Removed nan-handlers for torch tensors. Updated the unit tests.

[drewoldag]

Change Description

Closes #737
Removed nan handlers for torch tensors and updated the associated test file.
The unit tests were simplified as well. There was an unused (or hardly used) dataset class that would produce data with nans, but that was previously packed into the HyraxRandomDataset. So I removed the old RandomNanDataset that was exclusively used for testing this feature. This is ok because HyraxRandomDataset can produce nans in it's random data and exercise the same functionality.

[Copilot]

Pull request overview

Removes the torch-specific NaN handling paths from DataProvider now that batches are expected to remain NumPy until they reach the model, and updates the NaN-related unit tests accordingly.

Changes:

• Removed the torch.Tensor-specific _handle_nans registration and torch-based NaN helper functions.
• Updated test_nan.py to use NumPy-based NaN assertions and removed the test-only NaN dataset subclass.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
src/hyrax/data_sets/data_provider.py	Drops torch tensor NaN-handling logic and simplifies tuple/list handling to NumPy arrays only.
tests/hyrax/test_nan.py	Removes torch dependency and adapts tests to validate NaN handling with NumPy arrays and HyraxRandomDataset.

---

You can also share your feedback on Copilot code review. Take the survey.

[Copilot]

_handle_nans_tuple is registered for both tuple and list, but it always returns tuple(handled_elements). This changes the type for list inputs (and contradicts the backward-compatibility comment). Consider preserving the input type (e.g., return list for list inputs) and updating the docstring/comment to reflect that it handles both tuples and lists.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 64.62%. Comparing base (6156e07) to head (f447fe0).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #751      +/-   ##
==========================================
+ Coverage   64.45%   64.62%   +0.17%     
==========================================
  Files          61       61              
  Lines        5925     5875      -50     
==========================================
- Hits         3819     3797      -22     
+ Misses       2106     2078      -28     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mtauraso]

LGTM 🚢 it

[github-actions]

Before [6156e07]	After [54da373]	Ratio	Benchmark (Parameter)
failed	failed	n/a	data_cache_benchmarks.DataCacheBenchmarks.time_preload_cache_hsc1k
failed	failed	n/a	data_cache_benchmarks.DataCacheBenchmarks.track_cache_hsc1k_hyrax_size_undercount
failed	failed	n/a	data_request_benchmarks.DatasetRequestBenchmarks.time_request_all_data
39.1±0.7ms	40.5±1ms	1.04	benchmarks.time_nb_obj_dir
38.2±0.3ms	38.5±0.4ms	1.01	benchmarks.time_nb_obj_construct
1.98±0.01s	1.99±0.02s	1.01	benchmarks.time_train_help
3.62±0s	3.64±0.02s	1.01	vector_db_benchmarks.VectorDBInsertBenchmarks.time_load_vector_db(256, 'chromadb')
2.01±0.01s	2.01±0.02s	1.00	benchmarks.time_infer_help
1.99±0.03s	1.99±0.01s	1.00	benchmarks.time_lookup_help
2.03±0.01s	2.02±0.03s	1.00	benchmarks.time_rebuild_manifest_help

Click here to view all benchmarks.