xPDT

MIT License

This software aims to make Persistent Data Topology (PDT) progamming easy for new commers to understand and to apply with a Python program. It comes along with the book xPDT---A Toolkit for Persistent Data Topology freely available on this link: https://doi.org/10.24355/dbbs.084-202312061146-0

Introduction

Persistant Data Topology (PDT) is a generalization of discrete scalar-field topology for topological features identification in high dimensions. Starting point is Topologically Assisted Topology (TAO) [https://pubs.aip.org/aip/pof/article-abstract/35/5/055105/2887659/Topologically-assisted-optimization-for-rotor?redirectedFrom=fulltext] xPDT is a Python implementation of PDT. xPDT includes three parts, they are extrema extraction, data smoothing, and some additional features. Up to now, the addition features include boundary detection and steepest descent/ascent trajectory extraction.

For refernce

@article{WangTY2023, author = {Wang, T. and Yang, Y. and Chen, X. and Li, P. and Iollo, A. and Cornejo Maceda, G.~Y. and Noack, B.~R.}, journal = {Physics of Fluids}, title = {Topologically assisted optimization for rotor design}, volume = {35}, pages = {055105}, year = {2023}, doi = {10.1063/5.0145941} }

Code

The code is written in Python version 3.6.5. A special convex hull calculation tool for higher dimensional problem written in MATLAB is included. See [https://github.com/yikaizhang/AVTA] for details.

main.py: The main function of PDT.

pdt_tools.py: The main tool functions for PDT.

load_data.py: The tool for loading data.

Specially, for higher dimensional problem, we have,

pdt_tool_high_dimension.py: The tool functions for higher dimensional problems.

For uniformly distributed data, we have,

main_uniformly.py: The main function for uniformly distributed data.

pdt_tool_uniform.py: The tool functions for uniformly distributed data.

Running the code

1. Create a folder "\PDT_author".
2. Copy the downloaded latest version of the PDT code to "\PDT_author".
3. Run "main.py" or "main_uniformly.py" needed.

Versioning

Version 0.10 - First public GitHub release. If you experience compatibility issues with previous versions, please contact the first author.

Authors

• Tianyu Wang WangTianyu@stu.hit.edu.cn
• Guy Y. Cornejo Maceda Yoslan@hit.edu.cn
• Bernd R.Noack Bernd.Noack@hit.edu.cn

Acknowledgments

• The authors thanks Bernd R. Noack (http://berndnoack.com/) for his precious advice and guidance.