Investigations of Data Quality in LIGO Data, Techniques

The existing data quality (DQ) flag veto pipeline procedure used to categorize DQ flag channels and calculate their associated metric attributes relies heavily on MATLAB scripts. The scripts have been built upon each other ad-hoc as users have required new information, sometimes with limited documentation. We are restructuring the calculation process in an object-oriented language, Python, for two primary purposes. The first is to allow users to retrieve specific isolated information without having to run the entire pipeline. The second is to calculate additional summary statistics regarding DQ flag interaction, which may prove useful in evaluating the effectiveness of the categorization process. The nature of the Python vetopipeline code allows the addition of such new features with greater simplicity. The presentation will detail the framework of the Python vetopipeline, and introduce the DQ flag interaction statistics.