Re: [Matplotlib-users] colormap: values to colors

[Benjamin R. <ben...@ou...>]

On Fri, Sep 17, 2010 at 8:59 AM, Martinho MA <mm...@ua...> wrote:

> Hello
>
> I am trying to calculate the facecolors for the Axes3D.plot_surface
> function, to have something similar to matlab surf 4th argument!!
> Something like plot_surface(x,y,z,facecolors=calc_colors(v))
>
> So, I need to obtain the colors corresponding to the values of v! I
> created the function calc_colors:
>
>
> import pylab
> import numpy as np
> from matplotlib.colors import rgb2hex
>
> def calc_colors(var):
>  colors = np.empty(var.shape, dtype='S7')
>  mm=pylab.cm.ScalarMappable()
>  mm.set_clim((var.min(),var.max()))
>  mm.set_cmap(pylab.cm.jet)
>  colors.flat= [rgb2hex(mm.to_rgba(k)[:-1]) for k in var.flat]
>  return colors
>
>
> Well I believe this is not the smartest way to calculate the colors !!
> At least it is quite slow !
> Can someone give me a better solution for this problem?
>
> than you
> mma
>
>
I might be wrong, but can't mm.to_rgba() take var.flat as an input, thereby
giving you an array of colors as output?  Also, I don't think it is
necessary to call rgb2hex() because matplotlib will just convert that hex
back into rgba.  I believe you should be able to just return the result of
mm.to_rgba(var.flat).

I hope that helps,
Ben Root