Upgraded some libraries like numpy,pandas and small refactorings.

CHANGES.txt

@@ -1 +1,2 @@
-v<0.0.0>, <08/15/2020> -- Initial release.
+v<0.1.0>, <08/15/2020> -- Initial release.
+v<0.2.0> <06/10/2023> -- Upgrade numpy version and small refactorings

examples/example_ensemble.py

@@ -26,7 +26,7 @@
     ensembler = AverageScoreEnsembler()  # Ensembler module.
 
     for X, y in tqdm(iterator.iter(X_all, y_all)):  # Iterate over examples.
-        model_scores = np.empty(len(models), dtype=np.float)
+        model_scores = np.empty(len(models), dtype=np.float64)
 
         # Fit & Score via for each model.
         for i, model in enumerate(models):

pysad/core/base_model.py

@@ -12,7 +12,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit.
+            X (np.float64 array of shape (num_features,)): The instance to fit.
             y (int): The label of the instance (Optional for unsupervised models, default=None).
 
         Returns:
@@ -25,7 +25,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.
@@ -36,7 +36,7 @@ def fit_score_partial(self, X, y=None):
         """Applies fit_partial and score_partial to the next instance, respectively.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit and score.
+            X (np.float64 array of shape (num_features,)): The instance to fit and score.
             y (int): The label of the instance (Optional for unsupervised models, default=None).
 
         Returns:
@@ -48,7 +48,7 @@ def fit(self, X, y=None):
         """Fits the model to all instances in order.
 
         Args:
-            X (np.float array of shape (num_instances, num_features)): The instances in order to fit.
+            X (np.float64 array of shape (num_instances, num_features)): The instances in order to fit.
             y (int): The labels of the instances in order to fit (Optional for unsupervised models, default=None).
 
         Returns:
@@ -63,12 +63,12 @@ def score(self, X):
         """Scores all instaces via score_partial iteratively.
 
         Args:
-            X (np.float array of shape (num_instances, num_features)): The instances in order to score.
+            X (np.float64 array of shape (num_instances, num_features)): The instances in order to score.
 
         Returns:
-            np.float array of shape (num_instances,): The anomalousness scores of the instances in order.
+            np.float64 array of shape (num_instances,): The anomalousness scores of the instances in order.
         """
-        y_pred = np.empty(X.shape[0], dtype=np.float)
+        y_pred = np.empty(X.shape[0], dtype=np.float64)
         for i, (xi, _) in enumerate(_iterate(X)):
             y_pred[i] = self.score_partial(xi)
 
@@ -78,13 +78,13 @@ def fit_score(self, X, y=None):
         """This helper method applies fit_score_partial to all instances in order.
 
         Args:
-            X (np.float array of shape (num_instances, num_features)): The instances in order to fit.
-            y (np.int array of shape (num_instances, )): The labels of the instances in order to fit (Optional for unsupervised models, default=None).
+            X (np.float64 array of shape (num_instances, num_features)): The instances in order to fit.
+            y (np.int32 array of shape (num_instances, )): The labels of the instances in order to fit (Optional for unsupervised models, default=None).
 
         Returns:
-            np.float array of shape (num_instances,): The anomalousness scores of the instances in order.
+            np.float64 array of shape (num_instances,): The anomalousness scores of the instances in order.
         """
-        y_pred = np.zeros(X.shape[0], dtype=np.float)
+        y_pred = np.zeros(X.shape[0], dtype=np.float64)
         for i, (xi, yi) in enumerate(_iterate(X, y)):
             y_pred[i] = self.fit_score_partial(xi, yi)
 

pysad/core/base_postprocessor.py

@@ -46,12 +46,12 @@ def transform(self, scores):
         """Shortcut method that iteratively applies transform_partial to all instances in order.
 
         Args:
-            np.float array of shape (num_instances,): Input scores.
+            np.float64 array of shape (num_instances,): Input scores.
 
         Returns:
-            np.float array of shape (num_instances,): Processed scores.
+            np.float64 array of shape (num_instances,): Processed scores.
         """
-        processed_scores = np.empty(scores.shape[0], dtype=np.float)
+        processed_scores = np.empty(scores.shape[0], dtype=np.float64)
         for i, (score, _) in enumerate(_iterate(scores)):
             processed_scores[i] = self.transform_partial(score)
 
@@ -61,7 +61,7 @@ def fit(self, scores):
         """Shortcut method that iteratively applies fit_partial to all instances in order.
 
         Args:
-            np.float array of shape (num_instances,): Input scores.
+            np.float64 array of shape (num_instances,): Input scores.
 
         Returns:
             object: self.
@@ -75,12 +75,12 @@ def fit_transform(self, scores):
         """Shortcut method that iteratively applies fit_transform_partial to all instances in order.
 
         Args:
-            np.float array of shape (num_instances,): Input scores.
+            np.float64 array of shape (num_instances,): Input scores.
 
