barrel: Ordination visualization in R

barrel

A tidy and flexible framework for visualizing multivariate ordinations in R

Overview

barrel is an R package that enhances the visualization of ordination analyses (e.g. NMDS, RDA, dbRDA) using ggplot2. It provides a modular set of tools to add ellipses, centroids, environmental vectors, and annotations — all compatible with tidyverse workflows.

Installation

From CRAN:

install.packages("barrel")

Development version from GitHub:

# install.packages("devtools")
devtools::install_github("BarrancoElena/barrel")

Quick example

library(vegan)
library(barrel)
library(ggplot2)

data(dune)
data(dune.env)

ord <- metaMDS(dune)
ord <- barrel_prepare(ord, dune.env)

autoplot(ord, group = "Management", data = dune)

# barrel

Key features

• autoplot(): single-function plotting of NMDS, RDA, dbRDA, CCA, etc.
• Support for method = "classic" and "robust" covariance estimation
• Ellipses (stat_barrel()), centroids (stat_barrel_centroid()), vectors (stat_barrel_arrows())
• Annotated variance or stress via stat_barrel_annotate()
• Functions to extract group summaries and environmental fits
• Customizable with standard ggplot2 syntax

Vignette

A full user guide is available:

browseVignettes("barrel")

Dependencies

• vegan

• ggplot2

• ggrepel

• robustbase

• grid

• rlang

Author

Diego Barranco-Elena
@BarrancoElena

License

MIT © 2025 Diego Barranco-Elena