Splitting the dataset for training and testing

Let's see how to split our data properly into training and testing datasets.

How to do it…

1. Add the following code snippet into the same Python file as the previous recipe:

   from sklearn import cross_validation
   
   X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y, test_size=0.25, random_state=5)
   classifier_gaussiannb_new = GaussianNB()
   classifier_gaussiannb_new.fit(X_train, y_train)

   Here, we allocated 25% of the data for testing, as specified by the test_size parameter. The remaining 75% of the data will be used for training.

2. Let's evaluate the classifier on test data:

   y_test_pred = classifier_gaussiannb_new.predict(X_test)

3. Let's compute the accuracy of the classifier:

   accuracy = 100.0 * (y_test == y_test_pred).sum() / X_test.shape[0]
   print "Accuracy of the classifier =", round(accuracy, 2), "%"

4. Let's plot the datapoints and the boundaries on test data:

   plot_classifier(classifier_gaussiannb_new...