         Returns:
-            np.float array of shape (num_instances,): Processed scores.
+            np.float64 array of shape (num_instances,): Processed scores.
         """
-        processed_scores = np.empty(scores.shape[0], dtype=np.float)
+        processed_scores = np.empty(scores.shape[0], dtype=np.float64)
         for i, (score, _) in enumerate(_iterate(scores)):
             processed_scores[i] = self.fit_transform_partial(score)
 

pysad/core/base_transformer.py

@@ -15,7 +15,7 @@ def fit_partial(self, X):
         """Fits particular (next) timestep's features to train the transformer.
 
         Args:
-            X (np.float array of shape (num_components,)): Input feature vector.
+            X (np.float64 array of shape (num_components,)): Input feature vector.
         Returns:
             object: self.
         """
@@ -26,35 +26,35 @@ def transform_partial(self, X):
         """Transforms particular (next) timestep's vector.
 
         Args:
-            X (np.float array of shape (num_features,)): Input feature vector.
+            X (np.float64 array of shape (num_features,)): Input feature vector.
 
         Returns:
-            transformed_X (np.float array of shape (num_components,)): Projected feature vector.
+            transformed_X (np.float64 array of shape (num_components,)): Projected feature vector.
         """
         pass
 
     def fit_transform_partial(self, X):
         """Shortcut method that iteratively applies fit_partial and transform_partial, respectively.
 
         Args:
-            X (np.float array of shape (num_components,)): Input feature vector.
+            X (np.float64 array of shape (num_components,)): Input feature vector.
 
         Returns:
-            transformed_X (np.float array of shape (num_components,)): Projected feature vector.
+            transformed_X (np.float64 array of shape (num_components,)): Projected feature vector.
         """
         return self.fit_partial(X).transform_partial(X)
 
     def transform(self, X):
         """Shortcut method that iteratively applies transform_partial to all instances in order.
 
         Args:
-            X (np.float array of shape (num_instances, num_features)): Input feature vectors.
+            X (np.float64 array of shape (num_instances, num_features)): Input feature vectors.
 
         Returns:
-            np.float array of shape (num_instances, num_components): Projected feature vectors.
+            np.float64 array of shape (num_instances, num_components): Projected feature vectors.
         """
         output_dims = self.output_dims if self.output_dims > 0 else X.shape[1]
-        transformed_X = np.empty((X.shape[0], output_dims), dtype=np.float)
+        transformed_X = np.empty((X.shape[0], output_dims), dtype=np.float64)
         for i, (xi, _) in enumerate(_iterate(X)):
             transformed_X[i] = self.transform_partial(xi)
 
@@ -64,7 +64,7 @@ def fit(self, X):
         """Shortcut method that iteratively applies fit_partial to all instances in order.
 
         Args:
-            X (np.float array of shape (num_instances, num_features)): Input feature vectors.
+            X (np.float64 array of shape (num_instances, num_features)): Input feature vectors.
 
         Returns:
             object: The fitted transformer
@@ -78,13 +78,13 @@ def fit_transform(self, X):
         """Shortcut method that iteratively applies fit_transform_partial to all instances in order.
 
         Args:
-            X (np.float array of shape (num_instances, num_components)): Input feature vectors.
+            X (np.float64 array of shape (num_instances, num_components)): Input feature vectors.
 
         Returns:
-            np.float array of shape (num_instances, num_components): Projected feature vectors.
+            np.float64 array of shape (num_instances, num_components): Projected feature vectors.
         """
         output_dims = self.output_dims if self.output_dims > 0 else X.shape[1]
-        transformed_X = np.empty((X.shape[0], output_dims), dtype=np.float)
+        transformed_X = np.empty((X.shape[0], output_dims), dtype=np.float64)
         for i, (xi, _) in enumerate(_iterate(X)):
             transformed_X[i] = self.fit_transform_partial(xi)
 

pysad/models/exact_storm.py

@@ -22,7 +22,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance. Simply, adds the instance to the window.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit.
+            X (np.float64 array of shape (num_features,)): The instance to fit.
             y (int): Ignored since the model is unsupervised (Default=None).
 
         Returns:
@@ -36,7 +36,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.

pysad/models/half_space_trees.py

@@ -7,12 +7,12 @@ class HalfSpaceTrees(BaseModel):
     """Half-Space Trees method :cite:`tan2011fast`.
 
