ROL
ROL_BundleStep.hpp
Go to the documentation of this file.
1 // @HEADER
2 // ************************************************************************
3 //
4 // Rapid Optimization Library (ROL) Package
5 // Copyright (2014) Sandia Corporation
6 //
7 // Under terms of Contract DE-AC04-94AL85000, there is a non-exclusive
8 // license for use of this work by or on behalf of the U.S. Government.
9 //
10 // Redistribution and use in source and binary forms, with or without
11 // modification, are permitted provided that the following conditions are
12 // met:
13 //
14 // 1. Redistributions of source code must retain the above copyright
15 // notice, this list of conditions and the following disclaimer.
16 //
17 // 2. Redistributions in binary form must reproduce the above copyright
18 // notice, this list of conditions and the following disclaimer in the
19 // documentation and/or other materials provided with the distribution.
20 //
21 // 3. Neither the name of the Corporation nor the names of the
22 // contributors may be used to endorse or promote products derived from
23 // this software without specific prior written permission.
24 //
25 // THIS SOFTWARE IS PROVIDED BY SANDIA CORPORATION "AS IS" AND ANY
26 // EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
28 // PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SANDIA CORPORATION OR THE
29 // CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
30 // EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
31 // PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
32 // PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
33 // LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
34 // NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
35 // SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
36 //
37 // Questions? Contact lead developers:
38 // Drew Kouri (dpkouri@sandia.gov) and
39 // Denis Ridzal (dridzal@sandia.gov)
40 //
41 // ************************************************************************
42 // @HEADER
43 
44 #ifndef ROL_BUNDLE_STEP_H
45 #define ROL_BUNDLE_STEP_H
46 
47 #include "ROL_Bundle_AS.hpp"
48 #include "ROL_Bundle_TT.hpp"
49 #include "ROL_Types.hpp"
50 #include "ROL_Step.hpp"
51 #include "ROL_Vector.hpp"
52 #include "ROL_Objective.hpp"
53 #include "ROL_BoundConstraint.hpp"
54 #include "ROL_LineSearch.hpp"
55 
56 #include "ROL_ParameterList.hpp"
57 #include "ROL_Ptr.hpp"
58 
64 namespace ROL {
65 
66 template <class Real>
67 class BundleStep : public Step<Real> {
68 private:
69  // Bundle
70  ROL::Ptr<Bundle<Real> > bundle_; // Bundle of subgradients and linearization errors
71  ROL::Ptr<LineSearch<Real> > lineSearch_; // Line-search object for nonconvex problems
72 
73  // Dual cutting plane solution
74  unsigned QPiter_; // Number of QP solver iterations
75  unsigned QPmaxit_; // Maximum number of QP iterations
76  Real QPtol_; // QP subproblem tolerance
77 
78  // Step flag
79  int step_flag_; // Whether serious or null step
80 
81  // Additional storage
82  ROL::Ptr<Vector<Real> > y_;
83 
84  // Updated iterate storage
85  Real linErrNew_;
86  Real valueNew_;
87 
88  // Aggregate subgradients, linearizations, and distance measures
89  ROL::Ptr<Vector<Real> > aggSubGradNew_; // New aggregate subgradient
90  Real aggSubGradOldNorm_; // Old aggregate subgradient norm
91  Real aggLinErrNew_; // New aggregate linearization error
