GaussianBasis.update_priors#

GaussianBasis.update_priors(priors)#

Update the priors for a function after initialization.

Uses {prefix}_{parameter_name} as the key for the priors instead of the parameter name in order to be used in the larger MMM.

Parameters:

priorsdict[str, Prior]: Dictionary with the new priors for the parameters of the function.

Examples

Update the priors for a transformation after initialization.

from pymc_marketing.mmm.components.base import Transformation
from pymc_marketing.prior import Prior

class MyTransformation(Transformation):
    lookup_name: str = "my_transformation"
    prefix: str = "transformation"
    function = lambda x, lam: x * lam
    default_priors = {"lam": Prior("Gamma", alpha=3, beta=1)}

transformation = MyTransformation()
transformation.update_priors(
    {"transformation_lam": Prior("HalfNormal", sigma=1)},
)