presets

Preset scenario builders.

pysatl_cpd.data.generator.presets.get_preset_scenario(name)

Retrieve a preset scenario specification by name.

Parameters:

name (str) – Name of the preset scenario to retrieve.

Returns:

The preset scenario specification.

Return type:

ScenarioSpec

Raises:

ValueError – If the preset name is unknown.

pysatl_cpd.data.generator.presets.build_preset_scenario(name, *, feature_names, series_length, num_segments)

Build a preset scenario specification with specified parameters.

Parameters:

• name (str) – Name of the preset scenario.

• feature_names (Sequence[str]) – Names of the features in the series.

• series_length (int) – Total length of each series.

• num_segments (int) – Number of segments in the series.

Returns:

The constructed scenario specification.

Return type:

ScenarioSpec

Raises:

ValueError – If series_length or num_segments is not positive, or feature_names is empty.

pysatl_cpd.data.generator.presets.preset_dataset(preset, *, n_series, seed, series_length=1200, n_features=None, num_segments=3)

Build a pandas-backed dataset from a preset scenario specification.

Parameters:

• preset (str) – Name of the preset scenario to use.

• n_series (int) – Number of series to generate.

• seed (int) – Random seed for reproducibility.

• series_length (int) – Length of each generated series.

• n_features (int | None) – Number of features. Defaults to 3 for 3d_mean_shifts and 2 for all other presets.

• num_segments (int) – Number of segments per series.

Returns:

Dataset containing the generated labeled series.

Return type:

Dataset