In thomasp85/farver: High Performance Colour Space Manipulation

farver

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-"
)

The goal of farver is to provide very fast, vectorised functions for conversion of colours between different colour spaces, colour comparisons (distance between colours), encoding/decoding, and channel manipulation in colour strings. To this end it provides an interface to a modified version of the ColorSpace C++ library developed by Berendea Nicolae.

Installation

farver can be installed from CRAN using install.packages('farver'). The development version can be installed from Github using devtools:

# install.packages('devtools')
devtools::install_github('thomasp85/farver')

Use

farver provides an alternative to the grDevices::rgb() and grDevices::col2rgb() for encoding and decoding colours strings. The farver functions are superficially equivalent but provides a uniform output format, and the option to encode and decode directly from/to other colour spaces.

library(farver)

codes <- rainbow(10)
codes

spectrum <- decode_colour(codes)
spectrum

encode_colour(spectrum)

It also provides an alternative to grDevices::convertColor() to switch between colours spaces. If the origin is a colour string it is possible to decode directly into the given colour space. Conversely, if the endpoint is a colour string it is also possible to encode directly from a given colour space.

spectrum_lab <- convert_colour(spectrum, 'rgb', 'lab')
spectrum_lab

decode_colour(codes, to = 'lab')

encode_colour(spectrum_lab, from = 'lab')

If colours are given as strings, manipulation of channels will normally require decoding, conversion to the correct colour space, manipulation of the given channel, converting back to rgb and the encoding to string. farver provides a range of functions that allow you to change any channel in the supported spaces directly in colour strings:

# Add a value to the channel
add_to_channel(codes, channel = 'l', value = 1:10, space = 'lab')

# Set a channel to a specific value
set_channel(codes, 'alpha', c(0.3, 0.7))

# Limit a channel to a given value
cap_channel(codes, 'r', 200)

Lastly, farver also provides utilities for calculating the distance between colours, based on a range of different measures

spectrum2 <- t(col2rgb(heat.colors(10)))

compare_colour(spectrum, spectrum2, 'rgb', method = 'cie2000')[1:6, 1:6]

Supported colour spaces

farver currently supports the following colour spaces:

• CMY
• CMYK
• HSL
• HSB
• HSV
• CIE L*AB
• Hunter LAB
• OK LAB
• LCH(ab)
• LCH(uv)
• LCH(OK)
• LUV
• RGB
• XYZ
• YXY

Supported distance measures

farver supports the following colour distance metrics

• Euclidean
• CIE1976
• CIE94
• CIE2000
• CMC

White References

farver allows you to set the white point for relative colour spaces, either based on a standard illuminant (A-F series supported) or by specifying chromaticity coordinates or tristimulus values directly

Benchmark

farver is faster than its grDevices counterpart but less so than it was at its first release, as the colour conversion in grDevices has been improved since.

library(ggplot2)
test <- matrix(runif(300000, min = 0, max = 255), ncol = 3)
timing <- bench::mark(
  farver = convert_colour(test, 'rgb', 'lab'),
  grDevices = convertColor(test, 'sRGB', 'Lab', scale.in = 255), 
  check = FALSE,
  min_iterations = 100
)
plot(timing, type = 'ridge')

Still, if the start- and/or endpoint are colour strings the ability to decode and encode directly from/to any colour space will give a huge speed up.

colour_strings <- colours()
timing <- bench::mark(
  farver = decode_colour(colour_strings, to = 'lab'),
  grDevices = convertColor(t(col2rgb(colour_strings)), 'sRGB', 'Lab', scale.in = 255), 
  check = FALSE,
  min_iterations = 100
)
plot(timing, type = 'ridge')

Code of Conduct

Please note that the 'farver' project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

thomasp85/farver documentation built on May 15, 2024, 12:46 p.m.