Hi there,
- as of 3/2016, I've uploaded dense tables, which should make it much easier to work with the data.
- the "group_id" in the tables is the field that holds the county FIPS code, a unqiue ID for every county. That's the key that all the tables can be joined on.
- the "group_norm" fields hold the actual (normalized) 1to3gram, dictionary, or topic frequencies.