Basix 0.9.0.0

Home     Installation     Demos     C++ docs     Python docs

finite-element.h
1 // Copyright (c) 2020-2024 Chris Richardson, Matthew Scroggs and Garth . Wells
2 // FEniCS Project
3 // SPDX-License-Identifier: MIT
4 
5 #pragma once
6 
7 #include "cell.h"
8 #include "element-families.h"
9 #include "maps.h"
10 #include "mdspan.hpp"
11 #include "polyset.h"
12 #include "precompute.h"
13 #include "sobolev-spaces.h"
14 #include <array>
15 #include <concepts>
16 #include <cstdint>
17 #include <functional>
18 #include <map>
19 #include <numeric>
20 #include <span>
21 #include <string>
22 #include <tuple>
23 #include <utility>
24 #include <vector>
25 
27 namespace basix
28 {
29 
30 namespace impl
31 {
32 template <typename T, std::size_t d>
33 using mdspan_t = MDSPAN_IMPL_STANDARD_NAMESPACE::mdspan<
34  T, MDSPAN_IMPL_STANDARD_NAMESPACE::dextents<std::size_t, d>>;
35 template <typename T, std::size_t d>
36 using mdarray_t
37  = MDSPAN_IMPL_STANDARD_NAMESPACE::MDSPAN_IMPL_PROPOSED_NAMESPACE::mdarray<
38  T, MDSPAN_IMPL_STANDARD_NAMESPACE::dextents<std::size_t, d>>;
39 
42 template <typename T>
43 std::array<std::vector<mdspan_t<const T, 2>>, 4>
44 to_mdspan(std::array<std::vector<mdarray_t<T, 2>>, 4>& x)
45 {
46  std::array<std::vector<mdspan_t<const T, 2>>, 4> x1;
47  for (std::size_t i = 0; i < x.size(); ++i)
48  for (std::size_t j = 0; j < x[i].size(); ++j)
49  x1[i].emplace_back(x[i][j].data(), x[i][j].extents());
50 
51  return x1;
52 }
53 
56 template <typename T>
57 std::array<std::vector<mdspan_t<const T, 4>>, 4>
58 to_mdspan(std::array<std::vector<mdarray_t<T, 4>>, 4>& M)
59 {
60  std::array<std::vector<mdspan_t<const T, 4>>, 4> M1;
61  for (std::size_t i = 0; i < M.size(); ++i)
62  for (std::size_t j = 0; j < M[i].size(); ++j)
63  M1[i].emplace_back(M[i][j].data(), M[i][j].extents());
64 
65  return M1;
66 }
67 
70 template <typename T>
71 std::array<std::vector<mdspan_t<const T, 2>>, 4>
72 to_mdspan(const std::array<std::vector<std::vector<T>>, 4>& x,
73  const std::array<std::vector<std::array<std::size_t, 2>>, 4>& shape)
74 {
75  std::array<std::vector<mdspan_t<const T, 2>>, 4> x1;
76  for (std::size_t i = 0; i < x.size(); ++i)
77  for (std::size_t j = 0; j < x[i].size(); ++j)
78  x1[i].push_back(mdspan_t<const T, 2>(x[i][j].data(), shape[i][j]));
79 
80  return x1;
81 }
82 
85 template <typename T>
86 std::array<std::vector<mdspan_t<const T, 4>>, 4>
87 to_mdspan(const std::array<std::vector<std::vector<T>>, 4>& M,
88  const std::array<std::vector<std::array<std::size_t, 4>>, 4>& shape)
89 {
90  std::array<std::vector<mdspan_t<const T, 4>>, 4> M1;
91  for (std::size_t i = 0; i < M.size(); ++i)
92  for (std::size_t j = 0; j < M[i].size(); ++j)
93  M1[i].push_back(mdspan_t<const T, 4>(M[i][j].data(), shape[i][j]));
94 
95  return M1;
96 }
97 
98 } // namespace impl
99 
100 namespace element
101 {
103 template <typename T, std::size_t d>
104 using mdspan_t = impl::mdspan_t<T, d>;
105 
121 template <std::floating_point T>
122 std::tuple<std::array<std::vector<std::vector<T>>, 4>,
123  std::array<std::vector<std::array<std::size_t, 2>>, 4>,
124  std::array<std::vector<std::vector<T>>, 4>,
125  std::array<std::vector<std::array<std::size_t, 4>>, 4>>
126 make_discontinuous(const std::array<std::vector<mdspan_t<const T, 2>>, 4>& x,
127  const std::array<std::vector<mdspan_t<const T, 4>>, 4>& M,
128  std::size_t tdim, std::size_t value_size);
129 
130 } // namespace element
131 
137 template <std::floating_point F>
139 {
140  template <typename T, std::size_t d>
141  using mdspan_t = MDSPAN_IMPL_STANDARD_NAMESPACE::mdspan<
142  T, MDSPAN_IMPL_STANDARD_NAMESPACE::dextents<std::size_t, d>>;
143 
144 public:
318  polyset::type poly_type, int degree,
319  const std::vector<std::size_t>& value_shape,
320  mdspan_t<const F, 2> wcoeffs,
