From: Wolfgang Bangerth Date: Mon, 28 Mar 2005 19:05:28 +0000 (+0000) Subject: Add accessors. Add a few more functions. X-Git-Tag: v8.0.0~14293 X-Git-Url: https://gitweb.dealii.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=f9ffbbddc9868f4b2fa0906fa4e06d915f1f107c;p=dealii.git Add accessors. Add a few more functions. git-svn-id: https://svn.dealii.org/trunk@10254 0785d39b-7218-0410-832d-ea1e28bc413d --- diff --git a/deal.II/base/include/base/symmetric_tensor.h b/deal.II/base/include/base/symmetric_tensor.h index 3af4dcd575..f2d4986ec3 100644 --- a/deal.II/base/include/base/symmetric_tensor.h +++ b/deal.II/base/include/base/symmetric_tensor.h @@ -15,8 +15,137 @@ template class SymmetricTensor; +template class SymmetricTensor<2,dim>; +namespace internal +{ + namespace SymmetricTensor + { + namespace Rank2Accessors + { + + /** + * Switch type to select a tensor of + * rank 2 and dimension dim, + * switching on whether the tensor + * should be constant or not. + */ + template + struct Types; + + /** + * Switch type to select a tensor of + * rank 2 and dimension dim, + * switching on whether the tensor + * should be constant or not. + * + * Specialization for constant tensors. + */ + template + struct Types + { + typedef + const typename ::SymmetricTensor<2,dim>::StorageType + base_tensor_type; + + typedef double reference; + }; + + /** + * Switch type to select a tensor of + * rank 2 and dimension dim, + * switching on whether the tensor + * should be constant or not. + * + * Specialization for non-constant + * tensors. + */ + template + struct Types + { + typedef + typename ::SymmetricTensor<2,dim>::StorageType + base_tensor_type; + + typedef double &reference; + }; + + + /** + * Accessor class to access the + * elements of individual rows in a + * symmetric tensor. Since the elements + * of symmetric tensors are not stored + * as in a table, the accessors are a + * little more involved. + * + * @author Wolfgang Bangerth, 2005 + */ + template + class RowAccessor + { + public: + /** + * Import which tensor we work on. + */ + typedef + typename Types::base_tensor_type + base_tensor_type; + + /** + * The type of a reference to an + * individual element of the + * symmetric tensor. If the tensor + * is constant, we can only return + * a value instead of a reference. + */ + typedef typename Types::reference reference; + + /** + * Constructor. Take the tensor to + * access as well as the row we + * point to as arguments. + */ + RowAccessor (const base_tensor_type &tensor, + const unsigned int row); + + /** + * Return a reference to an element + * of this row (if we point to a + * non-const tensor), or the value + * of the element (in case this is + * a constant tensor). + */ + reference operator[] (const unsigned int column) const; + + private: + /** + * Reference to the tensor we + * access. + */ + const base_tensor_type &base_tensor; + + /** + * Index of the row we access. + */ + const unsigned int row; + + /** + * Make the symmetric tensor + * classes a friend, since they are + * the only ones who can create + * objects like this. + */ + template class ::SymmetricTensor; + }; + + } + } +} + + + /** * Provide a class that stores symmetric tensors of rank 2 efficiently, @@ -59,6 +188,29 @@ class SymmetricTensor<2,dim> */ static const unsigned int rank = 2; + /** + * Number of independent components of a + * symmetric tensor of rank 2. We store + * only the upper right half of it. This + * information is probably of little + * interest to all except the accessor + * classes that need it. + */ + static const unsigned