crashes on X=None

[kootenpv]

No description provided.

[ryanpeach]

Shouldn't it? Also, see my most recent pull request for a maximize without data needed.

[kootenpv]

My apologies for the bad naming. I meant "it currently crashes complaining X=None". When I give X to the check_cv function, it stops crashing. How could you do cross validation without X? It just looks like you forgot to pass X to check_cv

[ryanpeach]

Hm, I never encountered that error, could you highlight the issue in code?

My "maximize" function is more similar to a scipy.optimize format than a sklearn format, no CV. However, it uses the same logic as the original code, and my hope is the sklearn estimator can use the maximize function to prevent code duplication. Sometimes, you just need to optimize a function's output based on parameters, other times you need to train a learner on data, but those two problems have an intersection with the same evolutionary parameter_grid search.