gmnl package in R
Authors: Mauricio Sarrias [aut, cre], Ricardo Daziano [aut]
gmnl is a package in R for estimating logit-type models with unobserved preference heterogeneity for cross-sectional and panel data. This package expands on mlogit and implements the maximum (simulated) likelihood estimator with analytical expressions of the gradient for the following models: multinomial or conditional logit (MNL), mixed multinomial logit (MIXL), scale heterogeneity multinomial logit (S-MNL), generalized multinomial logit (G-MNL), latent class logit (LC), and mixed-mixed multinomial logit (MM-MNL).
gmnl provides the ability of constructing preference and willingness-to-pay conditional estimates at the individual level.
Please cite as: Sarrias, MA and RA Daziano. 2017. Multinomial Logit Models with Continuous and Discrete Individual Heterogeneity in R: The gmnl Package. Journal of Statistical Software 79(2), 1-46.
FAQ
- How to constraint parameters using gmnl?
- Using gmnl to estimate Latent Class Multinomial Logit Models. The helper function can be downloaded here.
- My model is not converging? Why am I getting NAs?
Replication code
Paper: Daziano, RA, MA Sarrias and B Leard. 2017. Are consumers willing to pay to let cars drive for them? Analyzing response to autonomous vehicles. Transportation Research Part C: Emerging Technologies 78, 150-164.
Code: R code can be found here.
MATLAB code for the logit-mixed logit model in preference space
Code: posted on Mendeley
Papers:
Bansal, P, RA Daziano, M Achtnicht. 2018. Comparison of Parametric and Semiparametric Representations of Unobserved Taste Heterogeneity in Choice Modeling. Journal of Choice Modelling 27, 97-113.
Bansal, P, RA Daziano, M Achtnicht. 2018. Extending the logit-mixed logit model for a combination of random and fixed parameters. Journal of Choice Modelling 27, 88-96.
MATLAB code for parallel implementation of the kernel MNL model
Code: posted on Mendeley
Paper: Bansal, P, RA Daziano, and N Sunder. 2019. Arriving at a decision: a semi-parametric approach to institutional birth choice in India. Journal of Choice Modelling 31, 86-103.
MATLAB code for the mixture-of-normals logit
Coming soon.
