File: itkAmoebaOptimizerTest.cxx

package info (click to toggle)
insighttoolkit5 5.4.3-5
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 704,384 kB
  • sloc: cpp: 783,592; ansic: 628,724; xml: 44,704; fortran: 34,250; python: 22,874; sh: 4,078; pascal: 2,636; lisp: 2,158; makefile: 464; yacc: 328; asm: 205; perl: 203; lex: 146; tcl: 132; javascript: 98; csh: 81
file content (580 lines) | stat: -rw-r--r-- 16,320 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
/*=========================================================================
 *
 *  Copyright NumFOCUS
 *
 *  Licensed under the Apache License, Version 2.0 (the "License");
 *  you may not use this file except in compliance with the License.
 *  You may obtain a copy of the License at
 *
 *         https://www.apache.org/licenses/LICENSE-2.0.txt
 *
 *  Unless required by applicable law or agreed to in writing, software
 *  distributed under the License is distributed on an "AS IS" BASIS,
 *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 *  See the License for the specific language governing permissions and
 *  limitations under the License.
 *
 *=========================================================================*/

#include "itkAmoebaOptimizer.h"
#include "itkTestingMacros.h"
#include "vnl/vnl_vector_fixed.h"
#include "vnl/vnl_vector.h"
#include "vnl/vnl_matrix.h"
#include "itkMath.h"
#include <iostream>

/**
 *  The objective function is the quadratic form:
 *
 *  1/2 x^T A x - b^T x
 *
 *  Where A is represented as an itkMatrix and
 *  b is represented as an itkVector
 *
 *  The system in this example is:
 *
 *     | 3  2 ||x|   | 2|   |0|
 *     | 2  6 ||y| + |-8| = |0|
 *
 *
 *   the solution is the vector | 2 -2 |
 *
 *   and the expected final value of the function is 10.0
 *
 * \class amoebaTestF1
 */
class amoebaTestF1 : public itk::SingleValuedCostFunction
{
public:
  using Self = amoebaTestF1;
  using Superclass = itk::SingleValuedCostFunction;
  using Pointer = itk::SmartPointer<Self>;
  using ConstPointer = itk::SmartPointer<const Self>;
  itkNewMacro(Self);
  itkOverrideGetNameOfClassMacro(amoebaTestF1);

  enum
  {
    SpaceDimension = 2
  };

  using ParametersType = Superclass::ParametersType;
  using DerivativeType = Superclass::DerivativeType;
  using MeasureType = Superclass::MeasureType;

  using VectorType = vnl_vector<double>;
  using MatrixType = vnl_matrix<double>;


  amoebaTestF1()
    : m_A(SpaceDimension, SpaceDimension)
    , m_B(SpaceDimension)
  {
    m_A[0][0] = 3;
    m_A[0][1] = 2;
    m_A[1][0] = 2;
    m_A[1][1] = 6;

    m_B[0] = 2;
    m_B[1] = -8;
    m_Negate = false;
  }

  double
  GetValue(const ParametersType & parameters) const override
  {

    VectorType v(parameters.Size());
    for (unsigned int i = 0; i < SpaceDimension; ++i)
    {
      v[i] = parameters[i];
    }
    VectorType Av = m_A * v;
    double     val = (inner_product<double>(Av, v)) / 2.0;
    val -= inner_product<double>(m_B, v);
    if (m_Negate)
    {
      val *= -1.0;
    }
    return val;
  }

  void
  GetDerivative(const ParametersType & parameters, DerivativeType & derivative) const override
  {

    VectorType v(parameters.Size());
    for (unsigned int i = 0; i < SpaceDimension; ++i)
    {
      v[i] = parameters[i];
    }
    std::cout << "GetDerivative( " << v << " ) = ";
    VectorType gradient = m_A * v - m_B;
    std::cout << gradient << std::endl;
    derivative = DerivativeType(SpaceDimension);
    for (unsigned int i = 0; i < SpaceDimension; ++i)
    {
      if (!m_Negate)
      {
        derivative[i] = gradient[i];
      }
      else
      {
        derivative[i] = -gradient[i];
      }
    }
  }

  unsigned int
  GetNumberOfParameters() const override
  {
    return SpaceDimension;
  }

