🚧 R installation, dependencies update

- currently fails on M1/M2 amd64 architectures as no build is available for R=3.6.3 (check if R version could be updated)
RasmussenLab · Aug 10, 2023 · e68ac36 · e68ac36
1 parent 212e616
commit e68ac36
Show file tree

Hide file tree

Showing 5 changed files with 54 additions and 29 deletions.
diff --git a/.github/workflows/ci.yaml b/.github/workflows/ci.yaml
@@ -46,17 +46,18 @@ jobs:
     # # currently part of environment
     # - name: Install package and install library
     #   run: |
-    #     pip install pytest
-    - name: Run Tests on installed package
-      run: pytest . 
+    #     pip install pytest pytest-cov
+    - name: Run Unit tests on installed package
+      run: |
+        pytest . 
     - name: View papermill help message for notebooks (as scripts)
       run: |
         cd project
         papermill 01_0_split_data.ipynb --help-notebook
         papermill 01_1_train_VAE.ipynb --help-notebook
         papermill 01_1_train_DAE.ipynb --help-notebook
         papermill 01_1_train_CF.ipynb --help-notebook
-    - name: Run demo workflow
+    - name: Run demo workflow (integration test)
       run: | 
        cd project
        snakemake -p -c1 --configfile config/single_dev_dataset/example/config.yaml -n

diff --git a/environment.yml b/environment.yml
@@ -3,28 +3,27 @@ name: vaep
 channels:
   # - defaults
   - conda-forge
-  - bioconda
   - pytorch
   - fastai
+  - bioconda
   - plotly
 dependencies:
-  # - biopython # Aligner
   - python=3.8
-  - numpy=1.20
+  - numpy
   - pandas=1
   - scipy>=1.6
   # plotting
-  - matplotlib=3.3
+  - matplotlib
   - python-kaleido
   - plotly
   - seaborn
   - pip
   # ML
-  - pytorch=1.10
-  - scikit-learn=1.0
+  - pytorch=1
+  - scikit-learn
   - fastai
   - torchvision
-  - cudatoolkit=10.2
+  - cudatoolkit
   - tensorboard
   - umap-learn
   # stats
@@ -41,38 +40,40 @@ dependencies:
   - ipykernel
   - ipython
   - ipywidgets
-  - jupytext
   - jupyterlab # standalone jupyter installation
-  # - jupyterlab_code_formatter
-  # - jupyterlab-git
+  # - jupyter_contrib_nbextensions # delete configuration file if you see an error: https://github.com/jupyter/nbconvert/issues/526#issuecomment-277552771
   - jupyter-dash
   - papermill # execute ipynb's
   # R packages (listed in NAGuideR)
   - r-base=3.6
-  - r-irkernel  
+  - r-irkernel
   - r-reshape2
   - r-stringi # + rmarkdown hack for reshape2
   - r-stringr # reshape2
-  # - r-gdata
-  # - r-glmnet
+  - r-tidyverse
+  - r-gdata
+  - r-glmnet
   - r-e1071
   - r-norm
   - r-missforest
   - r-vim
   - r-mice
-  - r-tidyverse
   # - bioconductor-biocinstaller
   # - r-imputelcmd # bioconda
-  - bioconductor-impute
-  - bioconductor-pcamethods
+  # - bioconductor-impute
+  # - bioconductor-pcamethods
   # - rrcovNA, GMSimpute
   # SeqKnn, pcaMethods, DreamAI # bioconductor
   # dev
   - pytest
   - pytest-cov
-  - jupytext
-  - flake8
-  - flake8-bugbear
+  # - jupytext
+  # - flake8
+  # - flake8-bugbear
+  # - build
+  # - pre-commit
+  # - jupyterlab_code_formatter
+  # - jupyterlab-git
   - pip:
     - -e .
     - mrmr-selection

diff --git a/project/01_1_train_NAGuideR_methods.R b/project/01_1_train_NAGuideR_methods.R
@@ -6,7 +6,7 @@
 #       extension: .R
 #       format_name: light
 #       format_version: '1.5'
-#       jupytext_version: 1.14.5
+#       jupytext_version: 1.15.0
 #   kernelspec:
 #     display_name: R
 #     language: R
@@ -25,7 +25,7 @@ packages_base_R <- c("BiocManager", "reshape2", "data.table", "readr", "tibble")
 install_rpackage  <- function(pkg){
     # If not installed, install the package
     if (!require(pkg, character.only = TRUE)) {
-        install.packages(pkg)
+        install.packages(pkg) #, dependencies = TRUE)
         library(pkg, character.only = TRUE)
     }
 
