online

No-reset online metrics (ARL and delay metrics).

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

No-reset ARL metric using the no-change policy.

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

__init__(*, strict=True)[source]

class pysatl_cpd.benchmark.online.noreset.metrics.online.NoResetMeanDelayMetric(*, max_delay, strict=True)[source]

No-reset mean delay metric using mixed policy semantics.

Parameters:

max_delay (int) – Maximum allowed delay (in steps) for a true positive.
strict (bool) – Whether to use strict inequality when comparing detection function values against the threshold (default True).

__init__(*, max_delay, strict=True)[source]

Parameters:

Return type:

None

class pysatl_cpd.benchmark.online.noreset.metrics.online.NoResetMedianDelayMetric(*, max_delay, strict=True)[source]

No-reset median delay metric using mixed policy semantics.

Parameters:

max_delay (int) – Maximum allowed delay (in steps) for a true positive.
strict (bool) – Whether to use strict inequality when comparing detection function values against the threshold (default True).

__init__(*, max_delay, strict=True)[source]

Parameters:

Return type:

None