Tim,
Boxplots are nice, but I find that they can be somewhat
misleading when applied to small groups, especially in the
suggestion of spread differences. (Your real data may well
be more extensive than the ToothGrowth data and so the
following may be moot.)
I prefer stripcharts for small groups. Compare the
boxplot figure to the following (using Dennis's code
as basis):
stripchart(len ~ dose, data = subset(ToothGrowth, supp == 'VC'),
method = "jitter", jitter = .03,
vertical = TRUE, at = 1:3 - 0.1,
pch = 21, bg = "yellow", cex = 1.5,
main = "Guinea Pigs' Tooth Growth",
axes = FALSE,
xlab = "Vitamin C dose mg",
ylab = "Tooth length",
xlim = c(0.5, 3.5), ylim = c(0, 35), yaxs = "i")
stripchart(len ~ dose, data = subset(ToothGrowth, supp == 'OJ'),
add = TRUE,
method = "jitter", jitter = .03,
vertical = TRUE, at = 1:3 + 0.1,
axes = FALSE,
pch = 21, bg = "orange", cex = 1.5)
axis(1, at = 1:3, labels = c(0.5, 1, 2))
axis(2)
box()
legend(2, 9, c("Ascorbic acid", "Orange juice"),
pch = 21, pt.bg = c("yellow", "orange"), pt.cex = 1.5)
-Peter Ehlers
On 2010-09-26 4:48, Dennis Murphy wrote:
Hi:
Here are a couple of ways:
(1) Base graphics: add argument axes = FALSE to both plots, then add axes
boxplot(len ~ dose, data = subset(ToothGrowth, supp == 'VC'),
boxwex = 0.25, at = 1:3 - 0.2,
col = "yellow",
main = "Guinea Pigs' Tooth Growth",
axes = FALSE,
xlab = "Vitamin C dose mg",
ylab = "Tooth length",
xlim = c(0.5, 3.5), ylim = c(0, 35), yaxs = "i")
boxplot(len ~ dose, data = subset(ToothGrowth, supp == 'OJ'),
add = TRUE,
boxwex = 0.25, at = 1:3 + 0.2,
axes = FALSE,
col = "orange")
axis(1, at = 1:3, labels = c(0.5, 1, 2))
axis(2)
legend(2, 9, c("Ascorbic acid", "Orange juice"),
fill = c("yellow", "orange"))
box()
(2) ggplot2 (just for fun :)
library(ggplot2)
g<- ggplot(ToothGrowth, aes(x = factor(dose), y = len))
g + geom_boxplot(aes(fill = supp), position = position_dodge(width = 0.9)) +
scale_fill_manual('', breaks = c('OJ', 'VC'),
values = c('orange', 'yellow'),
labels = c('Orange juice', 'Ascorbic acid')) +
theme_bw() +
opts(legend.position = c(0.8, 0.2), legend.text = theme_text(size = 12))
+
opts(panel.grid.major = theme_blank()) +
xlab('Vitamin C dose (mg)') + ylab('Tooth length') +
opts(title = "Guinea pigs' tooth growth") +
opts(plot.title = theme_text(size = 16))
HTH,
Dennis
On Sat, Sep 25, 2010 at 11:13 PM, Tim Clark<mudiver1...@yahoo.com> wrote:
Dear List,
I am creating a boxplot with two subsets, very similar to the example by
Roger
Bivand at ?boxplot (reproduced below). I am trying to change the labels on
the
x-axis to have one number to cover both subsets. I can do this in other
plots
by using axis=FALSE followed by a separate axis() command. I have also
tried
variations in the names= argument but can't get it to work. Ideally I
would
like tickmarks on either side of each factor with the number for the level
centered between the two tick marks. Any suggestions?
Thanks,
Tim
Example:
boxplot(len ~ dose, data = ToothGrowth,
boxwex = 0.25, at = 1:3 - 0.2,
subset = supp == "VC", col = "yellow",
main = "Guinea Pigs' Tooth Growth",
xlab = "Vitamin C dose mg",
ylab = "tooth length",
xlim = c(0.5, 3.5), ylim = c(0, 35), yaxs = "i")
boxplot(len ~ dose, data = ToothGrowth, add = TRUE,
boxwex = 0.25, at = 1:3 + 0.2,
subset = supp == "OJ", col = "orange")
legend(2, 9, c("Ascorbic acid", "Orange juice"),
fill = c("yellow", "orange"))
Tim Clark
Marine Ecologist
National Park of American Samoa
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
