pd_provider

Pandas-backed labeled data.

class pysatl_cpd.data.providers.labeled.implementations.pd_provider.PandasLabeledData(unlabeled, segment_info, annotation, *, segment_column='segment', segment_start_column='start', segment_end_column='end')[source]

Bases: LabeledData[NumericArray, AnnotationT], Generic

Pandas labeled data provider with configurable column names.

Parameters:

unlabeled (PandasDataProvider[UnlabeledTimeseriesAnnotation]) – Unlabeled data provider for the timeseries.
segment_info (Iterable[SegmentInfo]) – Iterable of segment information.
annotation (TypeVar(AnnotationT, bound= TimeseriesAnnotation)) – Annotation instance for labeling.
segment_column (str) – Name of the segment column.
segment_start_column (str) – Name of the segment start column.
segment_end_column (str) – Name of the segment end column.

__init__(unlabeled, segment_info, annotation, *, segment_column='segment', segment_start_column='start', segment_end_column='end')[source]

Parameters:

unlabeled (PandasDataProvider[UnlabeledTimeseriesAnnotation])
segment_info (Iterable[SegmentInfo])
annotation (AnnotationT)
segment_column (str)
segment_start_column (str)
segment_end_column (str)

Return type:

None

property unlabeled: PandasDataProvider[UnlabeledTimeseriesAnnotation]

Get the underlying unlabeled data provider.

Returns:: The underlying unlabeled data provider.
Return type:: unlabeled

classmethod from_unlabeled_data(unlabeled, segment_info, annotation)[source]

Create labeled data from unlabeled data provider.

Parameters:

unlabeled (DataProvider[ndarray[tuple[int, ...], dtype[double]], UnlabeledTimeseriesAnnotation]) – Unlabeled data provider for the timeseries.
segment_info (Iterable[SegmentInfo]) – Iterable of segment information.
annotation (TypeVar(A, bound= TimeseriesAnnotation)) – Annotation instance for labeling.

Returns:

New labeled data instance.

Return type:

PandasLabeledData[TypeVar(A, bound= TimeseriesAnnotation)]

Raises:

TypeError – If unlabeled is not a PandasDataProvider.

dataset(state_columns=None)[source]

Get the dataset with segment columns.

Parameters:: state_columns (dict[str, str] | None) – Optional mapping of column names to state keys.
Returns:: DataFrame with timeseries and segment columns.
Return type:: DataFrame

property feature_columns: Sequence[str]

Get the feature column names.

Returns:: Sequence of feature column names.
Return type:: feature_columns

select_columns(*, feature_columns, rename_provider=False)[source]

Select a subset of feature columns.

Parameters:

feature_columns (Sequence[str]) – Sequence of column names to select.
rename_provider (bool) – If True, updates the provider name to reflect the selected columns.

Returns:

New labeled data with selected columns.

Return type:

PandasLabeledData[TypeVar(AnnotationT, bound= TimeseriesAnnotation)]

create_feature_column(*, name, mapping, rename_provider=False)[source]

Append a derived feature column computed row-wise.

Parameters:

name (str) – Name of the new feature column.
mapping (Callable[[Series], object]) – Callable applied to each feature row.
rename_provider (bool) – If True, updates the provider name to reflect the new column.

Returns:

New labeled data with the appended feature column.

Return type:

PandasLabeledData[TypeVar(AnnotationT, bound= TimeseriesAnnotation)]