Merge branch 'development' into master

CHANGES.txt

@@ -187,4 +187,6 @@ v<1.1.3>, <02/07/2024> -- Minor fix for SUOD changes.
 v<1.1.4>, <02/09/2024> -- Drop support of six for Python 2.
 v<1.1.4>, <03/16/2024> -- Fix documentation and temporarily restrict to Keras 2 in testing.
 v<2.0.0>, <05/21/2024> -- Moving from TF to Torch -- reimplement SO_GAAL.
-v<2.0.0>, <05/21/2024> -- Moving from TF to Torch -- implement dl base with more utilities.
+v<2.0.0>, <05/21/2024> -- Moving from TF to Torch -- implement dl base with more utilities.
+v<2.0.1>, <06/16/2024> -- Moving from TF to Torch -- reimplement DeepSVDD.
+v<2.0.1>, <06/17/2024> -- Moving from TF to Torch -- reimplement dl_base.

docs/index.rst

@@ -73,7 +73,7 @@ Welcome to PyOD, a versatile Python library for detecting anomalies in multivari
 
 * **For graph outlier detection**, please use `PyGOD <https://pygod.org/>`_.
 
-* **Performance Comparison \& Datasets**: We have a 45-page, the most comprehensive `anomaly detection benchmark paper <https://www.andrew.cmu.edu/user/yuezhao2/papers/22-neurips-adbench.pdf>`_. The fully `open-sourced ADBench <https://github.com/Minqi824/ADBench>`_ compares 30 anomaly detection algorithms on 57 benchmark datasets.
+* **Performance Comparison \& Datasets**: We have a 45-page, the most comprehensive `anomaly detection benchmark paper <https://arxiv.org/pdf/2206.09426>`_. The fully `open-sourced ADBench <https://github.com/Minqi824/ADBench>`_ compares 30 anomaly detection algorithms on 57 benchmark datasets.
 
 * **Learn more about anomaly detection** \@ `Anomaly Detection Resources <https://github.com/yzhao062/anomaly-detection-resources>`_
 
@@ -362,4 +362,4 @@ Key Attributes of a fitted model:
 .. bibliography::
    :cited:
    :labelprefix: A
-   :keyprefix: a-
+   :keyprefix: a-

docs/pyod.models.rst

@@ -43,6 +43,7 @@ pyod.models.auto\_encoder\_torch module
 
 .. automodule:: pyod.models.auto_encoder_torch
     :members:
+    :exclude-members: InnerAutoencoder
     :exclude-members:
     :show-inheritance:
     :inherited-members:
@@ -100,7 +101,7 @@ pyod.models.deep\_svdd module
 
 .. automodule:: pyod.models.deep_svdd
     :members:
-    :exclude-members:
+    :exclude-members: InnerDeepSVDD
     :undoc-members:
     :show-inheritance:
     :inherited-members:
@@ -121,7 +122,7 @@ pyod.models.ecod module
 
 .. automodule:: pyod.models.ecod
     :members:
-    :exclude-members:
+    :exclude-members: InnerDeepSVDD,
     :undoc-members:
     :show-inheritance:
     :inherited-members:
@@ -197,6 +198,7 @@ pyod.models.kpca module
 
 .. automodule:: pyod.models.kpca
     :members:
+    :exclude-members: PyODKernelPCA
     :undoc-members:
     :show-inheritance:
     :inherited-members:
@@ -328,6 +330,7 @@ pyod.models.rod module
 
 .. automodule:: pyod.models.rod
     :members:
+    :exclude-members: angle, euclidean, geometric_median, mad, process_sub, rod_3D, rod_nD, scale_angles, sigmoid
     :undoc-members:
     :show-inheritance:
     :inherited-members:
@@ -356,6 +359,7 @@ pyod.models.so\_gaal module
 
 .. automodule:: pyod.models.so_gaal
     :members:
+    :exclude-members: Discriminator, Generator
     :undoc-members:
     :show-inheritance:
     :inherited-members:

examples/auto_encoder_example.py

@@ -1,11 +1,9 @@
 # -*- coding: utf-8 -*-
 """Example of using AutoEncoder for outlier detection
 """
-# Author: Yue Zhao <zhaoy@cmu.edu>
+# Author: Tiankai Yang <tiankaiy@usc.edu>
 # License: BSD 2 clause
 
