How to perform plots compositing / use subplots?

[henczati]

It might be just me, but I am unable to use subplot-like features on colour-generated plots.
Any pointers?
Thanks.

[KelSolaar]

You can take a look at this example for instance:

https://github.com/colour-science/colour/blob/develop/colour/plotting/phenomenon.py#L107,

The trick lies into setting standalone=False in the **kwargs: https://github.com/colour-science/colour/blob/develop/colour/plotting/phenomenon.py#L149

Let me know if it works!

[henczati]

Thanks.
I looked at it, and simplified the workflow for some trials in an IPython console:

# init:
import matplotlib.pyplot as plt
from colour.plotting import *

# trial (plotfunc1 and plotfunc2 are from colour.plotting):
plt.subplot(121)
plotfunc1(..., standalone=False)
plt.subplot(122)
plotfunc2(...)

It works only in some cases. If I don't choose both plotfunc1 and plotfunc2 (in any order) from a good set of functions, then I get a single empty figure. E.g. single_colour_plot was not amongst the good ones.

Then I noticed that the_blue_sky_plot uses single_colour_plot as a subplot, and it works there.
So I found out that there are some...I'll call them sensitive functions, for which if I set the kwarg aspect=None, they work as a subplot but not without it.

Plotfuncs that I tried:

• good: visible_spectrum_plot, single_spd_plot, multi_spd_plot
• sensitive: single_colour_plot, multi_colour_plot, colour_quality_scale_bars_plot

I realized that setting aspect=None also allows e.g. colour_quality_scale_bars_plot to be drawn as a non-square figure (solving the 3. point in my #198 request).

Although when I see the plotting kwargs used, their naming is usually quite self-explanatory, but up to now I did not find where any of them are documented, and I do not even know what are the choices and their possible values. Could you, please, point me somewhere I can find out?

[KelSolaar]

The kwargs argument are not documented yet, it is part of my TODO list which is not available online.

However they are self-contained in those four definitions:

• colour.plotting.canvas
• colour.plotting.camera
• colour.plotting.boundaries
• colour.plotting.display

[henczati]

Thanks. Good to know that I don't have to look elsewhere.

p.s.: Btw, I found those even by myself, but they were only of limited help. I guess I can thank my shallow knowledge of matplotlib's workings for that. :)

[KelSolaar]

I'll try to document them in the next release although you indeed need a bit of knowledge of matplotlib to understand properly what they do.

[sdbbs]

Hi all,

Just wanted to post an example of this, as I had a hard time finding anything similar.

With the code below, you can get a colour-science plot within a predefined matplotlib subplot, which looks like this:

This works in:

• colour.version 0.3.13 in Anaconda, Windows 10
• colour.version 0.3.14 in Ubuntu 18.04

... however, it will fail in colour.version 0.3.11 (which is the current version in MSYS2/MINGW64)

Here is the code:

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import os
import sys

import numpy as np
import matplotlib
print("matplotlib.__version__ {}".format(matplotlib.__version__))
import matplotlib.pyplot as plt

import colour
print("colour.__version__ {}".format(colour.__version__))

from colour.plotting import *

# this gets/sets the colour style for matplotlib plot, https://colour.readthedocs.io/en/develop/_modules/colour/plotting/common.html
colour_style()

RGB = np.random.random((4, 4, 3))

fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(9, 6), dpi=120, gridspec_kw={'height_ratios': [1, 1]}) # sharex=True,
fig.subplots_adjust(hspace=0)

# setup first subplot
pl21, = ax1.plot([0, 100], [0, 200], color="Red")

# via https://colour.readthedocs.io/en/develop/_modules/colour/plotting/common.html
settings = {
  'figure': fig,
  'axes': ax2,
  'uniform': False,
  'aspect': 'auto',
  'font.size': 7,
}

# do the plot of CIE diagram on the figure and axes (subplot) specified in settings:
plot_RGB_chromaticities_in_chromaticity_diagram_CIE1931(
  RGB,
  standalone=False,
  **settings,
)

# do the blackbody curve on top of previous CIE diagram in same figure and axes (subplot) specified in settings:
colour.plotting.temperature.plot_planckian_locus(planckian_locus_colours=None, method='CIE 1931', standalone=False, **settings)

plt.show()

[MichaelMauderer]

Thanks for the feedback. Could you describe what happens instead? Or if there is an exception, post the backtrace?

[sdbbs]

Hi @MichaelMauderer ,

Could you describe what happens instead? Or if there is an exception, post the backtrace?

Ah, yeah - sorry; about the:

fail in colour.version 0.3.11

... it is simply because of API changes, I think; first problem is:

$ python3 /tmp/test-colour.py
matplotlib.__version__ 3.1.2
colour.__version__ 0.3.11
C:/msys64/mingw64/lib/python3.8/site-packages\colour\utilities\verbose.py:141: ColourWarning: "x" independent variable is not uniform, unpredictable results may occur!
  warn(*args, **kwargs)
Traceback (most recent call last):
  File "C:/msys64/tmp/test-colour.py", line 18, in <module>
    colour_style()
NameError: name 'colour_style' is not defined

Then, if I comment the colour_style() line, I get:

$ python3 /tmp/test-colour.py
matplotlib.__version__ 3.1.2
colour.__version__ 0.3.11
C:/msys64/mingw64/lib/python3.8/site-packages\colour\utilities\verbose.py:141: ColourWarning: "x" independent variable is not uniform, unpredictable results may occur!
  warn(*args, **kwargs)
Traceback (most recent call last):
  File "C:/msys64/tmp/test-colour.py", line 38, in <module>
    plot_RGB_chromaticities_in_chromaticity_diagram_CIE1931(
NameError: name 'plot_RGB_chromaticities_in_chromaticity_diagram_CIE1931' is not defined

I presume plot_RGB_chromaticities_in_chromaticity_diagram_CIE1931 in v0.3.11 must have been called RGB_colourspaces_chromaticity_diagram_plot_CIE1931 ( via https://github.com/colour-science/colour/blob/v0.3.11/colour/plotting/models.py )

[KelSolaar]

colour.plotting.colour_style equivalent would be colour.plotting.colour_plotting_defaults in 0.3.11.

I presume plot_RGB_chromaticities_in_chromaticity_diagram_CIE1931 in v0.3.11 must have been called RGB_colourspaces_chromaticity_diagram_plot_CIE1931 ( via https://github.com/colour-science/colour/blob/v0.3.11/colour/plotting/models.py )

Yes! 0.3.12 was a rather large release and a ton of stuff has been renamed, we support forward migration, e.g. redirected the old object name to the new one and emit a warning but it only works one way :)