On 03/11/2015 8:58 AM, GINGINS Simon wrote:
Hi,
I am currently building graphs using dochart(). I plotted the points in two
different colors according to specific criteria from my dataset. Later, I
decided to also give them different symbols, for easier reading if printed. I
got quite a surprise when I realized that, even though I gave the same variable
to both color= and pch= , I got a mix of the colors and symbols. I wanted one
type of symbol to be only one color, and the second only the other color, and
it is clearly not what happened. From my data, I would say that the pch=
argument is not dealing properly with the variable I gave it. I tried using a
dataset from R to see if this was only in my dataset, and it did the same.
here’s a reproducible example, using beaver1:
dotchart(beaver1$temp, groups=factor(beaver1$day),
color=as.factor(beaver1$activ), pch=beaver1$activ)
Does anyone know why, with the same variable, the arguments color & pch give
different results?
That would be of great help, since now I am no longer sure which one isthe
correct representation of my data.
Best regards & many thanks for the help
Looks like a bug in dotchart. When you specify groups, it sorts the
data to put it into groups. It also sorts the colors, but it doesn't
sort the pch values, so they end up associated with the wrong points.
A workaround would be for you to do the sorting in advance. For example,
groups <- factor(beaver1$day)
o <- sort.list(as.numeric(groups), decreasing = TRUE)
tmp <- beaver1[o,]
dotchart(tmp$temp, groups=factor(tmp$day), color=as.factor(tmp$activ),
pch=tmp$activ)
Duncan Murdoch
______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
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.
