Notes
| This version includes several counties that were inadvertently missing from the previous version. It also improves the contact structure for people in large locations. The contacts analyzed here are derived from a synthetic population of the U.S. in which each person is assigned an activity template based on American Community survey data and locations for each activity are assigned based on location types, sizes, and distance from residence or catchment region. Contacts are inferred between people present at a location simultaneously. For locations with many occupants, a subgraph is constructed that interpolates between a minimum and maximum number of contacts, depending on the total number of occupants. Activities are assigned differently for weekdays and weekends. Hence the contact networks also differ. In addition, we analyzed an activity-independent network the includes only household contacts. See the README file for more information on directory and file structure and parameter values. For more information on the research behind the synthetic population, see https://biocomplexity.virginia.edu/nssac This revision fixes corrupted files for the weekend mixing rates for all counties in the states of California, Delaware, Maryland, Michigan, Minnesota, Montana, New_Jersey, New_Mexico, New_York, Ohio, Pennsylvania, Texas, and Washington. Note that there are three sparsely-populated counties with no mixing between some age groups. County, Hawaii, FIPS code 15005 Golden Valley County, Montana, FIPS code 30037 Loving County, Texas, FIPS code 48301 This may be an artifact of our contact network synthesis process, but it does not seem to be a bug |