I'm trying to overlay two sets of arrays with values between 0,1 with matplotlib pcolormesh. This is no problem doing:
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import matplotlib.colors as colors
%matplotlib inline
array1 = np.random.random([20,20])
array2 = np.random.random([20,20])
fig, ax = plt.subplots()
heatmap = ax.pcolormesh(array1, cmap=plt.cm.Reds, alpha=0.5)
heatmap2 = ax.pcolormesh(array2, cmap=plt.cm.Blues, alpha=0.5)
The result is correctly blended, except that the dynamic range is split in two (we cannot get full red or full blue values when one is high (close to 1) and the other is 0, as the alpha blending makes everything a little more dull).
What i'm after is similar to a multiplication of the color: values close to [1, 0] should be full red, values close to [0, 1] should be full blue, values close to [1, 1] full magenta and values close to [0, 0] white.
Is is possible to have such a 2D colormap?
Right now I can do it in Photoshop 'multiply' blend, but I need a Python solution