filter_idata_by_dims#

pymc_marketing.data.idata.utils.filter_idata_by_dims(idata, **dim_filters)[fuente]#

Filter InferenceData by dimension values.

Parameters:

idataaz.InferenceData: InferenceData object to filter
**dim_filters: Dimension filters, e.g., country=»US», channel=[«TV», «Radio»] Note: When filtering to a single value, the dimension is dropped (xarray’s default behavior). When filtering to multiple values, the dimension is preserved.

Returns:

az.InferenceData: New InferenceData with filtered groups

Raises:

ValueError: If a dimension in dim_filters doesn’t exist in any group

Examples

>>> filtered = filter_idata_by_dims(idata, country="US")
>>> # Dimension "country" is dropped (single value)
>>> filtered = filter_idata_by_dims(idata, channel=["TV", "Radio"])
>>> # Dimension "channel" is preserved (multiple values)