SubgrPlots (version 0.1.3)

ggplot_stepp: STEPP for subgroup effect size

Description

This function produces a plot of using the approach "Subpopulation Treatment Effect Pattern Plot". It shows the treatment effect size of subgroups, where subgruops are defined by certain ranges of a continuous covariate; each subgroup has a sample size close to a pre-specified value (N2) and any neighboring subgroups have an overlap size near another pre-specified value (N1). The plot shows the 95 a horizontal line). The y-coordinate of a point indicates the effect size within the corresponding subgroup; the x-coordinate shows the lower bound of the range which defines the subgroup. If part of the horizontal line is out of the simultaneous C.I., it may reveal hetergeneity across subgroup effects with repective to the overall effect. In addition, one can control the width of the C.I. by controlling Type I error rate in one of the function arguments. Note that the function uses log odd ratio and log hazard ratio for displaying subgroup effect sizes in binary and survival data, respectively. The actual subgroup sample sizes over the covariate are shown on the console window as well.

Usage

ggplot_stepp(
  dat,
  covari.sel,
  trt.sel,
  resp.sel,
  outcome.type,
  setup.ss,
  alpha,
  title = NULL,
  lab.y = NULL,
  subtitle = NULL
)

Arguments

dat

a data set

covari.sel

a vector of indices of the two covariates

trt.sel

a variate index specifying the treatment code

resp.sel

a variate index specifying the response variable

outcome.type

a string specifying the type of the response variable, it can be "continuous", or "binary" or "survival".

setup.ss

a vector specifying the approximate overlap size (N2) and subgroup sample size (N1).

alpha

the type I error rate

title

a string specifying the main title.

lab.y

a string specifying the labels of the y-axis.

subtitle

strings specifying the subtitle

Details

Contrary to plot_stepp, ggplot_stepp implements the proper x-axis. This function uses the ggplot2 package to draw the actual plot. To control font sizes, use theme with the resulting object.

See Also

plot_stepp

Examples

# NOT RUN {
# Load the data to be used
data(prca)
dat = prca

## 9. stepp Plot -----------------------------------------------------------
lab.y.title = paste("Treatment effect size (log-hazard ratio)");
setup.ss = c(30,40)
sub.title = paste0("(Subgroup sample sizes are set to ", setup.ss[2],
                   "; overlap sizes are set to ", setup.ss[1], ")" )
ggplot_stepp(dat,
           covari.sel = 8,
           trt.sel = 3,
           resp.sel = c(1, 2),
           outcome.type = "survival",
           setup.ss = c(30,40),
           alpha = 0.05,
           title = NULL,
           lab.y = lab.y.title,
           subtitle = sub.title)


# }