mc_chain_scale: Scale vectors in eigenchains of multi-companion matrices

In mcompanion: Objects and Methods for Multi-Companion Matrices

Scale vectors in eigenchains of multi-companion matrices

Description

Scale vectors in an eigenchain of a multi-companion matrix. Currently the factor is determined so that the seeding coefficient with the maximal value becomes equal to one. By default the seeding coefficients at the bottom are examined.

Usage

mc_chain_scale(ev, subset = NULL, fvec = NULL, fchain = NULL)

Arguments

ev	the eigenvectors, a list with elements "eigvec", "len.block".

subset	subset of the eigenvectors to examine, a string or a vector of integers suitable for indexing.
fvec	a function for the computation of the normailising factor, see Details.
fchain	can be used to specify a more general transformation, a function, see Details.

Details

All vectors in the chain are scaled by the same factor.

The factor is determined so that the seeding coefficient with the maximal modulus becomes equal to one. By default the seeding coefficients at the bottom of the eigenvector are examined but if argument subset is equal to "top" the tops are checked instead. subset may also be a vector of integers to be used as an index to select the desired subset.

Argument fvec may be used to specify a method other than modulus for the normalising factor. fvec is a function with two arguments, the first being the eigenvector and the second the index vector.

Argument fchain may be used to specify a completely different way to normailse the chain. fchain is a function with one argument, chain, which can do any transformation of the chain.

Value

a list as ev with the desired transformations done.

Author(s)

Georgi N. Boshnakov

mcompanion documentation built on Sept. 22, 2023, 5:12 p.m.