Random Walk Labs
  • SHOWCASE
  • HACKATHON
  • CONTACT

D3js


D3.js Visualization Hackathon
  • Date: May 2016
  • Category: Data Visualization, Web Development

Use D3.js to display data

Winning entry

Graph Visualization & Topic Modeling

The winning entry shows displays the output from a modified Latent Dirichlet Allocation (LDA) model run on articles scraped from various politically polar publications. Topic models were obtained as a result of LDA modeling. A graph was constructed using the correlations of topic assignments as edge weights. The picture below shows a snapshot of this graph. Check out the interactive demo and the corresponding GitHub repository.

This D3.js visualization has three features:

  1. Hive plot display of 3 publications (From 12:00, clockwise: Economist, Jacobin, WND) with links if a topicmodel correlation between documents is > 0.95.
  2. Clicking on a hive plot node will show in a separate panel that node and all connected nodes.
  3. Clicking on a node in the graph panel will open the article in a new page.

Other entries

  1. Dataframe Visualization (Demo, GitHub)
  2. Game of Life (Demo, GitHub)

Style adapted from Freelancer theme.