[BUG] Fix Benchmarking duplicate estimator label collisions

[ashum9]

Closes #297.

What this changes

• disambiguates duplicate estimator class names in Benchmarking results with deterministic 1-based suffixes
• preserves the original class name when there is only one estimator of that class
• adds a regression test proving two DummyClassifier instances no longer overwrite one another
• adds a regression test proving distinct estimator classes keep their original labels

Validation

• pytest -q pyaptamer/benchmarking/tests/test_benchmarking_core.py

[Copilot]

Pull request overview

Fixes a Benchmarking result-collision bug where multiple estimators of the same class overwrote each other by generating deterministic, disambiguated estimator display names (with 1-based suffixes only when needed), and adds regression tests for both duplicate and unique estimator naming.

Changes:

• Add _get_estimator_names() to generate stable estimator labels and suffix duplicates (e.g., DummyClassifier[1], DummyClassifier[2]).
• Update Benchmarking.run() to key results by these stable labels instead of raw class names.
• Add regression tests to ensure duplicate estimator classes stay distinct and unique classes keep original labels.

Reviewed changes

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

File	Description
pyaptamer/benchmarking/_base.py	Introduces stable estimator naming and uses it to prevent result-row overwrites for duplicate estimator classes.
pyaptamer/benchmarking/tests/test_benchmarking_core.py	Adds regression coverage for duplicate-name disambiguation and unique-name preservation.

---

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

[siddharth7113]

left the comment to change the suffix structure, otherwise looks good.

[siddharth7113]

this notation wouldn't be right, and could be confusing in downstream usage, user would have to make a call something like this df.loc["DummyClassifier[1]"] , I would suggest to index along with strategy something to make it either like DummyClassifer_1 or based on Pattern should be fine as well.

[ashum9]

@siddharth7113 Thanks for the feedback. I updated the duplicate-estimator label format to use underscore suffixes instead of brackets.

Now duplicate estimators of the same class are labeled deterministically as DummyClassifier_1, DummyClassifier_2 (1-based, only when needed), while unique estimator classes keep their original class name.

Updated the regression tests accordingly; pytest -q pyaptamer/benchmarking/tests/test_benchmarking_core.py passes.

[siddharth7113]

Please follow PR guidelines and make sure to install and run pre-commit , the current CI checks are failing.

[siddharth7113]

cc @satvshr , for your feedback

[satvshr]

Was this also a problem which you solved or are you adding a random test