• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // This file is part of Eigen, a lightweight C++ template library
2 // for linear algebra.
3 //
4 // Copyright (C) 2008-2014 Gael Guennebaud <gael.guennebaud@inria.fr>
5 // Copyright (C) 2009 Benoit Jacob <jacob.benoit.1@gmail.com>
6 //
7 // This Source Code Form is subject to the terms of the Mozilla
8 // Public License v. 2.0. If a copy of the MPL was not distributed
9 // with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
10 
11 #ifndef SVD_DEFAULT
12 #error a macro SVD_DEFAULT(MatrixType) must be defined prior to including svd_common.h
13 #endif
14 
15 #ifndef SVD_FOR_MIN_NORM
16 #error a macro SVD_FOR_MIN_NORM(MatrixType) must be defined prior to including svd_common.h
17 #endif
18 
19 #include "svd_fill.h"
20 #include "solverbase.h"
21 
22 // Check that the matrix m is properly reconstructed and that the U and V factors are unitary
23 // The SVD must have already been computed.
24 template<typename SvdType, typename MatrixType>
svd_check_full(const MatrixType & m,const SvdType & svd)25 void svd_check_full(const MatrixType& m, const SvdType& svd)
26 {
27   Index rows = m.rows();
28   Index cols = m.cols();
29 
30   enum {
31     RowsAtCompileTime = MatrixType::RowsAtCompileTime,
32     ColsAtCompileTime = MatrixType::ColsAtCompileTime
33   };
34 
35   typedef typename MatrixType::Scalar Scalar;
36   typedef typename MatrixType::RealScalar RealScalar;
37   typedef Matrix<Scalar, RowsAtCompileTime, RowsAtCompileTime> MatrixUType;
38   typedef Matrix<Scalar, ColsAtCompileTime, ColsAtCompileTime> MatrixVType;
39 
40   MatrixType sigma = MatrixType::Zero(rows,cols);
41   sigma.diagonal() = svd.singularValues().template cast<Scalar>();
42   MatrixUType u = svd.matrixU();
43   MatrixVType v = svd.matrixV();
44   RealScalar scaling = m.cwiseAbs().maxCoeff();
45   if(scaling<(std::numeric_limits<RealScalar>::min)())
46   {
47     VERIFY(sigma.cwiseAbs().maxCoeff() <= (std::numeric_limits<RealScalar>::min)());
48   }
49   else
50   {
51     VERIFY_IS_APPROX(m/scaling, u * (sigma/scaling) * v.adjoint());
52   }
53   VERIFY_IS_UNITARY(u);
54   VERIFY_IS_UNITARY(v);
55 }
56 
57 // Compare partial SVD defined by computationOptions to a full SVD referenceSvd
58 template<typename SvdType, typename MatrixType>
svd_compare_to_full(const MatrixType & m,unsigned int computationOptions,const SvdType & referenceSvd)59 void svd_compare_to_full(const MatrixType& m,
60                          unsigned int computationOptions,
61                          const SvdType& referenceSvd)
62 {
63   typedef typename MatrixType::RealScalar RealScalar;
64   Index rows = m.rows();
65   Index cols = m.cols();
66   Index diagSize = (std::min)(rows, cols);
67   RealScalar prec = test_precision<RealScalar>();
68 
69   SvdType svd(m, computationOptions);
70 
71   VERIFY_IS_APPROX(svd.singularValues(), referenceSvd.singularValues());
72 
73   if(computationOptions & (ComputeFullV|ComputeThinV))
74   {
75     VERIFY( (svd.matrixV().adjoint()*svd.matrixV()).isIdentity(prec) );
76     VERIFY_IS_APPROX( svd.matrixV().leftCols(diagSize) * svd.singularValues().asDiagonal() * svd.matrixV().leftCols(diagSize).adjoint(),
77                       referenceSvd.matrixV().leftCols(diagSize) * referenceSvd.singularValues().asDiagonal() * referenceSvd.matrixV().leftCols(diagSize).adjoint());
78   }
79 
80   if(computationOptions & (ComputeFullU|ComputeThinU))
81   {
82     VERIFY( (svd.matrixU().adjoint()*svd.matrixU()).isIdentity(prec) );
83     VERIFY_IS_APPROX( svd.matrixU().leftCols(diagSize) * svd.singularValues().cwiseAbs2().asDiagonal() * svd.matrixU().leftCols(diagSize).adjoint(),
84                       referenceSvd.matrixU().leftCols(diagSize) * referenceSvd.singularValues().cwiseAbs2().asDiagonal() * referenceSvd.matrixU().leftCols(diagSize).adjoint());
85   }
86 
87   // The following checks are not critical.
88   // For instance, with Dived&Conquer SVD, if only the factor 'V' is computedt then different matrix-matrix product implementation will be used
89   // and the resulting 'V' factor might be significantly different when the SVD decomposition is not unique, especially with single precision float.
90   ++g_test_level;
91   if(computationOptions & ComputeFullU)  VERIFY_IS_APPROX(svd.matrixU(), referenceSvd.matrixU());
92   if(computationOptions & ComputeThinU)  VERIFY_IS_APPROX(svd.matrixU(), referenceSvd.matrixU().leftCols(diagSize));
93   if(computationOptions & ComputeFullV)  VERIFY_IS_APPROX(svd.matrixV().cwiseAbs(), referenceSvd.matrixV().cwiseAbs());
94   if(computationOptions & ComputeThinV)  VERIFY_IS_APPROX(svd.matrixV(), referenceSvd.matrixV().leftCols(diagSize));
95   --g_test_level;
96 }
97 
98 //
99 template<typename SvdType, typename MatrixType>
svd_least_square(const MatrixType & m,unsigned int computationOptions)100 void svd_least_square(const MatrixType& m, unsigned int computationOptions)
101 {
102   typedef typename MatrixType::Scalar Scalar;
103   typedef typename MatrixType::RealScalar RealScalar;
104   Index rows = m.rows();
105   Index cols = m.cols();
106 
107   enum {
108     RowsAtCompileTime = MatrixType::RowsAtCompileTime,
109     ColsAtCompileTime = MatrixType::ColsAtCompileTime
110   };
111 
112   typedef Matrix<Scalar, RowsAtCompileTime, Dynamic> RhsType;
113   typedef Matrix<Scalar, ColsAtCompileTime, Dynamic> SolutionType;
114 
115   RhsType rhs = RhsType::Random(rows, internal::random<Index>(1, cols));
116   SvdType svd(m, computationOptions);
117 
118        if(internal::is_same<RealScalar,double>::value) svd.setThreshold(1e-8);
119   else if(internal::is_same<RealScalar,float>::value)  svd.setThreshold(2e-4);
120 
121   SolutionType x = svd.solve(rhs);
122 
123   RealScalar residual = (m*x-rhs).norm();
124   RealScalar rhs_norm = rhs.norm();
125   if(!test_isMuchSmallerThan(residual,rhs.norm()))
126   {
127     // ^^^ If the residual is very small, then we have an exact solution, so we are already good.
128 
129     // evaluate normal equation which works also for least-squares solutions
130     if(internal::is_same<RealScalar,double>::value || svd.rank()==m.diagonal().size())
131     {
132       using std::sqrt;
133       // This test is not stable with single precision.
134       // This is probably because squaring m signicantly affects the precision.
