fix: DIA-1740: 'int' not iterable when Rating GT empty

[hakan458]

Most control tags' labels are stored in a list, for some like Number and Rating it is just a single int. We were not handling the case where the annotation to compare against is emtpy, for a Rating / Number tag

Fixed DIA-1741 here as well. We were never calculating pairwise comparisons correctly. We were catching an exception in the case both annotation and pred have non-empty pairwise selection, but issue manifested when GT was empty
We were using the wrong key all along

[codecov]

Codecov Report

Attention: Patch coverage is 33.33333% with 2 lines in your changes missing coverage. Please review.

Project coverage is 70.26%. Comparing base (74e171a) to head (3dafebd).
Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
evalme/classification.py	33.33%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #163      +/-   ##
==========================================
- Coverage   70.32%   70.26%   -0.06%     
==========================================
  Files          26       26              
  Lines        2746     2748       +2     
==========================================
  Hits         1931     1931              
- Misses        815      817       +2     

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