Test-driven data analysis fuses and builds upon the ideas of test-driven development and reproducible research to support higher quality data analysis.

Foundational concepts are: * Level 0: Reference Tests * Level 1: Automatic constraint discovery and validation.

This talk will extend these to cover tight constraints on string fields with * automatically discovered regular expressions with rexpy * constraints between datasets and probably more.

Background material:

• PyCon UK Talk, Cardiff, Test-Driven Data Analysis https://www.youtube.com/watch?v=FIw_7aUuY50
• Blog: http://tdda.info, especially posts http://www.tdda.info/the-new-referencetest-class-for-tdda and http://www.tdda.info/constraint-discovery-and-verification-for-pandas-dataframes
• Overview: http://www.predictiveanalyticsworld.com/patimes/four-ways-data-science-goes-wrong-and-how-test-driven-data-analysis-can-help/ In terms of some of the new material that will be covered in this talk, see
• http://www.tdda.info/introducing-rexpy-automatic-discovery-of-regular-expressions
• http://rexpy.herokuapp.com