infomeasure-rs

infomeasure is a Rust library for computing information-theoretic measures such as entropy, mutual information, and transfer entropy. This project is a Rust implementation of the infomeasure, designed to provide superior performance while maintaining API compatibility.

Important

This crate is under development, features will be added step by step.

Features

• Multiple Estimation Techniques: Supports discrete and kernel-based approaches
• Future: all features from the parent pacakge, see infomeasure introduction
• High Performance: Leverages Rust's zero-cost abstractions for efficient computation
• Python Compatibility: Designed to potentially serve as a backend for the Python package
• Modular Design: Easily extensible architecture for adding new estimators and methods

Installation

Note

As of now infomeasure is not on crates.io, yet.

Add this to your Cargo.toml:

[dependencies]
infomeasure = "0.1.0"

Usage Examples

Calculating Entropy with Gaussian Kernel

use infomeasure::estimators::entropy::Entropy;
use ndarray::Array2;

fn main() {
    // Create or load your data
    let data = Array2::from_shape_vec((1000, 2), vec![/* your data */]).unwrap();

    // Calculate entropy with Gaussian kernel
    let bandwidth = 0.5;
    let entropy = Entropy::nd_kernel_with_type::<2>(
        data.clone(),
        "gaussian".to_string(),
        bandwidth
    ).global_value();

    println!("Entropy: {}", entropy);
}

Repository Structure

• src/ - Main source code
  • estimators/ - Estimation techniques implementations
    • approaches/ - Specific implementations (discrete, kernel, ...)
    • traits/ - Shared interfaces for estimators
• benches/ - Performance benchmarks using Criterion
• tests/ - Unit and integration tests
• examples/ - Example usage and demonstrations

Testing and Validation

The project includes a validation crate that compares results with the Python implementation to ensure compatibility and correctness. Run tests with:

cargo test

Benchmarks

Performance benchmarks are available for different estimation methods:

cargo bench

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.