create_distribution_from_config

pymc_marketing.model_config.create_distribution_from_config(name, config)

Wrapper around create_distribution that uses a configuration dictionary.

Parameters:

namestr

Name of the variable.

configDict

A configuration with the name mapping to parameter configuration.

Returns:

TensorVariable

A PyMC random variable.

Examples

import pymc as pm

from pymc_marketing.model_config import create_distribution_from_config

distribution = {
    "dist": "Normal",
    "kwargs": {"mu": 0, "sigma": 1},
}
config = {
    "alpha": distribution,
    "beta": distribution,
}

with pm.Model():
    alpha = create_distribution_from_config("alpha", config)