The minimal supported Rust version is 1.64.0 .

1. CatBoost Rust package uses catboost-sys crate inside that is a wrapper around libcatboostmodel library with an exposed C API. In order to build it some environment setup is necessary. Modern versions of CatBoost use CMake build system, build environment setup for CMake is described here, CatBoost versions before 1.2 used Ya Make build system, build environment setup for YaMake is described here.

2. This package uses bindgen to generate bindings that has its own requirements.

3. If you run into issues on Windows like stdbool.h not found you should set up the environment for the appropriate version of Microsoft Visual Studio C++ toolset before running cargo so bindgen can find this header.

   It can be done with these commands (in PowerShell):

   Import-Module "<VSDiskDrive>:\Program Files\Microsoft Visual Studio\<VSVersion>\<Edition>\Common7\Tools\Microsoft.VisualStudio.DevShell.dll
   Enter-VsDevShell -VsInstallPath "<VSDiskDrive>:\Program Files\Microsoft Visual Studio\<VSVersion>\<Edition>" -DevCmdArguments "-vcvars_ver=<VCVars>"
   

   where

   • <VSDiskDrive> is the disk drive where Visual Studio is installed, e.g. C
   • <VSVersion> is the version of Visual Studio, e.g. 2022
   • <Edition> is the edition of Visual Studio, e.g. Community, Enterprise
   • <VCVars> is the version of Visual C++ toolset, e.g. 14.40.

   Using the same toolset that has been used for building libcatboostmodel library mentioned above is recommended for consistency.

cargo build

CUDA support is available for Linux and Windows target platforms.

Inference on CUDA GPUs is currently supported only for models with exclusively numerical features.

It is disabled by default and can be enabled by adding gpu to cargo's features list, for example:

cargo build --features "gpu"

CUDA architectures to generate device code for are specified using CMAKE_CUDA_ARCHITECTURES variable, although the default value is non-standard, specified in cuda.cmake. The default value is intended to provide broad GPU compatibility and supported only when building with CUDA 11.8. The most convenient way to override the default value is to use CUDAARCHS environment variable.

1. Add a dependency to your Cargo.toml:

[dependencies]
catboost = { git = "https://github.com/catboost/catboost" }

2. Now you can apply pretrained model in your code:

// Bring catboost module into the scope
use catboost;

fn main() {
    // Load the trained model
    let model = catboost::Model::load("tmp/model.bin").unwrap();

    println!("Number of cat features {}", model.get_cat_features_count());
    println!("Number of float features {}", model.get_float_features_count());

    // Apply the model
    let prediction = model
        .calc_model_prediction(
            vec![
                vec![-10.0, 5.0, 753.0],
                vec![30.0, 1.0, 760.0],
                vec![40.0, 0.1, 705.0],
            ],
            vec![
                vec![String::from("north")],
                vec![String::from("south")],
                vec![String::from("south")],
            ],
        )
        .unwrap();
    println!("Prediction {:?}", prediction);
}

Run cargo doc --open in catboost/rust-package directory.

Run cargo test in catboost/rust-package directory.

Tests with GPU can be enabled by adding gpu to cargo's features list, for example:

cargo test --features "gpu"