Skip to content

This repository contains the codes for reproducing the results of the paper "Small singular values can increase in lower precision", https://arxiv.org/abs/2303.03547.

Notifications You must be signed in to change notification settings

cboutsikas/small_sigmas_increase

Repository files navigation

small_sigmas_increase

Repository Description:

Τhis repository contains the code for reproducing the results presented in the paper titled Small singular values can increase in lower precision .The purpose of this code is to generate matrices $A \in \mathbb{R}^{m \times n}$ with fixed singular values $\Sigma$ and subsequently compute the singular values in double, single, and half precision.

Installation Requirements:

  • Ensure you have Julia installed for performing major computations.
  • Python is required for generating plots.
  • If you're using Visual Studio Code, setting up a Python virtual environment is recommended. For instructions, please visit this link.

Before running the code, please review the description and examples provided at the beginning of the run.jl script. To reproduce the examples from the paper, you can use the run_fixed_cases.jl script. The corresponding data and figures from the paper are located in the data/, figures/ folders respectively.

About

This repository contains the codes for reproducing the results of the paper "Small singular values can increase in lower precision", https://arxiv.org/abs/2303.03547.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published