Struct AdjacencyMatrix 

pub struct AdjacencyMatrix<'a>(pub &'a [Vec<f64>]);

Row-major adjacency-matrix adapter.

Wraps an existing &[Vec<f64>] where matrix[i][j] is the weight of edge i -> j and 0.0 means no edge. Suitable for small or dense graphs where a matrix is the natural representation; for sparse graphs prefer an adjacency-list adapter like PetgraphRef.

Example

use graphops::{AdjacencyMatrix, Graph};

// 3-node directed cycle: 0 -> 1 -> 2 -> 0
let m = vec![
    vec![0.0, 1.0, 0.0],
    vec![0.0, 0.0, 1.0],
    vec![1.0, 0.0, 0.0],
];
let g = AdjacencyMatrix(&m);
assert_eq!(g.node_count(), 3);
assert_eq!(g.neighbors(0), vec![1]);

Tuple Fields

0: &'a [Vec<f64>]

Trait Implementations

impl<'a> Graph for AdjacencyMatrix<'a>

fn node_count(&self) -> usize

Number of nodes in the graph. Must be stable across calls.

fn neighbors(&self, node: usize) -> Vec<usize>

Out-neighbors of node as an owned Vec of node ids in 0..node_count().

fn out_degree(&self, node: usize) -> usize

Number of out-neighbors of node. Default calls neighbors(node).len(); override if out-degree is cheaper to compute without materializing the list.

impl<'a> WeightedGraph for AdjacencyMatrix<'a>

fn edge_weight(&self, source: usize, target: usize) -> f64

Weight of the edge (source, target). Return 0.0 when the edge is absent — callers treat zero as “no edge”, not as a zero-probability edge, to keep random-walk transition matrices well-defined.