Closed HaroldMills closed 1 year ago
I integrated the csv-exporting code you wrote to my research code, so I'm now exporting the calibration parameters as a csv file. We should not need to retain the scikit-learn dependency now.
Okay, I'm closing this issue. Our pinned scikit-learn dependency was removed from the pyproject.toml file in 156b2b6. (scikit-learn remains an indirect dependency, but we don't pin it.)
To avoid warnings from scikit-learn about potential probability calibration version incompatibilities, that package is currently pinned to version 1.0.1. That version is about a year and a half old, though, and using it causes installation of the Nighthawk package to trigger a build of a scikit-learn Python wheel, which slows installation considerably. It would be good to either be able to use a more recent version of scikit-learn that doesn't trigger such a build (the latest version, 1.2.2, does not), or perhaps eliminate the use of scikit-learn altogether by implementing our own version of probability calibration.