snap_trajectories_to_osm.py

# snap GPS trajectories to OSM road network
# expected csv format: tid,lon,lat
# where tid is the trip id, where all coordinates of the same trip share the same id
# the order of coordinates is expected to be in correct order

# setup OSRM server: https://hub.docker.com/r/osrm/osrm-backend/

import csv
import requests

INPUT_FILE_PATH = "data/input.csv"
OUTPUT_FILE_PATH = "data/output.csv"

# read in file
with open(INPUT_FILE_PATH, "rt") as fin:
    cr = csv.reader(fin)
    filecontents = [line for line in cr]

tids = [line[0] for line in filecontents]
lons = [line[1] for line in filecontents]
lats = [line[2] for line in filecontents]
lon_lats = [",".join((lon, lat)) for lon, lat in  list(zip(lons, lats))]

with open(OUTPUT_FILE_PATH, 'w') as csvfile: 
    write = csv.writer(csvfile) 
    write.writerow(["tid", "lon", "lat"])

    for tid in set(tids):
        indices_tid = [i for i in range(len(tids)) if tids[i] == tid]
        coords_query_format = ";".join(lon_lats[min(indices_tid):max(indices_tid)])

        # query osrm
        query_string = f"http://127.0.0.1:5000/match/v1/driving/{coords_query_format}?steps=true&geometries=geojson&overview=simplified&annotations=false"
        request = requests.get(query_string)
        
        if request == 400:
            coordinates = request.json()['matchings'][0]['geometry']['coordinates']

            # write to file
            for coord_pair in coordinates:
                write.writerow([tid, coord_pair[0], coord_pair[1]])