I am trying to link data from Google Trends with geocode data from Census. This is an example of the location data provided by Google Trends for Florida:
sub_code name
US-FL-571 Ft. Myers-Naples, FL
US-FL-592 Gainesville, FL
US-FL-561 Jacksonville, FL
US-FL-528 Miami-Ft. Lauderdale, FL
US-FL-534 Orlando-Daytona Beach-Melbourne, FL
US-FL-656 Panama City, FL
More details about this output can be found here .
From here, one can download the MSA shapefiles. I have downloaded the 2017 CBSAs data at 20m resolution. Here is the corresponding data for Florida:
[1] "Wauchula, FL"
[2] "Deltona-Daytona Beach-Ormond Beach, FL"
[3] "Port St. Lucie, FL"
[4] "Arcadia, FL"
[5] "Punta Gorda, FL"
[6] "Sebring, FL"
[7] "Homosassa Springs, FL"
[8] "Key West, FL"
[9] "Sebastian-Vero Beach, FL"
[10] "Tampa-St. Petersburg-Clearwater, FL"
[11] "Crestview-Fort Walton Beach-Destin, FL"
[12] "Okeechobee, FL"
[13] "Jacksonville, FL"
[14] "Tallahassee, FL"
[15] "Orlando-Kissimmee-Sanford, FL"
[16] "Miami-Fort Lauderdale-West Palm Beach, FL"
[17] "Gainesville, FL"
[18] "The Villages, FL"
[19] "Palatka, FL"
[20] "Lakeland-Winter Haven, FL"
[21] "Lake City, FL"
[22] "Ocala, FL"
[23] "North Port-Sarasota-Bradenton, FL"
[24] "Pensacola-Ferry Pass-Brent, FL"
[25] "Cape Coral-Fort Myers, FL"
[26] "Naples-Immokalee-Marco Island, FL"
[27] "Palm Bay-Melbourne-Titusville, FL"
[28] "Clewiston, FL"
[29] "Panama City, FL"
I understand Gtrends has a subset of all possible MSAs, but while some match perfectly (e.g. Panama City, FL), for others it is not very clear what should be merged with what. For example, Ft.Myers-Naples, FL from the first data could be merged with Cape Coral-Fort Myers, FL or with Naples-Immokalee-Marco Island, FL.
I would appreciate any guidance in dealing with such inconsistencies. Perhaps I am missing something obvious, so if you spot it, it'd be great to know!