135       if(internal::is_same<RealScalar,float>::value) ++g_test_level;
136 
137       VERIFY_IS_APPROX(m.adjoint()*(m*x),m.adjoint()*rhs);
138 
139       if(internal::is_same<RealScalar,float>::value) --g_test_level;
140     }
141 
142     // Check that there is no significantly better solution in the neighborhood of x
143     for(Index k=0;k<x.rows();++k)
144     {
145       using std::abs;
146 
147       SolutionType y(x);
148       y.row(k) = (RealScalar(1)+2*NumTraits<RealScalar>::epsilon())*x.row(k);
149       RealScalar residual_y = (m*y-rhs).norm();
150       VERIFY( test_isMuchSmallerThan(abs(residual_y-residual), rhs_norm) || residual < residual_y );
151       if(internal::is_same<RealScalar,float>::value) ++g_test_level;
152       VERIFY( test_isApprox(residual_y,residual) || residual < residual_y );
153       if(internal::is_same<RealScalar,float>::value) --g_test_level;
154 
155       y.row(k) = (RealScalar(1)-2*NumTraits<RealScalar>::epsilon())*x.row(k);
156       residual_y = (m*y-rhs).norm();
157       VERIFY( test_isMuchSmallerThan(abs(residual_y-residual), rhs_norm) || residual < residual_y );
158       if(internal::is_same<RealScalar,float>::value) ++g_test_level;
159       VERIFY( test_isApprox(residual_y,residual) || residual < residual_y );
160       if(internal::is_same<RealScalar,float>::value) --g_test_level;
161     }
162   }
163 }
164 
165 // check minimal norm solutions, the inoput matrix m is only used to recover problem size
166 template<typename MatrixType>
svd_min_norm(const MatrixType & m,unsigned int computationOptions)167 void svd_min_norm(const MatrixType& m, unsigned int computationOptions)
168 {
169   typedef typename MatrixType::Scalar Scalar;
170   Index cols = m.cols();
171 
172   enum {
173     ColsAtCompileTime = MatrixType::ColsAtCompileTime
174   };
175 
176   typedef Matrix<Scalar, ColsAtCompileTime, Dynamic> SolutionType;
177 
178   // generate a full-rank m x n problem with m<n
179   enum {
180     RankAtCompileTime2 = ColsAtCompileTime==Dynamic ? Dynamic : (ColsAtCompileTime)/2+1,
181     RowsAtCompileTime3 = ColsAtCompileTime==Dynamic ? Dynamic : ColsAtCompileTime+1
182   };
183   typedef Matrix<Scalar, RankAtCompileTime2, ColsAtCompileTime> MatrixType2;
184   typedef Matrix<Scalar, RankAtCompileTime2, 1> RhsType2;
185   typedef Matrix<Scalar, ColsAtCompileTime, RankAtCompileTime2> MatrixType2T;
186   Index rank = RankAtCompileTime2==Dynamic ? internal::random<Index>(1,cols) : Index(RankAtCompileTime2);
187   MatrixType2 m2(rank,cols);
188   int guard = 0;
189   do {
190     m2.setRandom();
191   } while(SVD_FOR_MIN_NORM(MatrixType2)(m2).setThreshold(test_precision<Scalar>()).rank()!=rank && (++guard)<10);
192   VERIFY(guard<10);
193 
194   RhsType2 rhs2 = RhsType2::Random(rank);
195   // use QR to find a reference minimal norm solution
196   HouseholderQR<MatrixType2T> qr(m2.adjoint());
197   Matrix<Scalar,Dynamic,1> tmp = qr.matrixQR().topLeftCorner(rank,rank).template triangularView<Upper>().adjoint().solve(rhs2);
198   tmp.conservativeResize(cols);
199   tmp.tail(cols-rank).setZero();
200   SolutionType x21 = qr.householderQ() * tmp;
201   // now check with SVD
202   SVD_FOR_MIN_NORM(MatrixType2) svd2(m2, computationOptions);
