使用自定义cmap的matplotlib散点图-颜色不正确

Question

So I'm following a pattern to create a custom color map and cbar for my scatter plot. I'm creating 4 subplots each covering a different range of a parameter that is used to give the dots their color. The values for this parameter range from 1e-10 to 1.0.

I do the following:

cmap = plt.cm.jet
# extract all colors from the .jet map
cmaplist = [cmap(i) for i in range(cmap.N)]
# force the first color entry to be grey
cmaplist[0] = (.5,.5,.5,1.0)
# create the new map
cmap = cmap.from_list('Custom cmap', cmaplist, cmap.N)

# define the bins and normalize
bounds = np.linspace(-10,0,11)
norm   = mpl.colors.BoundaryNorm(bounds, cmap.N)
...
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, sharex='col', sharey='row', figsize=(20,20))
...
ax1.scatter(xcoord[rng1], ycoord[rng1], s=massFilt[rng1]/dotNorm,
                    c=np.ma.log10(Zfilt[rng1]), cmap=cmap)
ax2.scatter(xcoord[rng2],ycoord[rng2], s=massFilt[rng2]/dotNorm,
                    c=np.ma.log10(Zfilt[rng2]), cmap=cmap)
ax3.scatter(xcoord[rng3], ycoord[rng3], s=massFilt[rng3]/dotNorm,
                    c=np.ma.log10(Zfilt[rng3]), cmap=cmap)
ax4.scatter(xcoord[rng4], ycoord[rng4], s=massFilt[rng4]/dotNorm,
                    c=np.ma.log10(Zfilt[rng4]), cmap=cmap)

where the rngs are defined as

rng1 = (Zfilt < 2.e-7)
rng2 = ((Zfilt >= 2.e-7) & (Zfilt < 1.e-4))
rng3 = ((Zfilt >= 1.e-4) & (Zfilt < 1.e-2))
rng4 = (Zfilt >= 1.e-2)

So the first panel includes only those less than 2e-7 and the points are all colored correctly... I've verified that the filter is working correctly: for instance Zfilt[rng4] only contains points where Zfilt > 1e-2 ... But somehow I'm seeing colors in rng2 and rng3 that should only be used for rng1 !!?? See attached.

Any ideas? What am I doing wrong?

Here's a sample of Zfilt[rng2] -

Z values rng2  [-4.23451696 -4.35974369 -5.18479833 -6.17304647 -4.48839191 -5.16774006
 -4.12047222 -6.11491263 -5.81392662 -4.6491248  -4.75038775 -5.06640103
 -4.20821705 -4.12556725 -4.58661378 -4.17023495 -5.40845781 -4.54981553
 -5.82830048 -4.11185471 -4.43155534 -6.17025186 -4.88154584 -5.00024704
 -4.13626926 -5.57797731 -6.0617742  -5.33182163 -5.44963247 -5.88409509
 -6.16903327 -5.70808154 -4.88578943 -4.00873256 -4.1457824  -4.45174817
 -5.43829583 -4.32470978 -4.11634754 -5.1141915  -5.13310282 -4.15469421
...

Answer 1

scatter() has no idea about the supposed range of your data, so it applies the color map for the full range it sees, that is from the minimum to the maximum value of the supplied array. Since you want a different behavior (same range for all the subplots, regardless of the data), you need to specify the range explicitly via vmin and vmax arguments:

ax1.scatter(xcoord[rng1], ycoord[rng1], s=massFilt[rng1]/dotNorm,
                c=np.ma.log10(Zfilt[rng1]), cmap=cmap,
                vmin=-10, vmax=0)
ax2.scatter(xcoord[rng2],ycoord[rng2], s=massFilt[rng2]/dotNorm,
                c=np.ma.log10(Zfilt[rng2]), cmap=cmap,
                vmin=-10, vmax=0)
# etc