Fix numpydoc validation errors for DataOps and Drop

[ayushidhete]

Summary

This PR fixes numpydoc validation errors introduced by stricter checks in #1753.

Changes

• Added missing Parameters sections to:
  • SkrubLearner
  • ParamSearch
  • OptunaParamSearch
• Added complete class and method docstrings to Drop
• Documented all constructor parameters to comply with numpydoc standards
• Removed corresponding entries from DOCSTRING_TEMP_IGNORE_SET

Validation

• All updated docstrings pass numpydoc.validate
• pytest skrub/tests/test_docstrings.py passes locally

This ensures the affected classes now fully comply with numpydoc validation requirements.

[rcap107]

Closing this because

1. Apply Repo-Review rules PP304 to PP308: improve pytest configuration #1753 hasn't even been merged yet
2. DOC - improve documentation for Drop #1756 #1825 is already working on part of this
3. You didn't ask in DOC - Improve the docs of various objects so they pass the numpydoc checks #1882 before starting, AND someone else has already asked there before
4. You didn't even make the effort of not committing the agent TODO list from the PR