choicer - Discrete Choice Models for Economic Applications
Fast estimation of discrete-choice models for applied
economics. Frequentist likelihoods, analytical gradients, and
Hessians are implemented in C++ with 'OpenMP' parallelism,
scaling efficiently to specifications with many
alternative-specific constants. Compiled Gibbs samplers provide
Bayesian multinomial probit and hierarchical models.
Post-estimation routines cover predicted shares, own- and
cross-price elasticities, diversion ratios, willingness to pay,
and welfare counterfactuals. Supports multinomial logit
('MNL'), mixed logit ('MXL'), nested logit ('NL'), Bayesian
multinomial probit ('MNP'), and hierarchical Bayesian
multinomial logit and probit ('HMNL', 'HMNP').