203   SolutionType x22 = svd2.solve(rhs2);
204   VERIFY_IS_APPROX(m2*x21, rhs2);
205   VERIFY_IS_APPROX(m2*x22, rhs2);
206   VERIFY_IS_APPROX(x21, x22);
207 
208   // Now check with a rank deficient matrix
209   typedef Matrix<Scalar, RowsAtCompileTime3, ColsAtCompileTime> MatrixType3;
210   typedef Matrix<Scalar, RowsAtCompileTime3, 1> RhsType3;
211   Index rows3 = RowsAtCompileTime3==Dynamic ? internal::random<Index>(rank+1,2*cols) : Index(RowsAtCompileTime3);
212   Matrix<Scalar,RowsAtCompileTime3,Dynamic> C = Matrix<Scalar,RowsAtCompileTime3,Dynamic>::Random(rows3,rank);
213   MatrixType3 m3 = C * m2;
214   RhsType3 rhs3 = C * rhs2;
215   SVD_FOR_MIN_NORM(MatrixType3) svd3(m3, computationOptions);
216   SolutionType x3 = svd3.solve(rhs3);
217   VERIFY_IS_APPROX(m3*x3, rhs3);
218   VERIFY_IS_APPROX(m3*x21, rhs3);
219   VERIFY_IS_APPROX(m2*x3, rhs2);
220   VERIFY_IS_APPROX(x21, x3);
221 }
222 
223 template<typename MatrixType, typename SolverType>
svd_test_solvers(const MatrixType & m,const SolverType & solver)224 void svd_test_solvers(const MatrixType& m, const SolverType& solver) {
225     Index rows, cols, cols2;
226 
227     rows = m.rows();
228     cols = m.cols();
229 
230     if(MatrixType::ColsAtCompileTime==Dynamic)
231     {
232       cols2 = internal::random<int>(2,EIGEN_TEST_MAX_SIZE);
233     }
234     else
235     {
236       cols2 = cols;
237     }
238     typedef Matrix<typename MatrixType::Scalar, MatrixType::ColsAtCompileTime, MatrixType::ColsAtCompileTime> CMatrixType;
239     check_solverbase<CMatrixType, MatrixType>(m, solver, rows, cols, cols2);
240 }
241 
242 // Check full, compare_to_full, least_square, and min_norm for all possible compute-options
243 template<typename SvdType, typename MatrixType>
svd_test_all_computation_options(const MatrixType & m,bool full_only)244 void svd_test_all_computation_options(const MatrixType& m, bool full_only)
245 {
246 //   if (QRPreconditioner == NoQRPreconditioner && m.rows() != m.cols())
247 //     return;
248   STATIC_CHECK(( internal::is_same<typename SvdType::StorageIndex,int>::value ));
249 
250   SvdType fullSvd(m, ComputeFullU|ComputeFullV);
251   CALL_SUBTEST(( svd_check_full(m, fullSvd) ));
252   CALL_SUBTEST(( svd_least_square<SvdType>(m, ComputeFullU | ComputeFullV) ));
253   CALL_SUBTEST(( svd_min_norm(m, ComputeFullU | ComputeFullV) ));
254 
255   #if defined __INTEL_COMPILER
256   // remark #111: statement is unreachable
257   #pragma warning disable 111
258   #endif
259 
260   svd_test_solvers(m, fullSvd);
261 
262   if(full_only)
263     return;
264 
265   CALL_SUBTEST(( svd_compare_to_full(m, ComputeFullU, fullSvd) ));
266   CALL_SUBTEST(( svd_compare_to_full(m, ComputeFullV, fullSvd) ));
267   CALL_SUBTEST(( svd_compare_to_full(m, 0, fullSvd) ));
268 
269   if (MatrixType::ColsAtCompileTime == Dynamic) {
270     // thin U/V are only available with dynamic number of columns
271     CALL_SUBTEST(( svd_compare_to_full(m, ComputeFullU|ComputeThinV, fullSvd) ));
272     CALL_SUBTEST(( svd_compare_to_full(m,              ComputeThinV, fullSvd) ));
273     CALL_SUBTEST(( svd_compare_to_full(m, ComputeThinU|ComputeFullV, fullSvd) ));
