173 fix refactor adex@main

[Melkiades]

Pull Request

This PR needs a lot of care in its review. The execution gives an error in tests that I can not reproduce. It is potentially necessary to reconvert the date formats into simple characters or system base date, it depends on how adex is used downstream. I was just checking the difference in the cache and the new data set seems way too distant from the original cashed-data. I will explore these better tomorrow morning! I want to highlight that this is a hotfix and I will tomorrow write down some new issues to fix ADSL times before this one (which is only modified at the last stage). I am sure there are better ways to modify the data set, and any idea is super welcome!

For some background on the new variables, I copy here my notes, that I took from talking with @malexthorpe.

# VAR 1: EXDOSFRQ --------------------------------------------------------------
# For the EXDOSFRQ that should be just set to 'ONCE' whenever the PARCAT1 is 'INDIVIDUAL' and NA for the 'OVERALL'
# We use to build ADEX contains each individual dose as one row per dose, we then also add OVERALL rows for each of the paramcd 
# values and for each drug but these OVERALL contains NAs for all the dose information as they're just summaries

# VAR 2: EXROUTE ------------------------------------------------------
# For the EXROUTE we can just set them all to INTRAVAEOUS as that is very common 
# for all the INDIVIDUAL doses (I kept 10% of INTRAVENOUS)

# VAR 3: EXENDY ----------------------------------------------------------------
# to add EXENDY we also need to add other columns like: VISIT, VISITDY, EXSTDTC, EXENDTC, EXSTDY.

# So the Exposure start date and end date should be the same with a small hour long 
# time gap between them while the patients is being dosed and then the EXSTDY and EXENDY can be set to the same values

# Yeah although also I guess also on the TRTSTDT, it should be the EXENDTC - TRTSDT rounded up in days

# I think after adding the other variables we also need to update the ASTDTM and AENDTM variables as they should be 
# using EXSTDTC and EXENDTC (intermediate) rather than the TRTSTM and TRTEDTM (general) that they are currently set on

# add VISIT column, the same as AVISIT (also with weeks)

# screening and baseline before the treatment start day/time

# ADSL should have treatment times bound to AVISIT (blocker)

# ASTDTM is from ADEX 

Fixes #173

[github-actions]

Unit Tests Summary

  1 files    1 suites   19s ⏱️
23 tests 23 ✔️ 0 💤 0 ❌
56 runs  56 ✔️ 0 💤 0 ❌

Results for commit 39e4f1e.

♻️ This comment has been updated with latest results.

[github-actions]

Code Coverage Summary

Filename         Stmts    Miss  Cover    Missing
-------------  -------  ------  -------  --------------------------------------------------------------------------------------
R/radab.R          215     215  0.00%    51-293
R/radae.R          173       3  98.27%   82, 101, 143
R/radaette.R       211       4  98.10%   52, 73, 79, 243
R/radcm.R          151      27  82.12%   53, 87, 173-199
R/raddv.R           91       3  96.70%   41, 56, 140
R/radeg.R          269       1  99.63%   185
R/radex.R          232       1  99.57%   226
R/radhy.R          136       1  99.26%   197
R/radlb.R          253       1  99.60%   221
R/radmh.R           94       2  97.87%   48, 87
R/radpc.R           60       1  98.33%   100
R/radpp.R           80       1  98.75%   141
R/radqs.R          107       1  99.07%   132
R/radrs.R          121       3  97.52%   49, 56, 156
R/radsaftte.R        1       1  0.00%    17
R/radsl.R          166       1  99.40%   238
R/radsub.R         142       1  99.30%   159
R/radtr.R          126       0  100.00%
R/radtte.R         144       4  97.22%   53, 73, 85, 128
R/radvs.R          160       1  99.38%   174
R/utils.R          154      59  61.69%   9-12, 56-59, 111-112, 117-118, 155-156, 214, 264, 266, 295-301, 305, 338-408, 466, 470
TOTAL             3086     331  89.27%

Results for commit: 18df990

Minimum allowed coverage is 80%

♻️ This comment has been updated with latest results

[ayogasekaram]

lgtm @Melkiades
I would double check the -1 relative date. Other than that, everything appears to work well.