PgRouting Functions Overview

Understanding a routing network

The following terminology is often used within pgRouting documentation:

• edge_table: the table containing the network;
• id: primary key column for the edge_table;
• source: the start node for calculations;
• target: the target node for calculations;
• cost: a user-calculated value for the 'cost' of travelling in the default direction along the line. Usually (eg the default with osm2po) this is a length/speed calculation;
• reverse_cost: a user-calculated value for the 'cost' of travelling in the reverse direction along a line. If one-way, it will generally be an arbitrarily high value, if two-way the default will be the same as the cost.

Routing Functions

These are well-documented at docs.pgrouting.org. The basic function format is:

select pgr_algorithm(SQL for edges, start, end, additional options)

This returns records of type pgr_costresult[] or pgr_geomresult[]. These are arrays (or graphs) of nodes and edges (running from source to target), with a cost or geometry accordingly.

To use these in a map, they must be joined to the original edge and node data.

Shortest Path

Several variations on shortest path functions are available within pgRouting. For example pgr_dijkstra:

pgr_dijkstra(sql, source, target, directed, has_rcost);

• sql: a SQL query, eg SELECT id, source, target, cost [,reverse_cost] FROM edge_table. This MUST return the columns, id, source, cost, target, reverse_cost from the edge_table;
• source: integer identifier of the source node;
• target: integer identifier of the target node;
• cost: floating point value of the edge cost. A negative cost will prevent the edge from being inserted in the graph.;
• reverse_cost: (optional) the cost for the reverse traversal of the edge. This is only used when directed = true and has_rcost = true;
• directed: true if you will be taking direction of travel along an edge into account;
• has_rcost: true if you wish to use the value of the reverse_cost column;

Returns pgr_costresult[]

Driving Distance

Note that the term "distance" is a misnomer, as it actually refers to a maximum cost of travelling from the source node.

pgr_drivingDistance(sql, source, distance, directed, has_rcost);

• sql: a SQL query, eg SELECT id, source, target, cost [,reverse_cost] FROM edge_table. This MUST return the columns, id, source, cost, reverse_cost from the edge_table;
• source: int4 identifier of the source vertex;
• cost: float8 value, of the edge traversal cost. A negative cost will prevent the edge from being inserted in the graph;
• reverse_cost: float8 (optional) the cost for the reverse traversal of the edge. This is only used when directed = true and has_rcost = true;
• directed: true if you will be taking direction of travel along an edge into account;
• has_rcost: true if you wish to use the value of the reverse_cost column.

Returns pgr_costresult[]

pgr_costresult[]

This is an array, so each row returned in the query result is a set of values in brackets:

• seq: sequential ID indicating the path order;
• id1: the node ID;
• id2: the edge ID;
• cost: cost, as defined by the function returning this result.