92  Real aggLinErrOld_; // Old aggregate linearization error
93  Real aggDistMeasNew_; // New aggregate distance measure
94 
95  // Algorithmic parameters
96  Real T_;
97  Real tol_;
98  Real m1_;
99  Real m2_;
100  Real m3_;
101  Real nu_;
102 
103  // Line-search parameters
105 
107  bool isConvex_;
108 
109  Real ftol_;
110 
112 
113 public:
114 
116  using Step<Real>::compute;
117  using Step<Real>::update;
118 
119  BundleStep(ROL::ParameterList &parlist)
120  : bundle_(ROL::nullPtr), lineSearch_(ROL::nullPtr),
121  QPiter_(0), QPmaxit_(0), QPtol_(0), step_flag_(0),
122  y_(ROL::nullPtr), linErrNew_(0), valueNew_(0),
123  aggSubGradNew_(ROL::nullPtr), aggSubGradOldNorm_(0),
125  T_(0), tol_(0), m1_(0), m2_(0), m3_(0), nu_(0),
126  ls_maxit_(0), first_print_(true), isConvex_(false),
127  ftol_(ROL_EPSILON<Real>()) {
128  Real zero(0), two(2), oem3(1.e-3), oem6(1.e-6), oem8(1.e-8), p1(0.1), p2(0.2), p9(0.9), oe3(1.e3), oe8(1.e8);
129  ROL::Ptr<StepState<Real> > state = Step<Real>::getState();
130  state->searchSize = parlist.sublist("Step").sublist("Bundle").get("Initial Trust-Region Parameter", oe3);
131  T_ = parlist.sublist("Step").sublist("Bundle").get("Maximum Trust-Region Parameter", oe8);
132  tol_ = parlist.sublist("Step").sublist("Bundle").get("Epsilon Solution Tolerance", oem6);
133  m1_ = parlist.sublist("Step").sublist("Bundle").get("Upper Threshold for Serious Step", p1);
134  m2_ = parlist.sublist("Step").sublist("Bundle").get("Lower Threshold for Serious Step", p2);
135  m3_ = parlist.sublist("Step").sublist("Bundle").get("Upper Threshold for Null Step", p9);
136  nu_ = parlist.sublist("Step").sublist("Bundle").get("Tolerance for Trust-Region Parameter", oem3);
137 
138  // Initialize bundle
139  Real coeff = parlist.sublist("Step").sublist("Bundle").get("Distance Measure Coefficient", zero);
140  Real omega = parlist.sublist("Step").sublist("Bundle").get("Locality Measure Coefficient", two);
141  unsigned maxSize = parlist.sublist("Step").sublist("Bundle").get("Maximum Bundle Size", 200);
142  unsigned remSize = parlist.sublist("Step").sublist("Bundle").get("Removal Size for Bundle Update", 2);
143  if ( parlist.sublist("Step").sublist("Bundle").get("Cutting Plane Solver",0) == 1 ) {
144  bundle_ = ROL::makePtr<Bundle_TT<Real>>(maxSize,coeff,omega,remSize);
145  //bundle_ = ROL::makePtr<Bundle_AS<Real>>(maxSize,coeff,omega,remSize);
146  }
147  else {
148  bundle_ = ROL::makePtr<Bundle_AS<Real>>(maxSize,coeff,omega,remSize);
149  }
150  isConvex_ = ((coeff == zero) ? true : false);
151 
152  // Initialize QP solver
153  QPtol_ = parlist.sublist("Step").sublist("Bundle").get("Cutting Plane Tolerance", oem8);
154  QPmaxit_ = parlist.sublist("Step").sublist("Bundle").get("Cutting Plane Iteration Limit", 1000);
155 
156  // Initialize Line Search
157  ls_maxit_
158  = parlist.sublist("Step").sublist("Line Search").get("Maximum Number of Function Evaluations",20);
159  if ( !isConvex_ ) {
160  lineSearch_ = LineSearchFactory<Real>(parlist);
161  }
162 
163  // Get verbosity level
164  verbosity_ = parlist.sublist("General").get("Print Verbosity", 0);
165  }
166 
167  void initialize( Vector<Real> &x, const Vector<Real> &g,
169  AlgorithmState<Real> &algo_state ) {
170  // Call default initializer, but maintain current searchSize
171  ROL::Ptr<StepState<Real> > state = Step<Real>::getState();
172  Real searchSize = state->searchSize;
173  Step<Real>::initialize(x,x,g,obj,con,algo_state);
