plot.TSVC: Plotting of Varying Coefficient Trees

Description

Visualization of trees of effects of covariates that vary with the values of one or several effect modifiers.

Usage

# S3 method for TSVC
plot(
  x,
  variable,
  ellipse_a = 0.8,
  ellipse_b = 0.2,
  ellipse_x = 0,
  ellipse_y = 0,
  branch_adj = 0,
  cex.lines = 2,
  cex.branches = 1,
  cex.coefs = 1,
  cex.main = 1,
  cex.numbers = 1,
  draw_numbers = TRUE,
  title = NULL,
  decimals = 3,
  confint = NULL,
  ...
)

Arguments

x: a fitted object of class TSVC.
variable: name of the variable, for which the tree shall be plotted.
ellipse_a: controls width of ellipses containing coefficient estimates.
ellipse_b: controls height of ellipses containing coefficient estimates.
ellipse_x: controls location on x-axis of ellipses containing coefficient estimates.
ellipse_y: controls location on y-axis of ellipses containing coefficient estimates.
branch_adj: vertical adjustment of branch labels.
cex.lines: width of branches of the tree.
cex.branches: size of the labels of the tree.
cex.coefs: size of the coefficients in the terminal nodes of the tree.
cex.main: size of the title of the tree.
cex.numbers: size of the internally used node number.
draw_numbers: if true, internally used node numbers are displayed.
title: optional title, which is addded to the tree; if title=NULL the title is the name of the variable in the data.
decimals: number of decimals of coefficient estimates. Per default the coefficient estimates are displayed with three decimals.
confint: optional fitted object of class confint.TSVC with confidence intervals to be plotted in the terminal nodes of the tree; if confint=NULL (default) only the coefficient estimates will be plotted.
...: further arguments passed to or from other methods.

Author

Moritz Berger <Moritz.Berger@imbie.uni-bonn.de>
https://www.imbie.uni-bonn.de/people/dr-moritz-berger/

References

Berger, M., G. Tutz and M. Schmid (2019). Tree-Structured Modelling of Varying Coefficients. Statistics and Computing 29, 217-229, https://doi.org/10.1007/s11222-018-9804-8.

Examples

Run this code

# Swiss Labour Market 
library(AER)
data("SwissLabor")

# recode factors 
sl <- SwissLabor
sl$participation <- as.numeric(sl$participation)-1
sl$foreign       <- as.numeric(sl$foreign)-1

if (FALSE) {
fit1 <- TSVC(participation~income+age, data=sl, family=binomial(link="logit"), 
             nperm=1000, trace=TRUE)
plot(fit1, "income")

fit2 <- TSVC(participation~income+age, data=sl, family=binomial(link="logit"), 
perm_test=FALSE, test_linear=FALSE, splits_max=3)
set.seed(20012025)
ci2  <- confint(fit2, bootstrap_n=500, alpha=0.05, trace=TRUE)
plot(fit2, variable="income", confint=ci2, ellipse_y=0, draw_numbers=FALSE)
}