chore: typos, formatting and refactor

[selmanozleyen]

Hi, there were some things bugging me and I wanted to clean some stuff and point out some inconsistencies but I also didn't want this to be in other unrelated PR's so here is a contained PR consisting of some of the things that bugged me. I will revert some commits based on whatever you guys would want to undo.

Here is an ai-based report of the changes by commits

Changes

• Breaking: Rename add_adatas to add_anndatas for consistent naming (selmanozleyen@2558b40, selmanozleyen@304dcbb and a60774c)

  • Renamed DatasetCollection.add_adatas -> DatasetCollection.add_anndatas
  • Updated the docstrings and tests accordingly. Might be also a different fix like doing vice versa

• Fix "Wether" typo in io.py (5232e12)

  • DatasetCollection.is_empty docstring: "Wether" -> "Whether"

• Define _collection_added as a class attribute in loader.py (5d14e5c)

  • _collection_added: bool = False is now declared on the Loader dataclass directly, instead of being accessed via getattr(self, "_collection_added", False).
  • For this warning: https://pylint.readthedocs.io/en/latest/user_guide/messages/warning/attribute-defined-outside-init.html which is reasonable imo

• Fix "who" typo in loader.py (d841e13)

  • Docstring in Loader.use_collection: "The collection who on-disk datasets" -> "The collection whose on-disk datasets"

• Ruff format test_dataset.py (36af588)

  • Reformatted a lambda expression's parameter list in test_dataset.py (whitespace-only change by ruff)

other mypy fixes but I don't mind them that much

• 4958749 -- add torch.* and h5py.* to mypy ignore_missing_imports
• 830f2d4 -- fix Mapping.copy() call in write_sharded callback (dict(dataset_kwargs))
• 6d6067a -- wrap categories in pd.Index for Categorical.from_codes
• 12830d5 -- replace match/case with if/elif for better mypy narrowing in _create_chunks_for_shuffling

[codecov]

Codecov Report

❌ Patch coverage is 92.85714% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 91.50%. Comparing base (d47240f) to head (50293fb).

Files with missing lines	Patch %	Lines
src/annbatch/io.py	90.90%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #145      +/-   ##
==========================================
- Coverage   93.71%   91.50%   -2.21%     
==========================================
  Files          11       11              
  Lines         811      812       +1     
==========================================
- Hits          760      743      -17     
- Misses         51       69      +18     

Files with missing lines	Coverage Δ
src/annbatch/loader.py	89.29% <100.00%> (-3.32%)	⬇️
src/annbatch/io.py	93.08% <90.90%> (ø)

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ilan-gold]

Just more places for consistency. Why go with anndata over adata? I'm really asking for feedback, I think adata is used for the actual object generally i.e., adata = read_h5ad but not anndata = read_h5ad

[ilan-gold]

Would move this argument also to load_anndata

[selmanozleyen]

Just more places for consistency. Why go with anndata over adata? I'm really asking for feedback, I think adata is used for the actual object generally i.e., adata = read_h5ad but not anndata = read_h5ad

For me it both was same so I wanted to minimize the breaking impact and guessed Loader was used more. Also thought, as you said, "anndata is a class name vs adata is an instance name". Like we take any anndata not a specific adata.

I would go with whatever is more adapted in the ecosystem but I don't have any insights on that.

[felix0097]

LGTM!

Another minor fix: Could you provide a the number of iterations here. This gets broken by using enumerate. Just set total=len(paths):

annbatch/src/annbatch/io.py

Line 181 in d47240f

for i, path in tqdm(enumerate(paths), desc="loading"):

this has been bugging me for ages

Moreover, the progress bar descriptions should get updated:
Here, I would use Creating collection
Here, I would use: Extending collection
Here, I would use Lazy loading anndatas
Here, I would start with a capital letter as well

Would be cool if you could add this to this PR. Otherwise, I can create a separate PR myself @selmanozleyen

[ilan-gold]

Also thought, as you said, "anndata is a class name vs adata is an instance name". Like we take any anndata not a specific adata.

"anndata" is the library name. "AnnData" is the class name. In any case, some spatial data folks concur with adata:

https://scverse.zulipchat.com/#narrow/channel/315789-data-structures/topic/API.20Terminology/near/575771345