174  state->searchSize = searchSize;
175  // Initialize bundle
176  bundle_->initialize(*(state->gradientVec));
177  // Initialize storage for updated iterate
178  y_ = x.clone();
179  // Initialize storage for aggregate subgradients
180  aggSubGradNew_ = g.clone();
181  aggSubGradOldNorm_ = algo_state.gnorm;
182  // Initialize line search
183  if ( !isConvex_ ) {
184  lineSearch_->initialize(x,x,g,obj,con);
185  }
186  }
187 
189  BoundConstraint<Real> &con, AlgorithmState<Real> &algo_state ) {
190  ROL::Ptr<StepState<Real> > state = Step<Real>::getState();
191  first_print_ = false; // Print header only on first serious step
192  QPiter_ = (step_flag_==1 ? 0 : QPiter_); // Reset QPiter only on serious steps
193  Real v(0), l(0), u = T_, gd(0); // Scalar storage
194  Real zero(0), two(2), half(0.5);
195  bool flag = true;
196  while (flag) {
197  /*************************************************************/
198  /******** Solve Dual Cutting Plane QP Problem ****************/
199  /*************************************************************/
200  QPiter_ += bundle_->solveDual(state->searchSize,QPmaxit_,QPtol_); // Solve QP subproblem
201  bundle_->aggregate(*aggSubGradNew_,aggLinErrNew_,aggDistMeasNew_); // Compute aggregate info
202  algo_state.aggregateGradientNorm = aggSubGradNew_->norm(); // Aggregate subgradient norm
203  if (verbosity_ > 0) {
204  std::cout << std::endl;
205  std::cout << " Computation of aggregrate quantities" << std::endl;
206  std::cout << " Aggregate subgradient norm: " << algo_state.aggregateGradientNorm << std::endl;
207  std::cout << " Aggregate linearization error: " << aggLinErrNew_ << std::endl;
208  std::cout << " Aggregate distance measure: " << aggDistMeasNew_ << std::endl;
209  }
210  /*************************************************************/
211  /******** Construct Cutting Plane Solution *******************/
212  /*************************************************************/
213  v = -state->searchSize*std::pow(algo_state.aggregateGradientNorm,two)-aggLinErrNew_; // CP objective value
214  s.set(aggSubGradNew_->dual()); s.scale(-state->searchSize); // CP solution
215  algo_state.snorm = state->searchSize*algo_state.aggregateGradientNorm; // Step norm
216  if (verbosity_ > 0) {
217  std::cout << std::endl;
218  std::cout << " Solve cutting plan subproblem" << std::endl;
219  std::cout << " Cutting plan objective value: " << v << std::endl;
220  std::cout << " Norm of computed step: " << algo_state.snorm << std::endl;
221  std::cout << " 'Trust-region' radius: " << state->searchSize << std::endl;
222  }
223  /*************************************************************/
224  /******** Decide Whether Step is Serious or Null *************/
225  /*************************************************************/
226  if (std::max(algo_state.aggregateGradientNorm,aggLinErrNew_) <= tol_) {
227  // Current iterate is already epsilon optimal!
228  s.zero(); algo_state.snorm = zero;
229  flag = false;
230  step_flag_ = 1;
231  algo_state.flag = true;
232  break;
233  }
234  else if (std::isnan(algo_state.aggregateGradientNorm)
235  || std::isnan(aggLinErrNew_)
236  || (std::isnan(aggDistMeasNew_) && !isConvex_)) {
237  s.zero(); algo_state.snorm = zero;
238  flag = false;
239  step_flag_ = 2;
240  algo_state.flag = true;
241  }
242  else {
243  // Current iterate is not epsilon optimal.