321  const std::array<std::vector<mdspan_t<const F, 2>>, 4>& x,
322  const std::array<std::vector<mdspan_t<const F, 4>>, 4>& M,
326  element::lagrange_variant lvariant,
327  element::dpc_variant dvariant,
328  std::vector<int> dof_ordering = {});
329 
331  FiniteElement(const FiniteElement& element) = default;
332 
334  FiniteElement(FiniteElement&& element) = default;
335 
337  ~FiniteElement() = default;
338 
340  FiniteElement& operator=(const FiniteElement& element) = default;
341 
343  FiniteElement& operator=(FiniteElement&& element) = default;
344 
349  bool operator==(const FiniteElement& e) const;
350 
352  std::size_t hash() const;
353 
363  std::array<std::size_t, 4> tabulate_shape(std::size_t nd,
364  std::size_t num_points) const
365  {
366  std::size_t ndsize = 1;
367  for (std::size_t i = 1; i <= nd; ++i)
368  ndsize *= (_cell_tdim + i);
369  for (std::size_t i = 1; i <= nd; ++i)
370  ndsize /= i;
371  std::size_t vs = std::accumulate(_value_shape.begin(), _value_shape.end(),
372  1, std::multiplies{});
373  std::size_t ndofs = _coeffs.second[0];
374  return {ndsize, num_points, ndofs, vs};
375  }
376 
398  std::pair<std::vector<F>, std::array<std::size_t, 4>>
399  tabulate(int nd, impl::mdspan_t<const F, 2> x) const;
400 
424  std::pair<std::vector<F>, std::array<std::size_t, 4>>
425  tabulate(int nd, std::span<const F> x,
426  std::array<std::size_t, 2> shape) const;
427 
453  void tabulate(int nd, impl::mdspan_t<const F, 2> x,
454  mdspan_t<F, 4> basis) const;
455 
481  void tabulate(int nd, std::span<const F> x, std::array<std::size_t, 2> xshape,
482  std::span<F> basis) const;
483 
486  cell::type cell_type() const { return _cell_type; }
487 
490  polyset::type polyset_type() const { return _poly_type; }
491 
494  int degree() const { return _degree; }
495 
500  int embedded_superdegree() const { return _embedded_superdegree; }
501 
505  int embedded_subdegree() const { return _embedded_subdegree; }
506 
512  const std::vector<std::size_t>& value_shape() const { return _value_shape; }
513 
518  int dim() const { return _coeffs.second[0]; }
519 
522  element::family family() const { return _family; }
523 
527  {
528  return _lagrange_variant;
529  }
530 
533  element::dpc_variant dpc_variant() const { return _dpc_variant; }
534 
537  maps::type map_type() const { return _map_type; }
538 
541  sobolev::space sobolev_space() const { return _sobolev_space; }
542 
547  bool discontinuous() const { return _discontinuous; }
548 
552  {
553  return _dof_transformations_are_permutations;
554  }
555 
558  {
559  return _dof_transformations_are_identity;
560  }
561 
576  std::pair<std::vector<F>, std::array<std::size_t, 3>>
577  push_forward(impl::mdspan_t<const F, 3> U, impl::mdspan_t<const F, 3> J,
578  std::span<const F> detJ, impl::mdspan_t<const F, 3> K) const;
579 
587  std::pair<std::vector<F>, std::array<std::size_t, 3>>
588  pull_back(impl::mdspan_t<const F, 3> u, impl::mdspan_t<const F, 3> J,
589  std::span<const F> detJ, impl::mdspan_t<const F, 3> K) const;
590 
622  template <typename O, typename P, typename Q, typename R>
623  std::function<void(O&, const P&, const Q&, F, const R&)> map_fn() const
624  {
625  switch (_map_type)
626  {
627  case maps::type::identity:
628  return [](O& u, const P& U, const Q&, F, const R&)
629  {
630  assert(U.extent(0) == u.extent(0));
631  assert(U.extent(1) == u.extent(1));
632  for (std::size_t i = 0; i < U.extent(0); ++i)
633  for (std::size_t j = 0; j < U.extent(1); ++j)
634  u(i, j) = U(i, j);
635  };
636  case maps::type::covariantPiola:
637  return [](O& u, const P& U, const Q& J, F detJ, const R& K)
638  { maps::covariant_piola(u, U, J, detJ, K); };
639  case maps::type::contravariantPiola:
640  return [](O& u, const P& U, const Q& J, F detJ, const R& K)
