I am trying to replicate this with R ggplot. I have exactly the same data:
<
I find it easier to work in rectangular coordinates first, and when that is correct, then switch to polar coordinates. The x coordinate becomes radius in polar. So, in rectangular coordinates, the inside plot goes from zero to a number, like 3, and the outer band goes from 3 to 4.
For example
ggplot(browsers) +
geom_rect(aes(fill=version, ymax=ymax, ymin=ymin, xmax=4, xmin=3)) +
geom_rect(aes(fill=browser, ymax=ymax, ymin=ymin, xmax=3, xmin=0)) +
xlim(c(0, 4)) +
theme(aspect.ratio=1)
Then, when you switch to polar, you get something like what you are looking for.
ggplot(browsers) +
geom_rect(aes(fill=version, ymax=ymax, ymin=ymin, xmax=4, xmin=3)) +
geom_rect(aes(fill=browser, ymax=ymax, ymin=ymin, xmax=3, xmin=0)) +
xlim(c(0, 4)) +
theme(aspect.ratio=1) +
coord_polar(theta="y")
This is a start, but may need to fine tune the dependency on y (or angle) and also work out the labeling / legend / coloring... By using rect for both the inner and outer rings, that should simplify adjusting the coloring. Also, it can be useful to use the reshape2::melt function to reorganize the data so then legend comes out correct by using group (or color).
you can get something similar using the package ggsunburst
# using your data without "ymax" and "ymin"
browsers <- structure(list(browser = structure(c(3L, 3L, 3L, 3L, 2L, 2L,
2L, 1L, 5L, 5L, 4L), .Label = c("Chrome", "Firefox", "MSIE",
"Opera", "Safari"), class = "factor"), version = structure(c(5L,
6L, 7L, 8L, 2L, 3L, 4L, 1L, 10L, 11L, 9L), .Label = c("Chrome 10.0",
"Firefox 3.5", "Firefox 3.6", "Firefox 4.0", "MSIE 6.0", "MSIE 7.0",
"MSIE 8.0", "MSIE 9.0", "Opera 11.x", "Safari 4.0", "Safari 5.0"
), class = "factor"), share = c(10.85, 7.35, 33.06, 2.81, 1.58,
13.12, 5.43, 9.91, 1.42, 4.55, 1.65)), .Names = c("parent", "node", "size")
, row.names = c(NA, -11L), class = "data.frame")
# add column browser to be used for colouring
browsers$browser <- browsers$parent
# write data.frame into csv file
write.table(browsers, file = 'browsers.csv', row.names = F, sep = ",")
# install ggsunburst
if (!require("ggplot2")) install.packages("ggplot2")
if (!require("rPython")) install.packages("rPython")
install.packages("http://genome.crg.es/~didac/ggsunburst/ggsunburst_0.0.9.tar.gz", repos=NULL, type="source")
library(ggsunburst)
# generate data structure
sb <- sunburst_data('browsers.csv', type = 'node_parent', sep = ",", node_attributes = c("browser","size"))
# add name as browser attribute for colouring to internal nodes
sb$rects[!sb$rects$leaf,]$browser <- sb$rects[!sb$rects$leaf,]$name
# plot adding geom_text layer for showing the "size" value
p <- sunburst(sb, rects.fill.aes = "browser", node_labels = T, node_labels.min = 15)
p + geom_text(data = sb$leaf_labels,
aes(x=x, y=0.1, label=paste(size,"%"), angle=angle, hjust=hjust), size = 2)
@rawr's solution is really nice, however, the labels will be overlapped if there are too many. Inspired by @user3969377 and @FlorianGD, I got a new solution using ggplot2
and ggrepel
.
