Conjugate gradient solver.
DESCRIPTION:
Solves a linear system of equation A(x)=b where A is a function
corresponding to a symmetric positive definite matrix.
USAGE:
cg.solve(A, b, cg.max.iter=length(b), rel.conv=0.1, verbose=F, ...)
REQUIRED ARGUMENTS:
- A
-
function corresponding to a real n by n symmetric positive definite
matrix operator. Example: wavelet tranform such as
nd.dwt
.
- b
-
numeric vector.
- cg.max.iter
-
real number giving the upper bound on the number of iterations.
- rel.conv
-
real number giving the convergence threshold.
- verbose
-
logical flag indicating whether or not information should be printed on the
screen at each iteration.
- ...
-
additional argument used by the A() function. Example:
wavelet
or
n.levels
for wavelet tranforms.
VALUE:
numeric vector. Solution to A(x)=b.
REFERENCES:
Golub, G.H. and Van Loan, C.F. (1996). Matrix Computations. Johns Hopkins
University Press, Baltimore, MD.
SEE ALSO:
.