prepare_sensitivity_data

pymc_marketing.mmm.summary.helpers.prepare_sensitivity_data(sa_da, data, dims=None, aggregation=None, x_sweep_axis='relative', apply_cost_per_unit=True)

Prepare sensitivity data for summary tables (steps 1-5 of sensitivity plots).

Parameters:

sa_daxr.DataArray

Raw sensitivity analysis values with a sweep dimension.

dataMMMIDataWrapper

Model data wrapper for spend/data scaling.

dimsdict, optional

Dimension filters.

aggregationdict, optional

Aggregation spec, e.g. {"sum": "channel"}.

x_sweep_axis{“relative”, “absolute”}, default “relative”

Whether x-values are sweep multipliers or absolute spend/data.

apply_cost_per_unitbool, default True

When x_sweep_axis="absolute", use spend (True) or channel data (False).

Returns:

tuple[xr.DataArray, xr.DataArray]

(sweep_x, sa_da_processed) ready for HDI summarization.