MODULE REFERENCE

adapters

anomx.data.base.adapters

Dataset adapters producing the canonical observation representation.

Adapters are the only place where storage formats (records, time-series batches, files, platform stores) are known. Everything downstream operates on ObservationSet.

classView source ↗

DatasetAdapter

DatasetAdapter

Bases: ABC

Read data from a concrete source and produce an observation set.

Members

loadmethod

# DatasetAdapter.load Source ↗

load() -> ObservationSet

Materialize the source into the canonical representation.

Returns ObservationSet

classView source ↗

RecordsAdapter

RecordsAdapter(records: list[dict[str, Any]] | pd.DataFrame, *, feature_columns: list[str] | None = None, entity_column: str | None = None, timestamp_column: str | None = None, metadata: dict[str, Any] | None = None)

Bases: DatasetAdapter

Adapt in-memory records (e.g. independent JSON samples) into observations.

Signature parameters
Parameter	Type	Default
`records`	`list[dict[str, Any]] \| pd.DataFrame`	`required`
`feature_columns`keyword only	`list[str] \| None`	`None`
`entity_column`keyword only	`str \| None`	`None`
`timestamp_column`keyword only	`str \| None`	`None`
`metadata`keyword only	`dict[str, Any] \| None`	`None`

Members

loadmethod

# RecordsAdapter.load Source ↗

load() -> ObservationSet

Returns ObservationSet

classView source ↗

TimeSeriesBatchAdapter

TimeSeriesBatchAdapter(batch: TimeSeriesBatch, *, metadata: dict[str, Any] | None = None)

Bases: DatasetAdapter

Adapt a TimeSeriesBatch into a temporal observation set.

Signature parameters
Parameter	Type	Default
`batch`	`TimeSeriesBatch`	`required`
`metadata`keyword only	`dict[str, Any] \| None`	`None`

Members

loadmethod

# TimeSeriesBatchAdapter.load Source ↗

load() -> ObservationSet

Returns ObservationSet