$ python
>>> from route import route
>>> G = {123: {456: 0.02, 789: 0.03},
... 456: {123: 0.02, 101112: 0.06},
... 789: {101112: 0.01},
... 101112: {123: 0.01}}
>>> D,P,v = route(G,123)
>>> D[101112] #distance from 123 to 101112
0.040000000000000001
>>> D
{456: 0.02, 101112: 0.040000000000000001, 123: 0, 789: 0.029999999999999999}
>>> P
{456: 123, 101112: 789, 789: 123}
>>> v
101112
Determines the distance from the source 1000001 to targets 1000002 and 1000003
$ python loadOsm.py file.osm
$ python
>>> from route import route
>>> from data import data
>>> source = 1000001
>>> D,P,v = route(data.G,source)
>>> target1 = 1000002
>>> D[target1]
19.250000666714794
>>> target2 = 1000003
>>> D[target2]
12.052146869717513
You need either some data from OpenStreetMap or weighted graphs in Python pickle format (second method not yet documented). If starting with OSM maps file.osm, run
python loadOsm.py file.osm
It is usually a good idea to pick out a single large connected component to work with. Here we pick the one containing the node with id 10000001.
python build.py graph 10000001
Then, to get the shortest path between nodes numbered 12345 and 67890, run
python route.py footd.pkl 12345 67890
More advanced features are not accessible via the command-line interface for the moment. From the python interactive shell,
$ python
>>> import route
>>> from data import data
>>> sources = [10000001, 10000002, 10000003, 10000005, 10000008]
>>> newnodes = route.preproute(data.G,data.nodes,data.rnodes,sources)
>>> newnodes
[10000001, 10000002, 10000003, 10000005, 10000008]
>>> D,P,v = route.route(data.G,sources)
>>> target = 10000013
>>> D[target]
19.250000666714794
This is the shortest distance from (the closest of) 10000001, 10000002, 10000003, 10000005, 10000008 to 10000013.
Any of the sources could also be a (latitude,longitude,name) tuple if we want the distance from a non-existent node.
Uses PADS (files included)
Originally based on PyrouteLib although not much of the library is left.