[BioC] Color space and pairs with hexbin panels
Nicholas Lewin-Koh
nikko at hailmail.net
Mon Jul 25 19:36:15 CEST 2005
Message: 6
Date: Tue, 19 Jul 2005 10:30:34 -0400
From: "Liaw, Andy" <andy_liaw at merck.com>
Hi,
Been off the list a while so I am late replying to this.
Colorspace: The colorspace package is needed by the current version of
hexbin, but I will
be changing that very soon as all the color functionality I needed is
now in graphics. Specifically
I need it for hexdiff plots with multiple groups to do color
intersections.
pairs Plots: I am working on functions and classes to do this. The big
problem is
a layout function that will set up the aspect ratios properly on the
page. I have written some
functions for lattice that do hexagon binning for the panel functions on
the fly, but it needs some
tweeking.
Nicholas
Subject: [BioC] pairs() with hexbin panels (was RE: hexbin needs
colorspac e - but where is colorspace?)
To: "'Martin Maechler'" <maechler at stat.math.ethz.ch>, "James
MacDonald" <jmacdon at med.umich.edu>
Cc: naomi at stat.psu.edu, bioconductor at stat.math.ethz.ch
Message-ID:
<39B6DDB9048D0F4DAD42CB26AAFF0AFA64EAC9 at usctmx1106.Merck.com>
Content-Type: text/plain
I recently ran into related problems. I was looking for a way to do
pairs()
with hexbin in each panel. I found pairs.CSP() in `rflowcyt', but even
the
example code in ?pairs.CSP failed. (That example code is wrapped in `if
(interactive()) ...', so R CMD check would not have picked it up.) The
culprit is that hexagons() is no where to be found. I suspect this is
caused by hexbin version incompatibility?
Can anyone suggest a way to do what I want (scatterplot matrix with
hexbin
panels)?
Best,
Andy
> From: Martin Maechler
>
> >>>>> "JimMcD" == James MacDonald <jmacdon at med.umich.edu>
> >>>>> on Mon, 18 Jul 2005 15:23:19 -0400 writes:
>
> JimMcD> library(reposTools)
> JimMcD> install.packages2("colorspace")
>
> JimMcD> HTH,
>
> JimMcD> Jim
>
> JimMcD> Also, if you are using the devel version you will
> probably need to be
> JimMcD> running R-2.2.0
>
> but really the grid-based hexbin is in the current Bioconductor
> (1.6), at least for me,
> so I am a little perplexed why Naomi first needs to download it.
>
> I guess, Naomi, you somehow managed to keep an old version of
> hexbin around? It may be because getBioC() has changed and
> doesn't install "hexbin" anymore AFAIK, unless you explicitly
> ask for
> getBioC("hexbin", lib = ...., destdir = ....)
>
> the need of which I had been wondering about as well..
>
> Also --- probably more a question to the
> bioconductor-release-managers :
> if you had used the above getBioC(..) shouldn't you have gotten
> "hexbin" *plus* all the packages it depends on, i.e., in your
> case "colorspace" as well?
> Naomi, can you tell us *how* you've installed hexbin?
>
> Regards,
> Martin Maechler, ETH Zurich
>
> >>>> Naomi Altman <naomi at stat.psu.edu> 7/18/2005 3:15:30 PM >>>
>
> Naomi> I downloaded hexbin from the developer site, since
> the release version
> Naomi> says
> Naomi> to do so.
>
> Naomi> But I then get the following message:
>
> >> library(hexbin)
> Naomi> Loading required package: grid
> Naomi> Loading required package: colorspace
> Naomi> Error: package 'colorspace' could not be loaded
> Naomi> In addition: Warning message:
> Naomi> there is no package called 'colorspace' in:
> library(pkg, character.only
> Naomi> = TRUE, logical = TRUE, lib.loc = lib.loc)
>
> Naomi> I thought this might be a version problem, so I also
> Naomi> reinstalled the most recent version of R - but to no
> Naomi> avail. I have not been able to locate colorspace.
>
> Naomi> As always, any assistance would be most appreciated.
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>
>
>
More information about the Bioconductor
mailing list