Add color vision deficiency, gradient, and palette display tools

[mezarque]

PR checklist

• Tag the issue(s) or milestones this PR fixes (e.g. Fixes #123, Resolves #456).
• Describe the changes you've made.
• Describe any tests you have conducted to confirm that your changes behave as expected.

Overview

This PR adds functionality for simulating color vision deficiency under the cvd module.

• addresses Add color vision deficiency functionality #15, enabling simulation of cvd in deuteranomaly, protanomaly, and tritanomaly spaces.
• includes a variety of methods for visualizing color vision deficiency of colors (as HexCode objects), palettes, and gradients under cvd.simulate_color, cvd.simulate_palette, and cvd.simulate_gradient. These methods return a transformed version of the color.
• includes a variety of methods for quickly displaying the original color + transformed colors under cvd.simulate_color_all, cvd_simulate_palette_all, and cvd.simulate_gradient_all. These methods return None but display all 4 color spaces (normal, d, p, and t).
• includes cvd.simulate_gradient_lightness() to display gradient lightness for all three major types of color vision deficiency.

In addition, this PR adds some helpful utilities to the Palette and Gradient classes:

• a swatch() method for Palette, which displays all the colors in the palette as a single strip of swatches prior to the specific color swatches, which has parallel functionality to the swatch() method for Gradient.
• a .display_all() method under each of the palettes and gradients modules, which displays a summary of all of the available colormaps within each module, as per Add function to display all palettes/ gradients #16.
• a reversed() method for Gradients, as per Add ability to reverse and slice gradients #17.
• automatic registration of reversed gradients into matplotlib's list of color gradients, with the suffix _r, as is conventional for matplotlib
• a resample() method for Gradients, which returns a Palette with a resampled list of colors from the original Gradient, allowing users to use Gradient colors for a categorical plotting purpose, as requested in Create function that returns resampled HEX codes from gradient #4.
• an interpolate() method for Gradients, which returns an updated Gradient with the values of colors adjusted so that they have uniformly increasing lightness in CAM02-UCS colorspace, addressing part of Add ability to reverse and slice gradients #17.

Testing

I confirmed that these functions work using usage_example.ipynb, where you can also see how these methods are meant to be used.

Not addressed

We are still finalizing the gradients, palettes, and font sizes, etc.; those changes will come in a future PR, including better names for unidirectional and bidirectional gradients. We're also working on making sure we have standardized sizing/ naming for plot sizes which can be drag/drop placed into Illustrator templates.

I also have not written tests for the mpl module or this new functionality yet... would love some help to do this if anyone has bandwidth.

Halp

Would love feedback on the naming and structure of things! I think we can be generally afford to be more verbose / less intuitive with some of the cvd stuff, since it's meant to be used primarily by color-obsessives like myself and Audrey.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[keithchev]

I did a first pass. Looks pretty good; just the usual comments about naming and clarity.

Some general comments:

• the names simulate_color, simulate_color_all, and simulate_colors are all very similar. I made some suggestions in inline comments about this, but one way or another, I think it's something we'd want to make a bit clearer (even if these are primarily meant for power users). In particular, I would make sure the names distinguish between methods that return simulated colors and methods that print/display simulated colors.
• For specifying color deficiencies, I would favor a more specific argument name than form like cvd_type.
• the docstrings of simulate_palette and simulate_palette_all are duplicated, as are the docstrings for simulate_gradient and simulate_gradient_all.
• In the cvd module, I would make the cvd_types dict a global and use it to instead of repeating the hard-coded list of ['d', 'p', 't'] in the subsequent methods.
• should Palette also have a reverse method?
• The type hints are incomplete (I flagged some of these, but not all).

[mertcelebi]

Very close, a couple suggestions + changes

[mertcelebi]

Maybe reverse instead of r

[mezarque]

Here, I'm trying to conform with matplotlib's syntax for designating reverse gradients. So, I think we should keep this in place.

[mertcelebi]

same here