For the NICAR 2016 conference, I participated in a panel entitled The way we look next: Mining past and future census data to predict diversity in race, income and aging
My part was mostly about implementing the USA Today Diversity Index using data from Census Reporter and either Excel or Python for the computation.
My slides from the session are online, and this repository contains a Jupyter notebook demonstrating the Python method in more detail.