plot.logicDT: Plot a logic decision tree

Description

This function plots a logicDT model on the active graphics device.

Usage

fancy.plot(x, cdot = FALSE, ...)
# S3 method for logicDT
plot(
  x,
  fancy = TRUE,
  x_scaler = 0.5,
  margin_scaler = 0.2,
  cex = 1,
  cdot = FALSE,
  ...
)

Value

No return value, called for side effects

Arguments

x: An object of the class logicDT
cdot: Should a centered dot be used instead of a logical and for depicting interactions?
...: Arguments passed to fancy plotting function
fancy: Should the fancy mode be used for plotting? Default is TRUE.
x_scaler: Scaling factor on the horizontal axis for deeper trees, i.e., x_scaler = 0.5 means that the horizontal distance between two adjacent nodes is halved for every vertical level.
margin_scaler: Margin factor. Smaller values lead to smaller margins.
cex: Scaling factor for the plotted text elements.

Details

There are two plotting modes:

fancy = FALSE which draws a tree with direct edges between the nodes. Leaves are represented by their prediction value which is obtained by the (observed) conditional mean.
fancy = TRUE plots a tree similar to those in the rpart (Therneau and Atkinson, 2019) and splinetree (Neufeld and Heggeseth, 2019) R packages. The trees are drawn in an angular manner and if leaf regression models were fitted, appropriate plots of the fitted curves are depicted in the leaves. Otherwise, the usual prediction values are shown.

References

Therneau, T. & Atkinson, B. (2019). rpart: Recursive Partitioning and Regression Trees. https://CRAN.R-project.org/package=rpart
Neufeld, A. & Heggeseth, B. (2019). splinetree: Longitudinal Regression Trees and Forests. https://CRAN.R-project.org/package=splinetree