[feature request] sklearn cross_val_score

[r0f1]

Hi, Often times I want to compare different classifiers on the same dataset and I find myself writing code that looks like this:

list_of_feature_list = [...] # could looks like this: [[0,1,2], [2,4,5], [0,1,2,3,4,5]]

for feature_list in list_of_feature_list:
    X = Xbig[features_list] # features
    y = ...                 # target

    model = LogisticRegression()
    scores = cross_val_score(model, X, y)
    dummy = DummyClassifier(strategy="most_frequent")
    dummys = np.mean(cross_val_score(dummy, X, y))

    print(scores)
    print(dummys)

I was wondering if ray could be used to speed up the process. Or more specifically: Can I use ray to do cross validation and use it instead of cross_val_score()? If not, I think that would be a useful feature to add.

Thank you.

[Yard1]

I think you could just use grid search with one possible combination of parameters to essentially get the equivalent of sklearn's CV on Ray