R package lcra
A user-friendly interface for doing joint Bayesian latent class and regression analysis
Overview
This R package provides a user-friendly interface for fitting Bayesian
joint lantend class and regression models. Unsing the standard R syntax,
the user can specify the form of the regression model and the desired
number of latent classes.
The technical details of the model implemented here are described in Elliott, Michael R., Zhao, Zhangchen, Mukherjee, Bhramar, Kanaya, Alka, Needham, Belinda L., “Methods to account for uncertainty in latent class assignments when using latent classes as predictors in regression models, with application to acculturation strategy measures” (2020) In press at Epidemiology. doi:10.1097/EDE.0000000000001139
Installation
This package uses the R interface for WinBUGS called R2WinBUGS. The
package R2WinBUGS in turn depends on the standalone Windows program
WinBUGS. Follow this link: University of Cambridge MRC Biostatistics
Unit
for WinBUGS download and installation instructions. WinBUGS can also
be used on Mac using the additional software
Wine. Plenty of tutorials that demonstrate
install of WinBUGS on non-Windows machines are available on Google. With
a bit of work, unix users can run this software, too.
Once the standalone WinBUGS program is installed, open R and run:
If the devtools package is not yet installed, install it first:
install.packages('devtools')
library(devtools)# install lcra from Github:
install_github('umich-biostatistics/lcra')
library(lcra)Example usage
The main function for the joint model is lcra(). Use ?lcra for the R
help file.
Here is an example Bayesian latent class analysis on simulated data with continuous and discrete outcomed:
# Data sets 1 and 2
data('paper_sim')
data('paper_sim_binary')
# Set initial values
inits =
list(
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2),
tau = 0.5
),
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2),
tau = 0.5
),
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2),
tau = 0.5
)
)
inits_binary =
list(
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2)
),
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2)
),
list(
theta = c(0.33, 0.33, 0.34),
beta = rep(0, length = 3),
alpha = rep(0, length = 2)
)
)
# Fit model 1
fit.gaus_paper =
lcra(
formula = Y ~ X1 + X2,
family = "gaussian",
data = paper_sim,
nclasses = 3,
manifest = paste0("Z", 1:10),
inits = inits,
dir = tempdir(),
n.chains = 3,
n.iter = 5000,
parameters.to.save = c("theta", "beta", "true", "alpha")
)
# Model 1 results
print(fit.gaus_paper, digits = 3)
plot(fit.gaus_paper)
# Extract results
fit.gaus_paper$median$true
fit.gaus_paper$mean$beta
fit.gaus_paper$mean$alpha
# Fit model 2
fit.binom_paper =
lcra(
formula = Y ~ X1 + X2,
family = "binomial",
data = paper_sim_binary,
nclasses = 3,
manifest = paste0("Z", 1:10),
inits = inits_binary,
dir = tempdir(),
n.chains = 3,
n.iter = 5000,
parameters.to.save = c("theta", "beta", "true", "alpha")
)
# Model 2 results
print(fit.binom_paper, digits = 3)
plot(fit.binom_paper)
# Extract results
fit.binom_paper$median$true
fit.binom_paper$mean$beta
fit.binom_paper$mean$alphaVignette
Once the package is installed, run the following to launch a vignette with detailed examples of the package functions.
vignette('learn-lcra')Current Suggested Citation
“Methods to account for uncertainty in latent class assignments when using latent classes as predictors in regression models, with application to acculturation strategy measures” (2020) In press at Epidemiology. doi:10.1097/EDE.0000000000001139