274     CALL_SUBTEST(( svd_compare_to_full(m, ComputeThinU             , fullSvd) ));
275     CALL_SUBTEST(( svd_compare_to_full(m, ComputeThinU|ComputeThinV, fullSvd) ));
276 
277     CALL_SUBTEST(( svd_least_square<SvdType>(m, ComputeFullU | ComputeThinV) ));
278     CALL_SUBTEST(( svd_least_square<SvdType>(m, ComputeThinU | ComputeFullV) ));
279     CALL_SUBTEST(( svd_least_square<SvdType>(m, ComputeThinU | ComputeThinV) ));
280 
281     CALL_SUBTEST(( svd_min_norm(m, ComputeFullU | ComputeThinV) ));
282     CALL_SUBTEST(( svd_min_norm(m, ComputeThinU | ComputeFullV) ));
283     CALL_SUBTEST(( svd_min_norm(m, ComputeThinU | ComputeThinV) ));
284 
285     // test reconstruction
286     Index diagSize = (std::min)(m.rows(), m.cols());
287     SvdType svd(m, ComputeThinU | ComputeThinV);
288     VERIFY_IS_APPROX(m, svd.matrixU().leftCols(diagSize) * svd.singularValues().asDiagonal() * svd.matrixV().leftCols(diagSize).adjoint());
289   }
290 }
291 
292 
293 // work around stupid msvc error when constructing at compile time an expression that involves
294 // a division by zero, even if the numeric type has floating point
295 template<typename Scalar>
zero()296 EIGEN_DONT_INLINE Scalar zero() { return Scalar(0); }
297 
298 // workaround aggressive optimization in ICC
sub(T a,T b)299 template<typename T> EIGEN_DONT_INLINE  T sub(T a, T b) { return a - b; }
300 
301 // This function verifies we don't iterate infinitely on nan/inf values,
302 // and that info() returns InvalidInput.
303 template<typename SvdType, typename MatrixType>
svd_inf_nan()304 void svd_inf_nan()
305 {
306   SvdType svd;
307   typedef typename MatrixType::Scalar Scalar;
308   Scalar some_inf = Scalar(1) / zero<Scalar>();
309   VERIFY(sub(some_inf, some_inf) != sub(some_inf, some_inf));
310   svd.compute(MatrixType::Constant(10,10,some_inf), ComputeFullU | ComputeFullV);
311   VERIFY(svd.info() == InvalidInput);
312 
313   Scalar nan = std::numeric_limits<Scalar>::quiet_NaN();
314   VERIFY(nan != nan);
315   svd.compute(MatrixType::Constant(10,10,nan), ComputeFullU | ComputeFullV);
316   VERIFY(svd.info() == InvalidInput);
317 
318   MatrixType m = MatrixType::Zero(10,10);
319   m(internal::random<int>(0,9), internal::random<int>(0,9)) = some_inf;
320   svd.compute(m, ComputeFullU | ComputeFullV);
321   VERIFY(svd.info() == InvalidInput);
322 
323   m = MatrixType::Zero(10,10);
324   m(internal::random<int>(0,9), internal::random<int>(0,9)) = nan;
325   svd.compute(m, ComputeFullU | ComputeFullV);
326   VERIFY(svd.info() == InvalidInput);
327 
328   // regression test for bug 791
329   m.resize(3,3);
330   m << 0,    2*NumTraits<Scalar>::epsilon(),  0.5,
331        0,   -0.5,                             0,
332        nan,  0,                               0;
333   svd.compute(m, ComputeFullU | ComputeFullV);
334   VERIFY(svd.info() == InvalidInput);
335 
336   m.resize(4,4);
337   m <<  1, 0, 0, 0,
338         0, 3, 1, 2e-308,
339         1, 0, 1, nan,
340         0, nan, nan, 0;
341   svd.compute(m, ComputeFullU | ComputeFullV);
342   VERIFY(svd.info() == InvalidInput);
343 }
344 
345 // Regression test for bug 286: JacobiSVD loops indefinitely with some
346 // matrices containing denormal numbers.
