Can function 'create_job_df' account for types of simulation set up?

[CaitlinA]

I am currently working on the project for the SOILWAT app, where each site-scenario combo gets its own row in both the input master and the treatment design file. If you recall, this is due to unequal simulation periods between historical and future data.

I am experiencing an issue with the 'create_job_df' function. The error is:

[1] "rSFSW2's 'recreate_dbWork': started at 2018-10-22 18:58:36"
Error in create_dbWork(path, jobs = create_job_df(infer_sim_size, infer_include_YN)) : 
  number of items to replace is not a multiple of replacement length.

More specifically, it is because create_job_df sees the simulation as having 25 jobs, when I only have 5:

  print(sim_size)

$runsN_master
[1] 5

$runsN_total
[1] 5

$expN
[1] 5

jobs[, "runID_sites"] <- rep(seq_len(sim_size[["runsN_master"]]),
    times = max(sim_size[["expN"]], 1L))
  temp <- rep(include_YN, times = max(sim_size[["expN"]], 1L))

I know the CDI project is unique, but I could see there being other projects where there is one to one relationship between sites and treatments (as the treatment design file is intended for), not multiplicative, as expN implies in this function. What do you think @dschlaep ?

[dschlaep]

There is a one-to-one relationship between sites and treatments. There is a multiplicative relationship between sites and experimental levels.

The expN = 5 suggests that you entered five experimental levels/rows in the file SWRuns_InputData_ExperimentalDesign_v08.csv -- these will be applied to each of the "run/site/rows" as defined by the SWRuns_InputMaster_YOURPROJECT_v11.csv as runsN_master = 5 which have to correspond to one row each in the file SWRuns_InputData_TreatmentDesign_v16.csv.

See also the question "How do I define experimental treatments?" in the vignette rSFSW2_FAQ.

[CaitlinA]

I do not have 5 rows in the experimental design and I understand the difference between the treatment and experimental design files. I have attached my current experimental, treatment, and input master below.

…

On Tue, Oct 23, 2018 at 10:37 AM daniel ***@***.***> wrote: There is a one-to-one relationship between sites and treatments. There is a multiplicative relationship between sites and experimental levels. The expN = 5 suggests that you entered five experimental levels/rows in the file SWRuns_InputData_ExperimentalDesign_v08.csv -- these will be applied to each of the "run/site/rows" as defined by the SWRuns_InputMaster_YOURPROJECT_v11.csv as runsN_master = 5 which have to correspond to one row each in the file SWRuns_InputData_TreatmentDesign_v16.csv. See also the question "How do I define experimental treatments?" in the vignette rSFSW2_FAQ. — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#342 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AM0e_w1_mcb3o4quEtyTGxuqfr3bTMuxks5unyl6gaJpZM4X0XTI> .

-- Caitlin Andrews Ecologist Southwest Biological Science Center U.S. Geological Survey Mobile: 802-922-3494 Office: 928-556-7036

[dschlaep]

Do you have outdated copies of SWRuns_InputAll_PreProcessed.rds and/or SFSW2_project_descriptions.rds?

The only way to get expN = 5 is to have five active rows in the experiment design input file. This object is set by the function determine_simulation_size:

expN <- NROW(sw_input_experimentals)

[CaitlinA]

No outdated copies of those .rds files... I never had 5 rows in the experimental design to begin with. Is there anything else I can send along which might help?

…

On Tue, Oct 23, 2018 at 12:54 PM daniel ***@***.***> wrote: Do you have outdated copies of SWRuns_InputAll_PreProcessed.rds and/or SFSW2_project_descriptions.rds? The only way to get expN = 5 is to have five active rows in the experiment design input file. This object is set by the function determine_simulation_size: expN <- NROW(sw_input_experimentals) — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#342 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AM0e_-Raxyg6EiHfnF-qVsVtkmDcHGhvks5un0nDgaJpZM4X0XTI> .

-- Caitlin Andrews Ecologist Southwest Biological Science Center U.S. Geological Survey Mobile: 802-922-3494 Office: 928-556-7036

[dschlaep]

That's weird.

It may be a good start to figure out how your run ends up with expN = 5 even though fnames_in[["fexpDesign"]] has only 1 row.