geomerrorbar and geomerrorbarh

Error bars display uncertainty in data values by showing a range around each point. geom_errorbar() creates vertical error bars, while geom_errorbarh() creates horizontal error bars. These are commonly used to show confidence intervals, standard errors, or other measures of variability around summary statistics.

Arguments

• plot::GGPlot (optional): a plot object to add this geom to. This is typically used to facilitate creating your ggplot as part of a @chain.
• data (DataFrame): Data to use for this geom. If not provided, the geom will inherit the data from ggplot.
• aes(...): the names of the columns in the DataFrame that will be used in the mapping
• inherit_aes: should the geom inherit aes from the ggplot?
• ...: options that are not mapped to a column (passed to Makie.Rangebars)

Vertical vs Horizontal

Use geom_errorbar() when your error bars should be vertical (most common case, e.g., when x is categorical and y is numeric). Use geom_errorbarh() when your error bars should be horizontal (e.g., when y is categorical and x is numeric).

Required Aesthetics (geom_errorbar)

• x - the x position of the error bar
• ymin - the lower end of the error bar
• ymax - the upper end of the error bar

Required Aesthetics (geom_errorbarh)

• y - the y position of the error bar
• xmin - the left end of the error bar
• xmax - the right end of the error bar

Optional Aesthetics (see aes)

• color / colour

Optional Arguments

• color / colour: color of the error bars
• linewidth: width of the error bar lines
• whiskerwidth / width: width of the whisker caps (in pixels)
• alpha

using TidierPlots
using TidierData
using DataFrames
using Statistics

penguins = penguins_data();

First, create a summary dataset with means and standard deviations:

df = @chain penguins begin
    @group_by(species)
    @summarize(
        mean_bill = mean(bill_length_mm),
        sd_bill = std(bill_length_mm),
        n = n()
    )
    @mutate(
        se_bill = sd_bill / sqrt(n)
    )
    @mutate(
        lower = mean_bill - 1.96 * se_bill,
        upper = mean_bill + 1.96 * se_bill
    )
end

3×7 DataFrame

Row	species	mean_bill	sd_bill	n	se_bill	lower	upper
	String	Float64	Float64	Int64	Float64	Float64	Float64
1	Adelie	38.824	2.6626	146	0.220358	38.3921	39.2559
2	Gentoo	47.5681	3.10612	119	0.284737	47.01	48.1262
3	Chinstrap	48.8338	3.33926	68	0.404944	48.0401	49.6275

Basic error bars showing 95% confidence intervals:

ggplot(df, @aes(x = species, ymin = lower, ymax = upper)) +
    geom_errorbar(linewidth = 2)

Add whisker caps to make the endpoints more visible:

ggplot(df, @aes(x = species, ymin = lower, ymax = upper)) +
    geom_errorbar(linewidth = 2, whiskerwidth = 20)

Combine error bars with points to show both the central value and uncertainty:

ggplot(df, @aes(x = species, y = mean_bill, ymin = lower, ymax = upper)) +
    geom_errorbar(linewidth = 1, whiskerwidth = 15) +
    geom_point(size = 15)

Combine with bar charts:

ggplot(df, @aes(x = species, y = mean_bill, ymin = lower, ymax = upper)) +
    geom_col(alpha = 0.5) +
    geom_errorbar(linewidth = 2, whiskerwidth = 20)

Create horizontal error bars using geom_errorbarh:

df_h = @chain penguins begin
    @group_by(species)
    @summarize(
        mean_bill = mean(bill_length_mm),
        sd_bill = std(bill_length_mm),
        n = n()
    )
    @mutate(
        se_bill = sd_bill / sqrt(n)
    )
    @mutate(
        lower = mean_bill - 1.96 * se_bill,
        upper = mean_bill + 1.96 * se_bill
    )
end

ggplot(df_h, @aes(y = species, xmin = lower, xmax = upper)) +
    geom_errorbarh(linewidth = 2)

Example with custom error bar data:

df_custom = DataFrame(
    trt   = [1, 1, 2, 2],
    resp  = [1, 5, 3, 4],
    group = ["A", "B", "A", "B"],
    lower = [0.8, 4.6, 2.4, 3.6],
    upper = [1.1, 5.3, 3.3, 4.2]
)

ggplot(df_custom, @aes(x = trt, y = resp, ymin = lower, ymax = upper, color = group)) +
    geom_errorbar(whiskerwidth = 20, linewidth = 2) +
    geom_point(size = 15)

---

This page was generated using Literate.jl.