[Matplotlib-users] non-linear color scale: best approach?

[Gerald J. M. M. <Ger...@jp...>]

Hi,

The attached png (generated from GMT) (also visible here:
http://sayatnova.jpl.nasa.gov/genesis/swhk_jmap_145.png )
shows a non-linear color scale which I am trying to replicate using
matplotlib.  Could anyone point me in the right direction as to the best
approach to do this?  I've seen posts where I can subclass the
colormap/normalize classes and customize them.  I've also seen
James Boyle's gmtColormap.py script to load GMT colormaps into
matplotlib colormaps however I would like to achieve a general
solution to this.  The following code is the generalized function
I use to generate maps:

def plotMap(pngFile,figsize,aspect,plotTitle,colorbarTitle,
             mapData,dataMin,dataMax,lonMin,lonMax,latMin,latMax,
             colorbarFormat='%1.1f',colorbarOrientation='horizontal',
             nanColor='k'):
     """Plot data on world map."""

     m=Basemap(projection='cyl',llcrnrlon=lonMin,llcrnrlat=latMin,
               urcrnrlon=lonMax,urcrnrlat=latMax,suppress_ticks=False)
     m.aspect=aspect
     palette=cm.jet
     palette.set_bad(nanColor,1.0)
     fig=m.createfigure(figsize=figsize)
     im=m.imshow(mapData,cmap=palette,norm=colors.normalize(vmin=dataMin,
                 vmax=dataMax,clip =False))
     cax=fig.colorbar(im,tickfmt=colorbarFormat,
                      orientation=colorbarOrientation)
     cax.set_xlabel(colorbarTitle)
     m.fillcontinents()
     m.drawmapboundary()
     gca().set_title(plotTitle)
     fig.savefig(pngFile)
     close(fig)

Any help is greatly appreciated.

Thanks,

Gerald