244  y_->set(x); y_->plus(s); // y is the candidate iterate
245  obj.update(*y_,true,algo_state.iter); // Update objective at y
246  valueNew_ = obj.value(*y_,ftol_); // Compute objective value at y
247  algo_state.nfval++;
248  obj.gradient(*(state->gradientVec),*y_,ftol_); // Compute objective (sub)gradient at y
249  algo_state.ngrad++;
250  // Compute new linearization error and distance measure
251  gd = s.dot(state->gradientVec->dual());
252  linErrNew_ = algo_state.value - (valueNew_ - gd); // Linearization error
253  // Determine whether to take a serious or null step
254  Real eps = static_cast<Real>(10)*ROL_EPSILON<Real>();
255  Real del = eps*std::max(static_cast<Real>(1),std::abs(algo_state.value));
256  Real Df = (valueNew_ - algo_state.value) - del;
257  Real Dm = v - del;
258  bool SS1 = false;
259  if (std::abs(Df) < eps && std::abs(Dm) < eps) {
260  SS1 = true;
261  }
262  else {
263  SS1 = (Df < m1_*Dm);
264  }
265  //bool SS1 = (valueNew_-algo_state.value < m1_*v);
266  //bool NS1 = (valueNew_-algo_state.value >= m1_*v);
267  bool NS2a = (bundle_->computeAlpha(algo_state.snorm,linErrNew_) <= m3_*aggLinErrOld_);
268  bool NS2b = (std::abs(algo_state.value-valueNew_) <= aggSubGradOldNorm_ + aggLinErrOld_);
269  if (verbosity_ > 0) {
270  std::cout << std::endl;
271  std::cout << " Check for serious/null step" << std::endl;
272  std::cout << " Serious step test SS(i): " << SS1 << std::endl;
273  std::cout << " -> Left hand side: " << valueNew_-algo_state.value << std::endl;
274  std::cout << " -> Right hand side: " << m1_*v << std::endl;
275  std::cout << " Null step test NS(iia): " << NS2a << std::endl;
276  std::cout << " -> Left hand side: " << bundle_->computeAlpha(algo_state.snorm,linErrNew_) << std::endl;
277  std::cout << " -> Right hand side: " << m3_*aggLinErrOld_ << std::endl;
278  std::cout << " Null step test NS(iib): " << NS2b << std::endl;
279  std::cout << " -> Left hand side: " << std::abs(algo_state.value-valueNew_) << std::endl;
280  std::cout << " -> Right hand side: " << aggSubGradOldNorm_ + aggLinErrOld_ << std::endl;
281  }
282  if ( isConvex_ ) {
283  /************* Convex objective ****************/
284  if ( SS1 ) {
285  bool SS2 = (gd >= m2_*v || state->searchSize >= T_-nu_);
286  if (verbosity_ > 0) {
287  std::cout << " Serious step test SS(iia): " << (gd >= m2_*v) << std::endl;
288  std::cout << " -> Left hand side: " << gd << std::endl;
289  std::cout << " -> Right hand side: " << m2_*v << std::endl;
290  std::cout << " Serious step test SS(iia): " << (state->searchSize >= T_-nu_) << std::endl;
291  std::cout << " -> Left hand side: " << state->searchSize << std::endl;
292  std::cout << " -> Right hand side: " << T_-nu_ << std::endl;
293  }
294  if ( SS2 ) { // Serious Step
295  step_flag_ = 1;
296  flag = false;
297  if (verbosity_ > 0) {
298  std::cout << " Serious step taken" << std::endl;
299  }
300  break;
301  }
302  else { // Increase trust-region radius
303  l = state->searchSize;
304  state->searchSize = half*(u+l);
305  if (verbosity_ > 0) {
306  std::cout << " Increase 'trust-region' radius: " << state->searchSize << std::endl;
307  }
308  }
309  }
310  else {
311  if ( NS2a || NS2b ) { // Null step
312  s.zero(); algo_state.snorm = zero;
313  step_flag_ = 0;
314  flag = false;
315  if (verbosity_ > 0) {
316  std::cout << " Null step taken" << std::endl;
317  }
318  break;
319  }
320  else { // Decrease trust-region radius
321  u = state->searchSize;
322  state->searchSize = half*(u+l);
323  if (verbosity_ > 0) {
324  std::cout << " Decrease 'trust-region' radius: " << state->searchSize << std::endl;
325  }
326  }
327  }
328  }
329  else {
330  /************* Nonconvex objective *************/
331  bool NS3 = (gd - bundle_->computeAlpha(algo_state.snorm,linErrNew_) >= m2_*v);
332  if (verbosity_ > 0) {
333  std::cout << " Null step test NS(iii): " << NS3 << std::endl;
334  std::cout << " -> Left hand side: " << gd - bundle_->computeAlpha(algo_state.snorm,linErrNew_) << std::endl;
335  std::cout << " -> Right hand side: " << m2_*v << std::endl;
336  }
337  if ( SS1 ) { // Serious step
338  step_flag_ = 1;
339  flag = false;
340  break;
341  }
342  else {
343  if ( NS2a || NS2b ) {
344  if ( NS3 ) { // Null step
345  s.zero();
346  step_flag_ = 0;
347  flag = false;
348  break;
349  }
350  else {
351  if ( NS2b ) { // Line search
352  Real alpha = zero;
353  int ls_nfval = 0, ls_ngrad = 0;
354  lineSearch_->run(alpha,valueNew_,ls_nfval,ls_ngrad,gd,s,x,obj,con);
355  if ( ls_nfval == ls_maxit_ ) { // Null step
356  s.zero();
357  step_flag_ = 0;
358  flag = false;
359  break;
360  }
361  else { // Serious step
362  s.scale(alpha);
363  step_flag_ = 1;
364  flag = false;
