Skip to content

A python package to provide with color palettes and auxiliary tools for data visualization.

License

Notifications You must be signed in to change notification settings

Chipdelmal/ArtSciColor

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ArtSciColor

PyPI version License: GPL v3 Open Source? Yes! DOI

Creating a python package with color palettes and utilities for their use in matplotlib, seaborn, plotly, and others.

🚧 WORK IN PROGRESS 🚧

R users or Python users who don't want to install the package but still want to use the palettes, can download them in CSV form from the dataset's permalink!

Installation

The package is available through pypi, so it can be installed by running:

pip install ArtSciColor

Usage

To use a color palette simply load the package and run:

import ArtSciColor as art

hexPalette = art.getSwatch(SWATCH_ID)

where the SWATCH_ID should match one of the palettes available in our package (see the following section for more info).

Available Swatches

Have a look at currently-available palettes by selecting your favorite artist or category, and use one through its ID!

Miro, Kandinsky, Kirchner, Matisse, Picasso, Signac, Warhol, Nolde, Monet, VanGogh, EdnaAndrade, DarbyBannard, UmbertoBoccioni, Magritte

Studio Ghibli, Disney

Splatoon1, Splatoon2, Splatoon3

chipdelmal, lospec, institutions, coolors, color-hex, and schemecolor

Full dataframe in CSV for available for download here!

How are the palettes generated?

Getting palette colors is a common exercise for people getting started into clustering methods. The most widely-used algorithm for this task is k-means, but in this package the algorithm and its parameters can be provided as long as they adhere to scikit-learn's standards. Most of the curated palettes were calculated through the agglomerative clustering algorithm as follows:

from sklearn.cluster import AgglomerativeClustering
# Read image and setup number of desired clusters
img = art.readCV2Image(fPath)
CLST_NUM=4
# Clustering algorithm
CLUSTERING = {
    'algorithm': AgglomerativeClustering, 
    'params': {'n_clusters': CLST_NUM} 
}
(pixels, labels) = art.calcDominantColors(
    img, 
    cFun=CLUSTERING['algorithm'], 
    cArgs=CLUSTERING['params']
)

Other algorithms such as DBSCAN and HDBSCAN, Spectral Clustering, OPTICS, etc; can also be used.

Notes and Sources

This package was initially inspired by Blake R Mills' R packages (MoMA Colors and MetBrewer). Most palettes or original artworks are sourced from: NGA, wikiart, staedelemuseum, filmartgallery, coolors, schemecolor, color-hex, inkipedia, lospec; so please visit and support their work!


Coded by: Héctor M. Sánchez C.