Installation

Requirements

Minimum Requirements:

Python ≥ 3.8
NumPy ≥ 1.19.0
SciPy ≥ 1.5.0

Optional Dependencies:

Matplotlib ≥ 3.3.0 (for plotting)
pytest ≥ 7.0.0 (for testing)
sphinx ≥ 4.0.0 (for documentation)

Installation Methods

1. Basic Installation

Install core package only (minimal dependencies):

pip install -e .

This installs PyMARS with NumPy and SciPy only.

2. Development Installation

Includes testing and code quality tools:

pip install -e ".[dev]"

Installs:

pytest, pytest-cov (testing)
black (code formatting)
flake8 (linting)
mypy (type checking)

3. With Plotting Support

Add visualization capabilities:

pip install -e ".[plot]"

Adds matplotlib for all plotting functions.

4. Complete Installation

Everything including documentation:

pip install -e ".[dev,plot,docs]"

Installs all optional dependencies.

From GitHub

Clone and install in development mode:

git clone https://github.com/abder111/pymars.git
cd pymars
pip install -e .

For development:

git clone https://github.com/abder111/pymars.git
cd pymars
pip install -e ".[dev,plot]"

Verification

Verify correct installation:

from pymars import MARS
import numpy as np

# Quick test
X = np.random.randn(100, 3)
y = X[:, 0] + np.sin(X[:, 1]) + np.random.randn(100)*0.1

model = MARS(max_terms=10)
model.fit(X, y)

print(f"R² score: {model.score(X, y):.4f}")
print("✅ PyMARS installed successfully!")

Expected output should show an R² score around 0.7-0.9 for this synthetic data.

Running Tests

Run the test suite:

cd pymars
pytest tests/ -v

With coverage report:

pytest tests/ --cov=pymars --cov-report=html

View coverage in your browser: open htmlcov/index.html

Running Examples

Interactive demonstration:

cd examples
python demo_regression.py

This runs interactive examples with real and synthetic datasets.

Troubleshooting

Import Error

If you get ModuleNotFoundError: No module named 'pymars':

Verify installation in editable mode:
```
pip install -e .
```
Check your Python interpreter:
```
which python
```
List installed packages:
```
pip list | grep pymars
```

NumPy/SciPy Build Issues

On Ubuntu/Debian:

sudo apt-get install python3-dev
pip install --upgrade numpy scipy

On macOS:

brew install openblas
pip install --upgrade numpy scipy

Test Failures

Check Python version (must be ≥ 3.8):
```
python --version
```
Clear pytest cache:
```
pytest --cache-clear tests/
```

Update all dependencies:

pip install --upgrade numpy scipy pytest

Conda Installation

If using Conda (alternative):

conda install numpy scipy matplotlib pytest
git clone https://github.com/abder111/pymars.git
cd pymars
pip install -e .

Virtual Environment Setup

Recommended: Use virtual environment

# Create virtual environment
python -m venv pymars_env

# Activate
source pymars_env/bin/activate  # Linux/macOS
# OR
pymars_env\Scripts\activate  # Windows

# Install
pip install -e .

Uninstalling

Remove PyMARS:

pip uninstall pymars

Development Setup

For contributing to PyMARS:

git clone https://github.com/abder111/pymars.git
cd pymars

# Install with all development tools
pip install -e ".[dev,plot,docs]"

# Create feature branch
git checkout -b feature/your-feature

# Make changes, run tests
pytest tests/ -v

# Format code
black pymars/ tests/

# Commit and push
git add .
git commit -m "Your message"
git push origin feature/your-feature

Building Documentation Locally

Build HTML documentation:

cd docs
pip install sphinx sphinx-rtd-theme sphinxcontrib-bibtex
make html

View in browser:

open _build/html/index.html  # macOS
xdg-open _build/html/index.html  # Linux
start _build/html/index.html  # Windows