int + n_tensor_components = (dim*dim + dim)/2; + + /** + * Declare the type in which we actually + * store the data. This information is + * probably of little interest to all + * except the accessor classes that need + * it. In particular, you shouldn't make + * any assumptions about the storage + * format in your application programs. + */ + typedef Tensor<1,n_tensor_components> StorageType; + + /** * Default constructor. Creates a zero * tensor. @@ -134,6 +286,22 @@ class SymmetricTensor<2,dim> */ SymmetricTensor operator - () const; + /** + * Access the elements of a row of this + * symmetric tensor. This function is + * called for constant tensors. + */ + internal::SymmetricTensor::Rank2Accessors::RowAccessor + operator [] (const unsigned int row) const; + + /** + * Access the elements of a row of this + * symmetric tensor. This function is + * called for non-constant tensors. + */ + internal::SymmetricTensor::Rank2Accessors::RowAccessor + operator [] (const unsigned int row); + /** * Return the Frobenius-norm of a tensor, * i.e. the square root of the sum of @@ -171,20 +339,6 @@ class SymmetricTensor<2,dim> private: - /** - * Number of independent components of a - * symmetric tensor of rank 2. We store - * only the upper right half of it. - */ - static const unsigned int - n_tensor_components = (dim*dim + dim)/2; - - /** - * Declare the type in which we actually - * store the data. - */ - typedef Tensor<1,n_tensor_components> StorageType; - /** * Data storage for a symmetric tensor. */ @@ -195,6 +349,74 @@ class SymmetricTensor<2,dim> // ------------------------- inline functions ------------------------ +namespace internal +{ + namespace SymmetricTensor + { + namespace Rank2Accessors + { + template + RowAccessor:: + RowAccessor (const base_tensor_type &base_tensor, + const unsigned int row) + : + base_tensor (base_tensor), + row (row) + { + Assert (row < dim, ExcIndexRange (row, 0, dim)); + } + + + + template + typename RowAccessor::reference + RowAccessor:: + operator[] (const unsigned int column) const + { + Assert (column < dim, ExcIndexRange (column, 0, dim)); + + // first treat the main diagonal + // elements, which are stored + // consecutively at the beginning + if (row == column) + return base_tensor[row]; + + // the rest is messier and requires a + // few switches. if someone has a + // better idea, help is welcome + switch (dim) + { + case 2: + Assert (((row==1) && (column==0)) || ((row==0) && (column==1)), + ExcInternalError()); + return base_tensor[2]; + + case 3: + if (((row==0) && (column==1)) || + ((row==1) && (column==0))) + return base_tensor[3]; + else if (((row==0) && (column==2)) || + ((row==2) && (column==0))) + return base_tensor[4]; + else if (((row==1) && (column==2)) || + ((row==2) && (column==1))) + return base_tensor[5]; + else + Assert (false, ExcInternalError()); + + default: + Assert (false, ExcNotImplemented()); + } + + Assert (false, ExcInternalError()); + return 0; + } + } + } +} + + + template inline SymmetricTensor<2,dim>::SymmetricTensor () @@ -363,6 +585,26 @@ SymmetricTensor<2,dim>::memory_consumption () +template +internal::SymmetricTensor::Rank2Accessors::RowAccessor +SymmetricTensor<2,dim>::operator [] (const unsigned int row) const +{ + return + internal::SymmetricTensor::Rank2Accessors::RowAccessor (data, row); +} + + + +template +internal::SymmetricTensor::Rank2Accessors::RowAccessor +SymmetricTensor<2,dim>::operator [] (const unsigned int row) +{ + return + internal::SymmetricTensor::Rank2Accessors::RowAccessor (data, row); +} + + + template <> double