browsers$ymax <- cumsum(browsers$share) # fed to geom_rect() in piedonut()
browsers$ymin <- browsers$ymax - browsers$share # fed to geom_rect() in piedonut()
browsers$share_browser <- sum(browsers$share[browsers$browser == unique(browsers$browser)[1]]) # "_browser" means at browser level
browsers$ymax_browser <- browsers$share_browser[browsers$browser == unique(browsers$browser)[1]][1]
for (z in 2:length(unique(browsers$browser))) {
browsers$share_browser[browsers$browser == unique(browsers$browser)[z]] <- sum(browsers$share[browsers$browser == unique(browsers$browser)[z]])
browsers$ymax_browser[browsers$browser == unique(browsers$browser)[z]] <- browsers$ymax_browser[browsers$browser == unique(browsers$browser)[z-1]][1] + browsers$share_browser[browsers$browser == unique(browsers$browser)[z]][1]
}
browsers$ymin_browser <- browsers$ymax_browser - browsers$share_browser
piedonut <- function(data, cols = c('cyan2','red','orange','green','dodgerblue2'), force = 80, nudge_x = 3, nudge_y = 10) { # force, nudge_x, nudge_y are parameters to fine tune positions of the labels by geom_label_repel.
nr <- nrow(data)
# width <- max(sqrt(data$share)) / 0.1
tbl <- with(data, table(browser)[order(unique(browser))])
cols <- unlist(Map(rep, cols, tbl))
col_subnum <- unlist(Map(rep, 255/tbl,tbl))
col <- rep(NA, nr)
col_browser <- rep(NA, nr)
for (i in 1:nr) {
## create color/shades
rgb <- col2rgb(cols[i])
col[i] <- rgb(rgb[1], rgb[2], rgb[3], col_subnum[i]*sequence(tbl)[i], maxColorValue = 255)
rgb <- col2rgb(cols[i])
col_browser[i] <- rgb(rgb[1], rgb[2], rgb[3], maxColorValue = 255)
}
#col
# set labels positions
x.breaks <- seq(1, 1.8, length.out = nr)
y.breaks <- cumsum(data$share)-data$share/2
ggplot(data) +
geom_rect(aes(ymax = ymax, ymin = ymin, xmax=4, xmin=1), fill=col) +
geom_rect(aes(ymax=ymax_browser, ymin=ymin_browser, xmax=1, xmin=0), fill=col_browser) +
coord_polar(theta = 'y') +
theme(axis.ticks = element_blank(),
axis.title = element_blank(),
axis.text = element_blank(),
panel.grid = element_blank(),
panel.background = element_blank()) +
geom_label_repel(aes(x = x.breaks, y = y.breaks, label = sprintf("%s: %s%%",data$version, data$share)),
force = force,
nudge_x = nudge_x,
nudge_y = nudge_y)
}
cols <- c('cyan2','red','orange','green','dodgerblue2')
pdf('~/Downloads/donuts.pdf', width = 10, height = 10, bg = "snow")
par(omi = c(0.5,0.5,0.75,0.5), mai = c(0.1,0.1,0.1,0.1), las = 1)
print(piedonut(data = browsers, cols = cols, force = 80, nudge_x = 3, nudge_y = 10))
dev.off()
Edit 2
My original answer is really dumb. Here is a much shorter version which does most of the work with a much simpler interface.
#' x numeric vector for each slice
#' group vector identifying the group for each slice
#' labels vector of labels for individual slices
#' col colors for each group
#' radius radius for inner and outer pie (usually in [0,1])
donuts <- function(x, group = 1, labels = NA, col = NULL, radius = c(.7, 1)) {
group <- rep_len(group, length(x))
ug <- unique(group)
tbl <- table(group)[order(ug)]
col <- if (is.null(col))
seq_along(ug) else rep_len(col, length(ug))
col.main <- Map(rep, col[seq_along(tbl)], tbl)
col.sub <- lapply(col.main, function(x) {
al <- head(seq(0, 1, length.out = length(x) + 2L)[-1L], -1L)
Vectorize(adjustcolor)(x, alpha.f = al)
})
plot.new()
par(new = TRUE)
pie(x, border = NA, radius = radius[2L],
col = unlist(col.sub), labels = labels)
par(new = TRUE)
pie(x, border = NA, radius = radius[1L],
col = unlist(col.main), labels = NA)
}
par(mfrow = c(1,2), mar = c(0,4,0,4))
with(browsers,
donuts(share, browser, sprintf('%s: %s%%', version, share),
col = c('cyan2','red','orange','green','dodgerblue2'))
)
with(mtcars,
donuts(mpg, interaction(gear, cyl), rownames(mtcars))
)
Original post
You guys don't have givemedonutsorgivemedeath
function? Base graphics are always the way to go for very detailed things like this. Couldn't think of an elegant way to plot the center pie labels, though.
