This project is one I did for a paper and modified it slightly for the General Assembly class on Data Science.

It reads the little files (about 130,000 of them) from the decompressed World of Warcraft Avatar History 
Dataset and makes them into one big (~3.7gb?) CSV file for analysis. Note there are some errors in the 
data (like missing guild data on occasion). 

URL for data: http://mmnet.iis.sinica.edu.tw/dl/wowah/