[SC-13583] Fix CumulativePredictionProbabilities test docstring

[juanmleng]

Pull Request Description

What and why?

What

Updated the CumulativePredictionProbabilities test docstring to state it analyzes a single dataset, not training and test datasets.

Why

The docstring is used by the LLM to generate result descriptions. The previous version incorrectly referenced training/test comparisons, which would mislead the LLM when interpreting results. The test only accepts one dataset parameter, so the docstring now matches the implementation.

How to test

Run this test and check that the LLM-based generated description does not reference train and test datasets if the test runs only against one dataset.

What needs special review?

Dependencies, breaking changes, and deployment notes

Release notes

Checklist

• What and why
• Screenshots or videos (Frontend)
• How to test
• What needs special review
• Dependencies, breaking changes, and deployment notes
• Labels applied
• PR linked to Shortcut
• Unit tests added (Backend)
• Tested locally
• Documentation updated (if required)
• Environment variable additions/changes documented (if required)

[github-actions]

PR Summary

This pull request refines the documentation and commentary of the function that visualizes cumulative prediction probabilities for classification models. The changes focus on simplifying the explanation by removing redundant references to separate training and testing datasets. Instead, the documentation now uniformly describes the evaluation process using a single dataset, where predicted probabilities are computed, added as a column, and then sorted to create cumulative distributions for both positive and negative classes.

The updated description also refines the explanation of the metric’s purpose, including its strengths, limitations, and potential calibration issues. The text now emphasizes that the visualization is used to identify abnormalities or imbalances in the model predictions, with a more streamlined narrative avoiding unnecessary dataset separation.

Test Suggestions

• Run the function with a known dataset and verify that the cumulative probability plot correctly shows the cumulative distributions for positive and negative classes.
• Check that the new column for predicted probabilities is added to the dataset as expected.
• Ensure that the plot layout and labels match the updated documentation.