Version:	0.4.0
Date:	2024-03-09
Title:	Variations of Parallel Coordinate Plots for Categorical Data
Description:	Create hammock plots, parallel sets, and common angle plots with 'ggplot2'.
License:	MIT + file LICENSE
Depends:	R (≥ 3.5.0), ggplot2 (≥ 3.5.0)
Imports:	reshape2 (≥ 1.4.4), plyr (≥ 1.8.9)
Suggests:	RColorBrewer, testthat (≥ 3.0.0)
URL:	https://github.com/heike/ggparallel/, https://heike.github.io/ggparallel/
BugReports:	https://github.com/heike/ggparallel/issues/
LazyData:	true
RoxygenNote:	7.2.3
Encoding:	UTF-8
Config/testthat/edition:	3
NeedsCompilation:	no
Packaged:	2024-03-09 20:07:53 UTC; hofmann
Author:	Heike Hofmann [aut, cre], Marie Vendettuoli [aut]
Maintainer:	Heike Hofmann <hofmann@mail.iastate.edu>
Repository:	CRAN
Date/Publication:	2024-03-09 22:00:02 UTC

Data linking genes and pathways.

Description

Table knownGene from track UCSC Genes was downloaded from the UCSC table browser for the human genome assembly (hg18, May 2006) and filtered for a selection of pathways associated with human metabolism was obtained from KEGG PATHWAY database. Bioconductor package KEGG.db was used to provide mappings between gene and pathway identifiers.

Usage

genes

Format

An object of class data.frame with 2768 rows and 14 columns.

References

Fujita PA, Rhead B, Zweig AS, Hinrichs AS, Karolchik D, Cline MS, Goldman M, Barber GP, Clawson H, Coelho A, Diekhans M, Dreszer TR, Giardine BM, Harte RA, Hillman-Jackson J, Hsu F, Kirkup V, Kuhn RM, Learned K, Li CH, Meyer LR, Pohl A, Raney BJ, Rosenbloom KR, Smith KE, Haussler D, Kent WJ. The UCSC Genome Browser database: update 2011. Nucleic Acids Res. 2010 Oct 18. http://genome.ucsc.edu/index.html?org=Human&db=hg19&hgsid=289810087

Marc Carlson, Seth Falcon, Herve Pages and Nianhua Li (). KEGG.db: A set of annotation maps for KEGG. R package version 2.6.1.

Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., and Tanabe, M.; KEGG for integration and interpretation of large-scale molecular datasets. Nucleic Acids Res. 40, D109-D114 (2012)

Kanehisa, M. and Goto, S.; KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 28, 27-30 (2000).

Examples

library(ggplot2)
library(RColorBrewer)
genes$chrom <- factor(genes$chrom, levels=c(paste("chr", 1:22, sep=""),
  "chrX", "chrY"))
ggparallel(
  list("path", "chrom"),
  text.offset = c(0.03, 0,-0.03),
  data = genes,
  width = 0.1,
  order = c(1, 0),
  angle = 0,
  color = "white",
  factorlevels =  c(sapply(unique(genes$chrom), as.character),unique(genes$path))
) +
 scale_fill_manual(
   values = c(brewer.pal("YlOrRd", n = 9), rep("grey80", 24)),
   guide = "none"
 ) +
 scale_colour_manual(
   values = c(brewer.pal("YlOrRd", n = 9), rep("grey80", 24)),
   guide = "none"
 ) +
 coord_flip()

Variations of parallel coordinate plots

Description

ggparallel implements and combines different types of parallel coordinate plots for categorical data: hammock plots, parallel sets plots, common angle plots, and common angle plots with a hammock-like adjustment for line widths.

Usage

ggparallel(
  vars = list(),
  data,
  weight = NULL,
  method = "angle",
  alpha = 0.5,
  width = 0.25,
  order = 1,
  ratio = NULL,
  asp = NULL,
  label = TRUE,
  label.colour = "grey90",
  label.size = 4,
  text.angle = 90,
  text.offset = NULL,
  same.level = FALSE,
  ...
)

Arguments

Details

Parallel sets have been suggested by kosara:2006 as a visualization technique to incorporate categorical variables into a parallel coordinate plot introduced by wegman:1990 and inselberg:1985. The parallel sets implemented here are reduced to representations of neighboring two-dimensional relationships only rather than the hierarchical version originally suggested.

Both versions, however, show perceptual problems with interpreting line widths, leading to potentially wrong conclusions about the data. The hammock display, introduced by schonlau:2003, and the common angle plots are two approaches at fixing this problem: in Hammock plots the linewidth is adjusted by a factor countering the strength of the illusion, in the common angle plot all lines are adjusted to show the same angle - making line widths again comparable across ribbons.

Additionally, we can also adjust ribbons in the common angle display for the angle, to make them appear having the same width (or height) across the display. We refer to this method as adj.angle.

