Global options for the QT toolbox

A few options can be set globally in the QT toolbox through the cqtoption command.

The supported options are the following:

'threshold': The truncation threshold in the operations. This is indicative of the accuracy that will be attained by the methods. The default value is 1e-12.
'inversion': Method used for the inversion of the symbol (and therefore of QT matrices); can be either 'cr' (for Cyclic Reduction) or 'fft' (Fast Fourier Transform).
'sqrt': Method used to compute the matrix square root. Can be either 'db' (Denman-Beavers) or 'cr' (Cyclic Reduction).
'compression': The compression algorithm for the compact correction; can be either 'lanczos' for the Golub-Kahan-Lanczos algorithm for the SVD, or 'random' to use a randomized sampling technique.
'wiener-hopf' selects the method to compute Wiener-Hopf factorizations; can be either 'cr' (Cyclic Reduction), or 'newton' (based on the Newton iteration).

Syntax

cqtoption('threshold')

ans =

   1.0000e-12