     Args:
-        feature_mins (np.float array of shape (num_features,)): Minimum boundary of the features.
-        feature_maxes (np.float array of shape (num_features,)): Maximum boundary of the features.
+        feature_mins (np.float64 array of shape (num_features,)): Minimum boundary of the features.
+        feature_maxes (np.float64 array of shape (num_features,)): Maximum boundary of the features.
         window_size (int): The size of the window (Default=100).
         num_trees (int): The number of treesint (Default=25).
         max_depth (int): Maximum depth of the trees (Default=15).
-        initial_window_X (np.float array of shape (num_initial_instances,num_features)): The initial window to fit for initial calibration period. If not `None`, we simply apply fit to these instances (Default=None).
+        initial_window_X (np.float64 array of shape (num_initial_instances,num_features)): The initial window to fit for initial calibration period. If not `None`, we simply apply fit to these instances (Default=None).
     """
 
     def __init__(
@@ -109,7 +109,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit.
+            X (np.float64 array of shape (num_features,)): The instance to fit.
             y (int): Ignored since the model is unsupervised (Default=None).
 
         Returns:
@@ -138,7 +138,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.

pysad/models/iforest_asd.py

@@ -6,7 +6,7 @@ class IForestASD(ReferenceWindowModel):
     """An Anomaly Detection Approach Based on Isolation Forest Algorithm for Streaming Data using Sliding Window :cite:`ding2013anomaly`. Note that concept drift is not implemented since it is a part of the simulation. See Algorithm 2 in "An Anomaly Detection Approach Based on Isolation Forest Algorithm for Streaming Data using Sliding Window" paper. This method is unsupervised so it is not needed to give y as parameter.
 
     Args:
-        initial_window_X (np.float array of shape (num_initial_instances,num_features)): The initial window to fit for initial calibration period. We simply apply fit to these instances (Default=None).
+        initial_window_X (np.float64 array of shape (num_initial_instances,num_features)): The initial window to fit for initial calibration period. We simply apply fit to these instances (Default=None).
         window_size (int): The size of the reference window and its sliding (Default=2048).
     """
 

pysad/models/integrations/one_fit_model.py

@@ -6,8 +6,8 @@ class OneFitModel(PYODModel):
 
         Args:
             model_cls (class):     The model class to be instantiated.
-            initial_X (np.float array of shape (num_initial_instances, num_features)): Initial instances to fit.
-            initial_y (np.int array of shape (num_initial_instances,): Initial window's ground truth labels. Used if not None. Needs to be `None` for the unsupervised `model_cls` models. (Default=None).
+            initial_X (np.float64 array of shape (num_initial_instances, num_features)): Initial instances to fit.
+            initial_y (np.int32 array of shape (num_initial_instances,): Initial window's ground truth labels. Used if not None. Needs to be `None` for the unsupervised `model_cls` models. (Default=None).
             **kwargs (Keyword arguments): Keyword arguments that are passed to the `model_cls`.
     """
 
@@ -43,7 +43,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.

pysad/models/integrations/reference_window_model.py

@@ -9,8 +9,8 @@ class ReferenceWindowModel(PYODModel):
             model_cls (class): The model class to be instantiated.
             window_size (int): The size of each window.
             sliding_size (int): The sliding length of the windows.
-            initial_X (np.float array of shape (num_initial_instances, num_features)): Initial instances to fit.
-            initial_y (np.int array of shape (num_initial_instances,)): Initial window's ground truth labels. Used if not None. Needs to be `None` for the unsupervised `model_cls` models. (Default=None).
+            initial_X (np.float64 array of shape (num_initial_instances, num_features)): Initial instances to fit.
+            initial_y (np.int32 array of shape (num_initial_instances,)): Initial window's ground truth labels. Used if not None. Needs to be `None` for the unsupervised `model_cls` models. (Default=None).
             **kwargs (Keyword arguments): Keyword arguments that is passed to the `model_cls`.
     """
 
@@ -52,7 +52,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit.
+            X (np.float64 array of shape (num_features,)): The instance to fit.
             y (int): The label of the instance (Optional for unsupervised models, default=None).
 
         Returns:
@@ -97,7 +97,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.

pysad/models/kitnet.py

@@ -33,7 +33,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance. Simply, adds the instance to the window.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to fit.
+            X (np.float64 array of shape (num_features,)): The instance to fit.
             y (int): Ignored since the model is unsupervised (Default=None).
 
         Returns:
@@ -57,7 +57,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (num_features,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.

pysad/models/knn_cad.py

@@ -35,7 +35,7 @@ def fit_partial(self, X, y=None):
         """Fits the model to next instance. Note that this model is univariate.
 
         Args:
-            X (np.float array of shape (1,)): The instance to fit.
+            X (np.float64 array of shape (1,)): The instance to fit.
             y (int): Ignored since the model is unsupervised (Default=None).
 
         Returns:
@@ -82,7 +82,7 @@ def score_partial(self, X):
         """Scores the anomalousness of the next instance.
 
         Args:
-            X (np.float array of shape (1,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
+            X (np.float64 array of shape (1,)): The instance to score. Higher scores represent more anomalous instances whereas lower scores correspond to more normal instances.
 
         Returns:
             float: The anomalousness score of the input instance.