Added testing to the module:

[Thematiq]

• Unit tests for PyBBT class
• Regression tests based on the recent article

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

[Copilot]

Pull request overview

Copilot reviewed 14 out of 16 changed files in this pull request and generated 6 comments.

Comments suppressed due to low confidence (3)

bbttest/bbt/py_bbt.py:234

• posterior_table() now accepts columns: Iterable[...], but columns is iterated in the validation loop and then used again in out_table[["pair", *columns]]. If a caller passes a one-shot iterable (e.g., a generator), it will be exhausted and the returned DataFrame will omit requested columns. Convert columns to a list once near the top of the method before iterating over it multiple times.
  bbttest/bbt/py_bbt.py:290
• rope_comparison_control_table() currently always appends row["left_model"] into the result buckets. However _get_pwin(..., control=...) emits comparisons in both directions (e.g. control > other for models ranked below the control; see bbttest/bbt/alg.py:217-222), so left_model can be the control itself. In that case this method will incorrectly classify the control as “better/equivalent/unknown/worse” instead of classifying the other model. Use row["left_model"] == control_model to decide whether to bucket right_model or left_model (and base the decision on the raw interpretation symbol rather than string-matching).
  bbttest/bbt/py_bbt.py:255
• The rope_comparison_control_table() docstring’s Args: section doesn’t match the function signature (it refers to model and ropes, but the parameters are rope_values, control_model, etc.). Please update the docstring argument names/descriptions so generated docs and IDE help are accurate.

---

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