DC2 Sample Data

by Mike Gleicher on October 6, 2018

(update 11/10/2018 – some students pointed out that the 2017 sample data has 4 groups not 5)

For each example data set there is a JSON, a CSV, and an SVG of each of the example designs (see DC2 Example Designs). These visualizations are VERY simple.

The simple synthetic graphs are good for testing things out – especially if you are trying to show graph structure.

The 2017 data is good for testing since its realistic. But its a bit boring.

The data is in http://graphics.cs.wisc.edu/Vis/CS765-2018-DC2-ExampleData/

Simple Single Synthetic Graphs

These aren’t realistic – but they are graphs of various sizes and patterns so you can test things. The times, users, and other features are assigned in a valid, but non-interesting manner. The main thing here is the tree structure.

For randomly generated things, there are a few of each, since each time, a different tree gets generated each time.

There are no swarmplots for these.

Data from Actual Assignments

These are from prior years assignments. The structure of the class was similar, but the assignments were not the same. Same basic idea (make an initial post and converse), but the timing was different, and there were oddities like requiring two initial postings on some assignments.

All the user IDs and assignment IDs have been changed, as have the times and dates. There was some randomness added to the message lengths as well.

This data is kindof boring – there wasn’t that much discussion.