| /petsc/src/tao/complementarity/impls/ssls/ |
| H A D | ssls.c | 31 PetscCall(VecNorm(ssls->ff, NORM_2, &ssls->merit)); in Tao_SSLS_Function()
32 *fcn = 0.5 * ssls->merit * ssls->merit; in Tao_SSLS_Function()
44 PetscCall(VecNorm(ssls->ff, NORM_2, &ssls->merit)); in Tao_SSLS_FunctionGradient()
45 *fcn = 0.5 * ssls->merit * ssls->merit; in Tao_SSLS_FunctionGradient()
|
| H A D | ssils.c | 60 …iter: %" PetscInt_FMT ", merit: %g, ndpsi: %g\n", tao->niter, (double)ssls->merit, (double)ndpsi)); in TaoSolve_SSILS()
62 PetscCall(TaoLogConvergenceHistory(tao, ssls->merit, ndpsi, 0.0, tao->ksp_its)); in TaoSolve_SSILS()
63 PetscCall(TaoMonitor(tao, tao->niter, ssls->merit, ndpsi, 0.0, t)); in TaoSolve_SSILS()
|
| H A D | ssfls.c | 47 …iter: %" PetscInt_FMT ", merit: %g, ndpsi: %g\n", tao->niter, (double)ssls->merit, (double)ndpsi)); in TaoSolve_SSFLS()
49 PetscCall(TaoLogConvergenceHistory(tao, ssls->merit, ndpsi, 0.0, tao->ksp_its)); in TaoSolve_SSFLS()
50 PetscCall(TaoMonitor(tao, tao->niter, ssls->merit, ndpsi, 0.0, t)); in TaoSolve_SSFLS()
|
| H A D | ssls.h | 54 PetscReal merit; /* merit function value (norm(fischer)) */ member
|
| /petsc/src/tao/complementarity/impls/asls/ |
| H A D | asils.c | 74 PetscCall(VecNorm(asls->ff, NORM_2, &asls->merit)); in Tao_ASLS_FunctionGradient()
75 *fcn = 0.5 * asls->merit * asls->merit; in Tao_ASLS_FunctionGradient()
134 …er %" PetscInt_FMT ", merit: %g, ||dpsi||: %g\n", tao->niter, (double)asls->merit, (double)ndpsi)); in TaoSolve_ASILS()
135 PetscCall(TaoLogConvergenceHistory(tao, asls->merit, ndpsi, 0.0, tao->ksp_its)); in TaoSolve_ASILS()
136 PetscCall(TaoMonitor(tao, tao->niter, asls->merit, ndpsi, 0.0, t)); in TaoSolve_ASILS()
167 asls->identifier = PetscMin(asls->merit, 1e-2) / (1 + asls->identifier); in TaoSolve_ASILS()
|
| H A D | asfls.c | 74 PetscCall(VecNorm(asls->ff, NORM_2, &asls->merit)); in Tao_ASLS_FunctionGradient()
75 *fcn = 0.5 * asls->merit * asls->merit; in Tao_ASLS_FunctionGradient()
137 …er %" PetscInt_FMT ", merit: %g, ||dpsi||: %g\n", tao->niter, (double)asls->merit, (double)ndpsi)); in TaoSolve_ASFLS()
138 PetscCall(TaoLogConvergenceHistory(tao, asls->merit, ndpsi, 0.0, tao->ksp_its)); in TaoSolve_ASFLS()
139 PetscCall(TaoMonitor(tao, tao->niter, asls->merit, ndpsi, 0.0, t)); in TaoSolve_ASFLS()
170 asls->identifier = PetscMin(asls->merit, 1e-2) / (1 + asls->identifier); in TaoSolve_ASFLS()
|
| /petsc/src/snes/impls/vi/ss/ |
| H A D | vissimpl.h | 15 PetscReal merit; /* Merit function */ member
|
| H A D | viss.c | 17 PetscErrorCode SNESVIComputeMeritFunction(Vec phi, PetscReal *merit, PetscReal *phinorm) in SNESVIComputeMeritFunction() argument
22 *merit = 0.5 * (*phinorm) * (*phinorm); in SNESVIComputeMeritFunction()
236 PetscCall(SNESVIComputeMeritFunction(vi->phi, &vi->merit, &vi->phinorm)); in SNESSolve_VINEWTONSSLS()
240 SNESCheckFunctionDomainError(snes, vi->merit); in SNESSolve_VINEWTONSSLS()
324 vi->merit = 0.5 * vi->phinorm * vi->phinorm; in SNESSolve_VINEWTONSSLS()
|
| /petsc/doc/manual/ |
| H A D | tao.md | 2046 where $L(x, \lambda_k)$ is the augmented Lagrangian merit function
2188 linearized constraints and improves the augmented Lagrangian merit
2202 sufficient descent for the merit function
2212 direction for the augmented Lagrangian merit function. We then find
2213 $\alpha$ to approximately minimize the augmented Lagrangian merit
2236 If the Newton direction computed does not provide descent for the merit
2332 we now approximately minimize the augmented Lagrangian merit function
2721 Furthermore, the natural merit function,
2732 of the merit function, $-\nabla \Psi(x^k)$, as the search
|
| /petsc/doc/ |
| H A D | petsc.bib | 28386 @Article{ facchinei.soares:merit,
32637 @InProceedings{ li:merit,
37507 title = {A new unconstrained differentiable merit function for box constrained variational
|