Grid definition that calls to Pathfinding.find_path and Pathfinding.find_walkable will search against. The grid is defined so its easy to make repeated searches across it without repeatedly reconstructing it.

Tiles, Walkability, Costs

%Pathfinding.Grid{
  tiles: [
    [1, 1, 0, 1, 1],
    [1, 1, 0, 1, 1],
    [1, 1, 0, 1, 1],
    [1, 1, 1, 1, 1],
    [1, 1, 1, 1, 1]
  ]
}

tiles is the tile definition that most of the functions in Pathfinding.Grid and Pathfinding use for traversal. Whatever numbers you choose for the tiles is unimportant, except that they are a superset of all potential values in walkable_tiles and keys in costs.

Important to mix with tiles is walkable_tiles because it determines what tiles are valid for traversal, without specifying walkable_tiles the grid will not be pathable at all.

%Pathfinding.Grid{
  tiles: [
    [1, 0, 1],
    [1, 0, 1],
    [1, 1, 1]
  ],
  walkable_tiles: [1]
}

In the example above, only 1 is walkable, so a path from (0, 0) to (0, 2) will avoid (0, 1).

Costs

%Pathfinding.Grid{
  costs: %{
    0 => 5
  }
  tiles: [
    [1, 0, 1],
    [1, 0, 1],
    [1, 1, 1]
  ],
  walkable_tiles: [0, 1]
}

Specifying a costs map will cause different tiles to have different weights, a detail that can be much better explained by the A-star algorithm. Any tiles not specified in the costs map, or if a costs map is not specified at all, will have a cost of 1.

Extra Costs

%Pathfinding.Grid{
  costs: %{
    1 => 3
  }
  extra_costs: %{
    1 => %{
      2 => 2
    }
  }
  tiles: [
    [1, 0, 1],
    [1, 0, 1],
    [1, 1, 1]
  ],
  walkable_tiles: [0, 1]
}

Similar to costs, extra_costs indicates a specific coordinate as an increased weight associated with it. Any coord with an extra cost will have its tile's cost overriden, so in the above example the cost of traversing (1, 2) is 2.

Specifying this map is programmatically cumbersome, so two alternatives exist:

• Pathfinding.Grid.add_extra_cost/4 and Pathfinding.Grid.remove_extra_cost/3 can be used to specify a single unwalkable coordinate.
• Pathfinding.Grid.to_coord_map/2 can convert a list of coordinates into an extra_costs map.

Unwalkable Coords

%Pathfinding.Grid{
  tiles: [
    [1, 0, 1],
    [1, 0, 1],
    [1, 1, 1]
  ],
  unwalkable_coords: %{
    1 => %{
      2 => true
    }
  },
  walkable_tiles: [1]
}

unwalkable_coords is a map that can be specified to mark a specific coordinate invalid for traversal, even though the tile is normally pathable. This is useful to simulate an obstruction that is not typically represented in the grid of tiles. In the above example, the right-most column is unreachable by the left-most column because (1, 2) is unwalkable.

Specifying this map is programmatically cumbersome, so two alternatives exist:

• Pathfinding.Grid.add_unwalkable_coord/3 and Pathfinding.Grid.remove_unwalkable_coord/3 can be used to specify a single unwalkable coordinate.
• Pathfinding.Grid.to_coord_map/2 can convert a list of coordinates into an unwalkable_coords map.

Unstoppable Coords

%Pathfinding.Grid{
  tiles: [
    [1, 0, 1],
    [1, 0, 1],
    [1, 1, 1]
  ],
  unstoppable_coords: %{
    1 => %{
      2 => true
    }
  },
  walkable_tiles: [1]
}

In the same vein as unwalkable_coords, unstoppable_coords are pathable, but not valid destinations. This is useful to simulate an obstruction can be moved through. In the above example, (1, 2) cannot be pathed to, but it can be passed through, making the the right-most column accessible from left-most column.

Specifying this map is programmatically cumbersome, so two alternatives exist:

• Pathfinding.Grid.add_unstoppable_coord/3 and Pathfinding.Grid.remove_unstoppable_coord/3 can be used to specify a single unstoppable coordinate.
• Pathfinding.Grid.to_coord_map/2 can convert a list of coordinates into an unstoppable_coords map.

Grid Type

The grid's type determines how the grid will be traversed. The examples below specify the order in which a coordinate's neighbors are traversed for each grid type.

:cardinal
  1
4 x 2
  3

:hex
  1 2
6 x 3
5 4

:intercardinal
8 1 2
7 x 3
6 5 4