347 template<typename>
svd_underoverflow()348 void svd_underoverflow()
349 {
350 #if defined __INTEL_COMPILER
351 // shut up warning #239: floating point underflow
352 #pragma warning push
353 #pragma warning disable 239
354 #endif
355   Matrix2d M;
356   M << -7.90884e-313, -4.94e-324,
357                  0, 5.60844e-313;
358   SVD_DEFAULT(Matrix2d) svd;
359   svd.compute(M,ComputeFullU|ComputeFullV);
360   CALL_SUBTEST( svd_check_full(M,svd) );
361 
362   // Check all 2x2 matrices made with the following coefficients:
363   VectorXd value_set(9);
364   value_set << 0, 1, -1, 5.60844e-313, -5.60844e-313, 4.94e-324, -4.94e-324, -4.94e-223, 4.94e-223;
365   Array4i id(0,0,0,0);
366   int k = 0;
367   do
368   {
369     M << value_set(id(0)), value_set(id(1)), value_set(id(2)), value_set(id(3));
370     svd.compute(M,ComputeFullU|ComputeFullV);
371     CALL_SUBTEST( svd_check_full(M,svd) );
372 
373     id(k)++;
374     if(id(k)>=value_set.size())
375     {
376       while(k<3 && id(k)>=value_set.size()) id(++k)++;
377       id.head(k).setZero();
378       k=0;
379     }
380 
381   } while((id<int(value_set.size())).all());
382 
383 #if defined __INTEL_COMPILER
384 #pragma warning pop
385 #endif
386 
387   // Check for overflow:
388   Matrix3d M3;
389   M3 << 4.4331978442502944e+307, -5.8585363752028680e+307,  6.4527017443412964e+307,
390         3.7841695601406358e+307,  2.4331702789740617e+306, -3.5235707140272905e+307,
391        -8.7190887618028355e+307, -7.3453213709232193e+307, -2.4367363684472105e+307;
392 
393   SVD_DEFAULT(Matrix3d) svd3;
394   svd3.compute(M3,ComputeFullU|ComputeFullV); // just check we don't loop indefinitely
395   CALL_SUBTEST( svd_check_full(M3,svd3) );
396 }
397 
398 // void jacobisvd(const MatrixType& a = MatrixType(), bool pickrandom = true)
399 
400 template<typename MatrixType>
svd_all_trivial_2x2(void (* cb)(const MatrixType &,bool))401 void svd_all_trivial_2x2( void (*cb)(const MatrixType&,bool) )
402 {
403   MatrixType M;
404   VectorXd value_set(3);
405   value_set << 0, 1, -1;
406   Array4i id(0,0,0,0);
407   int k = 0;
408   do
409   {
410     M << value_set(id(0)), value_set(id(1)), value_set(id(2)), value_set(id(3));
411 
412     cb(M,false);
413 
414     id(k)++;
415     if(id(k)>=value_set.size())
416     {
417       while(k<3 && id(k)>=value_set.size()) id(++k)++;
418       id.head(k).setZero();
419       k=0;
420     }
421 
422   } while((id<int(value_set.size())).all());
423 }
424 
425 template<typename>
svd_preallocate()426 void svd_preallocate()
427 {
428   Vector3f v(3.f, 2.f, 1.f);
429   MatrixXf m = v.asDiagonal();
430 
431   internal::set_is_malloc_allowed(false);
432   VERIFY_RAISES_ASSERT(VectorXf tmp(10);)
433   SVD_DEFAULT(MatrixXf) svd;
434   internal::set_is_malloc_allowed(true);
435   svd.compute(m);
436   VERIFY_IS_APPROX(svd.singularValues(), v);
437 
438   SVD_DEFAULT(MatrixXf) svd2(3,3);
