GreenInfo-Network / Stanford-BOE2017

0 stars 0 forks source link

Geocode 2017 BOE Retailers #1

Closed ychaiu closed 6 years ago

ychaiu commented 6 years ago

boe2017_12.xlsx

From Amanda: This is a list of stores we need to geocode for Stanford. We should use the ESRI street layer and track the match rate. We need the location to be by address, not allow a city/zip or other less accurate match be assigned. It will take credits from ESRI which is fine - we can try and sort out the cost to pass through to the client.

Bill to the root of Stanford PRC in QB ($70) - as I still need to sort out the new billing jobs we will need in 2018.

ychaiu commented 6 years ago

P:\proj_p_s\Stanford_PRC\BOE_2017\data\BOE2017_12_Geocoding_Results.shp

33,046 Addresses

Score>=90 : 32,687 (98.9%) 90>Score>=80 : 321 (0.97%) Score<80 : 38 (0.1%)

Sorting by ascending order in the matched address fields shows that are roughly 274 records that were matched to non-specific street/city/zip.

AmandaGIN commented 6 years ago

AR reviewed 45 records with a score less than 82. Included notes on if they were moved, found, not changed, etc.

Basic approach: Descending sort on Score from geocode Use Google Maps to locate (search on combos of name and address, as available) looking for confirmation of store (some cases dropped into street view). Note the final determination of the geocode quality/result/changes.

Generate file of stores for BOE 2017 with current date at end. This is the file to copy/edit if changes are made in the future.

AmandaGIN commented 6 years ago

Evaluated possible improvements on store locations:

AmandaGIN commented 6 years ago

Final shapefile as of 1/17/2018: P:\proj_p_s\Stanford_PRC\BOE_2017\data\BOEe2017_2018_01_17.shp with summary table for client produced & saved in same folder: BOE2017_withCityCountyLongLat_2018_01_17.xlsx BOE2017_withCityCountyLongLat_2018_01_17.xlsx