The goal: To provide an easy, unified interface to load external and custom colormaps in matplotlib. The focus is on .cpt files, which are a common standard in external colormaps [1] and are frequently featured in colormap repos such as cpt-city [2].
The background: Multiple tools to load .cpt files already exist (e.g. see here: [3],[4],[5]). The most encapsulated effort is pycpt [6], a well-written package that is able to load .cpt files from multiple sources. However, this package was originally made for python 2.x and has certain limitations and, sometimes, added complexity for many user groups.
The functionality:
- To create a simple, flexible interface where
.cptfiles can be loaded, building frompycpt. - To utilise the robust engine of
pycptfor parsing.cptfiles, after some tweaks and updates. - To allow multiple use cases of easy
.cptfile loads from the user (from full paths, from a preset repo, from a url). - To control and simplify the complexity
matplotlibintroduces with its multiple classes of colormaps, but offer added functionality to the user as well, in case he/she wants hex color lists, fewer colors,ListedColormapobjects instead ofLinearSegmentedones etc.
Requirements:
- python
3.6or newer - numpy
1.14or newer - matplotlib
2.2.2or newer
Tested in two environments: (py 3.7|np 1.18.1|mpl 3.1.3) and (py 3.6|np 1.14.2|mpl 2.2.2). Also uses urllib for the url handles.
Howto: Instructions are provided in the relevant Jupyter Notebook tutorial.
Closing remarks:
- Every modification or enhancement/redevelopment of the code is welcome.
- If you would like to use this module to your work, please refer to its Zenodo DOI (10.5281/zenodo.3703160)(https://zenodo.org/record/3703160#.XmfGXqhKiUk), as well as the original pycpt package (https://github.com/j08lue/pycpt).