ichol: An Incomplete Cholesky factorization

In abylayzhumekenov/lanczos: Lanczos algorithm implemented in R language

An Incomplete Cholesky factorization

Description

The function compues IC(0) factorization as a lower triangular matrix.

Usage

ichol(A, diag.comp = "aj", alpha = 1)

Arguments

A	A symmetric positive definite matrix to be factored. No default value.
diag.comp	A type of diagonal compensation. Either "aj" (default) for Ajiz-Jennings, "as" for absolute shift (diag(A) = diag(A) + alpha), "rs" for relative shift (diag(A) = diag(A) + alpha * diag(A)).
alpha	Value for diagonal compensation. Default is 1. For matrices not diagonally dominant, small values of the parameter could lead to negative pivots. When such are encountered, the pivot is replaced by 1.

Examples

A = matrix(c(1,-1,-1,1), ncol=2)
L = ichol(A, diag.comp="as", alpha=1)

abylayzhumekenov/lanczos documentation built on April 27, 2022, 9:54 a.m.