641  { maps::contravariant_piola(u, U, J, detJ, K); };
642  case maps::type::doubleCovariantPiola:
643  return [](O& u, const P& U, const Q& J, F detJ, const R& K)
644  { maps::double_covariant_piola(u, U, J, detJ, K); };
645  case maps::type::doubleContravariantPiola:
646  return [](O& u, const P& U, const Q& J, F detJ, const R& K)
647  { maps::double_contravariant_piola(u, U, J, detJ, K); };
648  default:
649  throw std::runtime_error("Map not implemented");
650  }
651  }
652 
660  const std::vector<std::vector<std::vector<int>>>& entity_dofs() const
661  {
662  return _edofs;
663  }
664 
674  const std::vector<std::vector<std::vector<int>>>& entity_closure_dofs() const
675  {
676  return _e_closure_dofs;
677  }
678 
760  std::pair<std::vector<F>, std::array<std::size_t, 3>>
761  base_transformations() const;
762 
767  std::map<cell::type, std::pair<std::vector<F>, std::array<std::size_t, 3>>>
769  {
770  return _entity_transformations;
771  }
772 
795  void permute(std::span<std::int32_t> d, std::uint32_t cell_info) const
796  {
797  if (!_dof_transformations_are_permutations)
798  {
799  throw std::runtime_error(
800  "The DOF transformations for this element are not permutations");
801  }
802 
803  if (_dof_transformations_are_identity)
804  return;
805  else
806  permute_data<std::int32_t, false>(d, 1, cell_info, _eperm);
807  }
808 
826  void permute_inv(std::span<std::int32_t> d, std::uint32_t cell_info) const
827  {
828  if (!_dof_transformations_are_permutations)
829  {
830  throw std::runtime_error(
831  "The DOF transformations for this element are not permutations");
832  }
833 
834  if (_dof_transformations_are_identity)
835  return;
836  else
837  permute_data<std::int32_t, true>(d, 1, cell_info, _eperm_rev);
838  }
839 
870  template <typename T>
871  void T_apply(std::span<T> u, int n, std::uint32_t cell_info) const;
872 
883  template <typename T>
884  void Tt_apply(std::span<T> u, int n, std::uint32_t cell_info) const;
885 
897  template <typename T>
898  void Tt_inv_apply(std::span<T> u, int n, std::uint32_t cell_info) const;
899 
910  template <typename T>
911  void Tinv_apply(std::span<T> u, int n, std::uint32_t cell_info) const;
912 
923  template <typename T>
924  void Tt_apply_right(std::span<T> u, int n, std::uint32_t cell_info) const;
925 
935  template <typename T>
936  void T_apply_right(std::span<T> u, int n, std::uint32_t cell_info) const;
937 
948  template <typename T>
949  void Tinv_apply_right(std::span<T> u, int n, std::uint32_t cell_info) const;
950 
961  template <typename T>
962  void Tt_inv_apply_right(std::span<T> u, int n, std::uint32_t cell_info) const;
963 
970  const std::pair<std::vector<F>, std::array<std::size_t, 2>>& points() const
971  {
972  return _points;
973  }
974 
1026  const std::pair<std::vector<F>, std::array<std::size_t, 2>>&
1028  {
1029  return _matM;
1030  }
1031 
1037  const std::pair<std::vector<F>, std::array<std::size_t, 2>>&
1038  dual_matrix() const
1039  {
1040  return _dual_matrix;
1041  }
1042 
1079  const std::pair<std::vector<F>, std::array<std::size_t, 2>>& wcoeffs() const
1080  {
1081  return _wcoeffs;
1082  }
1083 
1088  const std::array<
1089  std::vector<std::pair<std::vector<F>, std::array<std::size_t, 2>>>, 4>&
1090  x() const
1091  {
1092  return _x;
1093  }
1094 
1131  const std::array<
1132  std::vector<std::pair<std::vector<F>, std::array<std::size_t, 4>>>, 4>&
1133  M() const
1134  {
1135  return _M;
1136  }
1137 
1141  const std::pair<std::vector<F>, std::array<std::size_t, 2>>&
1143  {
1144  return _coeffs;
1145  }
1146 
1158  {
1159  return !_tensor_factors.empty();
1160  }
1161 
1174  std::vector<std::vector<FiniteElement<F>>>
1176  {
1178  throw std::runtime_error("Element has no tensor product representation.");
1179  return _tensor_factors;
1180  }
1181 
