raincloud

Version 1.0.0 | 2026-04-08

raincloud draws raincloud plots: a half-violin density, jittered raw points, and a box-and-whisker summary in one figure. The command is built for distributional comparisons where you want the overall shape and the observed values on the same graph.

It supports grouped displays through over(), horizontal or vertical orientation, mirror-mode split violins, analytic and frequency weights, and pass-through styling options for the cloud, points, and box layers.

Requirements

• Stata 16 or later

Installation

capture ado uninstall raincloud
net install raincloud, from("https://raw.githubusercontent.com/tpcopeland/Stata-Tools/main/raincloud") replace

Commands

Command	Description
raincloud	Draw a raincloud plot with density, scatter, and box elements

Quick Start

The easiest place to start is Stata's built-in auto dataset.

sysuse auto, clear
raincloud mpg, over(foreign)

This draws separate rainclouds for domestic and foreign cars so you can compare the full MPG distributions, not just the means.

How It Works

Each raincloud combines three pieces:

• The cloud is a half-violin kernel density.
• The rain is a jittered scatter of the observed values.
• The box is a quartile-and-whisker summary of the same distribution.

You can turn individual layers off with nocloud, norain, and nobox or noumbrella. Only one numeric outcome variable is allowed at a time. To compare multiple measures, reshape to long format and use over() for the grouping variable.

Worked Examples

1. Single distribution

sysuse auto, clear
raincloud mpg

2. Grouped comparison

sysuse auto, clear
raincloud mpg, over(foreign)

Use this when the main question is whether groups differ in spread, skewness, overlap, or outliers.

3. Vertical layout with a mean marker

sysuse auto, clear
raincloud price, over(foreign) vertical mean

4. Mirror mode and custom styling

sysuse auto, clear
raincloud mpg, over(foreign) mirror ///
    opacity(70) jitter(0.6) ///
    cloudopts(lwidth(medium)) ///
    pointopts(msymbol(d) msize(tiny)) ///
    boxopts(lwidth(thick)) ///
    colors(navy cranberry)

Common Options

Option	Description
over(varname)	Draw one raincloud per group
horizontal / vertical	Choose the plot orientation; horizontal is the default
mirror	Draw the cloud on both sides of center
nocloud	Suppress the half-violin density
norain	Suppress the jittered raw points
nobox / noumbrella	Suppress the box-and-whisker summary
bandwidth(#)	Set the kernel-density bandwidth; 0 uses Stata's default selector
jitter(#)	Control point jitter from 0 to 1
opacity(#)	Control cloud fill opacity from 0 to 100
colors(string)	Supply a space-separated custom palette
mean	Add a mean marker
seed(#)	Make the jitter reproducible

Returned Results

raincloud stores the following in r():

• r(N) for the number of observations used
• r(n_groups) for the number of groups
• r(varname) for the plotted variable
• r(over) for the grouping variable, when used
• r(stats) for the group-wise summary matrix containing n, mean, sd, median, q25, q75, iqr, and bandwidth

Gallery

Basic grouped comparison

Vertical orientation

Mirror layout

Reference

• Allen M, Poggiali D, Whitaker K, Marshall TR, Kievit RA. Raincloud plots: a multi-platform tool for robust data visualization. Wellcome Open Research. 2019;4:63.

Version History

• 1.0.0 (2026-04-08): Initial Stata-Tools release of raincloud

Author

Timothy P Copeland, Karolinska Institutet

License

MIT