  // Used to switch between maximization and minimization.
  void
  SetNegate(bool flag)
  {
    m_Negate = flag;
  }

private:
  MatrixType m_A;
  VectorType m_B;
  bool       m_Negate;
};


/**
 * Function we want to optimize, comprised of two parabolas with C0 continuity
 * at 0:
 * f(x) = if(x<0) x^2+4x; else 2x^2-8x
 *
 * Minima are at -2 and 2 with function values of -4 and -8 respectively.
 */
class amoebaTestF2 : public itk::SingleValuedCostFunction
{
public:
  using Self = amoebaTestF2;
  using Superclass = itk::SingleValuedCostFunction;
  using Pointer = itk::SmartPointer<Self>;
  using ConstPointer = itk::SmartPointer<const Self>;
  itkNewMacro(Self);
  itkOverrideGetNameOfClassMacro(amoebaTestF2);

  using ParametersType = Superclass::ParametersType;
  using MeasureType = Superclass::MeasureType;

  amoebaTestF2() = default;

  double
  GetValue(const ParametersType & parameters) const override
  {
    double val;
    if (parameters[0] < 0)
    {
      val = parameters[0] * parameters[0] + 4 * parameters[0];
    }
    else
    {
      val = 2 * parameters[0] * parameters[0] - 8 * parameters[0];
    }
    return val;
  }

  void
  GetDerivative(const ParametersType & itkNotUsed(parameters), DerivativeType & itkNotUsed(derivative)) const override
  {
    throw itk::ExceptionObject(__FILE__, __LINE__, "no derivative available");
  }

  unsigned int
  GetNumberOfParameters() const override
  {
    return 1;
  }
};

class CommandIterationUpdateAmoeba : public itk::Command
{
public:
  using Self = CommandIterationUpdateAmoeba;
  using Superclass = itk::Command;
  using Pointer = itk::SmartPointer<Self>;
  itkNewMacro(Self);

  void
  Reset()
  {
    m_IterationNumber = 0;
  }

  void
  Execute(itk::Object * caller, const itk::EventObject & event) override
  {
    Execute((const itk::Object *)caller, event);
  }

  void
  Execute(const itk::Object * object, const itk::EventObject & event) override
  {
    const auto * optimizer = static_cast<const itk::AmoebaOptimizer *>(object);
    if (dynamic_cast<const itk::FunctionEvaluationIterationEvent *>(&event))
    {
      std::cout << m_IterationNumber++ << ":  ";
      std::cout << "x: " << optimizer->GetCachedCurrentPosition() << "  ";
      std::cout << "f(x): " << optimizer->GetCachedValue() << std::endl;
    }
  }

protected:
  CommandIterationUpdateAmoeba() { m_IterationNumber = 0; }

private:
  unsigned long m_IterationNumber;
};

/**
 * Test Amoeba with a 2D quadratic function - happy day scenario.
 */
int
AmoebaTest1();

/**
 * Test Amoeba and Amoeba with restarts on a function with two minima.
 */
int
AmoebaTest2();

int
itkAmoebaOptimizerTest(int, char *[])
{
  int result1 = AmoebaTest1();
  int result2 = AmoebaTest2();

  std::cout << "All Tests Completed." << std::endl;

  if (result1 == EXIT_FAILURE || result2 == EXIT_FAILURE)
  {
    std::cerr << "[FAILURE]\n";
    return EXIT_FAILURE;
  }
  std::cout << "[SUCCESS]\n";
  return EXIT_SUCCESS;
}

int
AmoebaTest1()
{

  std::cout << "Amoeba Optimizer Test 1\n \n";

  using OptimizerType = itk::AmoebaOptimizer;

