Contents Previous Next Subchapters

Cholesky Factoring Of A Matrix
Syntax cholesky(B)
[C, p] = cholesky(B)
See Also svd , eigen , qred

Description
Returns an upper triangular matrix C that is a Cholesky factor for the matrix B; i.e.,
      H
     C  C = B
where B is a real, double-precision, or complex conjugate symmetric matrix and the superscript H denotes the complex conjugate transpose. The return value has the same type as B.

The matrix B is positive definite if
      H
     x  B x > 0
whenever x is a nonzero column vector with row dimension equal to the column dimension of B. If the argument p is not present and the matrix B is not positive definite, and error message will result. If the argument p is present and the matrix B is positive definite, p is the integer scalar zero. Otherwise, p is the smallest integer scalar such that B(1::p, 1::p) is not positive definite. In this case the return value C is a Cholesky factor for the upper left (p-1) by (p-1) block of B.

Example
     B = {[2., 1.], [1., 2.]}
     cholesky(B)
returns
     {
     [ 1.41421 , 0.707107 ]
     [ 0 , 1.22474 ]
     }

Mlmode
In Mlmode , this function is called chol instead of cholesky. If you continue the example above by entering
     mlmode
     chol(B)
O-Matrix will respond
     {
     [ 1.41421 , 0.707107 ]
     [ 0 , 1.22474 ]
     }
You can return to O-Matrix mode by entering
     omatrix