Skip to contents

Plot Cohen's d Effect Sizes

Source: R/plt_cohen.R
plt_cohen.Rd

Visualise Cohen's d values as a Cleveland dot plot (lollipop chart). Variables are ordered by the effect size of a reference group. Dashed lines at +/-0.2 indicate the small effect threshold.

Usage

plt_cohen(
  res,
  ref_group = NULL,
  palette = NULL,
  threshold = c(-0.2, 0.2),
  add_line = TRUE,
  facet = FALSE,
  base_size = 14
)

Arguments

res

Data frame from stat_cohen(): first column is Variable, remaining columns are Cohen's d per group.

ref_group

Character, name of the reference group column used for sorting. Default NULL uses the first group column.

palette

Colour palette name or character vector of colours. Default uses pal_lancet.

threshold

Numeric, dashed reference line positions for small effect. Default c(-0.2, 0.2).

add_line

Logical, connect dots with lines. Default TRUE.

facet

Logical, facet by group. Default FALSE.

base_size

Base font size. Default 14.

Value

A ggplot object.

See also

Other plot: PlotButterfly(), PlotButterfly2(), PlotRankCor(), plt_cat(), plt_con(), plt_dist(), plt_radar(), plt_sankey(), plt_upset()

Examples

# Using iris data
res <- stat_cohen(iris, group = "Species",
                  vars = c("Sepal.Length", "Sepal.Width",
                           "Petal.Length", "Petal.Width"))

# Basic dot plot
plt_cohen(res)


# Sort by a specific group
plt_cohen(res, ref_group = "versicolor")


# With facet per group
plt_cohen(res, facet = TRUE)


# Without connecting lines
plt_cohen(res, add_line = FALSE)


# Custom threshold and palette
plt_cohen(res, threshold = c(-0.5, 0.5), palette = "Paired")


# Simulated multi-group data
df <- data.frame(
  group = factor(sample(c("Control","Treatment","Placebo"), 150, TRUE)),
  age = rnorm(150, 50, 10),
  bmi = rnorm(150, 25, 5),
  score = rnorm(150, 100, 15)
)
res2 <- stat_cohen(df, "group", c("age", "bmi", "score"))
plt_cohen(res2)

plt_cohen(res2, facet = TRUE, palette = "Set2")