givemedonutsorgivemedeath('~/desktop/donuts.pdf')
Gives me
Note that in ?pie
you see
Pie charts are a very bad way of displaying information.
code:
browsers <- structure(list(browser = structure(c(3L, 3L, 3L, 3L, 2L, 2L,
2L, 1L, 5L, 5L, 4L), .Label = c("Chrome", "Firefox", "MSIE",
"Opera", "Safari"), class = "factor"), version = structure(c(5L,
6L, 7L, 8L, 2L, 3L, 4L, 1L, 10L, 11L, 9L), .Label = c("Chrome 10.0",
"Firefox 3.5", "Firefox 3.6", "Firefox 4.0", "MSIE 6.0", "MSIE 7.0",
"MSIE 8.0", "MSIE 9.0", "Opera 11.x", "Safari 4.0", "Safari 5.0"),
class = "factor"), share = c(10.85, 7.35, 33.06, 2.81, 1.58,
13.12, 5.43, 9.91, 1.42, 4.55, 1.65), ymax = c(10.85, 18.2, 51.26,
54.07, 55.65, 68.77, 74.2, 84.11, 85.53, 90.08, 91.73), ymin = c(0,
10.85, 18.2, 51.26, 54.07, 55.65, 68.77, 74.2, 84.11, 85.53,
90.08)), .Names = c("browser", "version", "share", "ymax", "ymin"),
row.names = c(NA, -11L), class = "data.frame")
browsers$total <- with(browsers, ave(share, browser, FUN = sum))
givemedonutsorgivemedeath <- function(file, width = 15, height = 11) {
## house keeping
if (missing(file)) file <- getwd()
plot.new(); op <- par(no.readonly = TRUE); on.exit(par(op))
pdf(file, width = width, height = height, bg = 'snow')
## useful values and colors to work with
## each group will have a specific color
## each subgroup will have a specific shade of that color
nr <- nrow(browsers)
width <- max(sqrt(browsers$share)) / 0.8
tbl <- with(browsers, table(browser)[order(unique(browser))])
cols <- c('cyan2','red','orange','green','dodgerblue2')
cols <- unlist(Map(rep, cols, tbl))
## loop creates pie slices
plot.new()
par(omi = c(0.5,0.5,0.75,0.5), mai = c(0.1,0.1,0.1,0.1), las = 1)
for (i in 1:nr) {
par(new = TRUE)
## create color/shades
rgb <- col2rgb(cols[i])
f0 <- rep(NA, nr)
f0[i] <- rgb(rgb[1], rgb[2], rgb[3], 190 / sequence(tbl)[i], maxColorValue = 255)
## stick labels on the outermost section
lab <- with(browsers, sprintf('%s: %s', version, share))
if (with(browsers, share[i] == max(share))) {
lab0 <- lab
} else lab0 <- NA
## plot the outside pie and shades of subgroups
pie(browsers$share, border = NA, radius = 5 / width, col = f0,
labels = lab0, cex = 1.8)
## repeat above for the main groups
par(new = TRUE)
rgb <- col2rgb(cols[i])
f0[i] <- rgb(rgb[1], rgb[2], rgb[3], maxColorValue = 255)
pie(browsers$share, border = NA, radius = 4 / width, col = f0, labels = NA)
