[ENH] start implementing BIDS stats model transformers

[Remi-Gau]

• add support for Add, Subtract, Divide, Multiply operations (many more to go)

List of transforms to implement

[codecov]

Codecov Report

Merging #493 (b300fd5) into dev (70761b0) will increase coverage by 0.19%.
The diff coverage is 89.28%.

❗ Current head b300fd5 differs from pull request most recent head 10e97ff. Consider uploading reports for the commit 10e97ff to get more accurate results

@@            Coverage Diff             @@
##              dev     #493      +/-   ##
==========================================
+ Coverage   73.33%   73.52%   +0.19%     
==========================================
  Files         190      192       +2     
  Lines        3945     4004      +59     
==========================================
+ Hits         2893     2944      +51     
- Misses       1052     1060       +8     

Flag	Coverage Δ
unittests	73.52% <89.28%> (+0.19%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/bids_model/getHRFderivatives.m	70.58% <0.00%> (ø)
src/utils/rmTrialTypeStr.m	66.66% <50.00%> (-33.34%)	⬇️
src/workflows/stats/bidsConcatBetaTmaps.m	75.38% <60.00%> (-1.67%)	⬇️
src/bids_model/applyTransformersToEventsTsv.m	86.66% <86.66%> (ø)
src/bids_model/getBidsTransformers.m	88.88% <88.88%> (ø)
src/subject_level/convertOnsetTsvToMat.m	100.00% <100.00%> (ø)
src/subject_level/createAndReturnOnsetFile.m	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 70761b0...10e97ff. Read the comment docs.

[marcobarilari]

lol

[CerenB]

Now i'm afraid!

[Remi-Gau]

In brief: the BIDS stats model specification should in theory allow you to define a bunch of operation you want to perform on your "inputs" before you include them in the models.

The simplest example I have started implementing is: say you have an difference between the stimuli onsets written in your events.tsv and the "real" time when stimuli actually occurred (say the stimulus computer messed up the time stamps and you know they all came one second earlier than what was recorded).

You can either:

• rewrite ALL your events.tsv to change this (Oh joy!)
• add a few lines in your model JSON to remove that offset for all the conditions you are using.

Which one do you take?

[Remi-Gau]

Added a bit of doc about this in here (scroll down):

https://cpp-spm.readthedocs.io/en/dev/set_up.html#bids-model-json-files