Description

GridSearchCV and RandomizedSearchCV do not allow for passing parameters to the scorer function. This could be made possible by adding an extra scorer_params, similar to the fit_params argument.

For consistency with fit_params, special care would have to be paid to sample weights. Weights fed through fit_params are currently well-distributed across training folds. Similarly, weights fed through scorer_params would have to be distributed across validation folds.

Is adding this feature under consideration?

If not, should I give it a try?

(Nota: follow-up issue to the closed #8156. I believed the change in focus required a new thread.)