This software is designed for identifying classification patterns in time series data. Patterns are defined based on a limited set of regular expression skeletons. These skeletons are exhaustively searched and scored based on a Fisher's exact test. The final scores are corrected using a Bonferroni correction to identify possible over fitting.

Download and run the synthetic case:

 python RegularTaskAnalysis.py -f synthetic_data.txt

Note that the code is parallelized, so will take advantage of all of the cores present on the machine. Note also that the code may take quite a while to run, depending on your dataset and machine.