365  break;
366  }
367  }
368  else { // Decrease trust-region radius
369  u = state->searchSize;
370  state->searchSize = half*(u+l);
371  }
372  }
373  }
374  else { // Decrease trust-region radius
375  u = state->searchSize;
376  state->searchSize = half*(u+l);
377  }
378  }
379  }
380  }
381  } // End While
382  /*************************************************************/
383  /******** Update Algorithm State *****************************/
384  /*************************************************************/
385  algo_state.aggregateModelError = aggLinErrNew_;
388  } // End Compute
389 
390  void update( Vector<Real> &x, const Vector<Real> &s, Objective<Real> &obj,
391  BoundConstraint<Real> &con, AlgorithmState<Real> &algo_state ) {
392  ROL::Ptr<StepState<Real> > state = Step<Real>::getState();
393  state->flag = step_flag_;
394  state->SPiter = QPiter_;
395  if ( !algo_state.flag ) {
396  /*************************************************************/
397  /******** Reset Bundle If Maximum Size Reached ***************/
398  /*************************************************************/
399  bundle_->reset(*aggSubGradNew_,aggLinErrNew_,algo_state.snorm);
400  /*************************************************************/
401  /******** Update Bundle with Step Information ****************/
402  /*************************************************************/
403  if ( step_flag_==1 ) {
404  // Serious step was taken
405  x.plus(s); // Update iterate
406  Real valueOld = algo_state.value; // Store previous objective value
407  algo_state.value = valueNew_; // Add new objective value to state
408  bundle_->update(step_flag_,valueNew_-valueOld,algo_state.snorm,*(state->gradientVec),s);
409  }
410  else if ( step_flag_==0 ) {
411  // Null step was taken
412  bundle_->update(step_flag_,linErrNew_,algo_state.snorm,*(state->gradientVec),s);
413  }
414  }
415  /*************************************************************/
416  /******** Update Algorithm State *****************************/
417  /*************************************************************/
418  algo_state.iterateVec->set(x);
419  algo_state.gnorm = (state->gradientVec)->norm();
420  if ( step_flag_==1 ) {
421  algo_state.iter++;
422  }
423  } // End Update
424 
425  std::string printHeader( void ) const {
426  std::stringstream hist;
427  hist << " ";
428  hist << std::setw(6) << std::left << "iter";
429  hist << std::setw(15) << std::left << "value";
430  hist << std::setw(15) << std::left << "gnorm";
431  hist << std::setw(15) << std::left << "snorm";
432  hist << std::setw(10) << std::left << "#fval";
433  hist << std::setw(10) << std::left << "#grad";
434  hist << std::setw(15) << std::left << "znorm";
435  hist << std::setw(15) << std::left << "alpha";
436  hist << std::setw(15) << std::left << "TRparam";
437  hist << std::setw(10) << std::left << "QPiter";
438  hist << "\n";
439  return hist.str();
440  }
441 
442  std::string printName( void ) const {
443  std::stringstream hist;
444  hist << "\n" << "Bundle Trust-Region Algorithm \n";
445  return hist.str();
446  }
447 
448  std::string print( AlgorithmState<Real> &algo_state, bool print_header = false ) const {
449  const ROL::Ptr<const StepState<Real> > state = Step<Real>::getStepState();
450  std::stringstream hist;
451  hist << std::scientific << std::setprecision(6);
452  if ( algo_state.iter == 0 && first_print_ ) {
453  hist << printName();
454  if ( print_header ) {
455  hist << printHeader();
456  }
457  hist << " ";
458  hist << std::setw(6) << std::left << algo_state.iter;
459  hist << std::setw(15) << std::left << algo_state.value;
460  hist << std::setw(15) << std::left << algo_state.gnorm;
461  hist << "\n";
462  }
463  if ( step_flag_==1 && algo_state.iter > 0 ) {
464  if ( print_header ) {
465  hist << printHeader();
466  }
467  else {
468  hist << " ";
469  hist << std::setw(6) << std::left << algo_state.iter;
470  hist << std::setw(15) << std::left << algo_state.value;
471  hist << std::setw(15) << std::left << algo_state.gnorm;
472  hist << std::setw(15) << std::left << algo_state.snorm;
473  hist << std::setw(10) << std::left << algo_state.nfval;
474  hist << std::setw(10) << std::left << algo_state.ngrad;
475  hist << std::setw(15) << std::left << algo_state.aggregateGradientNorm;
476  hist << std::setw(15) << std::left << algo_state.aggregateModelError;
477  hist << std::setw(15) << std::left << state->searchSize;
478  hist << std::setw(10) << std::left << QPiter_;
479  hist << "\n";
480  }
481  }
482  return hist.str();
483  };
484 
485 }; // class BundleStep
486 
487 } // namespace ROL
488 
489 #endif
Provides the interface to evaluate objective functions.