  // Declaration of an itkOptimizer
  auto itkOptimizer = OptimizerType::New();

  ITK_EXERCISE_BASIC_OBJECT_METHODS(itkOptimizer, AmoebaOptimizer, SingleValuedNonLinearVnlOptimizer);


  ITK_TEST_EXPECT_TRUE(itkOptimizer->CanUseScales());

  // set optimizer parameters
  typename OptimizerType::NumberOfIterationsType numberOfIterations = 10;
  itkOptimizer->SetMaximumNumberOfIterations(numberOfIterations);
  ITK_TEST_SET_GET_VALUE(numberOfIterations, itkOptimizer->GetMaximumNumberOfIterations());

  auto automaticInitialSimplex = true;
  ITK_TEST_SET_GET_BOOLEAN(itkOptimizer, AutomaticInitialSimplex, automaticInitialSimplex);

  auto optimizeWithRestarts = false;
  ITK_TEST_SET_GET_BOOLEAN(itkOptimizer, OptimizeWithRestarts, optimizeWithRestarts);

  double xTolerance = -0.01;
  itkOptimizer->SetParametersConvergenceTolerance(xTolerance);
  ITK_TEST_SET_GET_VALUE(xTolerance, itkOptimizer->GetParametersConvergenceTolerance());

  double fTolerance = -0.001;
  itkOptimizer->SetFunctionConvergenceTolerance(fTolerance);
  ITK_TEST_SET_GET_VALUE(fTolerance, itkOptimizer->GetFunctionConvergenceTolerance());

  auto costFunction = amoebaTestF1::New();
  itkOptimizer->SetCostFunction(costFunction);
  std::cout << "itkOptimizer->GetCostFunction(): " << itkOptimizer->GetCostFunction() << std::endl;

  OptimizerType::ParametersType initialValue(2); // constructor requires vector size

  initialValue[0] = 100; // We start not far from  | 2 -2 |
  initialValue[1] = -100;

  OptimizerType::ParametersType currentValue(2);

  currentValue = initialValue;

  itkOptimizer->SetInitialPosition(currentValue);

  // Test exceptions
  ITK_TRY_EXPECT_EXCEPTION(itkOptimizer->StartOptimization());

  xTolerance = 0.01;
  itkOptimizer->SetParametersConvergenceTolerance(xTolerance);
  ITK_TEST_SET_GET_VALUE(xTolerance, itkOptimizer->GetParametersConvergenceTolerance());

  ITK_TRY_EXPECT_EXCEPTION(itkOptimizer->StartOptimization());

  fTolerance = 0.001;
  itkOptimizer->SetFunctionConvergenceTolerance(fTolerance);
  ITK_TEST_SET_GET_VALUE(fTolerance, itkOptimizer->GetFunctionConvergenceTolerance());

  try
  {

    std::cout << "Run for " << itkOptimizer->GetMaximumNumberOfIterations();
    std::cout << " iterations or less." << std::endl;

    itkOptimizer->StartOptimization();


    itkOptimizer->SetMaximumNumberOfIterations(100);
    std::cout << "Continue for " << itkOptimizer->GetMaximumNumberOfIterations();
    std::cout << " iterations or less." << std::endl;
    itkOptimizer->SetInitialPosition(itkOptimizer->GetCurrentPosition());
    itkOptimizer->StartOptimization();
  }
  catch (const itk::ExceptionObject & e)
  {
    std::cerr << "Exception thrown ! " << std::endl;
    std::cerr << "An error occurred during Optimization" << std::endl;
    std::cerr << "Location    = " << e.GetLocation() << std::endl;
    std::cerr << "Description = " << e.GetDescription() << std::endl;
    std::cerr << "[TEST 1 FAILURE]\n";
    return EXIT_FAILURE;
  }


  std::cout << "Optimizer: " << itkOptimizer;

