Parallel Processing Examples

Leverage multi-core CPUs for large graph analysis using Rayon.

Parallel Feature

Make sure the parallel feature is enabled in your Cargo.toml.

Parallel PageRank

Calculates PageRank significantly faster on large graphs.

use graphina::core::types::Digraph;
use graphina::parallel::pagerank_parallel;

fn main() {
    let mut graph = Digraph::<&str, f64>::new();

    // Imagine adding 100,000 nodes and edges...
    // Create a larger graph
    for i in 0..100 {
        let n = graph.add_node("Node");
        if i > 0 {
            // Connect to previous to form a chain/line
            // In a real scenario, you'd add many edges
        }
    }

    // Arguments: graph, damping, iteration, tolerance, nstart
    let scores = pagerank_parallel(&graph, 0.85, 100, 1e-6, None);
    println!("Computed {} ranks.", scores.len());
}

Parallel BFS

Run BFS from multiple source nodes simultaneously. Useful for distance estimations or analyzing spread.

use graphina::core::types::Digraph;
use graphina::parallel::bfs::bfs_parallel;

fn main() {
    let mut graph = Digraph::<&str, f64>::new();
    let n1 = graph.add_node("A");
    let n2 = graph.add_node("B");
    graph.add_edge(n1, n2, 1.0);

    let start_nodes = vec![n1];

    // visited is a Vec<Vec<NodeId>> containing search order for each starting node
    let visited = bfs_parallel(&graph, &start_nodes);
    println!("Reachable nodes from starting nodes: {:?}", visited);
}