Skip to content

Instantly share code, notes, and snippets.

@kohske
Created August 11, 2011 14:48
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save kohske/1139848 to your computer and use it in GitHub Desktop.
Save kohske/1139848 to your computer and use it in GitHub Desktop.
ggplot2: example codes for matrix-like legend guide
p <- qplot(1:20, 1:20, col=letters[1:20])
p + guides(col=guide_legend(nrow=8))
p + guides(col=guide_legend(ncol=8))
p + guides(col=guide_legend(nrow=8, byrow=T))
p + guides(col=guide_legend(ncol=8, byrow=T))
# output plots: http://yfrog.com/klhjpwp
# how to embed images in gist...?
# you can install very very easily the develop version of ggplot2 on github thanks to Hadley's devtools package.
install.packages('devtools')
install.packages(c("RColorBrewer", "stringr", "dichromat", "munsell", "plyr", "colorspace")) # probably windows only
library(devtools)
dev_mode(TRUE)
install_github("scales")
install_github("ggplot2", "kohske", "feature/new-guides-with-gtable")
library(ggplot2)
@GregoryEady
Copy link

Hi Kohske,

I installed the developer version of ggplot2 with the following:

install_github("ggplot2", "kohske", "feature/new-guides-with-gtable")

I did so because I need to avail myself of your guides() functions, which works as it should (thanks!)

I need, however, to use coord_cartesian(xlim(1,5)), but when I use this function with the developer version of ggplot2 as installed as you've instructed, I get the following error:

Error in coord$wise : object of type 'closure' is not subsettable

This error can be reproduced with the following code:

x1 <- c(1,1,1,3,3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5,6,6)
Dataset <- data.frame(x1)
ggplot(Dataset) + stat_density(aes(x = x1), geom = "path", position =
"identity") + coord_cartesian(xlim = c(1,5))

Any help would be appreciated! Thanks!

@kohske
Copy link
Author

kohske commented Nov 18, 2011

Hi,

Thanks for reporting the bug.

As for the bug, it has been fixed, but not yet merged into the "develop" branch.
Could you please try the "cutting-edge" branch?

install_github("ggplot2", "kohske", "cutting-edge")

This branch includes some bug fixes.
But please note that this is unofficial version, and hence I strongly recommend to use the official version on CRAN after it will be released.

@GregoryEady
Copy link

Excellent. Thanks for the quick response.

@GregoryEady
Copy link

The "cutting-edge" version seemed to fix the problem.

Two small(ish) problems with the ways the legend is now constructed with guides():

It seems that the legend labels are now centered rather than left-justified as they were before. I wonder if there is a parameter to put the labels back to left justification.

Also, when I set the legend.position to "bottom", the legend ends up in the bottom right hand corner of the plot window, rather than at the bottom and centered as it is in the official version.

My code is as follows:

ggplot(Dataset) +
my.theme(base_size = textsize) +
stat_density(aes(x = theAnswer, colour = theGroup, order = as.factor(theGroup)), size = 1, adjust = smooth, geom = "path", position = "identity") +
#facet_wrap(~question) +
coord_cartesian(xlim = c(0.5,5.5)) +
scale_x_continuous(breaks = c(1,2,3,4,5),
labels = answerLabels[(answer_5-4):(answer_5)]) +
scale_y_continuous(breaks = NA) +
labs(x = NULL, y = NULL, colour = "") +
opts(legend.position = "bottom",
legend.direction = "horizontal",
legend.key.size = unit((textsize/13), "lines"),
panel.background = theme_rect(fill = "grey14", colour = "grey14"),
plot.background = theme_rect(colour = "grey14"),
panel.border = theme_rect(fill = NA, colour = "grey90", size = 1.25),
plot.margin = unit(c(0, 7, 0, 5.5), "lines")) +
scale_colour_manual(breaks = legendOrder, values = color) +
guides(col=guide_legend(ncol=3))

Thanks again!

@kohske
Copy link
Author

kohske commented Nov 18, 2011

could you please provide a reproducible example?
?guide_legend will show you help page.

@liamxg
Copy link

liamxg commented Oct 17, 2022

how can i set the legend row into one?

@liamxg
Copy link

liamxg commented Oct 17, 2022

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment