RobustRandomCutForestAnomalyDetector

class dtaianomaly.anomaly_detection.RobustRandomCutForestAnomalyDetector(window_size: int | Literal['fft', 'acf', 'mwf', 'suss'], stride: int = 1, online_learning: bool = False, n_estimators: int = 100, max_samples: float | int | Literal['auto'] = 'auto', precision: int = 9, random_state: int = None)[source]

Detect anomalies using robust random cut forest [14].

A random cut tree is a binary tree similar to an isolation free. The main differerence is how the dimension to split on is selected, and how the anomaly scores are comupted. For a random cut tree, the dimension is chosen based on the difference between the minimum and maximum value along that dimension, to prioritise dimensions with wider spread. The anomaly score is computed based on collusive displacement: a sample is more anomalous if the height of other samples within the tree is substantially smaller when the sample is removed from the tree. This is based on the assumption that anomalies make it harder to explain the dataset as a whole. A robust random cut tree then consists of multiple random cut trees. Because the samples can be dynamically removed from and added to the tree, the model can also deal with streaming data.

Parameters:

window_sizeint or str

The window size to use for extracting sliding windows from the time series. This value will be passed to compute_window_size().

strideint, default=1

The stride, i.e., the step size for extracting sliding windows from the time series.

online_learningbool, default=True

Whether to perform online learning, i.e., update the trees when detecting anomalies.

n_estimatorsint, default=100

The number of base trees in the ensemble.

max_samplesint or float, default=’auto’

The number of samples to draw for training each base estimator:

if int: Draw at most max_samples samples.
if float: Draw at most max_samples percentage of the samples.
if 'auto': Set max_samples=min(256, n_windows).

precisionint, default=9

Floating-point precision for distinguishing duplicate points.

random_stateint, default=None

The seed used to create a random number generator from numpy.

Attributes:

window_size_int: The effectively used window size for this anomaly detector
max_samples_int: The effectively used maximum number of samples.
forest_list of RCTree: The trees which is used to isolate the observations.

Examples

>>> from dtaianomaly.anomaly_detection import RobustRandomCutForestAnomalyDetector
>>> from dtaianomaly.data import demonstration_time_series
>>> x, y = demonstration_time_series()
>>> rrcf = RobustRandomCutForestAnomalyDetector(10, random_state=0).fit(x)
>>> rrcf.decision_function(x)
array([3.72400345, 3.92008391, 4.05253107, ..., 3.74515887, 3.64998068,
       3.26297901]...)

check_is_fitted() → None

Raise an error if this object is not fitted.

Check whether this object is fitted, and if it is not fitted, an exception is thrown.

Raises:

NotFittedError: If this object is not fitted.

decision_function(X: ndarray) → array

Compute anomaly scores.

Compute the anomaly scores for the given time series using this detector.

Parameters:

Xarray-like of shape (n_samples, n_attributes): Input time series.

Returns:

array-like of shape (n_samples): The computed anomaly scores.

fit(X: ndarray, y: ndarray = None, **kwargs) → BaseDetector

Fit this detector.

Fit this detector to the given data.

Parameters:

Xarray-like of shape (n_samples, n_attributes): Input time series.
yarray-like, default=None: Ground-truth information.
**kwargs: Additional parameters to be used to fit the anomaly detector.

Returns:

BaseDetector: Returns the instance itself.

is_fitted() → bool

Check whether this object is fitted.

Check whether all the attributes of this object that end with an underscore (‘_’) has been initialized.

Returns:

bool: True if and only if all the attributes of this object ending with ‘_’ are initialized.

predict_confidence(X: ndarray, X_train: ndarray = None, contamination: float = 0.05, decision_scores_given: bool = False)

Predict the confidence of the anomaly scores on the test given test data [26].

This method implements ExCeeD (Example-wise Confidence of anomaly Detectors) to estimate the confidence. ExCeed transforms the predicted decision scores to probability estimates using a Bayesian approach, which enables to assign a confidence score to each prediction which captures the uncertainty of the anomaly detector in that prediction.

Parameters:

Xarray-like of shape (n_samples, n_attributes): The test time series for which the confidence of anomaly scores should be predicted.
X_trainarray-like of shape (n_samples_train, n_attributes), default=None: The training time series, which can be used as reference. If X_train=None, the test set is used as reference set.
contaminationfloat, default=0.05: The (estimated) contamination rate for the data, i.e., the expected percentage of anomalies.
decision_scores_givenbool, default=False: Whether the given X and X_train represent time series data or decision scores. If decision_scores_given=False (default), then the given arrays are interpreted as time series. Otherwise, they are interpreted as decision scores, as computed by decision_function().

Returns:

array-like of shape (n_samples): The confidence of this anomaly detector in each prediction in the given test time series.

predict_proba(X: ndarray) → ndarray

Predict anomaly probabilities.

Estimate the probability of a sample of X being anomalous, based on the anomaly scores obtained from decision_function by rescaling them to the range of [0, 1] via min-max scaling.

Parameters:

Xarray-like of shape (n_samples, n_attributes): Input time series.

Returns:

array-like of shape (n_samples): 1D array with the same length as X, with values in the interval [0, 1], in which a higher value implies that the instance is more likely to be anomalous.

Raises:

ValueError: If scores is not a valid array.
ValueError: If the prediction scores from ‘decision_function’ are constant, but not in the interval [0, 1], because these values can not unambiguously be transformed to an anomaly probability.

requires_fitting() → bool

Check whether this object requires fitting.

Check whether any of the attributes of this object ends with an underscore (‘_’), which indicates that the attribute is set when the object is fitted. Note that this method does not check whether the object is fitted, i.e., whether the attributes have been set.

Returns:

bool: True if and only if this object has attributes that end with ‘_’.

save(path: str | Path) → None

Save this detector.

Save detector to disk as a pickle file with extension .dtai. If the given path consists of multiple subdirectories, then the not existing subdirectories are created.

Parameters:

pathstr or Path: Location where to store the detector.