xref: /petsc/src/tao/bound/impls/bnk/bntr.c (revision 0875260307e97403c513ee2ad2ca062f01ff5a8a)
1 #include <../src/tao/bound/impls/bnk/bnk.h>
2 #include <petscksp.h>
3 
4 /*
5  Implements Newton's Method with a trust region approach for solving
6  bound constrained minimization problems.
7 
8  ------------------------------------------------------------
9 
10  initialize trust radius (default: BNK_INIT_INTERPOLATION)
11  x_0 = VecMedian(x_0)
12  f_0, g_0 = TaoComputeObjectiveAndGradient(x_0)
13  pg_0 = VecBoundGradientProjection(g_0)
14  check convergence at pg_0
15  niter = 0
16  step_accepted = true
17 
18  while niter <= max_it
19     if step_accepted
20       niter += 1
21       H_k = TaoComputeHessian(x_k)
22       if pc_type == BNK_PC_BFGS
23         add correction to BFGS approx
24         if scale_type == BNK_SCALE_AHESS
25           D = VecMedian(1e-6, abs(diag(H_k)), 1e6)
26           scale BFGS with VecReciprocal(D)
27         end
28       end
29     end
30 
31     if pc_type = BNK_PC_BFGS
32       B_k = BFGS
33     else
34       B_k = VecMedian(1e-6, abs(diag(H_k)), 1e6)
35       B_k = VecReciprocal(B_k)
36     end
37     w = x_k - VecMedian(x_k - 0.001*B_k*g_k)
38     eps = min(eps, norm2(w))
39     determine the active and inactive index sets such that
40       L = {i : (x_k)_i <= l_i + eps && (g_k)_i > 0}
41       U = {i : (x_k)_i >= u_i - eps && (g_k)_i < 0}
42       F = {i : l_i = (x_k)_i = u_i}
43       A = {L + U + F}
44       I = {i : i not in A}
45 
46     generate the reduced system Hr_k dr_k = -gr_k for variables in I
47     if pc_type == BNK_PC_BFGS && scale_type == BNK_SCALE_PHESS
48       D = VecMedian(1e-6, abs(diag(Hr_k)), 1e6)
49       scale BFGS with VecReciprocal(D)
50     end
51     solve Hr_k dr_k = -gr_k
52     set d_k to (l - x) for variables in L, (u - x) for variables in U, and 0 for variables in F
53 
54     x_{k+1} = VecMedian(x_k + d_k)
55     s = x_{k+1} - x_k
56     prered = dot(s, 0.5*gr_k - Hr_k*s)
57     f_{k+1} = TaoComputeObjective(x_{k+1})
58     actred = f_k - f_{k+1}
59 
60     oldTrust = trust
61     step_accepted, trust = TaoBNKUpdateTrustRadius(default: BNK_UPDATE_REDUCTION)
62     if step_accepted
63       g_{k+1} = TaoComputeGradient(x_{k+1})
64       pg_{k+1} = VecBoundGradientProjection(g_{k+1})
65       count the accepted Newton step
66     else
67       f_{k+1} = f_k
68       x_{k+1} = x_k
69       g_{k+1} = g_k
70       pg_{k+1} = pg_k
71       if trust == oldTrust
72         terminate because we cannot shrink the radius any further
73       end
74     end
75 
76     check convergence at pg_{k+1}
77  end
78 */
79 
80 static PetscErrorCode TaoSolve_BNTR(Tao tao)
81 {
82   PetscErrorCode               ierr;
83   TAO_BNK                      *bnk = (TAO_BNK *)tao->data;
84   KSPConvergedReason           ksp_reason;
85 
86   PetscReal                    resnorm, oldTrust, prered, actred, stepNorm, steplen;
87   PetscBool                    cgTerminate, stepAccepted = PETSC_TRUE, shift = PETSC_FALSE;
88   PetscInt                     stepType = BNK_NEWTON;
89 
90   PetscFunctionBegin;
91   /* Initialize the preconditioner, KSP solver and trust radius/line search */
92   tao->reason = TAO_CONTINUE_ITERATING;
93   ierr = TaoBNKInitialize(tao, bnk->init_type, &stepAccepted);CHKERRQ(ierr);
94   if (tao->reason != TAO_CONTINUE_ITERATING) PetscFunctionReturn(0);
95 
96   /* Have not converged; continue with Newton method */
97   if (!stepAccepted) {tao->niter = 1;}
98   while (tao->reason == TAO_CONTINUE_ITERATING) {
99 
100     if (stepAccepted) {
101       tao->niter++;
102       tao->ksp_its=0;
103       /* Compute the hessian and update the BFGS preconditioner at the new iterate */
104       ierr = TaoBNKComputeHessian(tao);CHKERRQ(ierr);
105     }
106 
107     /* Take BNCG steps (if enabled) to trade-off Hessian evaluations for more gradient evaluations */
108     ierr = TaoBNKTakeCGSteps(tao, &cgTerminate);CHKERRQ(ierr);
109     if (cgTerminate) {
110       tao->reason = bnk->bncg->reason;
111       PetscFunctionReturn(0);
112     }
113 
114     /* Use the common BNK kernel to compute the Newton step (for inactive variables only) */
115     ierr = TaoBNKComputeStep(tao, shift, &ksp_reason);CHKERRQ(ierr);
116 
117     /* Store current solution before it changes */
118     oldTrust = tao->trust;
119     bnk->fold = bnk->f;
120     ierr = VecCopy(tao->solution, bnk->Xold);CHKERRQ(ierr);
121     ierr = VecCopy(tao->gradient, bnk->Gold);CHKERRQ(ierr);
122     ierr = VecCopy(bnk->unprojected_gradient, bnk->unprojected_gradient_old);CHKERRQ(ierr);
123 
124     /* Temporarily accept the step and project it into the bounds */
125     ierr = VecAXPY(tao->solution, 1.0, tao->stepdirection);CHKERRQ(ierr);
126     ierr = VecMedian(tao->XL, tao->solution, tao->XU, tao->solution);CHKERRQ(ierr);
127 
128     /* Check if the projection changed the step direction */
129     ierr = VecCopy(tao->solution, tao->stepdirection);CHKERRQ(ierr);
130     ierr = VecAXPY(tao->stepdirection, -1.0, bnk->Xold);CHKERRQ(ierr);
131     ierr = VecNorm(tao->stepdirection, NORM_2, &stepNorm);CHKERRQ(ierr);
132     if (stepNorm != bnk->dnorm) {
133       /* Projection changed the step, so we have to recompute predicted reduction.
