[DOC] Improve logging of component table-based manual classification

[tsalo]

Closes #848.

Changes proposed in this pull request:

• Add a manual classification step in the four-echo integration test that uses the component table for the classifications.
• Change the classifications in the five-echo integration test's manual classification step.
• Add a logger message when a component table is provided.

[codecov]

Codecov Report

Merging #852 (64c9f29) into main (eaa0920) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #852   +/-   ##
=======================================
  Coverage   93.27%   93.28%           
=======================================
  Files          27       27           
  Lines        2217     2218    +1     
=======================================
+ Hits         2068     2069    +1     
  Misses        149      149           

Impacted Files	Coverage Δ
tedana/workflows/tedana.py	89.96% <100.00%> (+0.03%)	⬆️

---

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 eaa0920...64c9f29. Read the comment docs.

[tsalo]

I'm unable to reproduce the following problem from #848:

the png's in the 'figures' folder created by tedana appear to be unchanged (e.g., a component that was accepted via RICA is still shown as rejected).

First pass

Note that the selected component is ignored.

Manual classification

All ignored components were reclassified as accepted in this step, via the component table.

[eurunuela]

I will try to reproduce it sometime this week. I believe we have another user reporting the same issue in Neurostars.

[tsalo]

@eurunuela it looks like ME-ICA/rica#41 solves the component classification issue, correct? If so, then I think this PR can be reviewed purely as an improvement to documentation/logging.

[eurunuela]

@eurunuela it looks like ME-ICA/rica#41 solves the component classification issue, correct? If so, then I think this PR can be reviewed purely as an improvement to documentation/logging.

Indeed. I'm sorry I didn't have time to report back, it's been a busy evening.

As you say, we should improve documentation/logging.

[eurunuela]

#855 should fix a related issue on tedana's side.

[eurunuela]

I think these changes should be enough.

Should we add something to the docs? And if so, would you rather do that on this PR or open a new one?

[tsalo]

I think improving the documentation is a good idea, and I'm happy to incorporate that into this PR. Do you have anything specific in mind?

[handwerkerd]

LGTM
I don't have any strong opinions on adding documentation here vs a future PR.
My one general concern is how many different things are being tested within some of these integration tests (i.e. manual classification could be a separate test). This is going to create a lot of "fun" when we get to the point of fully reviewing #756 I suspect non-trivial changes to some of these existing tests will be required. It's beyond the scope of this PR, but I'll think about this more, and possibly open up a new issue so that we can start identifying & cleaning up tests that will interface with the modularized decision tree.