ylim-question

[july]

Hi there,

I've created a plot with the following commands:

p <- ggplot(plotData, aes(x = variable, y = value, colour = catVar,
group = catVar))
p <- p + stat_summary(fun.y = mean, geom = "line")
p <- p + stat_summary(fun.ymax = errorUpper, fun.ymin = errorLower,
geom = "errorbar", width=0.25)

Even if I print the plot without errorbars, the y-limit is too high.
It looks like it takes the raw data to calculate the y-limit instead
of the summarized data. Is it true? Is there any other solution beside
changing the viewport?
Or is there any mistake I'm not aware of?

[Brandon Hurr]

You could set your limits manually using coord_cartesian. http://had.co.nz/ggplot2/coord_cartesian.html

p + coord_cartesian(ylim = c(325, 500), wise=TRUE) 

I'm not sure if it is taking into account the data range or the range of what is being plotted though. 

B

--
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

[july]

Okay, I was not yet aware of this option (I reached this chapter of
the book today :-)).
Anyhow, it's a bit annoying to group the data and calculate the ranges
of the summarized values by myself. This is exactly what ggplot tries
to avoid, no?

A



On Jan 17, 10:13 am, Brandon Hurr <brandon.h...@gmail.com> wrote:
> You could set your limits manually using coord_cartesian.http://had.co.nz/ggplot2/coord_cartesian.html

>
> p + coord_cartesian(ylim = c(325, 500), wise=TRUE)
>
> I'm not sure if it is taking into account the data range or the range of
> what is being plotted though.
>
> B
>

> On Mon, Jan 17, 2011 at 08:34, july <antje.niederl...@googlemail.com> wrote:
> > Hi there,
>
> > I've created a plot with the following commands:
>
> > p <- ggplot(plotData, aes(x = variable, y = value, colour = catVar,
> > group = catVar))
> > p <- p + stat_summary(fun.y = mean, geom = "line")
> > p <- p + stat_summary(fun.ymax = errorUpper, fun.ymin = errorLower,
> > geom = "errorbar", width=0.25)
>
> > Even if I print the plot without errorbars, the y-limit is too high.
> > It looks like it takes the raw data to calculate the y-limit instead
> > of the summarized data. Is it true? Is there any other solution beside
> > changing the viewport?
> > Or is there any mistake I'm not aware of?
>
> > --
> > 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<ggplot2%2Bunsu...@googlegroups.com>
> > More options:http://groups.google.com/group/ggplot2

[Hadley Wickham]

This should be fixed in the next version (hopefully)
Hadley

> More options: http://groups.google.com/group/ggplot2
>

--
Assistant Professor / Dobelman Family Junior Chair
Department of Statistics / Rice University
http://had.co.nz/