Skip to content

StanfordHCI/semantic-colors

Repository files navigation

Selecting Semantically-Resonant Colors for Data Visualization

Semantically-resonant color assignments (such as mapping "ocean" to "blue") are useful in categorical visualizations, and can facilitate value lookup and comparisons.

This repository contains C# code that assigns a set of categorical concept names to semantically-resonant colors.
The algorithm makes use of images to determine frequent and discirminative colors for each concept name. It uses Google Image Search to query for relevant images. Project page: link

Main file: SemanticColors.sln

Dependencies

The code has been tested on Windows 7.

Usage

Usage: SemanticColors [inFile] [outFile] [paramFile] [renderDir]
  • inFile - file with list of categories to assign colors. See sampleCategories.csv for an example
  • outFile - output file where color assignments are saved
  • paramFile - config file. See sampleParams.txt for an example where
  • imageDir - is where images of concept names will be saved
  • cacheDir - is where computed histograms will be saved and re-used
  • jsonFile - is the location of the c3_data.json file
  • apiKey - is your Google CustomSearch API key
  • cx - is your Google CustomSearch engine id
  • renderDir (optional)- If included, images of the color assignments and affinity scores for each category are saved in this directory.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages