python - Calculating speed using public GPS data -

i'm scraping web api displays latitude , longitude of bus.

the web-service doesn't seems have fixed update time gps position, can take 1 second 30 seconds.

when update takes long reasonable speeds (10km/h~80km/h), when update happens in less 10 seconds, unreal speeds, 1000 km/h.

def haversine(lon1, lat1, lon2, lat2):     """calculate distance between 2 points"""     lon1, lat1, lon2, lat2 = map(radians, [lon1, lat1, lon2, lat2])     dlon = lon2 - lon1     dlat = lat2 - lat1     = sin(dlat/2)**2 + cos(lat1) * cos(lat2) * sin(dlon/2)**2     c = 2 * asin(sqrt(a))     r = 6371 # radius of earth in kilometers. use 3956 miles     return c * r  def get_speed(new_lat, new_lng, cur_time):     old_lat, old_lng, old_time = buses[bus_prefix]     if (new_lat, new_lng) != (old_lat, old_lng):         distance = haversine(old_lng, old_lat, new_lng, new_lat)         speed = distance / (cur_time - old_time) * 3600         _speed = "%.1f km/h" % speed         updated_time = "%02d:%02d"%(divmod(cur_time-old_time, 60))         return _speed, updated_time      return none, none  url = 'http://api.plataforma.cittati.com.br/m3p/js/vehicles/service/22233'  buses = {}  requests.session() s:     while true:         response = s.get(url)         t = time.time()         content = json.loads(response.content)         bus in content:             bus_prefix = bus['prefix']             latitude = bus['lat']             longitude = bus['lng']             if bus_prefix in buses:                 speed, update_time = get_speed(latitude, longitude, t)                 if speed , update_time:                     print "bus {bus_prefix} traveling @ {speed}\t" \                     "last update in: {update_time}".format(**locals())             buses[bus_prefix] = latitude, longitude, t         time.sleep(1) 

maybe i'm doing math wrong, or buses in city racing, double checked everything, including racing!

here's debug did try figure out what's going on:

old_lat = -32.0916777778 new_lat = -32.0937277778  new_lng = -52.1608333333 old_lng = -52.1598611111  distance = 0.245660526035 in kms  old_time = 1436580324.66 in epoch seconds cur_time =  1436580325.94 in epoch seconds  delta time = 1.28700017929 in seconds speed = 687.162214861 in km/h  bus 1145 traveling @ 687.2 km/h last update in: 00:01 

can spot mistake? or right approach? haversine right tool job?

from posted url, got follwoing output:

[{"plate":"iuz4600","prefix":"1310","ts":1436583122000,"lat":-32.061505555555556,"lng":-52.147172222222224,"bearing":37},{"plate":"iso8600","prefix":"1145","ts":1436583099000,"lat":-32.15838333333333,"lng":-52.19098888888889,"bearing":309}]

this list of dictionaries, assume each dictionary referring 1 bus,

in each dictionary, there key called ts, should time stamp each updated lat , long, it's corresponding value should used calculations speed give more accurate results:

t = bus["ts"]