I have continued reading Varian’s paper tonight. He mentioned using R to make predictive models, using rpart (short for recursive partitioning and regression trees) to make a tree. My only experience with R was using it to compute Gini coefficients on Census 2000 data (I have issues with people’s interpretations of a Gini coefficient but that is a subject for another time). I am excited to hit Data Camp once I wrap up the Python course on Code Academy to learn R.
I also am impressed with Varian practice of making his “source code” available. This is a great practice and there needs to be more of it. I will be following his lead and making mine available on this blog (I would share my Gini coefficient code but it is lost to time).