Skip to content

Latest commit

 

History

History
14 lines (9 loc) · 997 Bytes

README.md

File metadata and controls

14 lines (9 loc) · 997 Bytes

tensor-spacing

This is a Python package for the article Second maximum of a Gaussian random field and exact (t-)spacing test by Jean-Marc Azaïs, Federico Dalmao and Yohann De Castro. It illustrates the method of testing for the existence of low-rank tensors within a noisy observation. While our procedure is not specifically designed to detect low rank structures, it is applicable to ALL alternatives. However, it is particularly powerful when applied to low rank alternatives.

Main module

  1. tensor_spacing.py: implementation of spacing test.

Jupyter Notebooks

  1. spacing_tensors.ipynb : Demonstrates spacing test in the case of symmetric tensors.

Reference

This notebook is based on the article:

  1. Azaïs J.-M., Dalmao F., De Castro Y., Second maximum of a Gaussian random field and exact (t-)spacing test, arXiv:2406.18397.