Life, as we all know it, is bound to throw at us some hurdle because of which we end up opening a Pandora’s box of new(sometimes wholesome) experiences while looking for its solution. For me, this hurdle presented itself in the form of my Intro to Machine Learning course final assignment. I was required to test and compare the Bayesian vs frequentist methods of learning. I had some experience with the latter method, but the Bayesian approach was utterly new to me. My search for getting a working model led me to PyMC3. After playing around with pymc3 and getting a model working, my next job was to present my data. Here is when I came across ArviZ, and the timing couldn’t have been better.

Soon my semester came to an end, and I ended up spending the better part of my two-week long winter-break knowing more about ArviZ. The open source environment in my university is reasonably prevalant, and I like many of my peers, wanted to enter the Google Summer of Code program for the upcoming summer. I saw that ArviZ was planning to put up a project under the NUMFOCUS umbrella, and I made up my mind as to which project I was going to contribute. While going through their experimental projects, the one about applying Numba to speed up the code caught my interest. Numba is an open source JIT compiler that translates a subset of Python and NumPy code into fast machine code. Forget the developers or the sophisticated engineers; I am yet to come across a normal human being who would say no to a computational speedup. Plus ArviZ, which handles large datasets, could surely benefit from a gain of a few milliseconds or so.

So here I am finally, after two months of contributing and writing proposals as an accepted student of the GSoC 2019 program with NUMFOCUS. All I can say is I am excited about this summer. Wish me luck!!