Simplify the multivariate case

[mayer79]

This PR introduced the possibility to pass a univariate response to perm_importance() and average_loss() when the predictions are multivariate (e.g., represent predictions of multiple models).

[codecov-commenter]

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.14% 🎉

Comparison is base (278845b) 97.80% compared to head (33e6dba) 97.95%.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #62      +/-   ##
==========================================
+ Coverage   97.80%   97.95%   +0.14%     
==========================================
  Files          14       14              
  Lines         730      734       +4     
==========================================
+ Hits          714      719       +5     
+ Misses         16       15       -1     

Files Changed	Coverage Δ
R/average_loss.R	100.00% <ø> (ø)
R/perm_importance.R	94.25% <ø> (-0.13%)	⬇️
R/utils.R	97.34% <ø> (-0.03%)	⬇️
R/losses.R	100.00% <100.00%> (+2.04%)	⬆️

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