AdstockTransformation#

class pymc_marketing.mmm.components.adstock.AdstockTransformation(l_max, normalize=True, mode=ConvMode.After, priors=None, prefix=None)[source]#

Subclass for all adstock functions.

In order to use a custom saturation function, inherit from this class and define:

function: a function that takes x to adstock x
default_priors: dictionary with priors for every parameter in function

Consider the predefined subclasses as examples.

Methods

`AdstockTransformation.__init__`(l_max[, ...])
`AdstockTransformation.apply`(x[, dims])	Called within a model context.
`AdstockTransformation.plot_curve`(curve[, ...])	Plot curve HDI and samples.
`AdstockTransformation.plot_curve_hdi`(curve)	Plot the HDI of the curve.
`AdstockTransformation.plot_curve_samples`(curve)	Plot samples from the curve.
`AdstockTransformation.sample_curve`(parameters)	Sample the adstock transformation given parameters.
`AdstockTransformation.sample_prior`([coords])	Sample the priors for the transformation.
`AdstockTransformation.update_priors`(priors)	Helper to update the priors for a function after initialization.

Attributes

`model_config`	Mapping from variable name to prior for the model.
`prefix`
`variable_mapping`	Mapping from parameter name to variable name in the model.
`lookup_name`
`default_priors`
`function`