The videos from PyData NYC are now up. In particular, you can now watch the video for my IPython Notebook + pandas tutorial. I’m not sure how well it translates to video, since it was a pretty interactive tutorial and I spent a lot of the time running around in the audience answering questions. The IPython Notebook and the same notebook on Wakari should be useful, though.

Some talks I enjoyed at PyData NYC:

Probabilistic Data Structures for Realtime Analytics by Martin Laprise

I didn’t know too much about probabilistic data structures before this talk. He explained how Bloom filters work and when it would be appropriate to use them, and now I know! There’s also an IPython Notebook.

My main takeaway from this talk was that you can use Bloom filters to describe a huge amount of data, but not an unlimited amount of data – the size of your Bloom filter depends on how much data you’re going to put into it. He also described how to set up a Bloom filter where the elements expire after a certain amount of time.

IPython - The Attributes of Software and How They Affect Our Work by Brian Granger

I wasn’t able to make it to this talk, but everyone was abuzz about it on Twitter afterwards, so I’m definitely going to watch the video.

Python at Datadog - Building High-Volume Data Systems in the Python Ecosystem

At 25:27, he demos how Cython has a tool to automatically colour your Cython code and show how optimized it is. Whoa. That is the kind of stuff I go to conferences to learn :)

I didn’t make it to very many talks at the conference, so I’d love to hear about what else I missed.