All issues

Search results

predict_p_online, smoothing=False: p-value ranks the wrong object, and the Mondrian path crashes

Status: Open.

#52 In henrikbostrom/crepes;

· shivamlalakiya opened on Aug 18, 2026

Set membership uses p >= 1-confidence against a float threshold: is the boundary behaviour intentional?

Status: Open.

#48 In henrikbostrom/crepes;

· shivamlalakiya opened on Aug 2, 2026

ConformalClassifier gives no warning when the calibration set is too small for the confidence level

Status: Open.

#47 In henrikbostrom/crepes;

· shivamlalakiya opened on Aug 2, 2026

predict_set docstring states the inverse of the implemented set-membership rule

Status: Open.

#46 In henrikbostrom/crepes;

· shivamlalakiya opened on Aug 2, 2026

[FEATURE]: customized sklearn Pipeline wrapper object that includes predict_p and predict_set methods

Status: Open.

#43 In henrikbostrom/crepes;

· lukethomrichardson opened on Oct 29, 2025

CRPS when sigmas not 1

Status: Open.

#40 In henrikbostrom/crepes;

· LucasMMK opened on Jul 3, 2025

CRPS not implemented for semi-online conformal predictive systems

Status: Open.

#39 In henrikbostrom/crepes;

· henrikbostrom opened on Mar 31, 2025

Avoid generating the full cpds array when calculating CRPS

enhancementNew feature or requestNew feature or request

Status: Open.

#10 In henrikbostrom/crepes;

· henrikbostrom opened on Jun 20, 2023