Value

returns a ggplot2 object that can be plotted directly or used as base layer for additional modifications.

Examples

data(mtcars)

ggparallel(list("gear", "cyl"), data=mtcars)
ggparallel(list("gear", "cyl"), data=mtcars, method="hammock", ratio=0.25)

require(RColorBrewer)
require(ggplot2)
cols <- c(brewer.pal(4, "Reds")[-1], brewer.pal(4, "Blues")[-1])
ggparallel(list("gear", "cyl"), ratio=0.2, data=mtcars,
           method="hammock", text.angle=0) +
  scale_fill_manual(values=cols) + scale_colour_manual(values=cols) +
  theme_bw()

## combination of common angle plot and hammock adjustment:
ggparallel(list("gear", "cyl"), data=mtcars, method="adj.angle",
           ratio=2)

## compare with method='parset'
ggparallel(list("gear", "cyl"), data=mtcars, method='parset')

## flip plot and rotate text
ggparallel(list("gear", "cyl"), data=mtcars, text.angle=0) +
  coord_flip()

## change colour scheme
ggparallel(list("gear", "cyl"), data=mtcars, text.angle=0) +
  coord_flip() +
  scale_fill_brewer(palette="Set1") +
  scale_colour_brewer(palette="Set1")

## example with more than two variables:
titanic <- as.data.frame(Titanic)
ggparallel(names(titanic)[c(1,4,2,1)], order=0, titanic, weight="Freq") +
  scale_fill_brewer(palette="Paired", guide="none") +
  scale_colour_brewer(palette="Paired", guide="none")

## Not run: 
cols <- c(brewer.pal(5,"Blues")[-1], brewer.pal(3, "Oranges")[-1],
          brewer.pal(3, "Greens")[-1])
ggparallel(names(titanic)[c(1,4,2,1)], order=0, titanic, weight="Freq") +
  scale_fill_manual(values=cols, guide="none") +
  scale_colour_manual(values=cols, guide="none") + theme_bw()

## hammock plot with same width lines
ggparallel(names(titanic)[c(1,4,2,3)], titanic, weight=1, asp=0.5,
           method="hammock", ratio=0.2, order=c(0,0)) +
theme( legend.position="none") +
scale_fill_brewer(palette="Paired") +
scale_colour_brewer(palette="Paired")

## hammock plot with line widths adjusted by frequency
ggparallel(names(titanic)[c(1,4,2,3)], titanic, weight="Freq",
           asp=0.5, method="hammock", order=c(0,0), text.angle=0,
           width=0.45) +
  theme( legend.position="none")


## biological examples: genes and pathways
data(genes)
cols <- c(rep("grey80", 24), brewer.pal("YlOrRd", n = 9))
genes$chrom <- factor(genes$chrom, levels=c(paste("chr", 1:22, sep=""), "chrX", "chrY"))
ggparallel(list("path", "chrom"), text.offset=c(0.03, 0,-0.03),
           data = genes,  width=0.1, order=c(1,0), text.angle=0,
           color="white",
   factorlevels =  c(sapply(unique(genes$chrom), as.character),
     unique(genes$path))) +
   scale_fill_manual(values = cols, guide="none") +
   scale_colour_manual(values = cols, guide="none") +
   coord_flip()

## End(Not run)

data(Titanic)
titanic <- as.data.frame(Titanic)

titanic$SexSurvived <- with(titanic, interaction(Sex, Survived))
titanic$SexClassSurvived <- with(titanic, interaction(Sex,Class, Survived))

ggparallel(vars=list("Survived", "SexSurvived", "SexClassSurvived"), weight="Freq", data=titanic) +
  theme(legend.position="none") +
  scale_fill_manual(values = rep(c("Orange", "Steelblue"), 14)) +
  scale_colour_manual(values = rep(c("Orange", "Steelblue"), 14))

###########

p1 <- ggparallel(vars=list("Survived", "SexSurvived", "SexClassSurvived"),
                 weight="Freq", data=titanic, label = FALSE) +
  theme(legend.position="none") +
  scale_fill_manual(values = rep(c("Orange", "Steelblue"), 14)) +
  scale_colour_manual(values = rep(c("Orange", "Steelblue"), 14))
p2 <- ggparallel(vars=list("Survived", "SexSurvived", "SexClassSurvived"),
                 weight="Freq", data=titanic, label = TRUE) +
  theme(legend.position="none") +
  scale_fill_manual(values = rep(c("Orange", "Steelblue"), 14)) +
  scale_colour_manual(values = rep(c("Orange", "Steelblue"), 14))

ggplot_build(p2)$data[[4]]
p1 + geom_text(aes(x = x, y=y, label = label), data = ggplot_build(p2)$data[[4]])

A package for creating parallel coordinates for categorical data

Description

The main function ggparallel implements three types of parallel coordinate plots for categorical data: hammock plots, parallel sets plots, and common angle plots.