Struct graph::walkers::best::BestFirst

pub struct BestFirst<'a, I, N, E, Ty, Ix, F> {
    pub direction: Direction,
    pub h: F,
    /* private fields */
}

Best first implementation.

This is a key function that is later used to implement variants, this is possible due to the pattern all variants repeat, where only the heuristics function changes but te procedure is the same.

The following searching methods are implemented using this function: - Dijkstra - A* - A* Weighted - Greedy best first

The function takes a function h of type F for computing the heuristic value for each node. This function F has access to the node index on the graph, the edge index of the connection to its parent, the current state of the search instance and the index of its parent. With all these values, any heuristic can be calculated, I presume.

Fields

direction: Direction

h: F

Implementations

impl<'a, I, N, E, Ty: EdgeType, Ix: IndexType, F> BestFirst<'a, I, N, E, Ty, Ix, F>where F: Fn(NodeIndex<Ix>, EdgeIndex<Ix>, f32, NodeIndex<Ix>) -> f32,

pub fn new( graph: &'a Graph<I, N, E, Ty, Ix>, journey: (NodeIndex<Ix>, Option<NodeIndex<Ix>>), h: F, direction: Direction ) -> Self

Trait Implementations

impl<'a, I, N, E, Ty: EdgeType, Ix: IndexType, F> Walker<Ix> for BestFirst<'a, I, N, E, Ty, Ix, F>where F: Fn(NodeIndex<Ix>, EdgeIndex<Ix>, f32, NodeIndex<Ix>) -> f32,

fn step(&mut self) -> WalkerState<Ix>