[NO-MRG] Debug mid-Kappa classification

[tsalo]

References #226 (comment). The top component is being classified as mid-Kappa. This isn't happening in #226 or locally, so it is definitely fragile.

[codecov]

Codecov Report

Merging #229 into master will decrease coverage by 0.23%.
The diff coverage is 0%.

@@            Coverage Diff            @@
##           master    #229      +/-   ##
=========================================
- Coverage   50.34%   50.1%   -0.24%     
=========================================
  Files          32      32              
  Lines        1891    1900       +9     
=========================================
  Hits          952     952              
- Misses        939     948       +9

Impacted Files	Coverage Δ
tedana/selection/select_comps.py	4.86% <0%> (-0.33%)	⬇️

---

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 27bf65a...95f6e35. Read the comment docs.

[codecov]

Codecov Report

Merging #229 into master will decrease coverage by 0.23%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master     #229      +/-   ##
==========================================
- Coverage   50.31%   50.07%   -0.24%     
==========================================
  Files          32       32              
  Lines        1894     1903       +9     
==========================================
  Hits          953      953              
- Misses        941      950       +9

Impacted Files	Coverage Δ
tedana/selection/select_comps.py	4.86% <0%> (-0.33%)	⬇️

---

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 e44058c...ef061da. Read the comment docs.

[tsalo]

I don't think I'm going to be able to figure out much about issues with component selection at this point. The rationale behind each step is just missing. While I can see the value of most of the selection steps, I can't determine why each metric was chosen specifically, what alternatives were originally considered, or why they're in the order they're in. I think the best next step is to close this PR and to focus on making sure that the metrics used by the component selection algorithm are properly calculated. We have a few issues open that touch on that.

[emdupre]

Yes, I think component selection is the next big hurdle. Would you want to talk about that in the next developers's call, @tsalo ?

Thanks for looking into this !!