134          However, we deliberately do not change the step norm and the trust radius
135          in order for the safeguard to more closely mimic a piece-wise linesearch
136          along the bounds. */
137       ierr = TaoBNKRecomputePred(tao, tao->stepdirection, &prered);CHKERRQ(ierr);
138     } else {
139       /* Step did not change, so we can just recover the pre-computed prediction */
140       ierr = KSPCGGetObjFcn(tao->ksp, &prered);CHKERRQ(ierr);
141     }
142     prered = -prered;
143 
144     /* Compute the actual reduction and update the trust radius */
145     ierr = TaoComputeObjective(tao, tao->solution, &bnk->f);CHKERRQ(ierr);
146     actred = bnk->fold - bnk->f;
147     ierr = TaoBNKUpdateTrustRadius(tao, prered, actred, bnk->update_type, stepType, &stepAccepted);CHKERRQ(ierr);
148 
149     if (stepAccepted) {
150       /* Step is good, evaluate the gradient and the hessian */
151       steplen = 1.0;
152       ++bnk->newt;
153       ierr = TaoComputeGradient(tao, tao->solution, bnk->unprojected_gradient);CHKERRQ(ierr);
154       ierr = VecBoundGradientProjection(bnk->unprojected_gradient,tao->solution,tao->XL,tao->XU,tao->gradient);CHKERRQ(ierr);
155       ierr = VecNorm(tao->gradient, NORM_2, &bnk->gnorm);CHKERRQ(ierr);
156       if (PetscIsInfOrNanReal(bnk->gnorm)) SETERRQ(PETSC_COMM_SELF,1,"User provided compute function generated Not-a-Number");
157     } else {
158       /* Step is bad, revert old solution and re-solve with new radius*/
159       steplen = 0.0;
160       bnk->f = bnk->fold;
161       ierr = VecCopy(bnk->Xold, tao->solution);CHKERRQ(ierr);
162       ierr = VecCopy(bnk->Gold, tao->gradient);CHKERRQ(ierr);
163       ierr = VecCopy(bnk->unprojected_gradient_old, bnk->unprojected_gradient);CHKERRQ(ierr);
164       if (oldTrust == tao->trust) {
165         /* Can't change the radius anymore so just terminate */
166         tao->reason = TAO_DIVERGED_TR_REDUCTION;
167       }
168     }
169 
170     /*  Check for termination */
171     ierr = VecFischer(tao->solution, bnk->unprojected_gradient, tao->XL, tao->XU, bnk->Gwork);CHKERRQ(ierr);
172     ierr = VecNorm(bnk->Gwork, NORM_2, &resnorm);CHKERRQ(ierr);
173     ierr = TaoLogConvergenceHistory(tao, bnk->f, resnorm, 0.0, tao->ksp_its);CHKERRQ(ierr);
174     ierr = TaoMonitor(tao, tao->niter, bnk->f, resnorm, 0.0, steplen);CHKERRQ(ierr);
175     ierr = (*tao->ops->convergencetest)(tao, tao->cnvP);CHKERRQ(ierr);
176   }
177   PetscFunctionReturn(0);
178 }
179 
180 /*------------------------------------------------------------*/
181 
182 PETSC_INTERN PetscErrorCode TaoSetUp_BNTR(Tao tao)
183 {
184   TAO_BNK        *bnk = (TAO_BNK *)tao->data;
185   PetscErrorCode ierr;
186 
187   PetscFunctionBegin;
188   ierr = TaoSetUp_BNK(tao);CHKERRQ(ierr);
189   if (!bnk->is_nash && !bnk->is_stcg && !bnk->is_gltr) SETERRQ(PETSC_COMM_SELF,1,"Must use a trust-region CG method for KSP (KSPNASH, KSPSTCG, KSPGLTR)");
190   PetscFunctionReturn(0);
191 }
192 
193 /*------------------------------------------------------------*/
194 
195 PETSC_INTERN PetscErrorCode TaoCreate_BNTR(Tao tao)
196 {
197   TAO_BNK        *bnk;
198   PetscErrorCode ierr;
199 
200   PetscFunctionBegin;
201   ierr = TaoCreate_BNK(tao);CHKERRQ(ierr);
202   tao->ops->solve=TaoSolve_BNTR;
203   tao->ops->setup=TaoSetUp_BNTR;
204 
205   bnk = (TAO_BNK *)tao->data;
206   bnk->update_type = BNK_UPDATE_REDUCTION; /* trust region updates based on predicted/actual reduction */
207   PetscFunctionReturn(0);
208 }