vinereg
An R package for D-vine copula based mean and quantile regression.
How to install
-
the stable release from CRAN:
install.packages("vinereg") -
the latest development version:
# install.packages("remotes") remotes::install_github("tnagler/vinereg", build_vignettes = TRUE)
Functionality
See the package website.
Example
set.seed(5)
library(vinereg)
data(mtcars)
# declare factors and discrete variables
for (var in c("cyl", "vs", "gear", "carb"))
mtcars[[var]] <- as.ordered(mtcars[[var]])
mtcars[["am"]] <- as.factor(mtcars[["am"]])
# fit model
(fit <- vinereg(mpg ~ ., family = "nonpar", data = mtcars))
#> D-vine regression model: mpg | disp, qsec, hp
#> nobs = 32, edf = 24.35, cll = -56.11, caic = 160.93, cbic = 196.63
summary(fit)
#> var edf cll caic cbic p_value
#> 1 mpg 0.000000 -100.189867 200.379733 200.3797334 NA
#> 2 disp 12.663078 29.654825 -33.983493 -15.4227648 5.284917e-08
#> 3 qsec 2.447922 4.326359 -3.756874 -0.1688666 2.101665e-02
#> 4 hp 9.242147 10.096325 -1.708356 11.8381898 1.898984e-02
# show marginal effects for all selected variables
plot_effects(fit)
#> `geom_smooth()` using method = 'loess' and formula 'y ~ x'
# predict mean and median
head(predict(fit, mtcars, alpha = c(NA, 0.5)), 4)
#> mean 0.5
#> 1 22.39380 22.29866
#> 2 22.23677 22.08015
#> 3 25.33841 24.92450
#> 4 20.33708 20.16153
Vignettes
For more examples, have a look at the vignettes with
vignette("abalone-example", package = "vinereg")
vignette("bike-rental", package = "vinereg")
References
Kraus and Czado (2017). D-vine copula based quantile regression. Computational Statistics & Data Analysis, 110, 1-18. link, preprint
Schallhorn, N., Kraus, D., Nagler, T., Czado, C. (2017). D-vine quantile regression with discrete variables. Working paper, preprint.
