fdjac1.h
Go to the documentation of this file.
1 namespace Eigen {
2 
3 namespace internal {
4 
5 template<typename FunctorType, typename Scalar>
7  const FunctorType &Functor,
11  DenseIndex ml, DenseIndex mu,
12  Scalar epsfcn)
13 {
14  using std::sqrt;
15  using std::abs;
16 
17  typedef DenseIndex Index;
18 
19  /* Local variables */
20  Scalar h;
21  Index j, k;
22  Scalar eps, temp;
23  Index msum;
24  int iflag;
25  Index start, length;
26 
27  /* Function Body */
28  const Scalar epsmch = NumTraits<Scalar>::epsilon();
29  const Index n = x.size();
30  eigen_assert(fvec.size()==n);
33 
34  eps = sqrt((std::max)(epsfcn,epsmch));
35  msum = ml + mu + 1;
36  if (msum >= n) {
37  /* computation of dense approximate jacobian. */
38  for (j = 0; j < n; ++j) {
39  temp = x[j];
40  h = eps * abs(temp);
41  if (h == 0.)
42  h = eps;
43  x[j] = temp + h;
44  iflag = Functor(x, wa1);
45  if (iflag < 0)
46  return iflag;
47  x[j] = temp;
48  fjac.col(j) = (wa1-fvec)/h;
49  }
50 
51  }else {
52  /* computation of banded approximate jacobian. */
53  for (k = 0; k < msum; ++k) {
54  for (j = k; (msum<0) ? (j>n): (j<n); j += msum) {
55  wa2[j] = x[j];
56  h = eps * abs(wa2[j]);
57  if (h == 0.) h = eps;
58  x[j] = wa2[j] + h;
59  }
60  iflag = Functor(x, wa1);
61  if (iflag < 0)
62  return iflag;
63  for (j = k; (msum<0) ? (j>n): (j<n); j += msum) {
64  x[j] = wa2[j];
65  h = eps * abs(wa2[j]);
66  if (h == 0.) h = eps;
67  fjac.col(j).setZero();
68  start = std::max<Index>(0,j-mu);
69  length = (std::min)(n-1, j+ml) - start + 1;
70  fjac.col(j).segment(start, length) = ( wa1.segment(start, length)-fvec.segment(start, length))/h;
71  }
72  }
73  }
74  return 0;
75 }
76 
77 } // end namespace internal
78 
79 } // end namespace Eigen
IntermediateState sqrt(const Expression &arg)
iterative scaling algorithm to equilibrate rows and column norms in matrices
Definition: matrix.hpp:471
Holds information about the various numeric (i.e. scalar) types allowed by Eigen. ...
Definition: NumTraits.h:88
EIGEN_STRONG_INLINE const CwiseUnaryOp< internal::scalar_abs_op< Scalar >, const Derived > abs() const
Derived & setZero(Index size)
EIGEN_DEFAULT_DENSE_INDEX_TYPE DenseIndex
Definition: XprHelper.h:27
#define eigen_assert(x)
DenseIndex fdjac1(const FunctorType &Functor, Matrix< Scalar, Dynamic, 1 > &x, Matrix< Scalar, Dynamic, 1 > &fvec, Matrix< Scalar, Dynamic, Dynamic > &fjac, DenseIndex ml, DenseIndex mu, Scalar epsfcn)
Definition: fdjac1.h:6


acado
Author(s): Milan Vukov, Rien Quirynen
autogenerated on Mon Jun 10 2019 12:34:34