}
## extra labels on graph
## center labels, guess and check?
text(x = c(-.05, -.05, 0.15, .25, .3), y = c(.08, -.12, -.15, -.08, -.02),
labels = unique(browsers$browser), col = 'white', cex = 1.2)
mtext('Browser market share, April 2011', side = 3, line = -1, adj = 0,
cex = 3.5, outer = TRUE)
mtext('stackoverflow.com:::maryam', side = 3, line = -3.6, adj = 0,
cex = 1.75, outer = TRUE, font = 3)
mtext('/questions/26748069/ggplot2-pie-and-donut-chart-on-same-plot',
side = 1, line = 0, adj = 1.0, cex = 1.2, outer = TRUE, font = 3)
dev.off()
}
givemedonutsorgivemedeath('~/desktop/donuts.pdf')
Edit 1
width <- 5
tbl <- table(browsers$browser)[order(unique(browsers$browser))]
col.main <- Map(rep, seq_along(tbl), tbl)
col.sub <- lapply(col.main, function(x)
Vectorize(adjustcolor)(x, alpha.f = seq_along(x) / length(x)))
plot.new()
par(new = TRUE)
pie(browsers$share, border = NA, radius = 5 / width,
col = unlist(col.sub), labels = browsers$version)
par(new = TRUE)
pie(browsers$share, border = NA, radius = 4 / width,
col = unlist(col.main), labels = NA)
I used floating.pie instead of ggplot2 to create two overlapping pie charts:
library(plotrix)
# browser data without "ymax" and "ymin"
browsers <-
structure(
list(
browser = structure(
c(3L, 3L, 3L, 3L, 2L, 2L,
2L, 1L, 5L, 5L, 4L),
.Label = c("Chrome", "Firefox", "MSIE",
"Opera", "Safari"),
class = "factor"
),
version = structure(
c(5L,
6L, 7L, 8L, 2L, 3L, 4L, 1L, 10L, 11L, 9L),
.Label = c(
"Chrome 10.0",
"Firefox 3.5",
"Firefox 3.6",
"Firefox 4.0",
"MSIE 6.0",
"MSIE 7.0",
"MSIE 8.0",
"MSIE 9.0",
"Opera 11.x",
"Safari 4.0",
"Safari 5.0"
),
class = "factor"
),
share = c(10.85, 7.35, 33.06, 2.81, 1.58,
13.12, 5.43, 9.91, 1.42, 4.55, 1.65)
),
.Names = c("parent", "node", "size")
,
row.names = c(NA,-11L),
class = "data.frame"
)
# aggregate data for the browser pie chart
browser_data <-
aggregate(browsers$share,
by = list(browser = browsers$browser),
FUN = sum)
# order version data by browser so it will line up with browser pie chart
version_data <- browsers[order(browsers$browser), ]
browser_colors <- c('#85EA72', '#3B3B3F', '#71ACE9', '#747AE6', '#F69852')
# adjust these as desired (currently colors all versions the same as browser)
version_colors <-
c(
'#85EA72',
'#3B3B3F',
'#3B3B3F',
'#3B3B3F',
'#71ACE9',
'#71ACE9',
'#71ACE9',
'#71ACE9',
'#747AE6',
'#F69852',
'#F69852'
)
# format labels to display version and % market share
version_labels <- paste(version_data$version, ": ", version_data$share, "%", sep = "")
# coordinates for the center of the chart
center_x <- 0.5
center_y <- 0.5
plot.new()
# draw version pie chart first
version_chart <-
floating.pie(
xpos = center_x,
ypos = center_y,
x = version_data$share,
radius = 0.35,
border = "white",
col = version_colors
)
# add labels for version pie chart
pie.labels(
x = center_x,
y = center_y,
angles = version_chart,
labels = version_labels,
radius = 0.38,
bg = NULL,
cex = 0.8,
font = 2,
col = "gray40"
)
# overlay browser pie chart
browser_chart <-
floating.pie(
xpos = center_x,
ypos = center_y,
x = browser_data$x,
radius = 0.25,
border = "white",
col = browser_colors
)
# add labels for browser pie chart
pie.labels(
x = center_x,
y = center_y,
angles = browser_chart,
labels = browser_data$browser,
radius = 0.125,
bg = NULL,
cex = 0.8,
font = 2,
col = "white"
)
I created a general purpose donuts plot function to do this, which could
panel
and colorize each circular sector by given percentage pctr
and colors
cols. The ring width could be tuned by outradius
>radius
>innerradius
. The main function actually draw a bar chart and bend it into a ring, hence it is something between a pie chart and a bar chart.