  //
  // check results to see if it is within range
  //

  OptimizerType::ParametersType finalPosition;
  finalPosition = itkOptimizer->GetCurrentPosition();

  double trueParameters[2] = { 2, -2 };
  bool   pass = true;

  std::cout << "Right answer   = " << trueParameters[0] << " , " << trueParameters[1] << std::endl;
  std::cout << "Final position = " << finalPosition << std::endl;

  for (unsigned int j = 0; j < 2; ++j)
  {
    if (itk::Math::abs(finalPosition[j] - trueParameters[j]) > xTolerance)
    {
      pass = false;
    }
  }

  if (!pass)
  {
    std::cerr << "[TEST 1 FAILURE]\n";
    return EXIT_FAILURE;
  }

  // Get the final value of the optimizer
  std::cout << "Testing optimizers GetValue() : ";
  OptimizerType::MeasureType finalValue = itkOptimizer->GetValue();
  if (itk::Math::abs(finalValue + 9.99998) > 0.01)
  {
    std::cerr << "failed\n";
    std::cerr << "[TEST 1 FAILURE]\n";
    return EXIT_FAILURE;
  }
  else
  {
    std::cout << "succeeded\n";
  }

  // Set now the function to maximize
  //
  { // add a block-scope to have local variables

    std::cout << "Testing Maximization " << std::endl;

    currentValue = initialValue;

    itkOptimizer->SetInitialPosition(currentValue);

    auto observer = CommandIterationUpdateAmoeba::New();
    itkOptimizer->AddObserver(itk::FunctionEvaluationIterationEvent(), observer);

    try
    {
      // These two following statement should compensate each other
      // and allow us to get to the same result as the test above.
      costFunction->SetNegate(true);
      itkOptimizer->MaximizeOn();

      std::cout << "Run for " << itkOptimizer->GetMaximumNumberOfIterations();
      std::cout << " iterations or less." << std::endl;

      itkOptimizer->StartOptimization();

      itkOptimizer->SetMaximumNumberOfIterations(100);
      itkOptimizer->SetInitialPosition(itkOptimizer->GetCurrentPosition());

      std::cout << "Continue for " << itkOptimizer->GetMaximumNumberOfIterations();
      std::cout << " iterations or less, starting from previous position.";
      std::cout << std::endl;
      itkOptimizer->StartOptimization();
    }
    catch (const itk::ExceptionObject & e)
    {
      std::cerr << "Exception thrown ! " << std::endl;
      std::cerr << "An error occurred during Optimization" << std::endl;
      std::cerr << "Location    = " << e.GetLocation() << std::endl;
      std::cerr << "Description = " << e.GetDescription() << std::endl;
      std::cerr << "[TEST 1 FAILURE]\n";
      return EXIT_FAILURE;
    }

    finalPosition = itkOptimizer->GetCurrentPosition();
    std::cout << "Right answer   = " << trueParameters[0] << " , " << trueParameters[1] << std::endl;
    std::cout << "Final position = " << finalPosition << std::endl;

    for (unsigned int j = 0; j < 2; ++j)
    {
      if (itk::Math::abs(finalPosition[j] - trueParameters[j]) > xTolerance)
      {
        pass = false;
      }
    }

    if (!pass)
    {
      std::cerr << "[TEST 1 FAILURE]\n";
      return EXIT_FAILURE;
    }

    // Get the final value of the optimizer
    std::cout << "Testing optimizer's GetValue() [invokes additional function evaluation]: ";
    finalValue = itkOptimizer->GetValue();
    if (itk::Math::abs(finalValue + 9.99998) > 0.01)
    {
      std::cerr << "failed\n";
      std::cerr << "[TEST 1 FAILURE]\n";
      return EXIT_FAILURE;
    }
    else
    {
      std::cout << "succeeded\n";
    }
  }
  std::cout << "[TEST 1 SUCCESS]\n";
  return EXIT_SUCCESS;
}

int
AmoebaTest2()
{
  std::cout << "Amoeba Optimizer Test 2\n \n";

  using OptimizerType = itk::AmoebaOptimizer;
  auto itkOptimizer = OptimizerType::New();

