How to get the training accuracy in Scikit-learn RandomForest?

Question

I am following this tutorials: https://www.datacamp.com/community/tutorials/random-forests-classifier-python on using Scikit-learn with random forests. However, the current code only shows the test accuracy whereas I want to know the training accuracy as well since may dataset is very small.

The code to get the test accuracy is:

from sklearn import metrics
print("Accuracy:", metrics.accuracy_score(y_test, y_pred))

How would I modify this to get the training accuracy?

You can get the training set predictions as y_train_pred = clf.predict(X_train) where clf is your fitted RandomForestClassifier . After that you can use metrics.accuracy_score(y_train, y_train_pred) to get the training accuracy. Alternatively, you can use clf.score(X_train, y_train), which should give you the same result. — user11989081
– user11989081, Commented Nov 3, 2021 at 18:16
That works, please convert to an answer and I can accept it. — logankilpatrick
– logankilpatrick, Commented Nov 3, 2021 at 18:22

user11989081 · Accepted Answer · 2021-11-03 18:26:33Z

6

You can get the training set predictions as

y_train_pred = clf.predict(X_train)

where clf is your fitted RandomForestClassifier. After that you can use

metrics.accuracy_score(y_train, y_train_pred)

to calculate the training accuracy. Alternatively, you can use

clf.score(X_train, y_train)

which should give you the same result.

answered Nov 3, 2021 at 18:26

user11989081

8,75211 gold badges34 silver badges44 bronze badges

Sign up to request clarification or add additional context in comments.

1 Answer 1