Engee documentation

LU Factorization

LU decomposition of a square matrix.

lu factorizations

Description

Block LU Factorization decomposes a row-rearranged version of the square input matrix A into the following upper and lower triangular matrices:

where:

  • - lower triangular matrix with unit diagonal elements (unitriangular matrix);

  • - upper triangular matrix;

  • - matrix A with rows rearranged as shown by the rearrangement vector P.

The block uses the matrix , obtained by row permutation, instead of the exact input matrix A, as this improves the numerical accuracy of the computation.

Ports

Input

A - input matrix
`matrix M by M

Input square matrix M on M.

Data types: Float64

Output

LU - lower and upper triangular matrices
`matrix M by M

A composite matrix containing both lower elements , and upper elements . Usage of a composite matrix increases the efficiency of the block.

Data types: Float64

P is a permutation vector
vector

A permutation vector of length M. The block rearranges the rows of the input square matrix A as specified in the permutation vector P, and decomposes the rearranged matrix into a unitriangular matrix and an upper triangular matrix .

Consider the following input matrix:

With the permutation vector P equal to , then the matrix obtained after permutation takes the form:

The block divides this matrix into upper and lower triangular matrices.

Data types: Float64

Parameters

Main

Show singularity status - show singularity status
off (by default) | on

Specifies whether to output the input matrix singularities to port S, which outputs values of logical data type 1 or 0. An output of 1 indicates that the input matrix is singular and vice versa.

Rounding mode - rounding mode
Floor (By default) | Ceiling | Convergent | Nearest | Round | Simplest | Zero

Set the rounding mode:

  • Floor.

  • Ceiling

  • Convergent

  • `Nearest

  • `Round

  • `Simplest

  • `Zero