mixl-package: Estimate mixed multinomial logit models

In mixl: Simulated Maximum Likelihood Estimation of Mixed Logit Models for Large Datasets

Description

Estimate mixed multinomial logit models using (simulated) maximum likelihood estimation. The package supports standard mnl, mixed-logit and hybrid choice. Using compilation to C++, model estimation is significantly faster than in native R code.

Details

This section should provide a more detailed overview of how to use the package, including the most important functions.

Author(s)

Joe Molloy <[e-mail]>.

References

This optional section can contain literature or other references for background information.

See Also

Optional links to other man pages

Examples

  
  
  data("Train", package="mlogit")
  head(Train, 3)
  Train$ID <- Train$id
  Train$CHOICE <- as.numeric(Train$choice)

  mnl_test <- "
    ASC_B_RND 	= @ASC_B 	+ draw_2 * @SIGMA_B;

    U_A =             @B_price * $price_A / 1000 + @B_time * $time_A / 60 + @B_change * $change_A; 
    U_B = ASC_B_RND + @B_price * $price_B / 1000 + @B_timeB * $time_B / 60;
  "
  
  model_spec <- mixl::specify_model(mnl_test, Train)
  
  #only take starting values that are needed
  est <- stats::setNames(c(0,0,0,0,0,0), c("B_price", "B_time", "B_timeB", 
  "B_change", "ASC_B", "SIGMA_B"))
  
  availabilities <- mixl::generate_default_availabilities(Train, model_spec$num_utility_functions)
  
  model <- mixl::estimate(model_spec, est, Train, availabilities = availabilities, nDraws = 20)  
  
  summary(model)
  

mixl documentation built on Dec. 11, 2021, 9:07 a.m.