Skip to contents

ggrrr

Data presentation and visualisation hacks.

R-CMD-check DOI ggrrr status badge

see the package documentation

ggrrr is a repository of utility functions. It is not primarily designed to be installed, but rather as a source of standalone library files - ideally in conjunction with pkgtools which provides some infrastructure to render unit tests, and interact with local standalone files.

# Optional:
# Enable repository from terminological
options(repos = c(
  terminological = 'https://terminological.r-universe.dev',
  CRAN = 'https://cloud.r-project.org'))
install.packages("pkgtools")

# Use the `ggrrr` repository as a source of standalones
pkgtools::use_standalone("terminological/ggrrr")

# Alternatively:
usethis::use_standalone("terminological/ggrrr") 
# will work in a package project.

Most of the functions are available in standalone format for inclusion into your own package or analysis project. There are some functions that are only available in ggrrr as a package particularly font installation functions. For thes you will have to install ggrrr which can be done with:

# Enable repository from terminological
options(repos = c(
  terminological = 'https://terminological.r-universe.dev',
  CRAN = 'https://cloud.r-project.org'))
install.packages("ggrrr")

# Or the unstable github version
devtools::install_github("terminological/ggrrr")

ggrrr is not on CRAN. It is not likely to ever be and the use of standalones is the primary channel for distribution.

ggrrr has a focus on output from ggplot and huxtable with file size, image resolution, opinionated formatting of defaults, and consistent output and sizing between screen, html, pdf and png outputs.

It also addresses an eclectic mix of other issues that arose during my time developing with R.

  • Font handling - a million and one ways of doing it in R. Now a million and two.
  • writing files to date-versioned output directories, and constructing data supplements.
  • caching intermediate analysis stages.
  • outputting the right thing at the right time depending on whether you are running code in the terminal, knitting a markdown document etc.
  • converting tidy data to tabular content.
  • embedding tables into ggplot.
  • putting source code into html and pdf documents.
  • random scales I found useful.
  • mixture normal distributions for combining bootstraps
  • tidy data utility functions such as cutting dates into regular periods, cutting integer values into labelled categories.

Pre installation:

ggrrr will fallback to use of a headless chrome instance in some situations, particularly for rendering very complicated ggplots and huxtables to pdf. ggrrr is configured to use html2pdfr for html to pdf conversion if available. This uses rJava, which in turn requires a Java installation. In theory it is optional and will not be installed automatically, or everything will be installed automagically. In practice getting a working html2pdfr is probably done separately by following the html2pdfr installation instructions.