Example Pie Chart, two rings:
Browser Pie Chart
donuts_plot <- function(
panel = runif(3), # counts
pctr = c(.5,.2,.9), # percentage in count
legend.label='',
cols = c('chartreuse', 'chocolate','deepskyblue'), # colors
outradius = 1, # outter radius
radius = .7, # 1-width of the donus
add = F,
innerradius = .5, # innerradius, if innerradius==innerradius then no suggest line
legend = F,
pilabels=F,
legend_offset=.25, # non-negative number, legend right position control
borderlit=c(T,F,T,T)
){
par(new=add)
if(sum(legend.label=='')>=1) legend.label=paste("Series",1:length(pctr))
if(pilabels){
pie(panel, col=cols,border = borderlit[1],labels = legend.label,radius = outradius)
}
panel = panel/sum(panel)
pctr2= panel*(1 - pctr)
pctr3 = c(pctr,pctr)
pctr_indx=2*(1:length(pctr))
pctr3[pctr_indx]=pctr2
pctr3[-pctr_indx]=panel*pctr
cols_fill = c(cols,cols)
cols_fill[pctr_indx]='white'
cols_fill[-pctr_indx]=cols
par(new=TRUE)
pie(pctr3, col=cols_fill,border = borderlit[2],labels = '',radius = outradius)
par(new=TRUE)
pie(panel, col='white',border = borderlit[3],labels = '',radius = radius)
par(new=TRUE)
pie(1, col='white',border = borderlit[4],labels = '',radius = innerradius)
if(legend){
# par(mar=c(5.2, 4.1, 4.1, 8.2), xpd=TRUE)
legend("topright",inset=c(-legend_offset,0),legend=legend.label, pch=rep(15,'.',length(pctr)),
col=cols,bty='n')
}
par(new=FALSE)
}
## col- > subcor(change hue/alpha)
subcolors <- function(.dta,main,mainCol){
tmp_dta = cbind(.dta,1,'col')
tmp1 = unique(.dta[[main]])
for (i in 1:length(tmp1)){
tmp_dta$"col"[.dta[[main]] == tmp1[i]] = mainCol[i]
}
u <- unlist(by(tmp_dta$"1",tmp_dta[[main]],cumsum))
n <- dim(.dta)[1]
subcol=rep(rgb(0,0,0),n);
for(i in 1:n){
t1 = col2rgb(tmp_dta$col[i])/256
subcol[i]=rgb(t1[1],t1[2],t1[3],1/(1+u[i]))
}
return(subcol);
}
### Then get the plot is fairly easy:
# INPUT data
browsers <- structure(list(browser = structure(c(3L, 3L, 3L, 3L, 2L, 2L,
2L, 1L, 5L, 5L, 4L),
.Label = c("Chrome", "Firefox", "MSIE","Opera", "Safari"),class = "factor"),
version = structure(c(5L,6L, 7L, 8L, 2L, 3L, 4L, 1L, 10L, 11L, 9L),
.Label = c("Chrome 10.0", "Firefox 3.5", "Firefox 3.6", "Firefox 4.0", "MSIE 6.0",
"MSIE 7.0","MSIE 8.0", "MSIE 9.0", "Opera 11.x", "Safari 4.0", "Safari 5.0"),
class = "factor"),
share = c(10.85, 7.35, 33.06, 2.81, 1.58,13.12, 5.43, 9.91, 1.42, 4.55, 1.65),
ymax = c(10.85, 18.2, 51.26,54.07, 55.65, 68.77, 74.2, 84.11, 85.53, 90.08, 91.73),
ymin = c(0,10.85, 18.2, 51.26, 54.07, 55.65, 68.77, 74.2, 84.11, 85.53,90.08)),
.Names = c("browser", "version", "share", "ymax", "ymin"),
row.names = c(NA, -11L), class = "data.frame")
## data clean
browsers=browsers[order(browsers$browser,browsers$share),]
arr=aggregate(share~browser,browsers,sum)
### choose your cols
mainCol = c('chartreuse3', 'chocolate3','deepskyblue3','gold3','deeppink3')
donuts_plot(browsers$share,rep(1,11),browsers$version,
cols=subcolors(browsers,"browser",mainCol),
legend=F,pilabels = T,borderlit = rep(F,4) )
donuts_plot(arr$share,rep(1,5),arr$browser,
cols=mainCol,pilabels=F,legend=T,legend_offset=-.02,
outradius = .71,radius = .0,innerradius=.0,add=T,
borderlit = rep(F,4) )
###end of line