Re: [R] heatmap color distribution

From: Wiener, Matthew <>
Date: Fri 22 Jul 2005 - 01:01:28 EST

Breaks affects the binning into colors. Try this. Assume that temp is one of your data sets. It's values are restricted to 0.25 - 0.75, and we'll assume that the full data set goes from 0 to 1.

> temp <- matrix(runif(60, 0.25, 0.75), nc = 6)
> breaks <- seq(from = 0, to = 1, length = 11)
> image(temp2, col = heat.colors(10)) # full range of
> image(temp2, col = heat.colors(10), breaks = breaks) # muted colors

The second image is told about all the colors, and about the full range of data through breaks, and only uses the colors in the middle.

Is that what you mean?

HTH, Matt

-----Original Message-----
From: Jake Michaelson [] Sent: Thursday, July 21, 2005 10:45 AM
To: Wiener, Matthew
Subject: Re: [R] heatmap color distribution

Thanks for the reply. As I understand it, "breaks" only controls the binning. The problem I'm having is that each subset heatmap has slightly different min and max log2 intensities. I'd like the colors to be based on the overall (complete set) max and min, not the subsets' max and min -- I could be wrong, but I don't think "breaks" will help me there. And you're right - this might obscure some of the trends/features, but we'll also plot the "default" heatmaps.

Also (I should have specified) I'm using heatmap.2.



On Jul 21, 2005, at 8:09 AM, Wiener, Matthew wrote:

> You can use the "breaks" argument in image to do this. (You don't
> specify a
> function you're using, but other heatmap functions probably have a
> similar
> parameter.) Look across all your data, figure out the ranges you want
> to
> have different colors, and specify the appropriate break points in
> each call
> to image. Then you're using the same color set in each one. You run
> the
> risk, of course, that some of your images will have a very narrow color
> range, which might obscure interesting features. But nothing stops
> you from
> making more than one plot.
> Hope this helps.
> Regards,
> Matt Wiener
> -----Original Message-----
> From:
> [] On Behalf Of Jacob Michaelson
> Sent: Thursday, July 21, 2005 9:26 AM
> To:
> Subject: [R] heatmap color distribution
> Hi all,
> I've got a set of gene expression data, and I'm plotting several
> heatmaps for subsets of the whole set. I'd like the heatmaps to have
> the same color distribution, so that comparisons may be made
> (roughly) across heatmaps; this would require that the color
> distribution and distance functions be based on the entire dataset,
> rather than on individual subsets. Does anyone know how to do this?
> Thanks in advance,
> Jake
> ______________________________________________
> mailing list
> PLEASE do read the posting guide!
> -----------------------------------------------------------------------
> -------
> Notice: This e-mail message, together with any attachment...{{dropped}} mailing list PLEASE do read the posting guide! Received on Fri Jul 22 01:38:35 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:33:55 EST