MixedLogit.make_utility

MixedLogit.make_utility(X_data, B_full, alphas, W_contrib, price_error_contrib)

Compute total systematic utility for each alternative.

Combines contributions from: - Alternative-specific covariates with individual-specific coefficients - Fixed covariates with alternative-specific effects - Alternative-specific constants - Price endogeneity correction (if applicable)

Parameters:

X_dataTensorVariable

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

B_fullTensorVariable

Individual coefficient matrix, shape (n_obs, n_covariates)

alphasTensorVariable

Alternative-specific constants

W_contribTensorVariable

Fixed covariate contribution, shape (n_obs, n_alts)

price_error_contribTensorVariable

Control function contribution, shape (n_obs, n_alts)

Returns:

UTensorVariable

Systematic utility, shape (n_obs, n_alts)