  // set optimizer parameters
  unsigned int maxIterations = 100;
  itkOptimizer->SetMaximumNumberOfIterations(maxIterations);

  double xTolerance = 0.01;
  itkOptimizer->SetParametersConvergenceTolerance(xTolerance);

  double fTolerance = 0.001;
  itkOptimizer->SetFunctionConvergenceTolerance(fTolerance);

  // the initial simplex is constructed as:
  // x,
  // x_i = [x[0], ... , x[i]+initialSimplexDelta[i], ... , x[n]]
  //
  OptimizerType::ParametersType initialSimplexDelta(1);
  initialSimplexDelta[0] = 10;
  itkOptimizer->SetInitialSimplexDelta(initialSimplexDelta);
  ITK_TEST_SET_GET_VALUE(initialSimplexDelta, itkOptimizer->GetInitialSimplexDelta());

  OptimizerType::ParametersType initialParameters(1), finalParameters;
  // starting position
  initialParameters[0] = -100;

  itkOptimizer->SetInitialPosition(initialParameters);

  // the function we want to optimize
  auto costFunction = amoebaTestF2::New();
  itkOptimizer->SetCostFunction(costFunction);

  // observe the iterations
  auto observer = CommandIterationUpdateAmoeba::New();
  itkOptimizer->AddObserver(itk::IterationEvent(), observer);

  try
  {
    itkOptimizer->StartOptimization();
  }
  catch (const itk::ExceptionObject & e)
  {
    std::cerr << "Exception thrown ! " << std::endl;
    std::cerr << "An error occurred during Optimization" << std::endl;
    std::cerr << "Location    = " << e.GetLocation() << std::endl;
    std::cerr << "Description = " << e.GetDescription() << std::endl;
    std::cerr << "[TEST 2 FAILURE]\n";
    return EXIT_FAILURE;
  }

  // we should have converged to the local minimum, -2
  finalParameters = itkOptimizer->GetCurrentPosition();
  double knownParameters = -2.0;
  std::cout << "Standard Amoeba:\n";
  std::cout << "Known parameters   = " << knownParameters << "   ";
  std::cout << "Estimated parameters = " << finalParameters << std::endl;
  std::cout << "Converged to local minimum." << std::endl;
  if (itk::Math::abs(finalParameters[0] - knownParameters) > xTolerance)
  {
    std::cerr << "[TEST 2 FAILURE]\n";
    return EXIT_FAILURE;
  }

  // run again using multiple restarts
  observer->Reset();
  itkOptimizer->SetInitialPosition(initialParameters);
  itkOptimizer->OptimizeWithRestartsOn();

  try
  {
    itkOptimizer->StartOptimization();
  }
  catch (const itk::ExceptionObject & e)
  {
    std::cerr << "Exception thrown ! " << std::endl;
    std::cerr << "An error occurred during Optimization" << std::endl;
    std::cerr << "Location    = " << e.GetLocation() << std::endl;
    std::cerr << "Description = " << e.GetDescription() << std::endl;
    std::cerr << "[TEST 2 FAILURE]\n";
    return EXIT_FAILURE;
  }

  // we should have converged to the global minimum, 2
  finalParameters = itkOptimizer->GetCurrentPosition();
  knownParameters = 2.0;
  std::cout << "Amoeba with restarts:\n";
  std::cout << "Known parameters   = " << knownParameters << "   ";
  std::cout << "Estimated parameters = " << finalParameters << std::endl;
  std::cout << "Converged to global minimum." << std::endl;

  if (itk::Math::abs(finalParameters[0] - knownParameters) > xTolerance)
  {
    std::cerr << "[TEST 2 FAILURE]\n";
    return EXIT_FAILURE;
  }
  std::cout << "[TEST 1 SUCCESS]\n";
  return EXIT_SUCCESS;
}