Automation Update 8/8

[KELSEYDOWLING7]

Hi Jake, these weekly reports are currently failing in GCP in order of priority

• [ ] Biospecimen Report [R error]
• [x] Biospecimen QC [Successful]
• [x] Notifications Report [Successful]
• [ ] Weekly module Metrics [R error - Kelsey fixed, building now]
• [x] Recruitment Derived Variable QC [Successful]

All of these run locally for me, so anything you can see on GCP would be helpful to know. I'll have any PR's done before our meeting.

[KELSEYDOWLING7]

PRs completed for both repos!

[jacobmpeters]

All of these are built, so I'll begin testing now..

[KELSEYDOWLING7]

Great!

[jacobmpeters]

The Weekly Biospecimen Report failed with this R error:

[jacobmpeters]

The Module Metrics with the same "col" vs. "column" issue that we saw last time.

[jacobmpeters]

The Module Metrics with the same "col" vs. "column" issue that we saw last time.

I fixed this myself and I'll test it again...

[KELSEYDOWLING7]

Oh, thank you! I'll have to update all other reports to have column instead of col. That must be be a GCP update

[KELSEYDOWLING7]

Table 7.3 in the Biospecimen report keeps throwing me off because its not an issue locally

[jacobmpeters]

Oh, thank you! I'll have to update all other reports to have column instead of col. That must be be a GCP update

Yeah I don't understand why this is suddenly causing problems. I did not change the version of tbl_cross in the container

[KELSEYDOWLING7]

Its a quick fix regardless. I'll be pushing a new Weekly Biospecimen Report PR in just a minute

[jacobmpeters]

I made the fix to the Weekly Module Metrics and tested it, but I received this error. It seems like the opposite error of what we received before.

[jacobmpeters]

We're still getting this error with the Biospecimen report..

[KELSEYDOWLING7]

Jake it looks like one of the Biospecimen QC csvs didn't run. The Duplicates report csv should have been populated. It populated when I ran it locally

[KELSEYDOWLING7]

For the follow up his week:

• [ ] Biospecimen report
• [ ] Weekly Biospecimen CSV files (new)
• [x] Biospecimen QC- it ran but one of the output files (Duplicate_Baseline_Biospecimen_Collections) didn't post to box. It is the only xlsx file instead of csv. Does that matter?

[jacobmpeters]

We got a tbl_cross error in the Biospecimen Report.

[KELSEYDOWLING7]

Ok I'll fix that now and make sure there aren't any more of those instances. I thought I had caught them all... I hope its not a merge conflict

[KELSEYDOWLING7]

It's definitely a merge conflict. They're all col = in the tbl cross functions on my local drive but somehow pushing that to Github didn't change them... I work on a new PR and try manually to change them all again if need be

[jacobmpeters]

Ok! That's definitely annoying. I'll wait for your PR and work on the CSVs..

[jacobmpeters]

Seems like we're still getting the same error:

[KELSEYDOWLING7]

Huh I really don't know how thats possible... the main branch has the correct code that doesn't use 'column': tab___1 <- table1 %>%
mutate(biocol.appoint = factor(biocol.appoint, levels=c("Baseline Collection","No Baseline Collection"))) %>% dplyr::select(Site, biocol.appoint ) %>% tbl_cross( row = Site, col = biocol.appoint, digits=c(0,1), percent = "row", label=list(Site="Site", biocol.appoint = " "), missing="ifany", margin_text="Total")

[jacobmpeters]

Hmm.. let me take a look. It might not have built completely before I kicked off the test run.. I'll get back to you on that..

[jacobmpeters]

The Biospecimen QC ran and the .xlsx file should have delivered to box.. Can you check?

[KELSEYDOWLING7]

Yes, those look perfect!! That one in good to set to complete

[jacobmpeters]

The Biospecimen CSVs Report script is trying to use a local_drive variable that isn't defined..

[KELSEYDOWLING7]

Ah ok that's my fault, I didn't include the intro in this version but "{local_drive}" can just be removed. I can do a PR for that real quick but I understand you may not have time to build and run it

[jacobmpeters]

I'm getting a useless error for the Biospecimen Report. Do you have any thoughts on where this might be breaking? It doesn't even get to the first chunk so I'm confused..

[jacobmpeters]

Did this section of the header change recently?

[KELSEYDOWLING7]

Huh, that I don't know. Maybe its this part in the beginning? I think we can delete from the Report code since we no longer have the CSVs:

if (!is.null(Sys.getenv("USE_TEST_BOX_FOLDER")) && Sys.getenv("USE_TEST_BOX_FOLDER") != "") { use_test_box_folder <- as.logical(Sys.getenv("USE_TEST_BOX_FOLDER")) }

if (use_test_box_folder) { boxfolder <- 222593912729 # test box folder } else { boxfolder <- 221280601453 # destination of CSV files (not pdf) }

Making sure personal C drives aren't referenced if this code is being used by others

Change to FALSE if referencing this code

write_to_local_drive = F #F

local_drive="C:/Users/dowlingk2/Documents/Module-Missingness-and-Metrics/data/"

This function below is put before any write.csv functions, and "filename" is updated. It determines wheteher the file will be created locally or not.

filename=

local_drive= ifelse(write_to_local_drive, "C:/Users/dowlingk2/Documents/Module-Missingness-and-Metrics/data/", "")

[KELSEYDOWLING7]

The yaml part was part of the weird merge conflicts. It should look like this:

---

title: "Biospecimen Weekly Metrics RMD" author: "Kelsey Sanchez" date: "Data Extracted and Report Ran: r Sys.Date()" header-includes:
\usepackage[labelformat=empty]{caption} \usepackage{placeins} \usepackage{booktabs} \usepackage{pdflscape}

output: pdf_document: extra_dependencies: ["float"] toc: true keep_tex: yes fig_width: 7 fig_height: 5 fig_caption: true df_print: paged

[jacobmpeters]

The yaml part was part of the weird merge conflicts. It should look like this:

title: "Biospecimen Weekly Metrics RMD" author: "Kelsey Sanchez" date: "Data Extracted and Report Ran: r Sys.Date()" header-includes: \usepackage[labelformat=empty]{caption} \usepackage{placeins} \usepackage{booktabs} \usepackage{pdflscape}

output:

pdf_document: extra_dependencies: ["float"] toc: true keep_tex: yes fig_width: 7 fig_height: 5 fig_caption: true df_print: paged

Ok. I think I just fixed it.. there was an missing indent

output:
  pdf_document:
    extra_dependencies: ["float"]
    toc: true
    keep_tex: yes
    fig_width: 7
    fig_height: 5
    fig_caption: true
    df_print: paged 

[KELSEYDOWLING7]

Oh ok perfect!