MixedLogit.make_model#

MixedLogit.make_model(X, F, y, observed=True)[source]#

Build mixed logit model with random coefficients.

Parameters:
Xnp.ndarray

Alternative-specific covariates, shape (n_obs, n_alts, n_covariates)

Fnp.ndarray or None

Fixed covariates, shape (n_obs, n_fixed_covariates)

ynp.ndarray

Observed choices, shape (n_obs,)

observed: bool

Whether to include observed data in the model

Returns:
modelpm.Model

PyMC model