Livestock markets with an online presence may post data about cattle sales, which the companion roundup repository writes to local CSV files. This repository archives that data in a MySQL database, which the Bansal Lab uses to study movement of livestock between animal holding premises, such as a ranch, market or feed lot. The repository also includes post-processing steps that attempt to identify the source and destination counties for every recorded movement and aggregates external datasets relevant to cattle movement (e.g. the USDA Agriculural Census).

1. address - As much address information as we obtain from one source about a premises
2. premises - A unique identifier for a physical premises
3. association - An association table linking premises to addresses
4. geoname - Locations (including www.geonames.org identifiers) of premises
5. movement - Representative lots recorded from a market report
6. georef - The latitude and longitude of a verified premises
7. roundup_report - The report from which representative lots were read

1. ag_census
2. ams
3. ams_location
4. ams_quantity
5. ams_receipts
6. aphis
7. gipsa
8. lma
9. loadboard
10. roundup_market
11. roundup_website
12. state_code
13. county

See db_class.py for fields within tables and relationships between them.

• Do this before importing the first market report from the source.
• Ensure that roundup_website has an entry with the correct URL.
• Insert an address record for the market, with source = 'roundup_market'.
• Insert a roundup_market record, referencing the new addres_id and roundup_website_id
• Insert an association record to any known premises.
• Insert a new record into georef, after obtaining the market's latitude and longitude.