-from __future__ import division
-from __future__ import print_function
 
 import os
 import sys
@@ -14,6 +12,7 @@
 # if pyod is installed, no need to use the following line
 sys.path.append(
     os.path.abspath(os.path.join(os.path.dirname("__file__"), '..')))
+sys.path.append(os.path.abspath(os.path.dirname("__file__")))
 
 from pyod.models.auto_encoder import AutoEncoder
 from pyod.utils.data import generate_data
@@ -35,7 +34,7 @@
 
     # train AutoEncoder detector
     clf_name = 'AutoEncoder'
-    clf = AutoEncoder(epochs=30, contamination=contamination)
+    clf = AutoEncoder(epoch_num=30, contamination=contamination)
     clf.fit(X_train)
 
     # get the prediction labels and outlier scores of the training data

examples/deepsvdd_example.py

@@ -37,7 +37,7 @@
 
     # train DeepSVDD detector (Without-AE)
     clf_name = 'DeepSVDD'
-    clf = DeepSVDD(use_ae=use_ae, epochs=5, contamination=contamination,
+    clf = DeepSVDD(n_features=n_features, use_ae=use_ae, epochs=5, contamination=contamination,
                    random_state=random_state)
     clf.fit(X_train)
 

examples/perf_check.py

@@ -0,0 +1,54 @@
+# -*- coding: utf-8 -*-
+"""Example of combining multiple base outlier scores. Four combination
+frameworks are demonstrated:
+"""
+# Author: Yue Zhao <zhaoy@cmu.edu>
+# License: BSD 2 clause
+
+from __future__ import division
+from __future__ import print_function
+
+import os
+import sys
+
+# temporary solution for relative imports in case pyod is not installed
+# if pyod is installed, no need to use the following line
+sys.path.append(
+    os.path.abspath(os.path.join(os.path.dirname("__file__"), '..')))
+
+from scipy.io import loadmat
+
+from pyod.models.knn import KNN
+from pyod.utils.data import generate_data
+from pyod.utils.data import evaluate_print
+
+if __name__ == "__main__":
+
+    file_list = ['arrhythmia.mat', 'cardio.mat', 'ionosphere.mat',
+                 'letter.mat', 'pima.mat']
+    # Define data file and read X and y
+    # Generate some data if the source data is missing
+
+    for mat_file in file_list:
+
+        try:
+            mat = loadmat(os.path.join('data', mat_file))
+
+        except TypeError:
+            print('{data_file} does not exist. Use generated data'.format(
+                data_file=mat_file))
+            X, y = generate_data(train_only=True)  # load data
+        except IOError:
+            print('{data_file} does not exist. Use generated data'.format(
+                data_file=mat_file))
+            X, y = generate_data(train_only=True)  # load data
+        else:
+            X = mat['X']
+            y = mat['y'].ravel()
+
+        clf = KNN() # the algorithm you want to check
+        # clf = KNN_new()
+        clf.fit(X) # fit model
+
+        # print performance
+        evaluate_print(mat_file, y, clf.decision_scores_)

examples/so_gaal_example.py

@@ -5,8 +5,6 @@
 # Author: Winston Li <jk_zhengli@hotmail.com>
 # License: BSD 2 clause
 
-from __future__ import division
-from __future__ import print_function
 
 import os
 import sys

examples/auto_encoder_torch_example.py → examples/so_gaal_example_new.py

@@ -1,12 +1,10 @@
 # -*- coding: utf-8 -*-
-"""Example of using Autoencoder for outlier detection (pytorch)
-detection
+"""Example of using Single-Objective Generative Adversarial Active
+Learning (SO_GAAL) for outlier detection
 """
-# Author: Yue Zhao <zhaoy@cmu.edu>
+# Author: Tiankai Yang <tiankaiy@usc.edu>
 # License: BSD 2 clause
 
-from __future__ import division
-from __future__ import print_function
 
 import os
 import sys
@@ -16,15 +14,14 @@
 sys.path.append(
     os.path.abspath(os.path.join(os.path.dirname("__file__"), '..')))
 
