Skip to main content
Get up and running with bun-scikit by training your first models. This guide covers regression, classification, and decision trees with native Zig acceleration.

​
Linear regression

1

Import libraries

Import the essential components for regression.
import {
  LinearRegression,
  StandardScaler,
  trainTestSplit,
  meanSquaredError
} from "bun-scikit";
2

Prepare your data

Create sample data and split it for training and testing.
const X = [[1], [2], [3], [4], [5], [6]];
const y = [3, 5, 7, 9, 11, 13];

const scaler = new StandardScaler();
const XScaled = scaler.fitTransform(X);

const { XTrain, XTest, yTrain, yTest } = trainTestSplit(XScaled, y, {
  testSize: 0.33,
  randomState: 42
});
3

Train and evaluate

Fit the model and make predictions.
const model = new LinearRegression({ solver: "normal" });
model.fit(XTrain, yTrain);

const predictions = model.predict(XTest);
const mse = meanSquaredError(yTest, predictions);

console.log("MSE:", mse);
console.log("Coefficients:", model.coef_);

​
Classification

1

Import libraries

import {
  LogisticRegression,
  StandardScaler,
  accuracyScore
} from "bun-scikit";
2

Prepare classification data

const X = [[1], [2], [3], [4], [5], [6]];
const y = [0, 0, 0, 1, 1, 1];

const scaler = new StandardScaler();
const XScaled = scaler.fitTransform(X);
3

Train classifier

const classifier = new LogisticRegression({
  solver: "gd",
  learningRate: 0.8,
  maxIter: 100,
  tolerance: 1e-5
});

classifier.fit(XScaled, y);
const predictions = classifier.predict(XScaled);
const accuracy = accuracyScore(y, predictions);

console.log("Accuracy:", accuracy);
console.log("Predicted probabilities:", classifier.predictProba(XScaled));

​
Decision trees with Zig acceleration

Train a decision tree using native Zig backend for maximum performance. 
import { DecisionTreeClassifier } from "bun-scikit";

const X = [
  [0, 0],
  [0, 1],
  [1, 0],
  [2, 2],
  [2, 3],
  [3, 2]
];
const y = [0, 0, 0, 1, 1, 1];

const tree = new DecisionTreeClassifier({ 
  maxDepth: 3, 
  randomState: 42 
});

tree.fit(X, y);

// Check backend (should be "zig" on Linux/Windows)
console.log("Fit backend:", tree.fitBackend_);
console.log("Backend library:", tree.fitBackendLibrary_);

// Make predictions
const predictions = tree.predict([[1, 1], [3, 3]]);
console.log("Predictions:", predictions); // [0, 1]
The Zig backend provides 2-6x faster training compared to scikit-learn on Linux and Windows. See Performance Benchmarks for detailed comparisons.

​
Random forest ensemble

Random forests combine multiple decision trees for improved accuracy. 
import { RandomForestClassifier } from "bun-scikit";

const forest = new RandomForestClassifier({ 
  nEstimators: 25, 
  maxDepth: 4, 
  randomState: 42 
});

forest.fit(X, y);

console.log("Fit backend:", forest.fitBackend_);
const predictions = forest.predict([[1, 1], [3, 3]]);
console.log("Forest predictions:", predictions);

​
Complete example

Here’s a full workflow combining preprocessing, training, and evaluation: 
import {
  LinearRegression,
  LogisticRegression,
  StandardScaler,
  trainTestSplit,
  meanSquaredError,
  accuracyScore
} from "bun-scikit";

// Regression workflow
const XReg = [[1], [2], [3], [4], [5], [6]];
const yReg = [3, 5, 7, 9, 11, 13];

const scaler = new StandardScaler();
const XScaled = scaler.fitTransform(XReg);

const { XTrain, XTest, yTrain, yTest } = trainTestSplit(XScaled, yReg, {
  testSize: 0.33,
  randomState: 42
});

const regressor = new LinearRegression({ solver: "normal" });
regressor.fit(XTrain, yTrain);
console.log("Regression MSE:", meanSquaredError(yTest, regressor.predict(XTest)));

// Classification workflow
const yCls = [0, 0, 0, 1, 1, 1];
const classifier = new LogisticRegression({
  solver: "gd",
  learningRate: 0.8,
  maxIter: 100,
  tolerance: 1e-5
});

classifier.fit(XScaled, yCls);
console.log("Classification accuracy:", accuracyScore(yCls, classifier.predict(XScaled)));

​
Run your code

Save your code to train.ts and run: 
bun run train.ts

​
Next steps

Core Concepts

Learn about preprocessing, pipelines, and model evaluation

Guides

Explore in-depth guides for different algorithms

API Reference

Browse the complete API documentation

Performance

See benchmark comparisons vs scikit-learn

Build docs developers (and LLMs) love

Get started for freeTalk to us