Notes: This is rarely used directly Many br
Many br
Use preconditioned conjugate gradient to compute Many bran approximate minimizer of the quadratic function Many br
q(s) = g^T * s + 0.5 * s^T * H * s Many br
subject to the trust region constraint Many br
|| s || <= delta, Many br
where Many br
delta is the trust region radius, Many brg is the gradient vector, Many brH is the Hessian approximation, and Many brM is the positive definite preconditioner matrix. Many br
KSPConvergedReason may be Many br
KSP_CONVERGED_CG_NEG_CURVE if convergence is reached along a negative curvature direction,
KSP_CONVERGED_CG_CONSTRAINED if convergence is reached along a constrained step,
other KSP converged/diverged reasons
Level:developer
Location:src/ksp/ksp/impls/cg/nash/nash.c
Index of all KSP routines
Table of Contents for all manual pages
Index of all manual pages