Crate ldpc

A toolbox for classical (and soon quantum) LDPC codes.

For now, only classical linear codes are implemented.

There is also a generic implementation of noise model that can be used to generate random error for codes.

Example

use ldpc::LinearCode;
use ldpc::noise_model::{Probability, BinarySymmetricChannel};
use rand::thread_rng;

// This sample a random regular LDPC code.
// It may returns an error, thus the unwrap.
let code = LinearCode::random_regular_code()
    .block_size(40)
    .number_of_checks(20)
    .bit_degree(3)
    .check_degree(6)
    .sample_with(&mut thread_rng())
    .unwrap();

let noise = BinarySymmetricChannel::with_probability(Probability::new(0.1));

// The error is a sparse binary vector where each 1 represent a bit flip.
let error = code.random_error(&noise, &mut thread_rng());

Modules

decoders
noise_model	Tools to generate random errors.

Structs

Edge
Edges
LinearCode	An implementation of linear codes optimized for LDPC codes.
RandomRegularCode	A random regular ldpc code sampler.
SparseBinMat	A sparse binary matrix optimized for row operations.

Type Definitions

SparseBinSlice
SparseBinVec