How to do it...

1. Add the following code fragment into the same Python file:

from sklearn import cross_validation
from sklearn.naive_bayes import GaussianNB
import numpy as np
import matplotlib.pyplot as plt
in_file = 'data_multivar.txt'
a = []
b = []
with open(in_file, 'r') as f:
  for line in f.readlines():
    data = [float(x) for x in line.split(',')]
    a.append(data[:-1])
    b.append(data[-1])
a = np.array(a)
b = np.array(b)

2. Allocate 75% of data for training and 25% of data for testing:

a_training, a_testing, b_training, b_testing = cross_validation.train_test_split(a, b, test_size=0.25, random_state=5)
classification_gaussiannb_new = GaussianNB()
classification_gaussiannb_new.fit(a_training, b_training)

3. Evaluate the classifier performance on test data:

b_test_pred = classification_gaussiannb_new.predict(a_testing)

4. Compute the accuracy of the classifier system:

correctness = 100.0 * (b_testing == b_test_pred).sum() / a_testing.shape[0]
print "correctness of the classification =", round(correctness, 2), "%"

5. Plot the datapoints and the boundaries for test data:

def plot_classification(classification_gaussiannb_new, a_testing , b_testing):
  a_min, a_max = min(a_testing[:, 0]) - 1.0, max(a_testing[:, 0]) + 1.0
  b_min, b_max = min(a_testing[:, 1]) - 1.0, max(a_testing[:, 1]) + 1.0
  step_size = 0.01
  a_values, b_values = np.meshgrid(np.arange(a_min, a_max, step_size), np.arange(b_min, b_max, step_size))
  mesh_output = classification_gaussiannb_new.predict(np.c_[a_values.ravel(), b_values.ravel()])
  mesh_output = mesh_output.reshape(a_values.shape)
  plt.figure()
  plt.pcolormesh(a_values, b_values, mesh_output, cmap=plt.cm.gray)
  plt.scatter(a_testing[:, 0], a_testing[:, 1], c=b_testing , s=80, edgecolors='black', linewidth=1,cmap=plt.cm.Paired)
  # specify the boundaries of the figure
  plt.xlim(a_values.min(), a_values.max())
  plt.ylim(b_values.min(), b_values.max())
  # specify the ticks on the X and Y axes
  plt.xticks((np.arange(int(min(a_testing[:, 0])-1), int(max(a_testing[:, 0])+1), 1.0)))
  plt.yticks((np.arange(int(min(a_testing[:, 1])-1), int(max(a_testing[:, 1])+1), 1.0)))
  plt.show()
plot_classification(classification_gaussiannb_new, a_testing, b_testing)

The accuracy obtained while splitting the dataset is shown in the following screenshot: