Skip to content

Latest commit

 

History

History
97 lines (76 loc) · 4.05 KB

faq.md

File metadata and controls

97 lines (76 loc) · 4.05 KB

FAQ

How can I add a new colormap?

We welcome contributions!

To add a colormap: you first need to pick a namespace for your colormap. Namespaces are directories in the src/cmap/data folder.

  • If the colormap is part of known broader collection (e.g., cet, cmocean, colorbrewer, etc.), please check to see whether the collection already exists in the cmap/data directory. If so, please extend that collection. If not, please create a new namespace and directory for that collection.
  • If you would like to contribute a colormap that doesn't nicely fit into an broader collection of colormaps (regardless of whether that collection exists in cmap), please contribute to the cmap/contrib namespace.

If you are unsure where to put your colormap, please open an issue.

Once you have picked a namespace:

  1. Find the cmap/data/<namespace>/record.json file

  2. For each colormap you want to add, add a new item to the colormaps object in the record.json file. The key should be the name of the colormap, and the value should be an object with the following keys:

    • data: either a direct ColormapLike data entry, such as an array of RGB values; or a string pointing to the python-path of colormap data in the form of cmap.data.<namespace>:<colormap>.
    • category: one of "sequential", "diverging", "cyclic", "qualitative", or "miscellaneous"

    The following keys are optional:

    • tags: a list of tags to help categorize the colormap. These appear on the website.
    • info: a brief description of the colormap. This appears on the website (recommended).
    • interpolation: whether the colormap should be interpolated, between stops. If not provided, the assumption is True.
    • aliases: a list of alternative names for the colormap.
    • over: color to show when values are over the range.
    • under: color to show when values are under the range.
    • bad: color to show when values are NaN or masked.

  3. If your data object above is a module:attribute string, don't forget to add the data in the cmap/data/<namespace>/__init__.py file. For example:

    my_colormap = [
    [0.468775, 0.468876, 0.468851],
    [0.473809, 0.47391, 0.473885],
    [0.478873, 0.478974, 0.478949],
    [0.483952, 0.484053, 0.484028],
    ...
]

!!!tip It may be helpful to look at existing folders and files in the cmap/data directory for examples of how to structure the data

When opening a PR, please include a screenshot of the colormap, along with a brief description of the colormap design and its intended use.

!!!info "Licensing" Please ensure that the colormap you are adding is available under a permissive license. Note that namespaces share a license, so if you are adding a colormap to an existing namespace, make sure that the colormap is compatible with the existing license. New licenses may be added to the LICENSES directory at the root of the repo. If you are unsure about licensing, please open an issue.

How can I add support for exporting to another colormap format?

cmap exports to a variety of known third-party colormap formats.

If you are the author (or user) of a library that consumes colormaps, and you would like to have a to_your_lib() function in cmap, we welcome contributions!

Have a look at _external.py for examples of how to add support for your format.

Don't we already have enough colormap libraries?

Perhaps! But maybe just one more? 😂

The primary driver for this was to create a dependency-free library (save numpy) that could be used in a variety of visualization libraries. We will never depend on anything outside of numpy, and we export to a variety of third-party colormaps.