DecisionTree.rs 🌳

A lightweight, general-purpose Decision Tree Classifier & Multi-Language Code Generator built with Rust and featuring a modern web interface.

visualize the tree,This tool allows you to upload any standard JSON dataset, dynamically select input features and target labels, train a decision tree model on the fly, simulate/test model outputs with real-time speed benchmarking, and export the trained model directly into ready-to-use functions in JavaScript, Python, Rust, and PHP.

🌟 Key Features

• High Performance Engine: Core decision tree learning algorithm implemented in Rust with Gini impurity calculation.
• Flexible Data Parsing: Supports standard flat JSON array format as well as complex phpMyAdmin JSON exports.
• Dynamic Column Selection: Interactive checkboxes to toggle which columns are used as inputs (features) and a dropdown to choose the target classification column (target).
• On-Demand Code Generation: Instant client-side code generators for multiple languages:
  • Supported Languages: JavaScript, Python, Rust, PHP.
  • Supported Formats: Fast (Nested If-Else) for maximum execution speed, and Compact (Array/Node Table) for minimal file size.
• Interactive Playground Simulator:
  • Test predictions directly in the browser.
  • Measures model latency by executing the tree traversal 100,000 times to provide sub-microsecond or nanosecond speed benchmarks.
• Interactive Tree Graph (Zoom & Pan): Renders beautiful SVG decision tree graphs using Viz.js (Graphviz) with built-in mouse dragging, mouse wheel zoom, and reset controls.
• Model Evaluation Metrics: Displays a performance dashboard showing overall Accuracy %, Correct predictions count, and Incorrect count over the training dataset.
• Clean Modern UI: Sleek, light-theme minimalist developer tool design built with a slate blue palette.

---

🛠️ Technology Stack

• Backend: Rust with Axum web framework, Tokio asynchronous runtime, and Serde for serialization.
• Frontend: HTML5, CSS3, Vanilla JavaScript, and Viz.js for Graphviz rendering.

---

📂 Supported JSON Formats

1. Standard JSON Array (Recommended)

You can upload any simple array of objects:

[
  {"gender": "M", "height": 180, "weight": 80, "size": "L"},
  {"gender": "F", "height": 160, "weight": 55, "size": "S"}
]

2. phpMyAdmin JSON Export

Automatically parses phpMyAdmin database export structures.

---

🚀 Quick Start

Prerequisites

Make sure you have Rust and Cargo installed.

1. Clone & Navigate to the Project

git clone <your-repo-url>
cd decision_tree

2. Run the Application

Start the local server:

cargo run

3. Open Web Interface

Open your web browser and navigate to:

http://localhost:3000

---

📄 License

This project is open-source and available under the MIT License.