pathing v0.1.1

Dijkstra's algorithm pathfinding in Ruby

Features

  • Quick - written in C as a Ruby extension!
  • Flexible - use your own graph interface!
  • Simple - describe your graph, call one method, and you have your path!

Sample Usage

The easiest way to download the gem is to use Bundler with a Gemfile. In your Gemfile, include the line gem 'pathing'. Then you can run bundle install to automatically download and compile the gem for your system. To include the gem in a Ruby file, use the line require 'pathing'.

To do some pathfinding, first set up a graph interface object. This object defines the entire layout of your graph. The following example is one way of achieving a two-dimensional grid with neighbors only in the cardinal directions.

class Interface
  SIZE = 32

  # gets an array of neighbors for the given node key
  # for this example, the node key is passed in as an array like so: [x, y]
  def neighbors_for(node_key)
    x, y = *node_key
    neighbors = []

    neighbors << [x-1, y] if x > 0
    neighbors << [x, y-1] if y > 0
    neighbors << [x+1, y] if x < SIZE-1
    neighbors << [x, y+1] if y < SIZE-1

    neighbors
  end

  # gets the cost for moving from node1 to node2
  def edge_cost_between(node1_key, node2_key)
    1
  end
end

Now, create a Graph object referencing your interface object.

graph = Pathing::Graph.new(Interface.new)

You're all set. Simply call

graph.path([0,0], [15, 20])

to get your optimized path!

Credits

Copyright (c) 2015 Nick Lowery

View LICENSE for full license.