[BioC] non-symmetric heatmap colors

[Sean Davis]

On Feb 3, 2005, at 4:30 AM, <Arne.Muller at sanofi-aventis.com> wrote:

> Wolfgang and Sean,
>
> thanks for your hints. The "breaks" argument (described in ?image), 
> did what I wanted. I just had to set all fold change values < -3  to 
> -3 and > 3 to 3 to get a saturation at these limits, otherwise the 
> color beyond these "extreme" values would just be white.
>

If you use the heatmap.2 function in the gplots package (part of the 
gregmisc bundle), it deals with the saturation issue automatically and 
assigns all values above (below) the top (bottom) break to the most 
extreme color--no more white!  It also adds some other very nice 
bells-and-whistles to the standard heatmap function.  I now use 
heatmap.2 exclusively.

Sean