Threshold Detector

Detector that detects anomaly based on user-given threshold. This detector compares time series values with user-given thresholds, and identifies time points as anomalous when values are beyond the thresholds.

class darts.ad.detectors.threshold_detector.ThresholdDetector(low_threshold=None, high_threshold=None)[source]

Bases: Detector

Flags values that are either below or above the low_threshold and high_threshold, respectively.

If a single value is provided for low_threshold or high_threshold, this same value will be used across all components of the series.

If sequences of values are given for the parameters low_threshold and/or high_threshold, they must be of the same length, matching the dimensionality of the series passed to detect(), or have a length of 1. In the latter case, this single value will be used across all components of the series.

If either low_threshold or high_threshold is None, the corresponding bound will not be used. However, at least one of the two must be set.

Parameters

  • low_threshold (Union[int, float, Sequence[float], None]) – (Sequence of) lower bounds. If a sequence, must match the dimensionality of the series this detector is applied to.

  • high_threshold (Union[int, float, Sequence[float], None]) – (Sequence of) upper bounds. If a sequence, must match the dimensionality of the series this detector is applied to.

Methods

detect(series)

Detect anomalies on given time series.

eval_accuracy(actual_anomalies, anomaly_score)

Score the results against true anomalies.
detect(series)

Detect anomalies on given time series.

Parameters

series (Union[TimeSeries, Sequence[TimeSeries]]) – series on which to detect anomalies.

Returns

binary prediciton (1 if considered as an anomaly, 0 if not)

Return type

Union[TimeSeries, Sequence[TimeSeries]]

eval_accuracy(actual_anomalies, anomaly_score, window=1, metric='recall')

Score the results against true anomalies.

Parameters

  • actual_anomalies (Union[TimeSeries, Sequence[TimeSeries]]) – The ground truth of the anomalies (1 if it is an anomaly and 0 if not).

  • anomaly_score (Union[TimeSeries, Sequence[TimeSeries]]) – Series indicating how anomoulous each window of size w is.

  • window (int) – Integer value indicating the number of past samples each point represents in the anomaly_score.

  • metric (str) – Metric function to use. Must be one of “recall”, “precision”, “f1”, and “accuracy”. Default: “recall”

Returns

Metric results for each anomaly score

Return type

Union[float, Sequence[float], Sequence[Sequence[float]]]

previous

Quantile Detector

next

Anomaly Scorers