CGLS#

class cuqi.solver.CGLS(A, b, x0, maxit, tol=1e-06, shift=0)#

Conjugate Gradient method for unsymmetric linear equations and least squares problems.

See http://web.stanford.edu/group/SOL/software/cgls/ for the matlab version it is based on.

If SHIFT is 0, then CGLS is Hestenes and Stiefel’s conjugate-gradient method for least-squares problems. If SHIFT is nonzero, the system (A’*A + SHIFT*I)*X = A’*b is solved.

Solve Ax=b or minimize ||Ax-b||^2 or solve (A^TA+sI)x=A^Tb.

Parameters:

  • A (ndarray or callable f(x,*args).)

  • b (ndarray.)

  • x0 (ndarray. Initial guess.)

  • maxit (The maximum number of iterations.)

  • tol (The numerical tolerance for convergence checks.)

  • shift (The shift parameter (s) shown above.)

__init__(A, b, x0, maxit, tol=1e-06, shift=0)#

Methods

__init__(A, b, x0, maxit[, tol, shift])

solve()