1186  bool interpolation_is_identity() const { return _interpolation_is_identity; }
1187 
1189  int interpolation_nderivs() const { return _interpolation_nderivs; }
1190 
1192  const std::vector<int>& dof_ordering() const { return _dof_ordering; }
1193 
1194 private:
1201  template <typename T, bool post>
1202  void permute_data(
1203  std::span<T> data, int block_size, std::uint32_t cell_info,
1204  const std::map<cell::type, std::vector<std::vector<std::size_t>>>& eperm)
1205  const;
1206 
1207  using array2_t = std::pair<std::vector<F>, std::array<std::size_t, 2>>;
1208  using array3_t = std::pair<std::vector<F>, std::array<std::size_t, 3>>;
1209  using trans_data_t
1210  = std::vector<std::pair<std::vector<std::size_t>, array2_t>>;
1211 
1218  template <typename T, bool post, typename OP>
1219  void
1220  transform_data(std::span<T> data, int block_size, std::uint32_t cell_info,
1221  const std::map<cell::type, trans_data_t>& etrans, OP op) const;
1222 
1223  // Cell type
1224  cell::type _cell_type;
1225 
1226  // Polyset type
1227  polyset::type _poly_type;
1228 
1229  // Topological dimension of the cell
1230  std::size_t _cell_tdim;
1231 
1232  // Topological dimension of the cell
1233  std::vector<std::vector<cell::type>> _cell_subentity_types;
1234 
1235  // Finite element family
1236  element::family _family;
1237 
1238  // Lagrange variant
1239  element::lagrange_variant _lagrange_variant;
1240 
1241  // DPC variant
1242  element::dpc_variant _dpc_variant;
1243 
1244  // Degree that was input when creating the element
1245  int _degree;
1246 
1247  // Degree
1248  int _interpolation_nderivs;
1249 
1250  // Highest degree polynomial in element's polyset
1251  int _embedded_superdegree;
1252 
1253  // Highest degree space that is a subspace of element's polyset
1254  int _embedded_subdegree;
1255 
1256  // Value shape
1257  std::vector<std::size_t> _value_shape;
1258 
1260  maps::type _map_type;
1261 
1263  sobolev::space _sobolev_space;
1264 
1265  // Shape function coefficient of expansion sets on cell. If shape
1266  // function is given by @f$\psi_i = \sum_{k} \phi_{k}
1267  // \alpha^{i}_{k}@f$, then _coeffs(i, j) = @f$\alpha^i_k@f$. ie
1268  // _coeffs.row(i) are the expansion coefficients for shape function i
1269  // (@f$\psi_{i}@f$).
1270  std::pair<std::vector<F>, std::array<std::size_t, 2>> _coeffs;
1271 
1272  // Dofs associated with each cell (sub-)entity
1273  std::vector<std::vector<std::vector<int>>> _edofs;
1274 
1275  // Dofs associated with the closdure of each cell (sub-)entity
1276  std::vector<std::vector<std::vector<int>>> _e_closure_dofs;
1277 
1278  // Entity transformations
1279  std::map<cell::type, array3_t> _entity_transformations;
1280 
1281  // Set of points used for point evaluation
1282  // Experimental - currently used for an implementation of
1283  // "tabulate_dof_coordinates" Most useful for Lagrange. This may change or go
1284  // away. For non-Lagrange elements, these points will be used in combination
1285  // with _interpolation_matrix to perform interpolation
1286  std::pair<std::vector<F>, std::array<std::size_t, 2>> _points;
1287 
1288  // Interpolation points on the cell. The shape is (entity_dim, num
1289  // entities of given dimension, num_points, tdim)
1290  std::array<std::vector<std::pair<std::vector<F>, std::array<std::size_t, 2>>>,
1291  4>
1292  _x;
1293 
1295  std::pair<std::vector<F>, std::array<std::size_t, 2>> _matM;
1296 
1297  // Indicates whether or not the DOF transformations are all
1298  // permutations
1299  bool _dof_transformations_are_permutations;
1300 
1301  // Indicates whether or not the DOF transformations are all identity
1302  bool _dof_transformations_are_identity;
1303 
1304  // The entity permutations (factorised). This will only be set if
1305  // _dof_transformations_are_permutations is True and
