For each new crawl, we run the extract_comm_details.sh script. This script processes pcaps (into json files), pickles dataframes so they can be loaded faster, and runs leak detection.
cd src/analysis/scripts./extract_comm_details.sh PATH_TO_CRAWL_DATA/roku-data-2019XXXX-XXXXXX/
You can then run the notebooks using the pickled datatframes.