these tools are the main software for phylogenetic placement. To the best of your knowledge, no other implementation of phylogenetic placement algorithms are available (with a conda package). - +As of November 2020, these tools are the main software for phylogenetic placement. To the best of your knowledge, no other implementation of phylogenetic placement algorithms are available (with a conda package). If you implement a new method, you are welcome to contact us for requesting future support. You can also implement a new snakemake module and contribute to PEWO via pull requests (see the [documentation](https://github.com/phylo42/PEWO/wiki/Developer-instructions) for contribution guidelines). From 4b848fad86bb09ceec9279dac391c6cb4ec1e5b4 Mon Sep 17 00:00:00 2001 From: Benjamin Linard Date: Fri, 20 Nov 2020 12:28:18 +0100 Subject: [PATCH 41/43] * compute_nodedistance updated according to changes made on PEWO_java side, contributors should not have to modify PEWO_java anymore --- config.yaml | 4 ++-- rules/op/operate_nodedistance.smk | 11 +++-------- 2 files changed, 5 insertions(+), 10 deletions(-) diff --git a/config.yaml b/config.yaml index 1eddc7e..26c49f8 100755 --- a/config.yaml +++ b/config.yaml @@ -16,8 +16,8 @@ workdir: /home/nikolai/dev/pewo/run_accuracy1 #states used in analysis, either '0' for nucleotides or '1' for amino acids states: 0 -#which software to test, at least one of : epa, epang, pplacer, rappas, apples -test_soft: [epa, epang, rappas, pplacer] +#which software to test, at least one of : epa, epang, pplacer, rappas, apples, appspam +test_soft: [epa, epang, rappas, pplacer,appspam] # READ GENERATION # Read lengths to generate diff --git a/rules/op/operate_nodedistance.smk b/rules/op/operate_nodedistance.smk index f9d42dd..22e9fe9 100755 --- a/rules/op/operate_nodedistance.smk +++ b/rules/op/operate_nodedistance.smk @@ -7,6 +7,7 @@ __license__ = "MIT" import os import pewo.config as cfg +from pewo.software import PlacementSoftware _working_dir = cfg.get_work_dir(config) @@ -20,13 +21,7 @@ rule compute_nodedistance: os.path.join(_working_dir, "logs", "compute_nd.log") params: workdir=_working_dir, - compute_epa=1 if "epa" in config["test_soft"] else 0, - compute_epang=1 if "epang" in config["test_soft"] else 0, - compute_pplacer=1 if "pplacer" in config["test_soft"] else 0, - compute_rappas=1 if "rappas" in config["test_soft"] else 0, - compute_apples=1 if "apples" in config["test_soft"] else 0, - compute_appspam=1 if "appspam" in config["test_soft"] else 0, + software_dir_list=','.join(PlacementSoftware.get_by_value(soft_str).value.upper() for soft_str in config["test_soft"]), jar=config["pewo_jar"] shell: - "java -cp {params.jar} DistanceGenerator_LITE2 {params.workdir} " - "{params.compute_epa} {params.compute_epang} {params.compute_pplacer} {params.compute_rappas} {params.compute_apples} {params.compute_appspam} &> {log}" + "java -cp {params.jar} DistanceGenerator_LITE2 {params.workdir} {params.software_dir_list} &> {log}" From fc411c560af8c7c19a20f8dc9b0f45d50f936e4f Mon Sep 17 00:00:00 2001 From: Benjamin Linard Date: Tue, 24 Nov 2020 15:26:33 +0100 Subject: [PATCH 42/43] * updated PEWO_java module * added __pycache__ to gitignore --- .gitignore | 1 + scripts/java/PEWO_java | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/.gitignore b/.gitignore index 5c67671..6b0489c 100755 --- a/.gitignore +++ b/.gitignore @@ -5,6 +5,7 @@ logs run #snakemake temporary files .snakemake +__pycache__ #images/results built by workflows *.svg *.pdf diff --git a/scripts/java/PEWO_java b/scripts/java/PEWO_java index 1b3bfd7..9033133 160000 --- a/scripts/java/PEWO_java +++ b/scripts/java/PEWO_java @@ -1 +1 @@ -Subproject commit 1b3bfd7fa14c4167e1392e061b12b3bd25f5c90e +Subproject commit 9033133f70db2fe2fd07bf2db695ce6be183151d From 577d981ed0fea0b33d7cc47264eac87699f9964a Mon Sep 17 00:00:00 2001 From: "B. Linard h" Date: Mon, 8 Feb 2021 20:16:10 +0100 Subject: [PATCH 43/43] * corrected bug in RES where plots generated in R were crashing if only one software is tested. --- scripts/R/eval_resources_plots.R | 108 +++++++++++++++++-------------- scripts/java/PEWO_java | 2 +- 2 files changed, 62 insertions(+), 48 deletions(-) diff --git a/scripts/R/eval_resources_plots.R b/scripts/R/eval_resources_plots.R index 632c0fc..3b4c9b2 100644 --- a/scripts/R/eval_resources_plots.R +++ b/scripts/R/eval_resources_plots.R @@ -4,7 +4,7 @@ args = commandArgs(trailingOnly=TRUE) # test if there is at least one argument if (length(args)<1) { - stop("The directory containing benchmark results must be supplies as 1st argument.\n", call.=FALSE) + stop("The directory containing benchmark results must be supplied as 1st argument.\n", call.=FALSE) } library(RColorBrewer) @@ -43,7 +43,7 @@ soft_params<-list( "rappas-placement"=rappasplacement, "apples-placement"=apples, "appspam-placement"=appspam, - "hmm-align"=hmmbuild, + "hmmer-align"=hmmbuild, "ansrec"=ansrec ) @@ -277,6 +277,7 @@ for (i in 1:length(stats_to_plot)) { # ansrec + dbbuild + placement*sample for alignment-free approches #this section NEEDS improvements, it is not dynamic as previous plots +#TODO: rework section below to be dynamic #associate operations to analyses analyses<-list() @@ -291,63 +292,76 @@ analyses["appspam"]<-c("appspam-placement") analyses["rappas"]<-c("ansrec", "rappas-dbbuild","rappas-placement") results<-list() +op_merged<-c() +i<-1 -#rappas times in seconds: AR + dbbuild + placement -ansrec_and_dbbuild<-merge(results_per_op["ansrec"][[1]],results_per_op["rappas-dbbuild"][[1]],by=c("red","ar")) -all_op<-merge(ansrec_and_dbbuild,results_per_op["rappas-placement"][[1]],by=c("red","ar","o","k")) - -#time for 1 analysis: AR + dbbuild + placement -all_op["sample_x1"]<-all_op["s.x"]+all_op["s.y"]+all_op["s"] - -#time for 1000 analyses : AR + dbbuild + 100 placements -all_op["sample_x1000"]<-all_op["s.x"]+all_op["s.y"]+all_op["s"]*1000 -all_op["operation"]<-"rappas" - -#results= simplier table -results[[1]]<-all_op[,c("operation","labels","k","o","red","ar","sample_x1","sample_x1000")] -op_analyzed<-c("rappas") -#create label from parameter combination as last column -results[[1]]["labels_short"]<-rep("",dim(results[[1]])[1]) -for (line in 1:dim(results[[1]])[1]) { - label<-"" - elts<-strsplit(results[[1]][line,"labels"],"_")[[1]] - for ( idx in 2:length(elts)) { - label<-paste0(label,elts[idx],"_") - } - results[[1]][line,"labels_short"]<-label -} +#### RAPPAS case + +if ( ("rappas-dbbuild" %in% op_analyzed) & ("rappas-placement" %in% op_analyzed) & ("ansrec" %in% op_analyzed) ) { + + #rappas times in seconds: AR + dbbuild + placement + ansrec_and_dbbuild<-merge(results_per_op["ansrec"][[1]],results_per_op["rappas-dbbuild"][[1]],by=c("red","ar")) + all_op<-merge(ansrec_and_dbbuild,results_per_op["rappas-placement"][[1]],by=c("red","ar","o","k")) + + #time for 1 analysis: AR + dbbuild + placement + all_op["sample_x1"]<-all_op["s.x"]+all_op["s.y"]+all_op["s"] + + #time for 1000 analyses : AR + dbbuild + 100 placements + all_op["sample_x1000"]<-all_op["s.x"]+all_op["s.y"]+all_op["s"]*1000 + all_op["operation"]<-"rappas" -#alignment-based times in seconds: align + placement -i<-2 -for ( op in names(results_per_op)) { - if ( op=="ansrec" || op=="rappas-dbbuild" || op=="rappas-placement" || op=="hmmer-align") { - next - } - results_per_op[op][[1]]["sample_x1"]<-results_per_op[op][[1]]["s"] - for (line in 1:dim(results_per_op[op][[1]])[1]) { - results_per_op[op][[1]]["sample_x1"][line,]<-results_per_op[op][[1]]["s"][line,]+results_per_op[op][[1]]["s"][1,] - } - results_per_op[op][[1]]["sample_x1000"]<-results_per_op[op][[1]]["s"] - for (line in 1:dim(results_per_op[op][[1]])[1]) { - results_per_op[op][[1]]["sample_x1000"][line,]<-results_per_op[op][[1]]["s"][line,]*1000+results_per_op["hmmer-align"][[1]]["s"][1,]*1000 - } #results= simplier table - op_analyzed<-c(op_analyzed,op) - results[[i]]<-results_per_op[op][[1]][,c("operation","labels",soft_params[op][[1]],"sample_x1","sample_x1000")] + results[[i]]<-all_op[,c("operation","labels","k","o","red","ar","sample_x1","sample_x1000")] + op_merged<-c("rappas") #create label from parameter combination as last column - results[[i]]["labels_short"]<-rep("",dim(results[[i]])[1]) - for (line in 1:dim(results[[i]])[1]) { + results[[i]]["labels_short"]<-rep("",dim(results[[1]])[1]) + for (line in 1:dim(results[[1]])[1]) { label<-"" - elts<-strsplit(results[[i]][line,"labels"],"_")[[1]] + elts<-strsplit(results[[1]][line,"labels"],"_")[[1]] for ( idx in 2:length(elts)) { label<-paste0(label,elts[idx],"_") } results[[i]][line,"labels_short"]<-label } - i<-i+1 } -names(results)<-op_analyzed + +#### ALIGNMENT-based + +if ( "hmmer-align" %in% op_analyzed ) { + + #alignment-based times in seconds: align + placement + for ( op in names(results_per_op)) { + if ( op=="ansrec" || op=="rappas-dbbuild" || op=="rappas-placement" || op=="hmmer-align" ) { + next + } + results_per_op[op][[1]]["sample_x1"]<-results_per_op[op][[1]]["s"] + for (line in 1:dim(results_per_op[op][[1]])[1]) { + results_per_op[op][[1]]["sample_x1"][line,]<-results_per_op[op][[1]]["s"][line,]+results_per_op["hmmer-align"][[1]]["s"][1,] + } + results_per_op[op][[1]]["sample_x1000"]<-results_per_op[op][[1]]["s"] + for (line in 1:dim(results_per_op[op][[1]])[1]) { + results_per_op[op][[1]]["sample_x1000"][line,]<-results_per_op[op][[1]]["s"][line,]*1000+results_per_op["hmmer-align"][[1]]["s"][1,]*1000 + } + #results= simplier table + op_merged<-c(op_merged,op) + results[[i]]<-results_per_op[op][[1]][,c("operation","labels",soft_params[op][[1]],"sample_x1","sample_x1000")] + #create label from parameter combination as last column + results[[i]]["labels_short"]<-rep("",dim(results[[i]])[1]) + for (line in 1:dim(results[[i]])[1]) { + label<-"" + elts<-strsplit(results[[i]][line,"labels"],"_")[[1]] + for ( idx in 2:length(elts)) { + label<-paste0(label,elts[idx],"_") + } + results[[i]][line,"labels_short"]<-label + } + + i<-i+1 + } +} + +names(results)<-op_merged ########################################### ## PLOTS 1 : summary plot per operation diff --git a/scripts/java/PEWO_java b/scripts/java/PEWO_java index 9033133..1b3bfd7 160000 --- a/scripts/java/PEWO_java +++ b/scripts/java/PEWO_java @@ -1 +1 @@ -Subproject commit 9033133f70db2fe2fd07bf2db695ce6be183151d +Subproject commit 1b3bfd7fa14c4167e1392e061b12b3bd25f5c90e