Linear Regression with Gradient Descent: Rust Edition 🚀

Welcome to Linear Regression with Gradient Descent, a Rust-based project designed to make linear regression approachable and visually stunning! Whether you’re a data science enthusiast or a Rustacean, this project has something for you.

---

Old version

---

Project Overview

This repository implements a linear regression model using gradient descent. It includes:

1. Trainer: Train the model with your dataset.
2. Predictor: Use the trained model to predict values.
3. GUI: Visualize the data, regression line, and gradient descent process interactively.
4. Linear Regression: library to handle linear regression stuff.

---

Features

• Gradient Descent: Train the model iteratively.
• Interactive GUI: Adjust settings, visualize plots, and watch gradient descent in action.
• Makefile Integration: Simplify build and execution workflows.
• Environment Configuration: Store paths and configurations in a .env file.

---

Project Structure

Here’s the project layout:

.
├── Cargo.lock              # Dependency lock file
├── Cargo.toml              # Workspace configuration
├── data                    # Data folder
│   ├── data.csv            # Dataset for training
│   └── theta.txt           # Model parameters (theta0, theta1)
├── gui                     # GUI application
│   ├── Cargo.toml
│   └── src
│       ├── app.rs          # Main application logic
│       ├── components      # Reusable UI components
│       │   ├── mod.rs
│       │   ├── navbar.rs
│       │   ├── plane.rs
│       │   └── sidebar.rs
│       ├── main.rs         # GUI entry point
│       ├── settings        # Settings modules
│       │   ├── grid_settings.rs
│       │   ├── mod.rs
│       │   ├── plot_settings.rs
│       │   └── sidebar_settings.rs
│       └── utils.rs        # Helper functions
├── linear_regression       # Core library for regression logic
│   ├── Cargo.toml
│   └── src
│       ├── lib.rs
│       ├── linear_regression.rs
│       └── utils.rs
├── Makefile                # Build and execution shortcuts
├── predictor               # Predictor application
│   ├── Cargo.toml
│   └── src
│       └── main.rs         # Entry point for predictions
├── README.md               # This file!
├── trainer                 # Trainer application
│   ├── Cargo.toml
│   └── src
│       └── main.rs         # Entry point for training

---

Environment Variables

To keep things clean and configurable, the project uses a .env file. Make sure to create it in the root directory:

THETA_PATH=data/theta.txt
DATASET_PATH=data/data.csv

Explanation:

• THETA_PATH: Path to the file where model parameters (theta0, theta1) are stored.
• DATASET_PATH: Path to the CSV file containing your dataset.

---

Makefile Commands

The Makefile simplifies building and running the project. Here’s what you can do:

Build the Project

make build

Run the Predictor

make predictor

Run the Trainer

make trainer

Run the GUI

make gui

Run the GUI webapp

make web

Clean the Workspace

make clean

---

Example Usage

Train the Model

1. Ensure data/data.csv contains your dataset (e.g., mileage vs. price).
2. Run the trainer:

   make trainer

3. The trained parameters will be saved to data/theta.txt.

Predict Values

1. Use the predictor to estimate values based on the trained model:

   make predictor

2. Enter a mileage value when prompted, and the program will output the estimated price.

Visualize with GUI

1. Run the GUI:

   make gui
   # or
   make web

2. Explore your dataset, adjust parameters, and watch the regression line and gradient descent come to life!

---

Example Dataset

Your data/data.csv should look like this:

mileage,price
5000,20000
10000,18000
15000,16000
20000,14000
25000,12000

---

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests to enhance the project.

---

Enjoy using Linear Regression with Gradient Descent! 🚀 Let me know if you have any feedback or feature requests.