ggRandomForests (version 1.1.3)

gg_partial_coplot.ggRandomForests: Data structures for stratified partial coplots

Description

Data structures for stratified partial coplots

Usage

gg_partial_coplot.ggRandomForests(object, xvar, groups, surv_type = c("mort",
  "rel.freq", "surv", "years.lost", "cif", "chf"), time, ...)

Arguments

object

randomForestSRC::rfsrc object

xvar

list of partial plot variables

groups

vector of stratification variable.

surv_type

for survival random forests, c("mort", "rel.freq", "surv", "years.lost", "cif", "chf")

time

vector of time points for survival random forests partial plots.

...

extra arguments passed to randomForestSRC::plot.variable function

Value

• gg_partial_coplot object. An subclass of a gg_partial_list object

Examples

# Load the forest
data(rfsrc_pbc, package="ggRandomForests")

# Create the variable plot.
ggvar <- gg_variable(rfsrc_pbc, time = 1)

# Find intervals with similar number of observations.
copper_cts <-quantile_pts(ggvar$copper, groups = 6, intervals = TRUE)

# Create the conditional groups and add to the gg_variable object
copper_grp <- cut(ggvar$copper, breaks = copper_cts)

## We would run this, but it's expensive
partial_coplot_pbc <- gg_partial_coplot(rfsrc_pbc, xvar = "bili",
                                         groups = copper_grp,
                                         surv_type = "surv",
                                         time = 1,
                                         show.plots = FALSE)
## so load the cached set
data(partial_coplot_pbc, package="ggRandomForests")

# Partial coplot
plot(partial_coplot_pbc, se = FALSE)