PyData Delhi 2017 Recap

September 3, 2017    review conference

I attended PyData Delhi 2017 over this weekend which was the first conference organized by PyData Delhi chapter at IIIT Delhi. In fact, it’s one of the first PyData conferences in India. If you wish to be a part of the community, join the meetup group

The conference was organised at the beautiful campus of IIIT Delhi.

I had a lot of fun and learned a lot over the two days of conference. Conferences are my way of taking a busy-break in my own way. I attend for workshops, talks, conference-friends-reunion and most importantly — the  hallway tracks.

Day 1 started with an awesome opening keynote by Siraj Raval on Mathematics behind Deep learning. It was a good introduction with a walkthrough of a deep learning code in a jupyter notebook. It was fascinating how Siraj could abstract out complex stuff and made it simple to understand. Here's a tweet about it.

Another talk which I particularly enjoyed was by James Powell. He live refactored a Python code to generate Newton's Fractal on stage. The crux of refactoring involved around not using numba as a dependency and still maintain the same performance. He attempted to attain it by utilizing numpy methods which are mostly unknown to programmers who do not understand ins and outs of numpy library. I believe that doing that sort of refactoring live on stage is a commendable job. Post his talk, I had a good discussion around NUMFocus organization and PyData community at-large  and how working professionals can contribute back to the community like PyData.

Key Project Takeaways

Firefly — Firefly is a small library which exposes a function over an API. So essentially, one can call a function as one makes a API call over REST method. I’m definitely motivated to explore this sweet library further. Over the next one week, I would deep dive into the project to understand the components and perhaps blog about it as well.

NetworkX — In my discussions with James Powell, we discussed quite a few short- comings of NetworkX library. Since I am not completely familiar with it, this can be a good time to strengthen upon some of the basic knowledge I have about the library. By the way, if you’re looking to learn NetworkX, I highly recommend taking up this online course offered by DataCamp.

Overall I had an amazing time meeting so many smart people. I highly recommend everyone to attend PyData Delhi 2018 as well. If you’re looking to attend another super-awesome Python conference - look no further and attend PyCon India 2017 in November in New Delhi.


comments powered by Disqus