Pairwise visualization of simulated endpoints from a makeDataSim object

plot.makeDataSim() produces a pairwise plot matrix for the simulated endpoints stored in a "makeDataSim" object returned by makeData.

Usage

# S3 method for class 'makeDataSim'
plot(x, arm = 0, names = NULL, ...)

Arguments

x

An object of class "makeDataSim", created by makeData.

arm

Numeric scalar indicating which treatment arm to plot.

If the simulated dataset includes a trt column, arm must be one of 0, 1, ..., K - 1, where K is the number of arms.

If the object was generated in control-only mode (i.e., no trt column is present), the arm argument is ignored and all simulated observations are plotted.

names

Optional character vector used to relabel the plotted endpoint columns.

If supplied, names must have length equal to the number of simulated endpoints. This affects only the displayed labels in the plot and does not modify the underlying data.

...

Additional arguments passed directly to GGally::ggpairs(). This can be used to customize panels, labels, sizing, or other plot options.

Value

A GGally::ggpairs object, which is also a ggplot-compatible object and can be printed or further modified using standard ggplot2 syntax.

Details

The method uses GGally::ggpairs() to display scatterplots in the upper triangle and pairwise correlations in the lower triangle for the selected study arm. This is intended as a quick diagnostic tool for inspecting the joint structure of simulated endpoints.

Displayed data

The plotting method extracts the simulated endpoint columns recorded in x$meta$endpoint_names. These are the automatically renamed endpoint variables, such as:

Continuous endpoints: Cont_1, Cont_2, ...
Binary endpoints: Bin_1, Bin_2, ...
Count endpoints: Int_1, Int_2, ...
Time-to-event endpoints: TTE_1, TTE_2, ...

Columns such as trt, Status_*, and enrollTime are not included in the pair plot.

Panel layout

The returned ggpairs object is configured as follows:

upper triangle: scatterplots for continuous-style panels,
lower triangle: pairwise correlations,
title: "Arm <k>" for the selected arm.

A ggplot2::theme_bw() theme is applied by default.

Dependencies

This method requires both GGally and ggplot2. If either package is not installed, the method throws an error with an installation message.

Examples



ep1 <- list(
  endpoint_type = "continuous",
  baseline_mean = 10,
  sd            = 2,
  trt_effect    = -1
)

ep2 <- list(
  endpoint_type = "binary",
  baseline_prob = 0.30,
  trt_prob      = 0.45
)

ep3 <- list(
  endpoint_type = "count",
  baseline_mean = 8,
  trt_count     = 10,
  size          = 20
)

R3 <- corr_make(
  num_endpoints = 3,
  values = rbind(
    c(1, 2, 0.20),
    c(1, 3, 0.10),
    c(2, 3, 0.15)
  )
)

sim_obj <- makeData(
  correlation_matrix    = R3,
  sample_size_per_group = 500,
  SEED                  = 1,
  endpoint_details      = list(ep1, ep2, ep3)
)

## Plot control arm
plot(sim_obj)


## Plot treatment arm with custom labels
plot(sim_obj, arm = 1, names = c("Biomarker", "Responder", "Hospitalizations"))

Pairwise visualization of simulated endpoints from a `makeDataSim` object