TanhSaturationBaselined#

class pymc_marketing.mmm.components.saturation.TanhSaturationBaselined(priors=None, prefix=None)[source]#

Wrapper around tanh saturation function.

Multiplies pymc_marketing.mmm.transformers.tanh_saturation_baselined() by an extra scaling parameter beta so the response can reach an asymptote other than the gain-implied one.

Parameters:

x0tensor

Reference point on the input scale, as in tanh_saturation_baselined(). Default prior: Prior("HalfNormal", sigma=1).

gaintensor

Value of the curve at x0 divided by x0 (the ROAS at the baseline). Default prior: Prior("HalfNormal", sigma=1).

rtensor

Overspend fraction, the ratio of the response at x0 to the saturation level. Default prior: Prior("HalfNormal", sigma=1).

betatensor

Scaling factor applied to the baselined-tanh response (multiplies the gain-implied asymptote gain * x0 / r). Default prior: Prior("HalfNormal", sigma=1).

.. plot::

context:: close-figs

import matplotlib.pyplot as plt import numpy as np from pymc_marketing.mmm import TanhSaturationBaselined

rng = np.random.default_rng(0)

adstock = TanhSaturationBaselined() prior = adstock.sample_prior(random_seed=rng) curve = adstock.sample_curve(prior) adstock.plot_curve(curve, random_seed=rng) plt.show()

Methods

`TanhSaturationBaselined.__init__`([priors, ...])
`TanhSaturationBaselined.apply`(x, *[, ...])	Call within a model context.
`TanhSaturationBaselined.from_dict`(data)	Reconstruct a transformation from a dict.
`TanhSaturationBaselined.function`(x, x0, ...)	Tanh saturation function.
`TanhSaturationBaselined.plot_curve`(curve[, ...])	Plot curve HDI and samples.
`TanhSaturationBaselined.plot_curve_hdi`(curve)	Plot the HDI of the curve.
`TanhSaturationBaselined.plot_curve_samples`(curve)	Plot samples from the curve.
`TanhSaturationBaselined.sample_curve`([...])	Sample the curve of the saturation transformation given parameters.
`TanhSaturationBaselined.sample_prior`([coords])	Sample the priors for the transformation.
`TanhSaturationBaselined.set_dims_for_all_priors`(dims)	Set the dims for all priors.
`TanhSaturationBaselined.to_dict`([_orig])	Convert the transformation to a dictionary.
`TanhSaturationBaselined.update_priors`(priors)	Update the priors for a function after initialization.
`TanhSaturationBaselined.with_default_prior_dims`(dims)	Return a copy with default prior dims (dims=None) set to `dims` instead.
`TanhSaturationBaselined.with_updated_priors`(priors)	Return a copy with updated priors.

Attributes

`TanhSaturationBaselined.combined_dims`	Get the combined dims for all the parameters.
`TanhSaturationBaselined.default_priors`
`TanhSaturationBaselined.function_priors`	Get the priors for the function.
`TanhSaturationBaselined.model_config`	Mapping from variable name to prior for the model.
`TanhSaturationBaselined.prefix`
`TanhSaturationBaselined.priors`	Get the priors for the function.
`TanhSaturationBaselined.requires_unscaled_input`	Whether the saturation must receive raw (unscaled) channel inputs.
`TanhSaturationBaselined.variable_mapping`	Mapping from parameter name to variable name in the model.