sunpy: Unexpected SunPy-rendered RHESSI colormaps

Description

The colormap “rhessi” as in SunPy visualization (blue-green-red) which appears to be the standard matplotlib-jet colormap, is not the same as the “official” mission colormaps, when compare with the ones used on hesperia (black, blue, purple, red, orange, green/yellow, white).

Expected vs Actual behavior

Being able to plot the map with the expected mission colormap.

Steps to Reproduce

from sunpy.io.fits import read
from sunpy.time import parse_time
from sunpy.map import Map

url = "https://hesperia.gsfc.nasa.gov/rhessi_extras/imagecube_fits/2002/07/05/20020705_0801_0815/hsi_imagecube_clean_20020705_0801_14tx2e.fits"
f = read(url)
m = Map(f[0].data[0][0], f[0].header)
m.plot_settings['cmap'] = 'rhessi'
m.plot()

System Details

SunPy Version: 2.0
Astropy Version: 4.0
Python Version: 3.7.3
OS information: Win10

About this issue

Original URL
State: closed
Created 4 years ago
Reactions: 1
Comments: 35 (11 by maintainers)

Most upvoted comments

I’ve made the PR @Laubeee if you want to have a look over it #4665

hayesla on Nov 20, 2020

@hayesla In my case I’d be happy to be able to use the STD GAMMA-II as well, but HESSI could be assigned as default colormap for the RHESSIMap?

Laubeee on Nov 20, 2020

Thanks for raising this issue @Laubeee! 😃

I’d say the hessi one is the best - I have a draft of this @nabobalis if you want me to make the PR?

Go right ahead.

nabobalis on Nov 20, 2020

thanks both

I suggest to use the names Albert used, resp. GSFC uses, so I suggest std_gamma_II, hessi and hessi_black_white (if the last one even makes sense to include? I don’t think it’s really used in practice…)

I have no “hard” preference. You could take this one here, then I can focus on opening another PR I wanted to open for some time in sunkit-instruments 😃

Laubeee on Nov 20, 2020

I will let you take over this issue @nabobalis Thanks!

kakirastern on Nov 20, 2020

RHESSI does have a “mission” colormap; for example, it’s what’s used for the quicklook images in the RHESSI Browser. However, the HESSI color table is actually a little more difficult to load than it used to be. This binary file for loadct contains it as table 41. The HESSI color table is intended for a range that is symmetric about zero (i.e., black is at the middle of the range), whereas STD GAMMA-II is better for a range that is (mostly) all positive (i.e., black is at the start of the range). Here’s a comparison:

HESSI color table

hsi_qlimg_2070502_006012

STD GAMMA-II color table

hsi_image_20020705_080420_6_12kev_contours_12_25kev

That said, matplotlib-jet doesn’t match either of these. It looks like @ehsteve picked that colormap in https://github.com/sunpy/sunpy/commit/07f203a9ad822dbfbb28c49003bc4003427d149a (part of #22) for SunPy 0.2, and it hasn’t been given further consideration since 2011.

ayshih on Oct 27, 2020

I dont know if RHESSI has a ‘mission’ colormap per say (maybe @ayshih can comment) - RHESSI map plotted in the archive are usually plotted with loadct 5 in IDL from here (https://www.harrisgeospatial.com/docs/LoadingDefaultColorTables.html) which is “STD GAMMA-II”.

If you wanted something close - maybe cmap=plt.cm.CMRmap?

hayesla on Oct 27, 2020