Overview
This Python package, developed by Dilanjan DK, provides tools for analyzing 3D data (such as quantum mechanical properties or neuroimaging fields) using Non-Uniform Fast Fourier Transforms (NUFFT) via the FINUFFT library. It is designed for high-performance, scalable analysis of large datasets, with advanced features for gradient mapping, spectral metrics, and more.
Key Features
- NUFFT implementation for efficient transformation between non-uniform and uniform grids
- Map building and k-space masking
- Analytical and interpolated gradient calculation
- Batch processing and HDF5 output structure
- Enhanced features: spectral slope, entropy, anisotropy, higher-order moments, HRF deconvolution
- Interactive 3D visualization with Plotly
- Comprehensive documentation and usage guides
Performance
Tested on datasets up to 50,000+ points and 100+ time points, with up to 9x speedup for large-scale gradient calculations using skip-interpolation mode. Efficient memory usage and HDF5 compression are supported.
Documentation & Usage
Contact
For questions or collaboration, contact Dilanjan DK at [email protected].