geom_boxplot box width & alignment

[Jokel Meyer]

Dear ggplot2 experts,

I am new to this package and would have a question that I failed to
solve now for quiet some time:

I am plotting a boxplot (continues variable on y-axis) and the boxes
group by two grouping variables (one with 2 levels, one with 3
levels). Using geom_boxplot brings me quiet nice results. However
boxes are aligned in a way that they are directly next to each other.
How could I control the box alignment in a way that there is a little
gap between the boxes?

Also I would like to add a little star which would indicate
significant differences between groups as it is sometimes used in
boxplots. Which would be the easiest way to get this?


Many thanks for your help!
Jokel

[James Howison]

Minimal example please. It's just much easier to help if there is code to copy into R.

> --
> You received this message because you are subscribed to the ggplot2 mailing list.
> Please provide a reproducible example: http://gist.github.com/270442
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2

[Jokel Meyer]

Dear ggplot2 experts,

I am new to this package and would have a question that I failed to
solve now for quiet some time:

I am plotting a boxplot (continues variable on y-axis) and the boxes
group by two grouping variables (one with 2 levels, one with 3
levels). Using geom_boxplot brings me quiet nice results. However
boxes are aligned in a way that they are directly next to each other.
How could I control the box alignment in a way that there is a little
gap between the boxes?

R-code:

library(ggplot2)

ass<-structure(list(assort = c(0.1994, 0.3696, 0.4056, 0.3933, 0.4322, 

0.4192, 0.1923, 0.3957, 0.2347, 0.3576, 0.2381, 0.143, 0.2908

), task = c(0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1), diag = c(0, 

0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2)), .Names = c("assort", "task", 

"diag"), class = "data.frame", row.names = c(NA, -13L))

p<-ggplot(ass, aes(factor(diag),assort)) 

p + geom_boxplot(aes(fill = factor(task)))

[Kohske Takahashi]

Hi, here is an example:

ggplot(ass, aes(factor(diag),assort)) +
geom_boxplot(aes(fill = factor(task)), position=position_dodge(width=0.9)) +
geom_text(aes(x,y,label=v), data.frame(x=factor(0:2),
y=rep(0.45,3),v=c("*","","*")))

--
Kohske Takahashi <takahash...@gmail.com>

Research Center for Advanced Science and Technology,
The University of  Tokyo, Japan.
http://www.fennel.rcast.u-tokyo.ac.jp/profilee_ktakahashi.html