CODEWITHSUNDEEP
pythonscikit-learn
paraflou
paraflou

Reputation: 413

How to pass parameters in scoring attributes of sklearn cross_validate function?

I would like to change the average parameter of the precision metric because this error occurs

"ValueError: Target is multiclass but average='binary'. Please choose another average setting."

I have read the official website but I couldn't find an answer in terms of using cross_validate function.

clf = RandomForestClassifier()
scoring = ['accuracy', 'precision']

scores = cross_validate(clf, X, Y, scoring=scoring, cv=10, return_train_score=False, n_jobs=-1)

Any idea how to handle this?

Upvotes: 6

Views: 1503

Answers (1)

panktijk
panktijk

Reputation: 1614

Use make_scorer which allows you to specify parameters for your individual scoring metrics, then use a dictionary to map multiple metrics to names:

from sklearn.metrics import accuracy_score, precision_score, make_scorer
scoring = {'Accuracy': make_scorer(accuracy_score), 
           'Precision': make_scorer(precision_score, average='None')}

scores = cross_validate(clf, X, Y, scoring=scoring, ...)

Refer this example

Upvotes: 8

Related Questions