virtual void scale(const Real alpha)=0
Compute where .
virtual ROL::Ptr< Vector > clone() const =0
Clone to make a new (uninitialized) vector.
ROL::Ptr< Vector< Real > > y_
virtual void plus(const Vector &x)=0
Compute , where .
virtual Real value(const Vector< Real > &x, Real &tol)=0
Compute value.
Provides the interface to compute optimization steps.
Definition: ROL_Step.hpp:68
Contains definitions of custom data types in ROL.
BundleStep(ROL::ParameterList &parlist)
std::string printName(void) const
Print step name.
virtual void zero()
Set to zero vector.
Definition: ROL_Vector.hpp:167
Defines the linear algebra or vector space interface.
Definition: ROL_Vector.hpp:80
virtual Real dot(const Vector &x) const =0
Compute where .
Objective_SerialSimOpt(const Ptr< Obj > &obj, const V &ui) z0_ zero()
State for algorithm class. Will be used for restarts.
Definition: ROL_Types.hpp:143
virtual void gradient(Vector< Real > &g, const Vector< Real > &x, Real &tol)
Compute gradient.
ROL::Ptr< StepState< Real > > getState(void)
Definition: ROL_Step.hpp:73
void initialize(Vector< Real > &x, const Vector< Real > &g, Objective< Real > &obj, BoundConstraint< Real > &con, AlgorithmState< Real > &algo_state)
Initialize step with bound constraint.
ROL::Ptr< Vector< Real > > iterateVec
Definition: ROL_Types.hpp:157
std::string printHeader(void) const
Print iterate header.
Provides the interface to apply upper and lower bound constraints.
void update(Vector< Real > &x, const Vector< Real > &s, Objective< Real > &obj, BoundConstraint< Real > &con, AlgorithmState< Real > &algo_state)
Update step, if successful.
std::string print(AlgorithmState< Real > &algo_state, bool print_header=false) const
Print iterate status.
void compute(Vector< Real > &s, const Vector< Real > &x, Objective< Real > &obj, BoundConstraint< Real > &con, AlgorithmState< Real > &algo_state)
Compute step.
Provides the interface to compute bundle trust-region steps.
ROL::Ptr< Vector< Real > > aggSubGradNew_
virtual void initialize(Vector< Real > &x, const Vector< Real > &g, Objective< Real > &obj, BoundConstraint< Real > &con, AlgorithmState< Real > &algo_state)
Initialize step with bound constraint.
Definition: ROL_Step.hpp:88
virtual void set(const Vector &x)
Set where .
Definition: ROL_Vector.hpp:209
virtual void update(const Vector< Real > &x, bool flag=true, int iter=-1)
Update objective function.
Real ROL_EPSILON(void)
Platform-dependent machine epsilon.
Definition: ROL_Types.hpp:91
ROL::Ptr< Bundle< Real > > bundle_
ROL::Ptr< LineSearch< Real > > lineSearch_
const ROL::Ptr< const StepState< Real > > getStepState(void) const
Get state for step object.
Definition: ROL_Step.hpp:211