[R] Lattice: Customizing point-sizes with groups

[Paul C. Boutros]

Yup, that would be my work-around.

I was hoping for a cleaner way of doing this, though, because I am
calculating cex based on other properties of the data-points, so that it
becomes a continuous variable.

-----Original Message-----
From: Sundar Dorai-Raj [mailto:sdorairaj at gmail.com] 
Sent: Tuesday, March 10, 2009 12:01 PM
To: Paul C. Boutros
Cc: r-help at r-project.org
Subject: Re: [R] Lattice: Customizing point-sizes with groups

Sorry, I missed your point the first time. Why not create a group for
each subset then?

xyplot(y ~ x, temp, groups = interaction(cex, groups),
       par.settings = list(
         superpose.symbol = list(
           cex = c(1, 2, 3, 4),
           pch = 19,
           col = c("blue", "red", "green", "purple"))))


On Tue, Mar 10, 2009 at 8:11 AM, Paul C. Boutros
<Paul.Boutros at utoronto.ca> wrote:
> Hi Sundar,
>
> Thanks for your help!  Unfortunately your code seems to give the same
> result.  Compare this:
>
> temp <- data.frame(
>       x = 1:10,
>       y = 1:10,
>       cex = rep( c(1,3), 5),
>       col = c( rep("blue", 5), rep("red", 5) ),
>       groups = c( rep("A", 5), rep("B", 5) )
>       );
>
> xyplot(y ~ x, temp, groups = groups,
>       par.settings = list(
>         superpose.symbol = list(
>           cex = c(1, 3),
>           pch = 19,
>           col = c("blue", "red"))))
>
> And this:
> xyplot(y ~ x, temp, cex = temp$cex, col = temp$col, pch = 19);
>
> Once I introduce groups, I lose the ability to customize individual
> data-points and seem only to be able to customize entire groups.
>
> Paul
>
> -----Original Message-----
> From: Sundar Dorai-Raj [mailto:sdorairaj at gmail.com]
> Sent: Tuesday, March 10, 2009 5:49 AM
> To: Paul.Boutros at utoronto.ca
> Cc: r-help at r-project.org
> Subject: Re: [R] Lattice: Customizing point-sizes with groups
>
> Try this:
>
> xyplot(y ~ x, temp, groups = groups,
>       par.settings = list(
>         superpose.symbol = list(
>           cex = c(1, 3),
>           pch = 19,
>           col = c("blue", "red"))))
>
> See:
>
> str(trellis.par.get())
>
> for other settings you might want to change.
>
> Also, you should drop the ";" from all your scripts.
>
> HTH,
>
> --sundar
>
> On Mon, Mar 9, 2009 at 6:49 PM, Paul Boutros <paul.boutros at utoronto.ca>
> wrote:
>> Hello,
>>
>> I am creating a scatter-plot in lattice, and I would like to customize
the
>> size of each point so that some points are larger and others smaller.
>>  Here's a toy example:
>>
>> library(lattice);
>>
>> temp <- data.frame(
>>        x = 1:10,
>>        y = 1:10,
>>        cex = rep( c(1,3), 5),
>>        groups = c( rep("A", 5), rep("B", 5) )
>>        );
>>
>> xyplot(y ~ x, temp, cex = temp$cex, pch = 19);
>>
>> This works just fine if I create a straight xy-plot, without groups.
>>  However when I introduce groupings the cex argument specifies the
>> point-size for the entire group.  For example:
>>
>> xyplot(y ~ x, temp, cex = temp$cex, pch = 19, group = groups);
>>
>> Is it possible to combine per-spot sizing with groups in some way?  One
>> work-around is to manually specify all graphical parameters, but I
thought
>> there might be a better way than this:
>>
>> temp$col <- rep("blue", 10);
>> temp$col[temp$groups == "B"] <- "red";
>> xyplot(y ~ x, temp, cex = temp$cex, pch = 19, col = temp$col);
>>
>> Any suggestions/advice is much appreciated!
>> Paul
>>
>> ______________________________________________
>> R-help at 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.
>>
>
>