Work with directed graphs.

Installation

Add the following to your Cargo.toml:

[dependencies]
graaf = "0.50.1"

Overview

Operations

Build and query digraphs.

use {
    graaf::{
        gen::EmptyConst,
        op::*,
    },
    std::collections::BTreeSet,
};

let mut digraph = <[BTreeSet<usize>; 3]>::empty();

digraph.add_arc(0, 1);
digraph.add_arc(0, 2);

assert_eq!(digraph.degree(0), 2);
assert_eq!(digraph.degree(1), 1);
assert_eq!(digraph.degree(2), 1);

Algorithms

Search, traverse, and analyze digraphs.

use graaf::algo::bfs::single_pair_shortest_path as spsp;

// 0  ←  1
// ↑     ↑
// 3  →  2

let digraph = [Vec::new(), vec![0], vec![1], vec![0, 2]];

assert_eq!(spsp(&digraph, 3, 0), Some(vec![3, 0]));
assert_eq!(spsp(&digraph, 3, 1), Some(vec![3, 2, 1]));
assert_eq!(spsp(&digraph, 3, 2), Some(vec![3, 2]));
assert_eq!(spsp(&digraph, 0, 3), None);

Generators

Generate parameterized digraphs.

use graaf::gen::*;

assert_eq!(
    Vec::<Vec<usize>>::empty(2),
    vec![Vec::new(), Vec::new()]
);

assert_eq!(
    Vec::<Vec<usize>>::cycle(3),
    vec![vec![1], vec![2], vec![0]]
);

assert_eq!(
    <[Vec::<usize>; 3]>::complete(),
    [vec![1, 2], vec![0, 2], vec![0, 1]]
);

Generate random digraphs.

use {
    graaf::{
        gen::RandomTournament,
        op::*,
    },
    std::collections::BTreeSet,
};

let tournament = Vec::<BTreeSet<usize>>::random_tournament(4);

assert_eq!(tournament.size(), 6);
assert_eq!(tournament.order(), 4);

for s in tournament.iter_vertices() {
    assert_eq!(tournament.degree(s), 3);
    assert!((0..3).contains(&tournament.outdegree(s)));
    assert!((0..3).contains(&tournament.indegree(s)));
}

Representations

An adjacency matrix representation is available with the adjacency_matrix feature.

use graaf::{
    op::*,
    repr::AdjacencyMatrix,
};

let mut digraph = AdjacencyMatrix::<3>::new();

digraph.add_arc(0, 1);
digraph.add_arc(1, 1);

assert!(!digraph.is_simple());

Project goals

• A flexible API for digraph operations
• A comprehensive set of algorithms
• Generators for common digraphs
• Competitive performance
• Complete documentation
• Full unit test coverage
• Full benchmark coverage
• Extensive property tests

Features

These features require nightly Rust.

• adjacency_matrix enables the adjacency matrix representation.

Changelog

See CHANGELOG.md for a list of changes.

License

Licensed under Apache License, Version 2.0 or MIT license at your option.