439   internal::set_is_malloc_allowed(false);
440   svd2.compute(m);
441   internal::set_is_malloc_allowed(true);
442   VERIFY_IS_APPROX(svd2.singularValues(), v);
443   VERIFY_RAISES_ASSERT(svd2.matrixU());
444   VERIFY_RAISES_ASSERT(svd2.matrixV());
445   svd2.compute(m, ComputeFullU | ComputeFullV);
446   VERIFY_IS_APPROX(svd2.matrixU(), Matrix3f::Identity());
447   VERIFY_IS_APPROX(svd2.matrixV(), Matrix3f::Identity());
448   internal::set_is_malloc_allowed(false);
449   svd2.compute(m);
450   internal::set_is_malloc_allowed(true);
451 
452   SVD_DEFAULT(MatrixXf) svd3(3,3,ComputeFullU|ComputeFullV);
453   internal::set_is_malloc_allowed(false);
454   svd2.compute(m);
455   internal::set_is_malloc_allowed(true);
456   VERIFY_IS_APPROX(svd2.singularValues(), v);
457   VERIFY_IS_APPROX(svd2.matrixU(), Matrix3f::Identity());
458   VERIFY_IS_APPROX(svd2.matrixV(), Matrix3f::Identity());
459   internal::set_is_malloc_allowed(false);
460   svd2.compute(m, ComputeFullU|ComputeFullV);
461   internal::set_is_malloc_allowed(true);
462 }
463 
464 template<typename SvdType,typename MatrixType>
465 void svd_verify_assert(const MatrixType& m, bool fullOnly = false)
466 {
467   typedef typename MatrixType::Scalar Scalar;
468   Index rows = m.rows();
469   Index cols = m.cols();
470 
471   enum {
472     RowsAtCompileTime = MatrixType::RowsAtCompileTime,
473     ColsAtCompileTime = MatrixType::ColsAtCompileTime
474   };
475 
476   typedef Matrix<Scalar, RowsAtCompileTime, 1> RhsType;
477   RhsType rhs(rows);
478   SvdType svd;
479   VERIFY_RAISES_ASSERT(svd.matrixU())
480   VERIFY_RAISES_ASSERT(svd.singularValues())
481   VERIFY_RAISES_ASSERT(svd.matrixV())
482   VERIFY_RAISES_ASSERT(svd.solve(rhs))
483   VERIFY_RAISES_ASSERT(svd.transpose().solve(rhs))
484   VERIFY_RAISES_ASSERT(svd.adjoint().solve(rhs))
485   MatrixType a = MatrixType::Zero(rows, cols);
486   a.setZero();
487   svd.compute(a, 0);
488   VERIFY_RAISES_ASSERT(svd.matrixU())
489   VERIFY_RAISES_ASSERT(svd.matrixV())
490   svd.singularValues();
491   VERIFY_RAISES_ASSERT(svd.solve(rhs))
492 
493   svd.compute(a, ComputeFullU);
494   svd.matrixU();
495   VERIFY_RAISES_ASSERT(svd.matrixV())
496   VERIFY_RAISES_ASSERT(svd.solve(rhs))
497   svd.compute(a, ComputeFullV);
498   svd.matrixV();
499   VERIFY_RAISES_ASSERT(svd.matrixU())
500   VERIFY_RAISES_ASSERT(svd.solve(rhs))
501 
502   if (!fullOnly && ColsAtCompileTime == Dynamic)
503   {
504     svd.compute(a, ComputeThinU);
505     svd.matrixU();
506     VERIFY_RAISES_ASSERT(svd.matrixV())
507     VERIFY_RAISES_ASSERT(svd.solve(rhs))
508     svd.compute(a, ComputeThinV);
509     svd.matrixV();
510     VERIFY_RAISES_ASSERT(svd.matrixU())
511     VERIFY_RAISES_ASSERT(svd.solve(rhs))
512   }
513   else
514   {
515     VERIFY_RAISES_ASSERT(svd.compute(a, ComputeThinU))
516     VERIFY_RAISES_ASSERT(svd.compute(a, ComputeThinV))
517   }
518 }
519 
520 #undef SVD_DEFAULT
521 #undef SVD_FOR_MIN_NORM
522