fix: join unit metrics docstrings for composite tests doc

[juanmleng]

Internal Notes for Reviewers

Fixed static descriptions handling for composite tests by combining docstrings from individual unit metrics. Previously, this was throwing an error in the notebook run_unit_metrics.ipynb as it was assuming a non-existing test_id for composite unit metrics tests. Now the docstring is built from a combination of the docstrings of the individual unit metrics.

External Release Notes

[johnwalz97]

thanks

[github-actions]

PR Summary

This pull request introduces an enhancement to the validmind/tests/run.py file, specifically in the _run_comparison_test function. The change improves the generation of test documentation by checking if unit_metrics are present. If they are, it constructs a detailed documentation string by iterating over each unit_metric, retrieving its name and documentation, and joining them with newlines. If unit_metrics are not present, it defaults to the previous behavior of loading the documentation for the test_id. This enhancement allows for more comprehensive documentation when multiple unit metrics are involved in a test.

The version number in pyproject.toml and validmind/__version__.py has been incremented from 2.7.1 to 2.7.2 to reflect this change.

Test Suggestions

• Test the _run_comparison_test function with a non-empty unit_metrics list to ensure the documentation is generated correctly for each metric.
• Test the _run_comparison_test function with an empty unit_metrics list to verify that it falls back to the original behavior.
• Verify that the output format of the documentation string is as expected when multiple unit metrics are provided.
• Check for any performance impacts when processing a large number of unit metrics.