1306  // _dof_transformations_are_identity is False
1307  std::map<cell::type, std::vector<std::vector<std::size_t>>> _eperm;
1308 
1309  // The reverse entity permutations (factorised). This will only be set
1310  // if _dof_transformations_are_permutations is True and
1311  // _dof_transformations_are_identity is False
1312  std::map<cell::type, std::vector<std::vector<std::size_t>>> _eperm_rev;
1313 
1314  // The entity transformations in precomputed form
1315  std::map<cell::type, trans_data_t> _etrans;
1316 
1317  // The transposed entity transformations in precomputed form
1318  std::map<cell::type, trans_data_t> _etransT;
1319 
1320  // The inverse entity transformations in precomputed form
1321  std::map<cell::type, trans_data_t> _etrans_inv;
1322 
1323  // The inverse transpose entity transformations in precomputed form
1324  std::map<cell::type, trans_data_t> _etrans_invT;
1325 
1326  // Indicates whether or not this is the discontinuous version of the
1327  // element
1328  bool _discontinuous;
1329 
1330  // The dual matrix
1331  std::pair<std::vector<F>, std::array<std::size_t, 2>> _dual_matrix;
1332 
1333  // Dof reordering for different element dof layout compatibility.
1334  // The reference basix layout is ordered by entity, i.e. dofs on
1335  // vertices, followed by edges, faces, then internal dofs.
1336  // _dof_ordering stores the map to the new order required, e.g.
1337  // for a P2 triangle, _dof_ordering=[0 3 5 1 2 4] will place
1338  // dofs 0, 3, 5 on the vertices and 1, 2, 4, on the edges.
1339  std::vector<int> _dof_ordering;
1340 
1341  // Tensor product representation
1342  // Entries of tuple are (list of elements on an interval, permutation
1343  // of DOF numbers)
1344  // @todo: For vector-valued elements, a tensor product type and a
1345  // scaling factor may additionally be needed.
1346  std::vector<std::vector<FiniteElement>> _tensor_factors;
1347 
1348  // Is the interpolation matrix an identity?
1349  bool _interpolation_is_identity;
1350 
1351  // The coefficients that define the polynomial set in terms of the
1352  // orthonormal polynomials
1353  std::pair<std::vector<F>, std::array<std::size_t, 2>> _wcoeffs;
1354 
1355  // Interpolation matrices for each entity
1356  using array4_t
1357  = std::vector<std::pair<std::vector<F>, std::array<std::size_t, 4>>>;
1358  std::array<array4_t, 4> _M;
1359 };
1360 
1386 template <std::floating_point T>
1387 FiniteElement<T> create_custom_element(
1388  cell::type cell_type, const std::vector<std::size_t>& value_shape,
1389  impl::mdspan_t<const T, 2> wcoeffs,
1390  const std::array<std::vector<impl::mdspan_t<const T, 2>>, 4>& x,
1391  const std::array<std::vector<impl::mdspan_t<const T, 4>>, 4>& M,
1392  int interpolation_nderivs, maps::type map_type,
1393  sobolev::space sobolev_space, bool discontinuous, int embedded_subdegree,
1394  int embedded_superdegree, polyset::type poly_type);
1395 
1407 template <std::floating_point T>
1408 FiniteElement<T> create_element(element::family family, cell::type cell,
1409  int degree, element::lagrange_variant lvariant,
1410  element::dpc_variant dvariant,
1411  bool discontinuous,
1412  std::vector<int> dof_ordering = {});
1413 
1425 std::vector<int> tp_dof_ordering(element::family family, cell::type cell,
1426  int degree, element::lagrange_variant lvariant,
1427  element::dpc_variant dvariant,
1428  bool discontinuous);
1429 
1442 template <std::floating_point T>
1443 std::vector<std::vector<FiniteElement<T>>>
1444 tp_factors(element::family family, cell::type cell, int degree,
1446  bool discontinuous, std::vector<int> dof_ordering);
1447 
1459 template <std::floating_point T>
1460 FiniteElement<T>
1461 create_tp_element(element::family family, cell::type cell, int degree,
