Color map that more closely matches the NWS mapping about metpy

[w142236]

I know this was posted a long time ago, but
ctable.get_colortable()
is no longer a function of ctable for the latest version. I got it to work by instead using:
cmap = ctable.registry.get_with_steps('NWSReflectivity',5,5)
I found this from here:

[dopplershift]

@w142236 Yes, the correct call for get_colortable should be:

from metpy.plots import ctables
ctables.registry.get_colortable('NWSReflectivity')

If you find this doesn't work for you, please open a new issue and we'll try to get you sorted out.

[shoyer]

@rankinstudio matplotlib has the gist_ncar colormap built in:

[dopplershift]

The metpy colormap for NWS Reflectivity was "painstakingly" :) hand-pulled from the colors on the NWS website long ago, they're not the problem. What you need is what matplotlib calls a "norm", or an instance of the Normalize class.

Matplotlib's colormaps handle mapping numbers in the range [0, 1] to colors. The norm handles mapping your actual range of colors to the range [0, 1]. If you don't explicitly pass a norm to something like pcolor, using the norm keyword argument, matplotlib creates one based on the range of data you're plotting; this is great for maximizing the dynamic range, bad for matching familiar plots. You probably want something like:

from matplotlib.colors import Normalize
import matplotlib.pyplot as plt
from metpy.plots import ctables
# READ AND PARSE HERE
cmap = ctables.get_colortable('NWSReflectivity')
plt.pcolor(x, y, data, cmap=cmap, norm=Normalize(-25, 75))

I'm not sure if I got the range exactly right, but I think that should get you started.

[rankinstudio]

Thanks!

Looks like -1, 80 put me in the right ballpark. Figured it was a usability thing. Appreciate the support.

Cheers,

[dopplershift]

No problem. Glad you're finding this useful.

[w142236]

@w142236 Yes, the correct call for get_colortable should be:

from metpy.plots import ctables
ctables.registry.get_colortable('NWSReflectivity')

If you find this doesn't work for you, please open a new issue and we'll try to get you sorted out.

Nah it's fine. I think the creator of metpy might have moved things around and changed some of the function names. That isn't too difficult for people with a good IDE like Spyder, but I have to run my code from a command line so I have to look a lot of this stuff up.