The D-Lab finance project was initiated as a part of the BIDS Collaborative and was described in a paper published in the Data Science Journal. Ironically, while that paper extolls the virtues of developing academic data science projects in the open, we are fairly certain that this stance has gotten at least one of our members scooped.
Therefore, we have adopted a new model of developing analyses in private, with the intention of publishing public versions in the future. At this point, we are in the process of developing an efficient Python codebase for reading the TAQ dataset from NYSE in the marketflow repo.
If you are interested in joining our efforts, please get in touch on our public issue tracker!