Adding r2_mlm() and possible update to r2() to work with it

[jluchman]

Fixes #758

This pull request includes a new function r2_mlm() that generates two $R^2$ values for multivariate linear regression objects/objects of class c("mlm", "lm") and is associated with issue #758.

This pull request also adds an multivariate logical argument to the r2() function for its "mlm" S3 method that can invoke the computation of the r2_mlm but is by default 'off'/FALSE and returns the current $R^2$ metric separated by response. I see value in this addition to the r2() function but do understand if it is not in the spirit of the function and it is not deemed a value add.

[DominiqueMakowski]

what I consider is the opposite: running r2(mlm) would return the multivariate R2 (i.e., would run r2_mlm()), but there would be an argument (documented in r2_mlm()) to set multivariate=FALSE (true by default) that would return the r2s per response.

[jluchman]

Update just pushed that incorporates the multivariate=TRUE default for r2(mlm) along with a few edits made by the linter and some minor documentation updates.

[codecov]

Codecov Report

Attention: Patch coverage is 25.53191% with 35 lines in your changes missing coverage. Please review.

Project coverage is 58.47%. Comparing base (852c1d9) to head (ee2b43f).

Files	Patch %	Lines
R/print-methods.R	0.00%	22 Missing ⚠️
R/r2.R	0.00%	13 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #764      +/-   ##
==========================================
- Coverage   58.52%   58.47%   -0.05%     
==========================================
  Files          88       89       +1     
  Lines        6597     6623      +26     
==========================================
+ Hits         3861     3873      +12     
- Misses       2736     2750      +14     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[strengejacke]

Thanks a lot, looks good to me!
Will wait for checks and then merge.