environmentaltools.common.outliers_detection

environmentaltools.common.outliers_detection(data, outliers_fraction, method='Local Outlier Factor', scaler_method='MinMaxScaler')[source]

Detect outliers in data using various sklearn algorithms.

Parameters:

  • data (array-like) – Input data (2D array expected).

  • outliers_fraction (float) – Fraction of outliers to detect (contamination parameter).

  • method (str, optional) – Outlier detection method. Options are: - “Robust covariance”: Uses EllipticEnvelope - “One-Class SVM”: Uses OneClassSVM - “Isolation Forest”: Uses IsolationForest - “Local Outlier Factor”: Uses LocalOutlierFactor (default)

  • scaler_method (str, optional) – Scaling method to apply before detection. If None, no scaling is applied. Defaults to “MinMaxScaler”.

Returns:

Boolean mask indicating outliers (True) and inliers (False).

Return type:

np.ndarray