Reputation: 913
When using randomforest in sklearn to do a binary classification.
I know I can use clf.predict(X)
to get the predicted class.
And when I use clf.predict_proba(X)
, I get an array look like this:
I think the first column indicates the probability of the prediction? How can I get a column that is the probability of the class being 1?
Upvotes: 2
Views: 5240
Reputation: 402553
From the RandomForest.predict_proba
docs:
predict_proba(X)
Predict class probabilities for X.
Returns:
p
: [...] The class probabilities of the input samples. The order of the classes corresponds to that in the attributeclasses_
.
You can look at the clf.classes_
attribute, see at what index your class 1
appears, and then just access the probabilities as so:
prob_class_1 = clf.predict_proba(X)[:, i]
Where i
is the index of class 1
in clf.classes_
.
Upvotes: 8