Add user configuration file /usr/preferences.json

[agahkarakuzu]

Description to be added.

[codecov]

Codecov Report

Merging #435 (d1de772) into master (8b8fb39) will decrease coverage by 0.91%.
The diff coverage is 46.57%.

❗ Current head d1de772 differs from pull request most recent head 09c7807. Consider uploading reports for the commit 09c7807 to get more accurate results

@@            Coverage Diff             @@
##           master     #435      +/-   ##
==========================================
- Coverage   47.40%   46.49%   -0.91%     
==========================================
  Files         260      272      +12     
  Lines        8647     9629     +982     
==========================================
+ Hits         4099     4477     +378     
- Misses       4548     5152     +604     

Flag	Coverage Δ
matlab	46.49% <46.57%> (-0.13%)	⬇️
octave	?

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

Impacted Files	Coverage Δ
src/Common/AbstractModel.m	71.25% <ø> (+5.00%)	⬆️
src/Common/FitBIDS.m	0.00% <0.00%> (ø)
src/Common/FitResultsSave_BIDS.m	0.00% <0.00%> (ø)
src/Common/qMRLab_static_Models.m	0.00% <0.00%> (ø)
src/Common/saveUserPreferences.m	0.00% <0.00%> (ø)
src/Common/tools/versionChecker.m	0.00% <0.00%> (-62.17%)	⬇️
src/Common/units/validatePanelUnits.m	0.00% <0.00%> (ø)
...els_Functions/MP2RAGE/func/T1B1correctpackageTFL.m	84.74% <ø> (ø)
src/Common/getUserPreferences.m	32.50% <32.50%> (ø)
src/Models/T1_relaxometry/vfa_t1.m	41.00% <42.85%> (-34.45%)	⬇️
... and 27 more

... and 17 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8b8fb39...09c7807. Read the comment docs.

[agahkarakuzu]

Purpose is to totally isolate user choices from native implementation choices. Config files provided by usr/devs will deal with necessary conversions.

• Replace qMRLab_static_models
• Usr UnifyOutputUnits infrastructure is ready
• Do the same for input params & for inputs with non-arbitrary units
• BIDS mappings are done
• Make changes operational at the usr GUI & Batch level
• Batch generation and GUI panels should reflect user selections
• When ready, switch unify on for Documentation
• Turn off on unit tests. Test the core in the form they were originally written.

[agahkarakuzu]

Function table is not implemented in freaking Octave. Dataframe does not give similar functionality, so change table with something else.

[agahkarakuzu]

@jvelazquez-reyes if you have time, maybe you can try having modelRegistry.m function the same way without using the table function. Should be pretty straightforward, just leaving it here if you can get that to it before I can. These are the relevant lines:

qMRLab/src/Common/modelRegistry.m

Line 70 in bd22f11

cur_lut = table(parent,xnames,suffix,isBIDS,folderBIDS,'VariableNames',{'Family','xname','suffixBIDS','isOfficialBIDS','folderBIDS'});

qMRLab/src/Common/modelRegistry.m

Line 98 in bd22f11

pre_out = table2struct(lut(idxs2,:));

qMRLab/src/Common/modelRegistry.m

Line 207 in bd22f11

cur_lut = table(parent,fieldnames(unitDefs.(fields{ii})),'VariableNames',{'Family','unit'});

If you can, please branch out from usrjson and implement on that branch.

[agahkarakuzu]

The main functionality to change input/output map units and input protocol parameters is ready. Here's an example, inversion_recovery in its original implementation works with (ms), here both input prots and output map are in second.

More fantastic configs are easily possible. For example, input parameters are in (usec) and the output map is in hrs 😁

Dealing with all these unit mappings is actually much trickier than it looks without inducing overhead to the voxelwise models. Tests will fail for a while, because I still need to go through classdefs of all the models. Then we'll decide on the default unified input/output units across models. Unified parameters will be enabled for users and documentation generation (because they'll be in the qMRgenBatch outputs or in GUI panel), but will be disabled for the unit tests.

I know that a json of SI units exists out there, but there are reasons why I come up with a customized unit definition schema.

The documentation will make it clear which units are available and changing which unit parents (e.g. Time) changes which input/output/protocol entities. And that page will be automatically parsed.

[agahkarakuzu]

https://qmrlab.readthedocs.io/en/v250/user.html

[mathieuboudreau]

@agahkarakuzu if this PR is ready for review, can you assign me and @jvelazquez-reyes ?

If it isn't ready for review, can you detail what is missing? Maybe one of us can work towards resolving this PR in that case.