UPG (version 0.2.2)

UPG: Gibbs Sampling for Bayesian discrete choice models

Description

UPG estimates Bayesian discrete choice models and returns the full posterior distribution for all parameters that can be used for further analysis and inference.

Usage

UPG(y,
    X,
    type,
    Ni         = NULL,
    baseline   = NULL,
    draws      = 1000,
    burnin     = 1000,
    A0         = 1,
    d0         = 0.5,
    D0         = 0.5,
    G0         = 99,
    verbose    = TRUE,
    BOOST      = TRUE,
    beta.start = NULL)

Arguments

y

a binary vector for probit and logit models. A character, factor or numeric vector for multinomial logit models. A vector of the number of successes for the binomial model.

X

a matrix of explanatory variables including an intercept in the first column. Rows are individuals, columns are variables.

type

indicates the model to be estimated. 'probit' for the probit model, 'logit' for the logit model, 'mnl' for the multinomial logit model or 'binomial' for the binomial logit model.

Ni

a vector containing the number of trials when estimating a binomial logit model.

baseline

a string that can be used to change the baseline category in MNL models. Default baseline is the most common category.

draws

number of saved Gibbs sampler iterations. Default is 1000 for illustration purposes, you should use more when estimating a model (e.g. 10,000)!

burnin

number of burned Gibbs sampler iterations. Default is 1000 for illustration purposes, you should probably use more when estimating a model (e.g. 2,000)!

A0

prior variance for coefficients, 1 is the default.

d0

prior shape for working parameter delta, 0.5 is the default.

D0

prior rate for working parameter delta, 0.5 is the default.

G0

prior variance for the intercept, 99 is the default.

verbose

logical variable indicating whether model output and progress should be printed during estimation.

BOOST

logical variable indicating whether MCMC boosting should be used.

beta.start

provide starting values for beta (e.g. for use within Gibbs sampler)

Value

Depending on the type of the model, UPG() returns an UPG.Probit, UPG.Logit, UPG.MNL or UPG.Binomial object.

See Also

summary.UPG.Probit to summarize the estimates of a discrete choice model from an UPG.Probit object and to create tables. predict.UPG.Logit to predict probabilities from a discrete choice model from an UPG.Logit object. plot.UPG.MNL to plot the results of a discrete choice model from an UPG.MNL object.

Examples

# NOT RUN {
# load package
library(UPG)

# estimate a probit model using example data
# warning: use more burn-ins, burnin = 100 is just for demonstration purposes
data(lfp)
y = lfp[,1]
X = lfp[,-1]
results.probit = UPG(y = y, X = X, type = "probit", verbose=TRUE, burnin = 100)

# estimate a logit model using example data
# warning: use more burn-ins, burnin = 100 is just for demonstration purposes
data(lfp)
y = lfp[,1]
X = lfp[,-1]
results.logit = UPG(y = y, X = X, type = "logit", verbose=TRUE, burnin = 100)

# estimate a MNL model using example data
# warning: use more burn-ins, burnin = 100 is just for demonstration purposes
data(program)
y = program[,1]
X = program[,-1]
results.mnl = UPG(y = y, X = X, type = "mnl", burnin = 100)

# estimate a binomial logit model using example data
# warning: use more burn-ins, burnin = 100 is just for demonstration purposes
data(titanic)
y  = titanic[,1]
Ni = titanic[,2]
X  = titanic[,-c(1,2)]
results.binomial = UPG(y = y, X = X, Ni = Ni, type = "binomial", burnin = 100)

# }