Skip to content

JJEWBresch/MatrixNormWithoutAdjoint

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This repository allows to:

  • calculate the operator norm of an operator A (real or complex) which allows the evaluation of the map/oracle: v -> A v.

  • detect orthogonal matrices w.r.t. some scaling parameter, i.e. A^* A = sigma * I_d form some sigma >= 0.

  • calculate the operator norm of an operator W = A - V (real or complex) for which only the evaluation of the map/oracle: v -> A v and u -> V^* u are available.

  • detect the case, where A = V.

The methods guarantee almost sure convergence to the maximal singular value, i.e. the operator norm. The method is a randomized ascend method for the Rayleigh quotient (or a generalization thereof, respectively) where the search direction is randomly sampled perpendicular to the current point and the stepsize is chosen optimally to maximize the Rayleigh quotient.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors