CatBoost Model Node package

A Node.js package for applying pretrained CatBoost models.

Installation

Install the package. You have two options:

• Install from npm registry:

  npm i catboost

• Build package from source.

  CatBoost source code is stored as a Git repository on GitHub at https://github.com/catboost/catboost/. You can obtain a local copy of this Git repository by running the following command from a command line interpreter (you need to have Git command line tools installed):

  git clone https://github.com/catboost/catboost.git

  Navigate to $PATH_TO_CATBOOST_REPO/catboost/node-package directory inside the repo and run:

  npm run install [-- <build_native arguments>]

  See build_native documentation about possible arguments. Don't specify already defined --target or --build-root-dir arguments.

  For example, build with CUDA support:

  npm run install -- --have-cuda

  Now you can link this package in your project via:

  npm install $PATH_TO_CATBOOST_REPO/catboost/node-package

Usage

Apply the pretrained model.

Example with numerical and categorical features (they must be passed in separate arrays containing features of each type for all samples):

catboost = require('catboost');

model = new catboost.Model();
model.loadModel('test_data/adult.cbm');

prediction = model.predict([
            [40., 85019., 16., 0., 0., 45.],
            [28., 85019., 13., 0., 0., 13.],
        ],
        [
            ["Private", "Doctorate", "Married-civ-spouce", "Prof-specialty", "Husband", "Asian-Pac-Islander", "Male", "nan"],
            ["Self-emp-not-inc", "Bachelors", "Married-civ-spouce", "Exec-managerial", "Husband", "White", "Male", "United-States"],
        ]
);
console.log(prediction);

Release procedure

See DEPLOYMENT.md.