The openrouteservice library gives you painless access to the openrouteservice (ORS) routing API's. It performs requests against our API's for
- directions
- isochrones
- matrix routing calculations
- places
- elevation
- Pelias geocoding
- Pelias reverse geocoding
- Pelias structured geocoding
- Pelias autocomplete
- Optimization
For further details, please visit:
We also have a repo with a few useful examples here.
For support, please ask our forum.
By using this library, you agree to the ORS terms and conditions.
openrouteservice-py is tested against CPython 2.7, 3.4, 3.5, 3.6, 3.7 and 3.8-dev, and PyPy3.5.
For setting up a testing environment, install requirements-dev.txt
:
pip install -r requirements-dev.txt
To install from PyPI, simply use pip:
pip install openrouteservice
To install the latest and greatest from source:
pip install git+git://github.com/GIScience/openrouteservice-py@development
For conda
users:
conda install -c nilsnolde openrouteservice
This command group will install the library to your global environment. Also works in virtual environments.
If you want to run the unit tests, see Requirements. cd
to the library directory and run:
nosetests -v
-v
flag for verbose output (recommended).
For an interactive Jupyter notebook have a look on mybinder.org.
import openrouteservice
coords = ((8.34234,48.23424),(8.34423,48.26424))
client = openrouteservice.Client(key='') # Specify your personal API key
routes = client.directions(coords)
print(routes)
For convenience, all request performing module methods are wrapped inside the client
class. This has the
disadvantage, that your IDE can't auto-show all positional and optional arguments for the
different methods. And there are a lot!
The slightly more verbose alternative, preserving your IDE's smart functions, is
import openrouteservice
from openrouteservice.directions import directions
coords = ((8.34234,48.23424),(8.34423,48.26424))
client = openrouteservice.Client(key='') # Specify your personal API key
routes = directions(client, coords) # Now it shows you all arguments for .directions
If you want to optimize the order of multiple waypoints in a simple Traveling Salesman Problem,
you can pass a optimize_waypoints
parameter:
import openrouteservice
coords = ((8.34234,48.23424),(8.34423,48.26424), (8.34523,48.24424), (8.41423,48.21424))
client = openrouteservice.Client(key='') # Specify your personal API key
routes = client.directions(coords, profile='cycling-regular', optimize_waypoints=True)
print(routes)
By default, the directions API returns encoded polylines.
To decode to a dict
, which is a GeoJSON geometry object, simply do
import openrouteservice
from openrouteservice import convert
coords = ((8.34234,48.23424),(8.34423,48.26424))
client = openrouteservice.Client(key='') # Specify your personal API key
# decode_polyline needs the geometry only
geometry = client.directions(coords)['routes'][0]['geometry']
decoded = convert.decode_polyline(geometry)
print(decoded)
Although errors in query creation should be handled quite decently, you can do a dry run to print the request and its parameters:
import openrouteservice
coords = ((8.34234,48.23424),(8.34423,48.26424))
client = openrouteservice.Client()
client.directions(coords, dry_run='true')
If you're hosting your own ORS instance, you can alter the base_url
parameter to fit your own:
import openrouteservice
coords = ((8.34234,48.23424),(8.34423,48.26424))
# key can be omitted for local host
client = openrouteservice.Client(base_url='http://localhost/ors')
# Only works if you didn't change the ORS endpoints manually
routes = client.directions(coords)
# If you did change the ORS endpoints for some reason
# you'll have to pass url and required parameters explicitly:
routes = client.request(
url='/new_url',
post_json={
'coordinates': coords,
'profile': 'driving-car',
'format': 'geojson'
})
For general support and questions, contact our forum.
For issues/bugs/enhancement suggestions, please use https://github.com/GIScience/openrouteservice-py/issues.
This library is based on the very elegant codebase from googlemaps.