nochange

No-change policy for no-reset ARL evaluation.

class pysatl_cpd.benchmark.online.noreset.metrics.policy.nochange.NoChangePolicy(*, strict=True)[source]

Policy that marks only the first qualifying detection for ARL evaluation.

Parameters:: strict (bool) – Whether to use strict inequality when comparing detection function values against the threshold (default True).

__init__(*, strict=True)[source]

apply(run, threshold)[source]

Apply the policy and keep only the first qualifying detection.

Validates the run as a no-change run, finds all points where the detection function exceeds the threshold, and retains at most the first such point.

Parameters:

run (SingleRun[OnlineDetectionTrace[TypeVar(StateT, bound= OnlineAlgorithmState)], TypeVar(ProviderT, bound= LabeledData[Any, Any])]) – Input run with an OnlineDetectionTrace and labeled data.
threshold (float) – Threshold applied to the detection function values.

Returns:

Run wrapping a no-reset trace with at most one detection point.

Return type: