GRASS logo

Note: A new GRASS GIS stable version has been released: GRASS GIS 7. Go directly to the new manual page here

NAME - Finds shortest path using timetables.


vector, network, shortest path

SYNOPSIS help input=name [layer=integer] output=name [walk_layer=integer] [path_layer=integer] route_id=name stop_time=name to_stop=name walk_length=name [--overwrite] [--verbose] [--quiet]


Allow output files to overwrite existing files
Verbose module output
Quiet module output


Name of input vector map
Layer number
A single vector map can be connected to multiple database tables. This number determines which table to use.
Default: 1
Name for output vector map
Layer number or name with walking connections or -1
A single vector map can be connected to multiple database tables. This number determines which table to use.
Default: -1
Layer number or name with route paths or -1
A single vector map can be connected to multiple database tables. This number determines which table to use.
Default: -1
Name of column name with route ids
Default: route_id
Name of column name with stop timestamps
Default: stop_time
Name of column name with stop ids
Default: to_stop
Name of column name with walk lengths
Default: length

DESCRIPTION finds the shortest path between two points using timetables. reads input, one query per line, from the standard input and writes output to the standard output as well as to the output map and to tables linked to layers 1 and 2. Each line of input must follow one of the following formats:
where PATH_ID is the identificator of a query that is used in the output map to differentiate between queries. Search begins at START_TIME. MIN_CHANGE gives the minimum number of time (inclusively) for a change from one route to another. MAX_CHANGES denotes the maximum number of changes allowed or -1 if infinity. WALK_CHANGE is 1 or 0 depending whether walking from a stop to another stop is considered a change or not. Finally, the path is found from FROM_STOP to TO_STOP in latter case and from the stop closest to (FROM_X, FROM_Y) coordinates to the stop closest to (TO_X, TO_Y) coordinates in former case.
For each input query, module outputs a description of the shortest path to the standard output. For example, using the tables given below, for the following input:
47 130 300 0 1 5 0
the following output is produced:
Route 15, from 130 leaving at 15 arriving to 250 at 22
Walk from 250 leaving at 22 arriving to 300 at 24
Moreover, the module writes the path to the output map and stores all the information necessary to reconstruct the path to the tables. Table corresponding to stops/points is linked to layer 1 and looks, after the query, as follows:
where CAT is the category of a point in the map, PATH_ID is the path identificator, STOP_ID is the identificator of the stop as used in the input map, INDEX is the index of the stop on the path (i.e, index=1 is the first stop visited, ...) and ARR_TIME and DEP_TIME denote the arrival time and departure time respectively. Arrival time for the first stop on the path is always equal to START_TIME and departure time for the last stop is always equal to the arrival time.
The table linked to the second layer corresponds to subroutes taken between stops. The following table is obtainedd for the above query:
where CAT is the category of lines of subroute between stops FROM_ID to TO_ID, ROUTE_ID is the identificator of the route taken or -1 if walking, INDEX and PATH_ID are as above and FROM_TIME and TO_TIME denote the times between which the route is taken.
The output map contains the points on the positions of used stops. If a subroute is taken between two stops then a line segment is added between two corresponding points. Finally, instead of straight line segment, the actual paths of routes can be given in paths layer. If this parameter is used then each line in the input map must contain identificators as category numbers of all routes passing through the line. The module then finds the path between two stops and writes this path instead. In case of walking from one stop to another, straight line between the stops is used.


Timetables are stored in a table linked to the given layer of the input map. Timetable consists of routes and each route is just a sequence of stops with specified arrival times. If two sequences of stops differ only in times then they still correspond to two routes. For example, if there is a bus line that leaves every 20 minutes and follow exactly the same path every time then there still needs to be a separate route for every time. For each stop (given by the category number of the point) the table storing information about the routes must contain the list of all routes stopping at the stop(given by route identificators) together with arrival times. That is, the table must contain three columns: stop - which is the key of the table, route_id and stop_time where each triple corresponds to a route arriving to a stop and a certain time. For example, a valid table might look as follows:
Note that stop_time is an integer and so you can use any units and offest to specify arrival times.
Also, walking connections between stops can be given by a table linked to walking layer of the input map. If this parameter is -1 then walking between stops is not allowed. The table must contain three columns: stop - which is the key of the table, to_stop and length. A record in the table says that it takes length units of time to walk from stop to to_stop. The following is a valid table:
Beware that this only means that it is possible to walk from stop 250 to stop 300 but not the other way round.


To find a path from stop with identificator 130 to stop with category 300, starting at time 0, with one time unit for change, maximum of 5 changes and with walking not considered a change of route, we use the following command:
echo "47 130 300 0 1 5 0" | \
     input=buses output=path layer=5 walking=6 path=7
If, on the other hand, we know the coordinates of the places then the following command might be used:
echo "47 10.31 54.31 90.21 28.21 0 1 5 0" | \
     input=buses output=path layer=5 walking=6 path=7



Daniel Bundala, Google Summer of Code 2009, Student
Wolf Bergenheim, Mentor

Last changed: $Date: 2013-05-23 13:01:55 -0700 (Thu, 23 May 2013) $

Main index - vector index - Full index

© 2003-2016 GRASS Development Team