inst/examples/assisted/plot_b01model.R

In rbiogeme: Interface to 'Biogeme'

#!/usr/bin/env Rscript

# b01model. Estimate three alternative Swissmetro choice-model structures.
#
# The catalog contains one logit and two nested-logit likelihood expressions.
# R defines those symbolic expressions, while native Python Biogeme enumerates
# the catalog, estimates every configuration, and compiles the result tables.

library(rbiogeme)

# prepare_swissmetro_example() is defined in ../swissmetro/example_utils.R.
# It handles data/Python paths and output location; the model is specified in
# this script so it remains readable and self-contained.
script_path <- commandArgs(trailingOnly = FALSE)
script_path <- sub("^--file=", "", script_path[startsWith(script_path, "--file=")][[1L]])
source(file.path(dirname(normalizePath(script_path)), "..", "swissmetro", "example_utils.R"))

build_b01model_model <- function(database) {
  # Parameter names, starts, bounds, and utility coding match native b01model.
  asc_car <- biogeme_beta("asc_car", start = 0)
  asc_train <- biogeme_beta("asc_train", start = 0)
  b_time <- biogeme_beta("b_time", start = 0)
  b_cost <- biogeme_beta("b_cost", start = 0)

  utilities <- list(
    `1` = asc_train + b_time * variable("TRAIN_TT_SCALED") +
      b_cost * variable("TRAIN_COST_SCALED"),
    `2` = b_time * variable("SM_TT_SCALED") +
      b_cost * variable("SM_COST_SCALED"),
    `3` = asc_car + b_time * variable("CAR_TT_SCALED") +
      b_cost * variable("CAR_CO_SCALED")
  )
  availability <- list(
    `1` = variable("TRAIN_AV_SP"),
    `2` = variable("SM_AV"),
    `3` = variable("CAR_AV_SP")
  )
  choice <- variable("CHOICE")
  logit <- logit_log_probability(utilities, availability, choice)

  # A native OneNestForNestedLogit is represented by nested_nest().
  mu_existing <- biogeme_beta("mu_existing", start = 1, lower = 1, upper = 10)
  existing <- nested_nest(mu_existing, c(1, 3), name = "Existing")
  existing_nests <- nested_nests(c(1, 2, 3), list(existing))
  nested_existing <- nested_log_probability(
    utilities, availability, existing_nests, choice
  )

  mu_public <- biogeme_beta("mu_public", start = 1, lower = 1, upper = 10)
  public <- nested_nest(mu_public, c(1, 2), name = "Public")
  public_nests <- nested_nests(c(1, 2, 3), list(public))
  nested_public <- nested_log_probability(
    utilities, availability, public_nests, choice
  )

  # The catalog controller is created and resolved by native Biogeme. The
  # branch names are preserved exactly for reports and configuration IDs.
  model_catalog <- catalog(
    "model_catalog",
    list(
      logit = logit,
      `nested existing` = nested_existing,
      `nested public` = nested_public
    )
  )
  biogeme_model(
    database = database,
    formula = model_catalog,
    control = biogeme_control(
    output_directory = prepared$output,
      model_name = "b01model",
      generate_html = FALSE,
      generate_yaml = FALSE,
      save_iterations = FALSE
    )
  )
}

prepared <- prepare_swissmetro_example(
  commandArgs(trailingOnly = TRUE),
  default_model = "b01model"
)

# Native biogeme.data.swissmetro.read_data() removes only CHOICE == 0.
database <- swissmetro_data(prepared$data, filter_purpose = FALSE)
model <- build_b01model_model(database)

# force = TRUE makes native Biogeme estimate every configuration afresh.
fit <- estimate_catalog(
  model,
  model_name = "b01model",
  control = model$control,
  force = TRUE
)

cat("A total of ", length(fit$results), " models have been estimated.\n", sep = "")
for (configuration in names(fit$results)) {
  result <- fit$results[[configuration]]
  cat(
    configuration,
    ": LL=",
    formatC(result$final_log_likelihood, digits = 2, format = "f"),
    " K=",
    length(result$beta_names),
    "\n",
    sep = ""
  )
}
print(fit$summary)
for (name in names(fit$description)) {
  if (!identical(name, unname(fit$description[[name]]))) {
    cat(name, "\t", fit$description[[name]], "\n", sep = "")
  }
}
cat("Non dominated models:\n")
for (configuration in fit$non_dominated) cat(configuration, "\n", sep = "")
print(fit$non_dominated_summary)
cat(fit$latex, "\n")

invisible(fit)