palettable

Scientific palettes in palettable are based on the work of Fabio Crameri. These palettes are designed for scientific visualization and are available with up to 256 discreetly defined colors.

They are organized into three sub-modules based on their color distribution type:

• palettable.scientific.diverging: For palettes that transition between two different colors, typically through a neutral midpoint.
• palettable.scientific.sequential: For palettes that represent data moving in one direction (e.g., low to high).
• palettable.scientific.qualitative: For palettes used to represent discrete, unordered categories.

Palettes are pre-built and loaded at import time. They follow a naming convention of <Name>_<number of colors>. For example, the Colorbrewer2 palette Dark2 with seven colors is named Dark2_7.

Every palette also has a reversed version with the same name plus the suffix _r (e.g., Dark2_7_r).

Palettes are organized into modules based on their source (e.g., cartocolors, cmocean, colorbrewer, plotly, tableau) and their type (diverging, qualitative, or sequential).

from palettable.colorbrewer.qualitative import Dark2_7

Palettable integrates deeply with matplotlib through several attributes:

Set the matplotlib Color Cycle

To change the default color cycle used when drawing plots, use the .mpl_colors attribute with ax.set_prop_cycle().

ax.set_prop_cycle('color', palettable.colorbrewer.qualitative.Dark2_8.mpl_colors)

Use as a Colormap

For functions that accept a cmap argument (like imshow), use the .mpl_colormap attribute.

from palettable.colorbrewer.sequential import Blues_8
ax.imshow(data, cmap=Blues_8.mpl_colormap)

Create a Discrete Colormap

Since .mpl_colormap is continuous/interpolated, use matplotlib.colors.ListedColormap with .mpl_colors to create a discrete version.

from matplotlib.colors import ListedColormap
cmap = ListedColormap(palettable.colorbrewer.qualitative.Dark2_7.mpl_colors)

To include a new Palettable module in the documentation, follow these steps:

1. Create a directory: Create a new directory dedicated to the module's documentation.
2. Create a template: Create an index.md.tpl file inside the new directory. This file must use Jinja templating to generate the table-of-contents and include palette preview images. You can use matplotlib/index.md.tpl as a reference.
3. Register the module: Update the MODULES dictionary in the gendocs.py script to map your new module to its corresponding directory.
4. Build and preview: Run the standard build process (make images, make compile, make build) to verify the changes.

The cmocean palettes in Palettable are sourced from the original cmocean package. They are organized into two sub-modules based on their color application type:

• palettable.cmocean.diverging: For diverging color schemes.
• palettable.cmocean.sequential: For sequential color schemes.

Install the palettable library using pip. It is compatible with Python 2.6, 2.7, and Python 3.

pip install palettable

Light and Bartlein palettes are sourced from the Department of Geography at the University of Oregon. They are organized into two distinct sub-modules based on their color scale type:

• Diverging scales: Use palettable.lightbartlein.diverging for scales that transition from one color to another through a neutral midpoint.
• Sequential scales: Use palettable.lightbartlein.sequential for scales that represent ordered data through a single color progression.

Colorbrewer palettes are available in palettable through three sub-modules, categorized by their intended use case:

• Diverging palettes: palettable.colorbrewer.diverging
• Qualitative palettes: palettable.colorbrewer.qualitative
• Sequential palettes: palettable.colorbrewer.sequential

CartoColors palettes are available in palettable and are categorized into three sub-modules based on their intended use case:

• Diverging: For data that has a meaningful midpoint (e.g., temperature, deviation from a mean).
• Qualitative: For categorical data where there is no inherent order (e.g., different types of land use, different species).
• Sequential: For data that ranges from low to high (e.g., population density, elevation).

These palettes are derived from the CartoColors node module.

The Palettable documentation is generated using a combination of hand-curated content and Jinja templates (.md.tpl files). The gendocs.py script compiles these templates into Markdown, which is then converted to HTML by Urubu.

Prerequisites

To build the documentation, you must have the following installed:

• palettable
• matplotlib
• Jinja
• Urubu
• tservice

Build Steps

1. Generate assets and compile templates:
   • Run make images to generate palette preview images.
   • Run make compile to compile the .md.tpl files into Markdown (if images are already up-to-date).
2. Generate HTML:
   • Urubu configuration is managed via _site.yml.
   • Run make build to trigger the Urubu build process.
3. Preview:
   • Run make serve to preview the documentation locally.

You can reverse the order of any pre-made Cubehelix color map by appending _r to its name (e.g., cubehelix.cubehelix3_16_r).

cubehelix.cubehelix3_16_r.show_continuous_image()