-from pyod.models.auto_encoder_torch import AutoEncoder
+from pyod.models.so_gaal_new import SO_GAAL
 from pyod.utils.data import generate_data
 from pyod.utils.data import evaluate_print
 
-
 if __name__ == "__main__":
     contamination = 0.1  # percentage of outliers
-    n_train = 20000  # number of training points
-    n_test = 2000  # number of testing points
+    n_train = 30000  # number of training points
+    n_test = 3000  # number of testing points
     n_features = 300  # number of features
 
     # Generate sample data
@@ -35,9 +32,9 @@
                       contamination=contamination,
                       random_state=42)
 
-    # train AutoEncoder detector
-    clf_name = 'AutoEncoder'
-    clf = AutoEncoder(epochs=10)
+    # train SO_GAAL detector
+    clf_name = 'SO_GAAL'
+    clf = SO_GAAL(epoch_num=6, contamination=contamination, verbose=2)
     clf.fit(X_train)
 
     # get the prediction labels and outlier scores of the training data

examples/vae_example.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 """Example of using Variational Auto Encoder for outlier detection
 """
-# Author: Andrij Vasylenko <andrij@liverpool.ac.uk>
+# Author: Tiankai Yang <tiankaiy@usc.edu>
 # License: BSD 2 clause
 
 from __future__ import division
@@ -14,6 +14,7 @@
 # if pyod is installed, no need to use the following line
 sys.path.append(
     os.path.abspath(os.path.join(os.path.dirname("__file__"), '..')))
+sys.path.append(os.path.abspath(os.path.dirname("__file__")))
 
 from pyod.models.vae import VAE
 from pyod.utils.data import generate_data
@@ -35,7 +36,8 @@
 
     # train VAE detector (Beta-VAE)
     clf_name = 'VAE'
-    clf = VAE(epochs=30, contamination=contamination, gamma=0.8, capacity=0.2)
+    clf = VAE(epoch_num=30, contamination=contamination, beta=0.8,
+              capacity=0.2)
     clf.fit(X_train)
 
     # get the prediction labels and outlier scores of the training data

pyod/__init__.py

@@ -2,6 +2,6 @@
 
 from . import models
 from . import utils
-# TODO: add version information here
+from .version import __version__
 
-__all__ = ['models', 'utils']
+__all__ = ['models', 'utils', '__version__']

pyod/models/alad.py

@@ -8,15 +8,34 @@
 
 import numpy as np
 import pandas as pd
+import tensorflow
 from matplotlib import pyplot as plt
 from sklearn.preprocessing import StandardScaler
 from sklearn.utils import check_array
 from sklearn.utils.validation import check_is_fitted
 
 from .base import BaseDetector
-from .base_dl import _get_tensorflow_version
 from ..utils.utility import check_parameter
 
+
+# Old function, deprecat this in the future
+def _get_tensorflow_version():  # pragma: no cover
+    """ Utility function to decide the version of tensorflow, which will
+    affect how to import keras models.
+
+    Returns
+    -------
+    tensorflow version : int
+
+    """
+
+    tf_version = str(tensorflow.__version__)
+    if int(tf_version.split(".")[0]) != 1 and int(
+            tf_version.split(".")[0]) != 2:
+        raise ValueError("tensorflow version error")
+
+    return int(tf_version.split(".")[0]) * 100 + int(tf_version.split(".")[1])
+
 # if tensorflow 2, import from tf directly
 if _get_tensorflow_version() < 200:
     raise NotImplementedError('Model not implemented for Tensorflow version 1')

pyod/models/anogan.py

@@ -10,14 +10,34 @@
 import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
+import tensorflow
 from sklearn.preprocessing import StandardScaler
 from sklearn.utils import check_array
 from sklearn.utils.validation import check_is_fitted
 
 from .base import BaseDetector
-from .base_dl import _get_tensorflow_version
 from ..utils.utility import check_parameter
 
+
+# Old function, deprecat this in the future
+def _get_tensorflow_version():  # pragma: no cover
+    """ Utility function to decide the version of tensorflow, which will
+    affect how to import keras models.
+
+    Returns
+    -------
+    tensorflow version : int
+
+    """
+
+    tf_version = str(tensorflow.__version__)
+    if int(tf_version.split(".")[0]) != 1 and int(
+            tf_version.split(".")[0]) != 2:
+        raise ValueError("tensorflow version error")
+
+    return int(tf_version.split(".")[0]) * 100 + int(tf_version.split(".")[1])
+
+
 # if tensorflow 2, import from tf directly
 if _get_tensorflow_version() < 200:
     raise NotImplementedError('Model not implemented for Tensorflow version 1')