icpack (version 0.1.0)

rasterplot: Plot probabilities as a raster`

Description

Plot probabilities as a raster`

Usage

rasterplot(
  icf,
  type = c("both", "R", "Y"),
  sel = NULL,
  label = NULL,
  show_label = FALSE,
  pow = 0.2,
  order = TRUE,
  do_plot = TRUE
)

Value

a ggplot object (Grob)

Arguments

icf

an object of class 'icfit'

type

a string giving the type of the plot. Accepted choices: 'R' (risk probabilities) and 'Y' (event probabilities)

sel

a vector of integers for selection of subject (rows of the matrix)

label

character vector containing labels for the individuals to be plotted in selection

show_label

Boolean, whether or not to show the labels

pow

a number, giving he power to which the probabilities will be raised, to improve the clarity of the plot

order

Boolean, default (TRUE) is to order according to first positive in Y, then first zero in Y, then first zero in R; if FALSE order of occurrence in data is used

do_plot

Boolean, default (TRUE) shows the plot, if FALSE object is returned but not plotted

Examples

# \donttest{
icf <- icfit(Surv(left, right, type='interval2') ~ period + gender + age,
  data=drugusers)
rasterplot(icf)
rasterplot(icf, type = 'R')
rasterplot(icf, type = 'Y')
rasterplot(icf, pow = 0.05) # very small power basically shows 0/1
sel <- c(
  11, 18,  # right-censored, event in (L, \infty)
  1:2,     # event in (0, R)
  115, 133 # event in (L, R)
)
rasterplot(icf, sel = sel)
rasterplot(icf, sel = sel, label = c("e", "p", "g", "c", "m", "n"), show_label = TRUE)
rasterplot(icf, sel = sel, label = c("e", "p", "g", "c", "m", "n"), show_label = TRUE,
  type = 'Y')
# }