python-vs-cpp-speed 0.0.1

A Python package comparing Python vs C++ performance with Pybind11

python-vs-cpp-speed

A Python package that benchmarks pure Python vs C++ loop performance using Pybind11.

Disclaimer: This repository is primarily for learning and templating purposes, demonstrating PyPI package development, Pybind11 bindings, and GitHub Actions workflows.

---

🔧 Features

• ⚡ C++ Speed Test: Measure loop execution time via a compiled Pybind11 extension.
• 🐍 Python Speed Test: Equivalent pure Python benchmark for direct comparison.
• 📊 Side-by-side Comparison: Returns Python time, C++ time, and speedup multiplier.

---

📦 Installation

pip install python-vs-cpp-speed

Requires Python ≥ 3.9. Prebuilt wheels available for Windows, macOS, and Linux.
Building from source requires cmake >= 3.15 and a C++17 compiler.

---

🚀 Quickstart

import python_vs_cpp_speed as pvc

results = pvc.compare_loop_speed(num_loops=1_000_000)

Output:

--- Python Version ---
Completed 1000000 loops in 42.3121 ms
--- C++ Version ---
Completed 1000000 loops in 0.6234 ms
--- Comparison ---
Python: 42.3121 ms
C++:    0.6234 ms
C++ is 67.88x faster

---

🧪 API Reference

compare_loop_speed

from python_vs_cpp_speed import compare_loop_speed

compare_loop_speed(
    num_loops: Optional[int] = None,
    verbose: bool = True
) -> dict

Runs both Python and C++ loop benchmarks and returns a results dictionary.

• num_loops: Number of loop iterations. If None, prompts for user input.
• verbose: If True, prints progress and results to stdout.

Returns:

{
    "python_ms": float,   # Python loop time in milliseconds
    "cpp_ms":    float,   # C++ loop time in milliseconds
    "speedup":   float    # python_ms / cpp_ms
}

num_loops must be >= 1.

---

🏗️ Building from Source

git clone https://github.com/AdamIH1/python-vs-cpp-speed.git
cd python-vs-cpp-speed
pip install .

Requirements: cmake >= 3.15, pybind11 >= 2.11, a C++17-compatible compiler.

---

📄 License

This project is licensed under the MIT License.

---

🧑‍💻 Contributing

Contributions welcome! Feel free to open issues or submit PRs on GitHub.