Preserving dtypes when running scrublet

[etlioglu]

Minor change for i) keeping track of data types within .col prior to feeding an AnnData object to sc.pp.scrublet(), ii) enforcing these types to the respective columns of the data frame returned by sc.pp.scrublet().

• Closes sc.pp.scrublet() is overwriting dtypes when being run with the "batch_key" parameter #3938
• Tests included: test_scrublet_dtypes() in scanpy/tests/test_scrublet.py

• Release notes not necessary because:

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 78.01%. Comparing base (45766a8) to head (c12fabb).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #3939   +/-   ##
=======================================
  Coverage   78.01%   78.01%           
=======================================
  Files         117      117           
  Lines       12466    12469    +3     
=======================================
+ Hits         9725     9728    +3     
  Misses       2741     2741           

Flag	Coverage Δ
hatch-test.low-vers	77.34% <100.00%> (+<0.01%)	⬆️
hatch-test.pre	76.96% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/scanpy/preprocessing/_scrublet/__init__.py	96.93% <100.00%> (+0.09%)	⬆️