Poster #19 Abstract (Paul Ivanov, UC Berkeley)

pyarbus is a library of tools for the analysis and visualization of time-series data from eyetracking experiments. pyarbus aims to be:

a Rosetta stone for different eyetracker data formats: one objective for pyarbus is to create a set of abstractions that holds the data, regardless of its original format, and to provide a unified interface to the data across different manufacturers. File formats from SR Research (EyeLink) and ~~SMI (iViewX and RED) are~~ is currently supported, or the user can create new data containers directly by passing in arrays.
temporally aware: data containers in pyarbus can be indexed with time points or time slices. Thus, indexing two different traces, perhaps collected at different rates, can be done in a unified manner, without having to remember to look up and verify what the appropriate sampling rate is, or having to hardcode assumptions into one's code.
collaboration-oriented: pyarbus integrates well with the IPython Notebook (http://ipython.org/notebook.html), a web-based interactive computational environment where you can combine code execution, text, mathematics, plots and rich media into a single document. The lead developer of pyarbus is also a core developer of IPython, and new interactive widgets for IPython will be developed with pyarbus as a testbed.
extensible: pyarbus is based on the nitime project (http://nipy.org/nitime/), which contains a core of numerical algorithms for time-series analysis both in the time and spectral domains.
test-driven: all new features proposed for inclusion in pyarbus must come with tests that verify their functional correctness.

https://github.com/ivanov/pyarbus

"Live" poster

Here are the notebooks to go along with Poster #19. I will be live-coding in these during the poster session.