SymmetricTensor<2,1>::norm () const @@ -392,25 +634,6 @@ SymmetricTensor<2,3>::norm () const /* ----------------- Non-member functions operating on tensors. ------------ */ -/** - * Compute the determinant of a tensor of rank one and dimension - * one. Since this is a number, the return value is, of course, the - * number itself. - * - * @relates Tensor - * @author Wolfgang Bangerth, 2005 - */ -inline -double determinant (const SymmetricTensor<1,1> &t) -{ - Assert (false, ExcNotImplemented()); - return 0; - -// return t[0]; -} - - - /** * Compute the determinant of a tensor or rank 2, here for dim==2. * @@ -420,11 +643,7 @@ double determinant (const SymmetricTensor<1,1> &t) inline double determinant (const SymmetricTensor<2,2> &t) { - Assert (false, ExcNotImplemented()); - return 0; - -// return ((t[0][0] * t[1][1]) - -// (t[1][0] * t[0][1])); + return (t[0][0] * t[1][1] - 2*t[0][1]*t[0][1]); } @@ -439,21 +658,14 @@ double determinant (const SymmetricTensor<2,2> &t) inline double determinant (const SymmetricTensor<2,3> &t) { - Assert (false, ExcNotImplemented()); - return 0; - -// // get this using Maple: -// // with(linalg); -// // a := matrix(3,3); -// // x := det(a); -// // readlib(C); -// // C(x, optimized); -// return ( t[0][0]*t[1][1]*t[2][2] -// -t[0][0]*t[1][2]*t[2][1] -// -t[1][0]*t[0][1]*t[2][2] -// +t[1][0]*t[0][2]*t[2][1] -// +t[2][0]*t[0][1]*t[1][2] -// -t[2][0]*t[0][2]*t[1][1] ); + // in analogy to general tensors, but + // there's something to be simplified for + // the present case + return ( t[0][0]*t[1][1]*t[2][2] + -t[0][0]*t[1][2]*t[1][2] + -t[1][1]*t[0][2]*t[0][2] + -t[2][2]*t[0][1]*t[0][1] + +2*t[0][1]*t[0][2]*t[1][2] ); } @@ -468,82 +680,10 @@ double determinant (const SymmetricTensor<2,3> &t) template double trace (const SymmetricTensor<2,dim> &d) { - Assert (false, ExcNotImplemented()); - return 0; - -// double t=0; -// for (unsigned int i=0; i -inline -SymmetricTensor<2,dim> -invert (const SymmetricTensor<2,dim> &t) -{ - Assert (false, ExcNotImplemented()); - return SymmetricTensor<2,dim>(); - -// SymmetricTensor<2,dim> return_tensor; -// switch (dim) -// { -// case 1: -// return_tensor[0][0] = 1.0/t[0][0]; -// return return_tensor; -// case 2: -// // this is Maple output, -// // thus a bit unstructured -// { -// const double t4 = 1.0/(t[0][0]*t[1][1]-t[0][1]*t[1][0]); -// return_tensor[0][0] = t[1][1]*t4; -// return_tensor[0][1] = -t[0][1]*t4; -// return_tensor[1][0] = -t[1][0]*t4; -// return_tensor[1][1] = t[0][0]*t4; -// return return_tensor; -// }; - -// case 3: -// { -// const double t4 = t[0][0]*t[1][1], -// t6 = t[0][0]*t[1][2], -// t8 = t[0][1]*t[1][0], -// t00 = t[0][2]*t[1][0], -// t01 = t[0][1]*t[2][0], -// t04 = t[0][2]*t[2][0], -// t07 = 1.0/(t4*t[2][2]-t6*t[2][1]-t8*t[2][2]+ -// t00*t[2][1]+t01*t[1][2]-t04*t[1][1]); -// return_tensor[0][0] = (t[1][1]*t[2][2]-t[1][2]*t[2][1])*t07; -// return_tensor[0][1] = -(t[0][1]*t[2][2]-t[0][2]*t[2][1])*t07; -// return_tensor[0][2] = -(-t[0][1]*t[1][2]+t[0][2]*t[1][1])*t07; -// return_tensor[1][0] = -(t[1][0]*t[2][2]-t[1][2]*t[2][0])*t07; -// return_tensor[1][1] = (t[0][0]*t[2][2]-t04)*t07; -// return_tensor[1][2] = -(t6-t00)*t07; -// return_tensor[2][0] = -(-t[1][0]*t[2][1]+t[1][1]*t[2][0])*t07; -// return_tensor[2][1] = -(t[0][0]*t[2][1]-t01)*t07; -// return_tensor[2][2] = (t4-t8)*t07; -// return return_tensor; -// }; - -// // if desired, take over the -// // inversion of a 4x4 tensor -// // from the FullMatrix - -// default: -// AssertThrow (false, ExcNotImplemented()); -// }; -// return return_tensor; + double t=0; + for (unsigned int i=0; i