Update similarity.py to improve test coverage to 99%

[mohini-aggarwal]

• Updated selector/measures/similarity.py to the 99% coverage version.
• Added comprehensive tests in test_diversity_comprehensive.py.
• Ready for review and integration into main repository.

[mohini-aggarwal]

Hi Mentors @FarnazH @FanwangM @marco-2023 !

I submitted this PR (#295 ) improving test coverage in similarity.py from 76% to 94%. This is part of my ongoing contributions to the Selector project for GSoC 2026.

Thank you for your guidance!

[FanwangM]

Thanks for the PR. The current PR adds an empty line after each docstring/code block, which should be avoided. This also makes the code unreadable. Please fix this and we will do further code review. Thanks.

@mohini-aggarwal

[mohini-aggarwal]

Hi @FanwangM, I’ve updated the PR to address the formatting issues you pointed out. Please let me know if anything else needs improvement, happy to make further changes.
Thank You!

[FanwangM]

Hi @mohini-aggarwal . Thanks for the updates. I have checked your current PR and noticed that you only added/deleted empty lines, without any actual content changes. Also, some changes violate the coding style we follow, e.g., PEP8. For example, you left only one empty line between two functions, whereas we usually have two.

[mohini-aggarwal]

Hi @FanwangM @FarnazH @marco-2023
I have refactored and improved the test suite for selector.measures.similarity.
Key improvement:

• Coverage increased from 76% → 99% for similarity module
• This was achieved by adding missing edge-case tests, improving validation coverage, and restructuring tests for better maintainability.
  Before vs After coverage screenshots are attached for reference.

Please let me know if any further changes are required. Thank you!

[Copilot]

Pull request overview

Adds a new unit test module for selector.measures.similarity to increase coverage of the similarity utilities (pairwise similarity, scaling, and similarity indices).

Changes:

• Introduces selector/measures/tests/test_similarity.py with tests for tanimoto, modified_tanimoto, pairwise_similarity_bit, scaled_similarity_matrix, and similarity_index.
• Adds negative-path tests for invalid dimensions, shape mismatches, and unsupported metric/index names.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[mohini-aggarwal]

Hi @FanwangM,
Thanks for the feedback.

• I’ve updated the test assertions to correctly handle NumPy scalar return types and synced the branch with the latest changes.
• I’ve also ensured the formatting follows PEP8 and updated the PR description to reflect the actual changes.
  Please let me know if anything else needs improvement.
  Thanks!

[FanwangM]

Looks great to me. Thanks.