magtrack 0.7.1

Convert images of beads into XYZ positions

MagTrack is a Python package for tracking symmetric beads in single-molecule magnetic tweezers experiments.

• Sub-pixel XYZ coordinates
• GPU accelerated (optional, requires a CUDA GPU)
• Python notebook included with examples
• Documented, tested, and benchmarked
• Only depends on NumPy, SciPy, and CuPy
• Actively maintained until at least end of 2026

Try a demo in a Google Colab notebook:

⏳ Install

Pre-requisites

• Python >=3.9
• NumPy >=1.26
• SciPy >=1.11.1
• (Optional, but needed for GPU acceleration) CuPy-CUDA12x >=13.0
• Windows or Linux or MacOS (MacOS does not support NVIDIA GPU acceleration)
• MagTrack can run on a CPU or GPU. But GPU execution requires a CUDA-compliant GPU with the CUDA Toolkit installed. This is free and easy to install for most NVIDIA GPUs.

Instructions

pip install magtrack[gpu]

Or without CuPy

pip install magtrack

Optional: For GPU acceleration on a computer with an NVIDIA CUDA GPU, you may need to install the CUDA Toolkit for CuPy. See details at https://docs.cupy.dev/en/stable/install.html

⚒ Usage

import magtrack

# Run the full default XYZ pipeline
x, y, z, profiles = magtrack.stack_to_xyzp(stack)

# Or make your own pipeline from algorithms you prefer
x, y = magtrack.center_of_mass(stack)
x, y = magtrack.auto_conv_sub_pixel(stack, x, y)
profiles = magtrack.fft_profile(stack)
...

More Examples

You can see more examples of how to use MagTrack in this notebook. You can download it and run it on your computer. Or try it out with Google Colab.

📖 Documentation

View the full documentation at magtrack.readthedocs.io

💬 Support

Having trouble? Need help? Have suggestions? Or want to contribute code? Report issues and make requests on the GitHub issue tracker
Or email us at [e-mail]