interpret
interpret copied to clipboard
Probability calibration
Hello!
Would you be interested in getting a contribution for probability calibration plots?
Similar functionality exists, e.g., as sklearn.calibrations CalibrationDisplay or a different approach can be found here.
An implementation within interpret could be relatively simple (similar to interpret.perf.ROC re-using the core metrics calculation from sklearn) but with a consistent plotting feel and style.
Please let me know if that sounds useful.
Hi @Krzys25 - That does sound useful, and it would be great if you wanted to submit a PR with this visualization!