https://github.com/davechallis/rust-xgboost

Rust bindings for XGBoost.
https://github.com/davechallis/rust-xgboost

ffi machine-learning rust xgboost

Last synced: 6 months ago
JSON representation

Rust bindings for XGBoost.

• Host: GitHub
• URL: https://github.com/davechallis/rust-xgboost
• Owner: davechallis
• License: mit
• Created: 2018-05-18T13:32:50.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2024-02-02T13:05:19.000Z (over 2 years ago)
• Last Synced: 2024-04-14T07:20:45.534Z (about 2 years ago)
• Topics: ffi, machine-learning, rust, xgboost
• Language: Rust
• Homepage:
• Size: 107 KB
• Stars: 95
• Watchers: 6
• Forks: 34
• Open Issues: 8
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-rust-ai-libraries - rust-xgboost - Rust language bindings for XGBoost, the popular gradient boosted tree library. Enables using XGBoost's optimized gradient boosting implementation within Rust applications with native Rust ergonomics. ([Read more](/details/rust-xgboost.md)) `Bindings` `Xgboost` `Gbdt` (Machine Learning)

README

          
# rust-xgboost

[![Travis Build Status](https://travis-ci.com/davechallis/rust-xgboost.svg?branch=master)](https://travis-ci.com/davechallis/rust-xgboost)

[![Documentation link](https://docs.rs/xgboost/badge.svg)](https://docs.rs/xgboost/badge.svg)

Rust bindings for the [XGBoost](https://xgboost.ai) gradient boosting library.

* [Documentation](https://docs.rs/xgboost)

Basic usage example:

```rust

extern crate xgboost;

use xgboost::{parameters, DMatrix, Booster};

fn main() {

    // training matrix with 5 training examples and 3 features

    let x_train = &[1.0, 1.0, 1.0,

                    1.0, 1.0, 0.0,

                    1.0, 1.0, 1.0,

                    0.0, 0.0, 0.0,

                    1.0, 1.0, 1.0];

    let num_rows = 5;

    let y_train = &[1.0, 1.0, 1.0, 0.0, 1.0];

    // convert training data into XGBoost's matrix format

    let mut dtrain = DMatrix::from_dense(x_train, num_rows).unwrap();

    // set ground truth labels for the training matrix

    dtrain.set_labels(y_train).unwrap();

    // test matrix with 1 row

    let x_test = &[0.7, 0.9, 0.6];

    let num_rows = 1;

    let y_test = &[1.0];

    let mut dtest = DMatrix::from_dense(x_test, num_rows).unwrap();

    dtest.set_labels(y_test).unwrap();

    // configure objectives, metrics, etc.

    let learning_params = parameters::learning::LearningTaskParametersBuilder::default()

        .objective(parameters::learning::Objective::BinaryLogistic)

        .build().unwrap();

    // configure the tree-based learning model's parameters

    let tree_params = parameters::tree::TreeBoosterParametersBuilder::default()

            .max_depth(2)

            .eta(1.0)

            .build().unwrap();

    // overall configuration for Booster

    let booster_params = parameters::BoosterParametersBuilder::default()

        .booster_type(parameters::BoosterType::Tree(tree_params))

        .learning_params(learning_params)

        .verbose(true)

        .build().unwrap();

    // specify datasets to evaluate against during training

    let evaluation_sets = &[(&dtrain, "train"), (&dtest, "test")];

    // overall configuration for training/evaluation

    let params = parameters::TrainingParametersBuilder::default()

        .dtrain(&dtrain)                         // dataset to train with

        .boost_rounds(2)                         // number of training iterations

        .booster_params(booster_params)          // model parameters

        .evaluation_sets(Some(evaluation_sets)) // optional datasets to evaluate against in each iteration

        .build().unwrap();

    // train model, and print evaluation data

    let bst = Booster::train(&params).unwrap();

    println!("{:?}", bst.predict(&dtest).unwrap());

}

```

See the [examples](https://github.com/davechallis/rust-xgboost/tree/master/examples) directory for

more detailed examples of different features.

## Status

Currently in a very early stage of development, so the API is changing as usability issues occur,

or new features are supported.

Builds against XGBoost 0.81.

### Platforms

Tested:

* Linux

* Mac OS

Unsupported:

* Windows