plotBart is a diagnostic and plotting package for bartCause and thinkCausal.
library(plotBart)
data(lalonde)
confounders <- c('age', 'educ', 'black', 'hisp', 'married', 'nodegr')
# fit BART model
model_results <- bartCause::bartc(
response = lalonde[['re78']],
treatment = lalonde[['treat']],
confounders = as.matrix(lalonde[, confounders]),
estimand = 'ate',
commonSup.rule = 'none',
verbose = FALSE,
keepTrees = TRUE
)
# plot common support
plot_common_support(.model = model_results)
# plot CATE and manipulate ggplot object
plot_CATE(
.model = model_results,
type = 'density',
ci_80 = TRUE,
ci_95 = TRUE,
.mean = TRUE
) +
labs(subtitle = 'My comments on the results') +
theme_classic()
plotBart is currently in development and is available to test by installing via:
# latest release on CRAN
install.packages('plotBart')
# latest development version
# install.packages("remotes")
remotes::install_github('priism-center/plotBart')