Inputs

GPS data

The system expects to find raw data in the directory specified by cvts.settings.RAW_PATH. The name of each file is <vehicle-id>.csv, and the vehicle id is inferred from the file name. So far, we have on file per vehicle per day, and the data for each day is contained in a subdirectory. All files for a vehicle are concatenated and the GPS points sorted by the time stamp. An input CSV looks like:

Index,PlateNumber,Latitude,Longitude,speed,Orientation,VehicleType,Weight,Time
1,1,108.189498901367,11.0894994735718,17,32,1,1,1501545628
2,1,108.189796447754,11.0899000167847,9,302.5,1,1,1501545658
3,1,108.189796447754,11.0899295806885,0,95.5,1,1,1501545779

Geographic Data

The system expects to find shape files for any geographies providing regionalisations of interest in the directory cvts.settings.BOUNDARIES_PATH.

That directory is expected to contain a file named geography-geom-field-names.json that contains a (JSON serialized) dictionary mapping from the names of the geographies to the fields that contain their geometry ids. For example:

{"District": "DistrictID", "trans_cities": "OBJECTID"}