1  # -*- coding: utf-8 -*-
 2  """Example of using KDE for outlier detection
 3  """
 4  
 5  # Author: Akira Tamamori
 6  # License: BSD 2 clause
 7  
 8  from __future__ import division, print_function
 9  
10  import os
11  import sys
12  
13  from pyod.models.kde import KDE
14  from pyod.utils.data import evaluate_print, generate_data
15  from pyod.utils.example import visualize
16  
17  # temporary solution for relative imports in case pyod is not installed
18  # if pyod is installed, no need to use the following line
19  sys.path.append(os.path.abspath(os.path.join(os.path.dirname("__file__"), "..")))
20  
21  if __name__ == "__main__":
22      contamination = 0.1  # percentage of outliers
23      n_train = 200  # number of training points
24      n_test = 100  # number of testing points
25  
26      # Generate sample data
27      X_train, X_test, y_train, y_test = \
28          generate_data(n_train=n_train,
29                        n_test=n_test,
30                        n_features=2,
31                        contamination=contamination,
32                        random_state=42)
33  
34      # train kNN detector
35      clf_name = "KDE"
36      clf = KDE()
37      clf.fit(X_train)
38  
39      # get the prediction labels and outlier scores of the training data
40      y_train_pred = clf.labels_  # binary labels (0: inliers, 1: outliers)
41      y_train_scores = clf.decision_scores_  # raw outlier scores
42  
43      # get the prediction on the test data
44      y_test_pred = clf.predict(X_test)  # outlier labels (0 or 1)
45      y_test_scores = clf.decision_function(X_test)  # outlier scores
46  
47      # evaluate and print the results
48      print("\nOn Training Data:")
49      evaluate_print(clf_name, y_train, y_train_scores)
50      print("\nOn Test Data:")
51      evaluate_print(clf_name, y_test, y_test_scores)
52  
53      # visualize the results
54      visualize(
55          clf_name,
56          X_train,
57          y_train,
58          X_test,
59          y_test,
60          y_train_pred,
61          y_test_pred,
62          show_figure=True,
63          save_figure=True,
64      )