SciPy 2016 Hallway Track

written by Eric J. Ma on 2016-07-16

scipy conferences python

Like PyCon 2016, I ended up doing the Hallway Track at SciPy 2016, which really meant hanging out in the hallway, talking with people, fixing code bugs, and writing. Oh, and not to mention, running/jogging/walking on the treadmill to burn off all the food they've fed us. Part of the rationale for doing the hallway track is because all of the talks will be online (they are already going up!) on YouTube, so I can selectively listen to the ones that I'm most interested in later. (We also have a grant deadline to meet today, so I've needed time to work on that.)

I re-connected with Hugo Bowne-Anderson, who works at DataCamp as a curriculum engineer. He was previously in the field of quantitative physical biology, and now works to bring data science skills to the workforce. A really funny giant Ozzylander (Australian)! We are working together on bringing some of my network analysis material into DataCamp - exciting stuff!

I got a chance to talk with Aric Hagberg, who is the lead developer for NetworkX. He shared with me some stories of how they got NetworkX started, and his thoughts on the future of NetworkX as a package. I definitely learned a lot from him on the software engineering side.

I had a chance to meet Greg Wilson, finally, in person! He led our software and data carpentry instructor training. He always has interesting stories to tell.

I re-connected with Carol Willing, and a few other MIT-related people (Noelle and Harriet). Carol is always telling me to stay out of trouble, just as a grandmother-like figure would :P. She also wrote a very encouraging comment for me at the end of the tutorial session that I led on Network Analysis. Such a grandma-type!

I had a lot of fun time chatting with my roommate, Tom Brander, who has a ton of industry experience and was once an employee of Citibank. He is a master networker, and connected me with tons of people at the conference.

I also met some of the financial aid recipients, such as Sartaj, Aman, and AMiT. The reason Amit spells his name that way is because Amit is a commonly-used names in India, so that’s his way of differentiation. I also re-connected with Harsh Gupta, whom I met last year - he has leveraged his experiences with SymPy and GSoC to do some really cool other things! I also connected with Nelson Liu (UW), who showed me quite a bit of the ropes behind Pokemon Go on the first night of SciPy. (Yes, I was the one responsible for that channel.) Turns out it wasn’t all games and play only; Chris Niemara put together bokehmongo - a Pokemon implementation that used good’ol physics code and Bokeh plots! (SoftEng nerds out there - do not read it as bokeh and mongo!)

During the sprints, I had a chance to sprint with the Bokeh team! I learned much more about the internals of Bokeh than I could have just reading the developer documentation alone (side note: it’s a complicated library of code, but very well organized). Bryan Van de Ven and Sarah Bird were also amazingly helpful during the sprint as well, just as Jens Nielsen, Thomas Caswell and Benjamin Root helped me during last year’s sprint with the matplotlib team. I recommended that Noelle, also a graduate student at MIT, sprint with the MPL team. In my opinion, they have a very mature and easy-to-follow workflow coupled with a dozen low-hanging fixes that a newcomer to open source can help with.

In parallel, with SciPy, there was another scientific conference ongoing somewhere else, and I was seeing the “new post-doc positions available!” (re-)tweets flooding my feed. I personally dislike those tweets because I think post-docs ought to be paid at least double the current NIH pay rate. SciPy conferences don’t have those tweets happening, and I think it is for the betterment of the community that way.

All-in-all, SciPy 2016 was yet another really fun community conference! (Apologies if I forgot the names of other people; there's just too many to remember!) Looking forward to helping out with next year’s conference!