Anomaly aggregators base classes

class darts.ad.aggregators.aggregators.Aggregator(*args, **kwargs)[source]

Bases: abc.ABC

Methods

eval_accuracy(actual_anomalies, series[, ...])

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

predict(series)

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.
eval_accuracy(actual_anomalies, series, window=1, metric='recall')[source]

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

Parameters

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

  • series (Sequence[TimeSeries]) – The (sequence of) multivariate binary series to aggregate

  • window (int) – (Sequence of) integer value indicating the number of past samples each point represents in the (sequence of) series. The parameter will be used by the function _window_adjustment_anomalies() in darts.ad.utils to transform actual_anomalies.

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

Returns

(Sequence of) score for the (sequence of) series

Return type

Union[float, Sequence[float]]

abstract predict(series)[source]

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.

Parameters

series (Union[TimeSeries, Sequence[TimeSeries]]) – The (sequence of) multivariate binary series to aggregate

Returns

(Sequence of) aggregated results

Return type

TimeSeries

class darts.ad.aggregators.aggregators.FittableAggregator[source]

Bases: darts.ad.aggregators.aggregators.Aggregator

Base class of Aggregators that do need training.

Methods

eval_accuracy(actual_anomalies, series[, ...])

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

fit(actual_anomalies, series)

Fit the aggregators on the (sequence of) multivariate binary series.

predict(series)

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.
eval_accuracy(actual_anomalies, series, window=1, metric='recall')

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

Parameters

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

  • series (Sequence[TimeSeries]) – The (sequence of) multivariate binary series to aggregate

  • window (int) – (Sequence of) integer value indicating the number of past samples each point represents in the (sequence of) series. The parameter will be used by the function _window_adjustment_anomalies() in darts.ad.utils to transform actual_anomalies.

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

Returns

(Sequence of) score for the (sequence of) series

Return type

Union[float, Sequence[float]]

fit(actual_anomalies, series)[source]

Fit the aggregators on the (sequence of) multivariate binary series.

If a list of series is given, they must have the same number of components.

Parameters

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

  • series (Union[TimeSeries, Sequence[TimeSeries]]) – The (sequence of) multivariate binary series

predict(series)[source]

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.

Parameters

series (Union[TimeSeries, Sequence[TimeSeries]]) – The (sequence of) multivariate binary series to aggregate

Returns

(Sequence of) aggregated results

Return type

TimeSeries

class darts.ad.aggregators.aggregators.NonFittableAggregator[source]

Bases: darts.ad.aggregators.aggregators.Aggregator

Base class of Aggregators that do not need training.

Methods

eval_accuracy(actual_anomalies, series[, ...])

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

predict(series)

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.
eval_accuracy(actual_anomalies, series, window=1, metric='recall')

Aggregates the (sequence of) multivariate series given as input into one (sequence of) series and evaluates the results against true anomalies.

Parameters

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

  • series (Sequence[TimeSeries]) – The (sequence of) multivariate binary series to aggregate

  • window (int) – (Sequence of) integer value indicating the number of past samples each point represents in the (sequence of) series. The parameter will be used by the function _window_adjustment_anomalies() in darts.ad.utils to transform actual_anomalies.

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

Returns

(Sequence of) score for the (sequence of) series

Return type

Union[float, Sequence[float]]

predict(series)[source]

Aggregates the (sequence of) multivariate binary series given as input into a (sequence of) univariate binary series.

Parameters

series (Union[TimeSeries, Sequence[TimeSeries]]) – The (sequence of) multivariate binary series to aggregate

Returns

(Sequence of) aggregated results

Return type

TimeSeries

previous

Anomaly Aggregators

next

AND Aggregator