This is the competitor's pack from the Differential Privacy Algorithms side of the challenge [https://deid.drivendata.org/]. Please feel free to make use of the resources in it, or adapt them to your own needs.
It has two resources you might find useful: 1. A simple differential privacy algorithm you can use to generate your own privatized data at varying levels of added noise (decreasing the epsilon parameter will increase the noise). 2. Pie chart metric code and the visualizer code that we used to produce the "deep dive" heatmaps of Baltimore for the pie chart metric (see the sample write-up).