The plot method for BinaryTree and mob objects are rather
flexible and can be extended by panel functions. Some pre-defined
panel-generating functions of class grapcon_generator
for the most important cases are documented here.
node_inner(ctreeobj, digits = 3, abbreviate = FALSE,
fill = "white", pval = TRUE, id = TRUE)
node_terminal(ctreeobj, digits = 3, abbreviate = FALSE,
fill = c("lightgray", "white"), id = TRUE)
edge_simple(treeobj, digits = 3, abbreviate = FALSE)
node_surv(ctreeobj, ylines = 2, id = TRUE, …)
node_barplot(ctreeobj, col = "black", fill = NULL, beside = NULL,
ymax = NULL, ylines = NULL, widths = 1, gap = NULL,
reverse = NULL, id = TRUE)
node_boxplot(ctreeobj, col = "black", fill = "lightgray",
width = 0.5, yscale = NULL, ylines = 3, cex = 0.5, id = TRUE)
node_hist(ctreeobj, col = "black", fill = "lightgray",
freq = FALSE, horizontal = TRUE, xscale = NULL, ymax = NULL,
ylines = 3, id = TRUE, …)
node_density(ctreeobj, col = "black", rug = TRUE,
horizontal = TRUE, xscale = NULL, yscale = NULL, ylines = 3,
id = TRUE)
node_scatterplot(mobobj, which = NULL, col = "black",
linecol = "red", cex = 0.5, pch = NULL, jitter = FALSE,
xscale = NULL, yscale = NULL, ylines = 1.5, id = TRUE,
labels = FALSE)
node_bivplot(mobobj, which = NULL, id = TRUE, pop = TRUE,
pointcol = "black", pointcex = 0.5,
boxcol = "black", boxwidth = 0.5, boxfill = "lightgray",
fitmean = TRUE, linecol = "red",
cdplot = FALSE, fivenum = TRUE, breaks = NULL,
ylines = NULL, xlab = FALSE, ylab = FALSE, margins = rep(1.5, 4), …)an object of class BinaryTree.
an object of class BinaryTree or mob.
an object of class mob.
integer, used for formating numbers.
logical indicating whether strings should be abbreviated.
a color for points and lines.
a color to filling rectangles.
logical. Should p values be plotted?
logical. Should node IDs be plotted?
number of lines for spaces in y-direction.
widths in barplots.
width in boxplots.
gap between bars in a barplot (node_barplot).
limits in y-direction
limits in x-direction
upper limit in y-direction
logical indicating if barplots should be side by side or stacked.
logical indicating whether the order of levels should be reversed for barplots.
logical indicating if the plots should be horizontal.
logical; if TRUE, the histogram graphic is a representation
of frequencies. If FALSE, probabilities are plotted.
logical indicating if a rug representation should be added.
numeric or character vector indicating which of the regressor variables should be plotted (default = all).
color for fitted model lines.
character extension of points in scatter plots.
plotting character of points in scatter plots.
logical. Should the points be jittered in y-direction?
logical. Should axis labels be plotted?
logical. Should the panel viewports be popped?
color for box plot borders.
fill color for box plots.
logical. Should lines for the predicted means from the model be added?
logical. Should CD plots (or spinograms) be used for visualizing the dependence of a categorical on a numeric variable?
logical. When using spinograms, should the five point summary of the explanatory variable be used for determining the breaks?
a (list of) numeric vector(s) of breaks for the spinograms. If set to NULL
(the default), the breaks are chosen according to the fivenum
argument.
character with x- and y-axis labels. Can also be logical: if FALSE
axis labels are surpressed, if TRUE they are taken from the underlying data.
Can be a vector of labels for xlab.
margins of the viewports.
additional arguments passed to callies.
The plot methods for BinaryTree and mob objects provide an
extensible framework for the visualization of binary regression trees. The
user is allowed to specify panel functions for plotting terminal and inner
nodes as well as the corresponding edges. The panel functions to be used
should depend only on the node being visualzied, however, for setting up
an appropriate panel function, information from the whole tree is typically
required. Hence, party adopts the framework of grapcon_generator
(graphical appearance control) from the vcd package (Meyer, Zeileis and
Hornik, 2005) and provides several panel-generating functions. For convenience,
the panel-generating functions node_inner and edge_simple
return panel functions to draw inner nodes and left and right edges.
For drawing terminal nodes, the functions returned by the other panel
functions can be used. The panel generating function node_terminal
is a terse text-based representation of terminal nodes.
Graphical representations of terminal nodes are available and depend on the kind of model and the measurement scale of the variables modelled.
For univariate regressions (typically fitted by ctree),
node_surv returns a functions that plots Kaplan-Meier curves in each
terminal node; node_barplot, node_boxplot, node_hist and
node_density can be used to plot bar plots, box plots, histograms and
estimated densities into the terminal nodes.
For multivariate regressions (typically fitted by mob),
node_bivplot returns a panel function that creates bivariate plots
of the response against all regressors in the model. Depending on the scale
of the variables involved, scatter plots, box plots, spinograms (or CD plots)
and spine plots are created. For the latter two spine and
cd_plot from the vcd package are re-used.
David Meyer, Achim Zeileis, and Kurt Hornik (2006). The Strucplot Framework: Visualizing Multi-Way Contingency Tables with vcd. Journal of Statistical Software, 17(3). 10.18637/jss.v017.i03
# NOT RUN {
set.seed(290875)
airq <- subset(airquality, !is.na(Ozone))
airct <- ctree(Ozone ~ ., data = airq)
## default: boxplots
plot(airct)
## change colors
plot(airct, tp_args = list(col = "blue", fill = hsv(2/3, 0.5, 1)))
## equivalent to
plot(airct, terminal_panel = node_boxplot(airct, col = "blue",
fill = hsv(2/3, 0.5, 1)))
### very simple; the mean is given in each terminal node
plot(airct, type = "simple")
### density estimates
plot(airct, terminal_panel = node_density)
### histograms
plot(airct, terminal_panel = node_hist(airct, ymax = 0.06,
xscale = c(0, 250)))
# }
Run the code above in your browser using DataLab