Compute method for MLPClassifierSpace

[Lilly-May]

PR Checklist

• [X] Referenced issue is linked (closes #562)
• [X] If you've fixed a bug or added code that should be tested, add tests!
• [X] Documentation in docs is updated

Description of changes

• Replaced three separate methods (load, train, and get_embeddings) in the MLPClassifierSpace with a single compute method. The motivation for this is to unify the API, aligning it with other perturbation space methods, which all feature a compute method.

• The load, train, and get_embeddings methods persist, however, they now trigger a DeprecationWarning when called that points users to the new compute method.

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 57.14286% with 3 lines in your changes are missing coverage. Please review.

Project coverage is 63.48%. Comparing base (80ef0f0) to head (2888906). Report is 3 commits behind head on main.

:exclamation: Current head 2888906 differs from pull request most recent head 8faad36. Consider uploading reports for the commit 8faad36 to get more accurate results

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #565 +/- ## ========================================== - Coverage 63.52% 63.48% -0.04% ========================================== Files 43 43 Lines 5091 5094 +3 ========================================== Hits 3234 3234 - Misses 1857 1860 +3 ``` | [Files](https://app.codecov.io/gh/theislab/pertpy/pull/565?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab) | Coverage Δ | | |---|---|---| | [.../\_perturbation\_space/\_discriminator\_classifiers.py](https://app.codecov.io/gh/theislab/pertpy/pull/565?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab#diff-cGVydHB5L3Rvb2xzL19wZXJ0dXJiYXRpb25fc3BhY2UvX2Rpc2NyaW1pbmF0b3JfY2xhc3NpZmllcnMucHk=) | `89.32% <57.14%> (-1.32%)` | :arrow_down: |

[Lilly-May]

We might want to wait for a fix for issue #566 before merging this (or rather include the fix in this PR), even though the issue is not directly related to this PR. It persists even when using the three separate methods (load, train, get_embeddings).

[Zethson]

Yup! I'll try n check it out ASAP!