@@ -111,17 +111,20 @@ nafunctions <- function(x,method="zero"){
       df <- norm::imp.norm(ss, thx, xxm)
     }
     else if(method=="qrilc"){
+      install_bioconductor("pcaMethods")  
       install_rpackage('imputeLCMD')
       xxm<-t(df1)
       data_zero1 <- imputeLCMD::impute.QRILC(xxm, tune.sigma = 1)[[1]]
       df<-t(data_zero1)
     }
     else if(method=="mindet"){
+      install_bioconductor("pcaMethods")
       install_rpackage('imputeLCMD')
       xxm<-as.matrix(df1)
       df <- imputeLCMD::impute.MinDet(xxm, q = 0.01)
     }
     else if(method=="minprob"){
+      install_bioconductor("pcaMethods")
       install_rpackage('imputeLCMD')
       xxm<-as.matrix(df1)
       df <- imputeLCMD::impute.MinProb(xxm, q = 0.01, tune.sigma = 1)
@@ -172,7 +175,7 @@ nafunctions <- function(x,method="zero"){
       df<-as.data.frame(t(df1x))
     }
     else if(method=="rf"){
-      install_rpackage('missForest')
+      install_rpackage("missForest")
       data_zero1 <- missForest(t(df1), maxiter =10,
                                ntree = 20 # input$rfntrees
                                ,mtry=floor(nrow(df1)^(1/3)),verbose = TRUE)
@@ -269,7 +272,7 @@ feat_name <- original_header[1]
 
 # Uncomment to test certain methods (only for debugging, as at least one method per package is tested using Github Actions)
 
-# +
+# + vscode={"languageId": "r"}
 # to_test <- c(
  # 'ZERO',
  # 'MINIMUM',

diff --git a/project/01_1_train_NAGuideR_methods.ipynb b/project/01_1_train_NAGuideR_methods.ipynb
@@ -139,17 +139,20 @@
     "      df <- norm::imp.norm(ss, thx, xxm)\n",
     "    }\n",
     "    else if(method==\"qrilc\"){\n",
+    "      install_bioconductor(\"pcaMethods\")  \n",
     "      install_rpackage('imputeLCMD')\n",
     "      xxm<-t(df1)\n",
     "      data_zero1 <- imputeLCMD::impute.QRILC(xxm, tune.sigma = 1)[[1]]\n",
     "      df<-t(data_zero1)\n",
     "    }\n",
     "    else if(method==\"mindet\"){\n",
+    "      install_bioconductor(\"pcaMethods\")\n",
     "      install_rpackage('imputeLCMD')\n",
     "      xxm<-as.matrix(df1)\n",
     "      df <- imputeLCMD::impute.MinDet(xxm, q = 0.01)\n",
     "    }\n",
     "    else if(method==\"minprob\"){\n",
+    "      install_bioconductor(\"pcaMethods\")\n",
     "      install_rpackage('imputeLCMD')\n",
     "      xxm<-as.matrix(df1)\n",
     "      df <- imputeLCMD::impute.MinProb(xxm, q = 0.01, tune.sigma = 1)\n",
@@ -200,7 +203,7 @@
     "      df<-as.data.frame(t(df1x))\n",
     "    }\n",
     "    else if(method==\"rf\"){\n",
-    "      install_rpackage('missForest')\n",
+    "      install_rpackage(\"missForest\")\n",
     "      data_zero1 <- missForest(t(df1), maxiter =10,\n",
     "                               ntree = 20 # input$rfntrees\n",
     "                               ,mtry=floor(nrow(df1)^(1/3)),verbose = TRUE)\n",
@@ -364,7 +367,11 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "162c5f7f-08f0-44ef-abf5-f0805ab58bb4",
-   "metadata": {},
+   "metadata": {
+    "vscode": {
+     "languageId": "r"
+    }
+   },
    "outputs": [],
    "source": [
     "# to_test <- c(\n",

diff --git a/requirements_dev.txt b/requirements_dev.txt
@@ -0,0 +1,13 @@
+# dev
+pytest
+pytest-cov
+jupytext
+flake8
+flake8-bugbear
+build
+wheel
+setuptools
+pre-commit
+pre-commit
+jupyterlab_code_formatter
+jupyterlab-git