For our September meetup, Andrew Donoho will give a presentation entitled "6 Million Tweets / Day with 200 Lines of Python". In his words: "Every new Pythonista needs to write an app based on something they know. As the author of Retweever for iPad, I know Twitter. Hence, this talk will be a review of how I built a simple system to sit on the Twitter 'sample' REST endpoint and save 6M tweets/day in the NoSQL database Couchbase. After this talk, you will know a bit about Twitter, multi-threaded queuing on I/O based tasks and Couchbase."
Additionally, Marian Nodine will give a short talk about using Latent Semantic Indexing for topic discovery, as a more advanced approach for textual data mining beyond just TF-IDF and term counts. She will introduce Latent Semantic Indexing, show how to use the GenSim package to find topics and analyze documents, and showcase an example of using this to make more accurate predictions in a specific machine learning problem.
We are always looking for lightning talks, so please get in touch if you have a short topic or a few slides you want to talk about.
Also, I'm eager to hear from folks who attended PyTexas - if anyone can give a recap, that'd be great!
See you all Wednesday!