Crate pagerank_rs

source ·
Expand description

pagerank_rs

Coverage Status

pagerank_rs is a Rust library for computing PageRank, designed to work with graphs of varying sizes. It’s a suitable choice for network analysis in different contexts, such as social networks or academic citations. The library aims to provide an efficient way to determine the significance of nodes within a network.

PageRank Example

For an in-depth look at the PageRank algorithm, visit its Wikipedia page.

Installation

Add pagerank_rs to your Cargo.toml:

[dependencies]
pagerank_rs = "0.1.0"

Usage

Here’s a simple example of how to use pagerank_rs:

use pagerank_rs::Pagerank;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut graph = Pagerank::new(3);
    graph.link(0, 1)?;
    graph.link(1, 2)?;

    let probability_of_following_a_link = 0.85;
    let tolerance = 0.0001;

    graph.rank(probability_of_following_a_link, tolerance, |node_id, rank| {
        println!("Node {} rank is {}", node_id, rank);
    });

    Ok(())
}

This code will output the PageRank of each node in the graph. Node IDs can be any usize integer. The rank values are returned through a closure to avoid allocating a large result object.

For more complex examples and usage patterns, please refer to the unit tests in the repository.

Performance

The pagerank_rs library’s performance was benchmarked on a graph with 100,000 nodes. Each node is connected to up to 400 other nodes, except for a few nodes that have a significantly higher number of inward links, simulating a more realistic distribution of links.

In tests on an Apple M2 Pro with 32 GB of RAM, the library processed the graph in approximately 1.9 seconds using typical PageRank parameters.

Performance will vary with different graph structures, system hardware, and other environmental factors.

Contributing

We welcome contributions to pagerank_rs! Here’s how to get started:

  • Fork the project on GitHub.
  • Create a new branch for your feature or bug fix.
  • Write code and add tests for your changes.
  • Ensure that all tests pass.
  • Submit a pull request against the main branch.

Please follow the Rust coding conventions and include appropriate documentation.

License

pagerank_rs is provided under the MIT License. See the LICENSE file for more details.

Author

Daniel Cadenas

Modules

Structs

  • The Pagerank struct supports adding nodes and directed edges, and provides a method to compute the PageRank scores for all nodes using the PageRank algorithm. It internally maintains mappings between node identifiers and their indices in vectors that store the graph’s adjacency information.