prep_sigmark: Prepare for geom_sigmark

In pstraforelli/tidytests: A tidy implementation of statistical testing

Prepare for geom_sigmark

Description

Helper function to prepare a summarized data frame for significance markers with geom_sigmark

Usage

prep_sigmark(df, test_df, subgroups, result, colours = NULL, percent, vs_rest)

Arguments

df	A data frame with the summarized data to be charted
test_df	The output of either pairwise_t_test or pairwise_prop_test. Note that you need to filter out that output to only include those comparisons to be shown as significant.
subgroups	Grouping vector
result	Result vector (either proportions or mean values)
colours	A vector of colours associated with the subgroup levels. Not needed if vs_rest argument is set to TRUE.
percent	A logical indicating whether the results are proportions or not
vs_rest	A logical indicating whether the vs_rest argument in ⁠pairwise_*_test()⁠ was set to TRUE or FALSE.

Value

A tibble ready for ggplot2 chart with a geom_sigmark() layer.

Examples

mydf <- data.frame(smokers = c(rbinom(100, 1, .8),
                               rbinom(70, 1, 0.7),
                               rbinom(50, 1, 0.6)),
                   region = c(rep("A", 100), rep("B", 70), rep("C", 50)))
test_res <- pairwise_prop_test(mydf, smokers, region)

colour_vec <- c("A" = "red", "B" = "blue", "C" = "green")

library(dplyr, warn.conflicts = FALSE)
mydf |>
  count(smokers, region) |>
  filter(smokers == 1) |>
  mutate(perc = n / sum(n)) |>
  prep_sigmark(test_res, region, perc, colour_vec, percent = TRUE, vs_rest = FALSE)

pstraforelli/tidytests documentation built on Oct. 28, 2023, 3:30 p.m.