Bayesian Model Average

[comane]

New module for Bayesian model averaging

New Features:

• Added a new module colibri.model_average for performing Bayesian model averaging, including functions for selecting fits and computing weights (colibri/model_average.py).

Enhancements:

• Introduced a new dataclass ColibriFitSpec to encapsulate fit specifications and Bayesian metrics (colibri/core.py).
• Added a new method parse_fits in colibri/config.py to parse a list of fits into ColibriFitSpec objects (colibri/config.py).

Refactoring:

• Refactored the ns_resampler script to use new utility functions ns_fit_resampler and write_resampled_ns_fit for improved clarity and reusability (colibri/scripts/ns_resampler.py). [1] [2]
• Updated colibri/utils.py with new utility functions for resampling and writing resampled fits, and included necessary imports (colibri/utils.py). [1] [2] [3]

Miscellaneous:

• Added pandas and new imports to colibri/config.py to support the new functionalities (colibri/config.py).
• Included the new module colibri.model_average in the colibri/app.py module list (colibri/app.py).

[codecov]

Codecov Report

❌ Patch coverage is 46.80851% with 50 lines in your changes missing coverage. Please review.
✅ Project coverage is 91.85%. Comparing base (19975dd) to head (b47f690).

Files with missing lines	Patch %	Lines
colibri/utils.py	5.76%	49 Missing ⚠️
colibri/model_average.py	96.29%	1 Missing ⚠️

❌ Your patch check has failed because the patch coverage (46.80%) is below the target coverage (100.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #293      +/-   ##
==========================================
- Coverage   95.01%   91.85%   -3.16%     
==========================================
  Files          29       30       +1     
  Lines        1344     1437      +93     
==========================================
+ Hits         1277     1320      +43     
- Misses         67      117      +50     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[comane]

• add example runcard
• Allow to specify that the fit is composed out of two: analytic and numeric -> allows for taking the correct evidence.