Predicting at the command line @ '14 [video]

Published on 16 February 2015

On November 17 of last year, I gave a tutorial at the first International Conference on Predictive APIs and Apps, also known as PAPi' '14. The recording is now available for your viewing pleasure:

After a myriad of API presentations, including those by Yhat, BigML, and GraphLab, I thought it would be nice to switch gears by demonstrating how such APIs can be employed from the command line (hence the title "Predicting at the Command Line").

Don't worry about the slides, they're quite boring. More interesting is the live demo, in which I demonstrate how to use the command line to:

  • connect to the Barcelona bicing API using curl (12:30),
  • convert the returned XML data to CSV using xml2json, jq, and json2csv (17:59),
  • plot the locations of bike stations using Rio and the ggmap package (27:18),
  • assign anomaly scores using BigML's predictive API (30:59), and as an additional impromptu demonstration,
  • apply indico's language detection API to tweets using t and parallel (38:21).

In case you'd like to try this out for yourself, you should check out my book Data Science at the Command Line and the associated Data Science Toolbox, which has all these command-line tools pre-installed.