Function reference • RprobitB

Data management

Use these functions for the choice data preparation or simulation.

check_form(): Check the model formula

overview_effects(): Effect overview

create_lagged_cov(): Create lagged choice covariates

as_cov_names(): Relabel the alternative specific covariates to the required format

missing_data(): Handle missing choice data

prepare_data(): Prepare empirical choice data for estimation

RprobitB_parameter(): Create object of class RprobitB_parameter

simulate_choices(): Simulate choice data

train_test(): Split choice data set in two parts

plot(<RprobitB_data>): Plot method for RprobitB_data

Model fitting

Use these function for fitting a probit model to choice data.

check_prior(): Check prior parameters

mcmc(): Probit model fitting via Markov chain Monte Carlo simulation

nested_model(): Estimating a nested model

transform(<RprobitB_fit>): Transform an RprobitB_fit object

Model evaluation

Use these functions for model evaluation.

coef(<RprobitB_fit>): Linear coefficients

cov_mix(): Estimated covariance matrix of the mixing distribution

point_estimates(): Compute point estimates based on Gibbs samples of an RprobitB_fit object

choice_probabilities(): Return choice probabilities of an RprobitB_fit.

preference_classification(): Classify deciders based on their preferences

get_cov(): Get covariates of choice situation

predict(<RprobitB_fit>): Choice prediction

plot(<RprobitB_fit>): Plot method for RprobitB_fit

Model selection

Use these functions for model selection.

model_selection(): Compare fitted models

AIC(): Akaike's Information Criterion

BIC(): Bayesian Information Criterion

logLik(): Log-likelihood value

nobs(): Number of observations

npar(): Number of model parameters

mml(): Marginal model likelihood

compute_p_si(): Compute probability for each observed choice at posterior samples

pred_acc(): Prediction accuracy

Datasets

The following datasets are included in the package.

choice_berserk: Choice of berserking

choice_chess_opening: Choice of a chess opening

Fitted model

The package contains the following fitted models.

model_elec: Mixed probit model for multivariate choice between electricity suppliers

model_train: Probit model for binary choice between Train trip alternatives

model_train_sparse: Probit model for binary choice between Train trip alternatives with the price as the only explanatory variable

Distributions

Implementation of distributions and number generating functions.

dmvnorm(): Density of multivariate normal distribution

rdirichlet(): Draw from Dirichlet distribution

rmvnorm(): Draw from multivariate normal distribution

rtnorm(): Draw from truncated normal

rwishart(): Draw from Wishart distribution

Posterior samplers

These functions draw from the conditional posterior distributions of the model parameters.

update_Omega(): Update class covariances

update_Sigma(): Update error term covariance matrix of multiple linear regression

update_U(): Update latent utility vector

update_b(): Update class means

update_m(): Update class sizes

update_reg(): Update coefficient vector of multiple linear regression

update_s(): Update class weight vector

update_z(): Update class allocation vector

Utilities

These functions may be useful beyond {RprobitB}.

R_hat(): Compute Gelman-Rubin statistic

delta(): Matrix difference operator

euc_dist(): Euclidean distance

is_covariance_matrix(): Check covariance matrix properties

pprint(): Print abbreviated matrices and vectors