obliqueRSF (version 0.1.2)

pdplot: Plot partial variable dependence using an oblique random survival forest

Description

Plot partial variable dependence using an oblique random survival forest

Usage

pdplot(
  object,
  xvar,
  xlab = NULL,
  xvar_units = NULL,
  xvals = NULL,
  nxpts = 10,
  ytype = "nonevent",
  event_lab = "death",
  nonevent_lab = "survival",
  fvar = NULL,
  flab = NULL,
  flvls = NULL,
  time_units = "years",
  xlvls = NULL,
  sub_times = NULL,
  separate_panels = TRUE,
  color_palette = "Dark2"
)

Value

A ggplot2 object showing partial dependence according to the oblique random survival forest object.

Arguments

object

an ORSF object (i.e. object returned from the ORSF function)

xvar

a string giving the name of the x-axis variable

xlab

the label to be printed describing the x-axis variable

xvar_units

the unit of measurement for the x-axis variable. For example, age is usually measured in years.

xvals

a vector containing the values that partial dependence will be computed with.

nxpts

instead of specifying xvals, you can specify how many points on the x-axis you would like to plot predicted responses for, and a set of nxpts equally spaced percentile values from the distribution of xvar will be used.

ytype

String. Use 'event' if you would like to plot the probability of the event, and 'nonevent' if you prefer to plot the probability of a non-event.

event_lab

string that describes the event

nonevent_lab

string that describes a non-event.

fvar

a string indicating a variable to facet the plot with

flab

a label describing the facet variable.

flvls

the labels to be printed describing the facet variable. For a facet variable with k categories, flab should be a vector with k labels, given in the same order as the levels of the facet variable.

time_units

the unit of time, e.g. days, since baseline.

xlvls

A character vector with descriptions of each category in the x-variable. This is only relevant if x is categorical.

sub_times

a vector of times to compute predicted survival probabilities. Note that the eval_times from the ORSF object are used to compute predictions, and sub_times must be a subset of those times.

separate_panels

true or false. If true, the plot will display predictions in two separate panels, determined by the facet variable.

color_palette

Palette to use for colors in the figure. Options are Diverging (BrBG, PiYG, PRGn, PuOr, RdBu, RdGy, RdYlBu, RdYlGn, Spectral), Qualitative (Accent, Dark2, Paired, Pastel1, Pastel2, Set1, Set2, Set3), Sequential (Blues, BuGn, BuPu, GnBu, Greens, Greys, Oranges, OrRd, PuBu, PuBuGn, PuRd, Purples, RdPu, Reds, YlGn, YlGnBu, YlOrBr, YlOrRd), and viridis.

Examples

if (FALSE) {
data("pbc",package='survival')
pbc$status[pbc$status>=1]=pbc$status[pbc$status>=1]-1
pbc$time=pbc$time/365.25
pbc$id=NULL
fctrs<-c('trt','ascites','spiders','edema','hepato','stage')
for(f in fctrs)pbc[[f]]=as.factor(pbc[[f]])
pbc=na.omit(pbc)

orsf=ORSF(data=pbc, eval_time=1:10,ntree=30)

pdplot(object=orsf, xvar='bili', xlab='Bilirubin', 
       xvar_units='mg/dl', sub_times=10)
}