model_builder

Base classes containing primitives and high-level API for model building, fitting, saving, and loading.

Module Attributes

SamplingMethod

Sampling methods supported by ModelFitter.fit().

Functions

create_idata_accessor(value, message)	Create a property accessor for a group of the model's DataTree.
create_sample_kwargs(sampler_config, ...)	Create the dictionary of keyword arguments for pm.sample.
requires_model(func)	Ensure that the model is built before calling a method.

Classes

ModelBuilder([model_config, sampler_config])	Base class for building PyMC-Marketing models.
ModelFitter()	Mixin providing a unified fitting API for all PyMC-Marketing models.
ModelIO()	Mixin to handle saving and loading of models.
RegressionModelBuilder([model_config, ...])	ModelBuilder class providing an easy-to-use API similar to scikit-learn for regression models.

Exceptions

DifferentModelError

Error raised when a model loaded is different than one saved.