Change `n_features_` to `n_features_in_` for sklearn compatibility

[eurunuela]

Closes #58.

This PR proposes:

• Change n_features_ for n_features_in_ to be compatibly with sklearn.
• ~Adds dependencies to the package (apparently they were not there).~
• Updates the necessary scikit-learn version to at least 1.2.

[eurunuela]

We may want to drop support for 3.5 and add tests for up to 3.10 or so.

[eurunuela]

I will open a new PR to work on the project.toml and update the minimum Python version for the package (needs to be at least 3.8).

At least now @marco7877 can work with maPCA.

[eurunuela]

It looks like @tsalo may be busy, @handwerkerd would you mind reviewing this? It's impacting tedana for users with a newer version of scikit-learn.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (70aa234) 90.96% compared to head (e89a44d) 90.96%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #59 +/- ## ======================================= Coverage 90.96% 90.96% ======================================= Files 3 3 Lines 310 310 Branches 32 32 ======================================= Hits 282 282 Misses 14 14 Partials 14 14 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[handwerkerd]

@eurunuela Will we also need to update the minimum mapca version in tedana to fully address this sklearn compatibility issue? It's currently mapca>=0.0.3 and I think this will now be mapca 0.0.5.

[eurunuela]

@eurunuela Will we also need to update the minimum mapca version in tedana to fully address this sklearn compatibility issue? It's currently mapca>=0.0.3 and I think this will now be mapca 0.0.5.

Yes, as soon as we cut a release for mapca. Shall we do that at the meeting on Thursday?

[handwerkerd]

With this and #64 merged, I think you can cut a release whenever you have time and then we can bump up the minimum mapca version in tedana.