1462  element::lagrange_variant lvariant,
1463  element::dpc_variant dvariant, bool discontinuous);
1464 
1467 std::string version();
1468 
1469 //-----------------------------------------------------------------------------
1470 template <std::floating_point F>
1471 template <typename T, bool post>
1472 void FiniteElement<F>::permute_data(
1473  std::span<T> data, int block_size, std::uint32_t cell_info,
1474  const std::map<cell::type, std::vector<std::vector<std::size_t>>>& eperm)
1475  const
1476 {
1477  if (_cell_tdim >= 2)
1478  {
1479  // This assumes 3 bits are used per face. This will need updating if 3D
1480  // cells with faces with more than 4 sides are implemented
1481  int face_start = _cell_tdim == 3 ? 3 * _edofs[2].size() : 0;
1482 
1483  // Permute DOFs on edges
1484  {
1485  auto& trans = eperm.at(cell::type::interval)[0];
1486  for (std::size_t e = 0; e < _edofs[1].size(); ++e)
1487  {
1488  // Reverse an edge
1489  if (cell_info >> (face_start + e) & 1)
1490  {
1491  precompute::apply_permutation_mapped(trans, data, _edofs[1][e],
1492  block_size);
1493  }
1494  }
1495  }
1496 
1497  if (_cell_tdim == 3)
1498  {
1499  // Permute DOFs on faces
1500  for (std::size_t f = 0; f < _edofs[2].size(); ++f)
1501  {
1502  auto& trans = eperm.at(_cell_subentity_types[2][f]);
1503 
1504  // Reflect a face (pre rotate)
1505  if (!post and cell_info >> (3 * f) & 1)
1506  {
1507  precompute::apply_permutation_mapped(trans[1], data, _edofs[2][f],
1508  block_size);
1509  }
1510 
1511  // Rotate a face
1512  for (std::uint32_t r = 0; r < (cell_info >> (3 * f + 1) & 3); ++r)
1513  {
1514  precompute::apply_permutation_mapped(trans[0], data, _edofs[2][f],
1515  block_size);
1516  }
1517 
1518  // Reflect a face (post rotate)
1519  if (post and cell_info >> (3 * f) & 1)
1520  {
1521  precompute::apply_permutation_mapped(trans[1], data, _edofs[2][f],
1522  block_size);
1523  }
1524  }
1525  }
1526  }
1527 }
1528 //-----------------------------------------------------------------------------
1529 template <std::floating_point F>
1530 template <typename T, bool post, typename OP>
1531 void FiniteElement<F>::transform_data(
1532  std::span<T> data, int block_size, std::uint32_t cell_info,
1533  const std::map<cell::type, trans_data_t>& etrans, OP op) const
1534 {
1535  if (_cell_tdim >= 2)
1536  {
1537  // This assumes 3 bits are used per face. This will need updating if
1538  // 3D cells with faces with more than 4 sides are implemented
1539  int face_start = _cell_tdim == 3 ? 3 * _edofs[2].size() : 0;
1540  int dofstart = 0;
1541  for (auto& edofs0 : _edofs[0])
1542  dofstart += edofs0.size();
1543 
1544  // Transform DOFs on edges
1545  {
1546  auto& [v_size_t, matrix] = etrans.at(cell::type::interval)[0];
1547  for (std::size_t e = 0; e < _edofs[1].size(); ++e)
1548  {
1549  // Reverse an edge
1550  if (cell_info >> (face_start + e) & 1)
1551  {
1552  op(std::span(v_size_t),
1553  mdspan_t<const F, 2>(matrix.first.data(), matrix.second), data,
1554  dofstart, block_size);
1555  }
1556  dofstart += _edofs[1][e].size();
1557  }
1558  }
1559 
1560  if (_cell_tdim == 3)
1561  {
1562  // Permute DOFs on faces
1563  for (std::size_t f = 0; f < _edofs[2].size(); ++f)
1564  {
1565  auto& trans = etrans.at(_cell_subentity_types[2][f]);
1566 
1567  // Reflect a face (pre rotation)
1568  if (!post and cell_info >> (3 * f) & 1)
1569  {
1570  const auto& m = trans[1];
1571  const auto& v_size_t = std::get<0>(m);
1572  const auto& matrix = std::get<1>(m);
1573  op(std::span(v_size_t),
1574  mdspan_t<const F, 2>(matrix.first.data(), matrix.second), data,
1575  dofstart, block_size);
1576  }
1577 
1578  // Rotate a face
1579  for (std::uint32_t r = 0; r < (cell_info >> (3 * f + 1) & 3); ++r)
1580  {
1581  const auto& m = trans[0];
1582  const auto& v_size_t = std::get<0>(m);
1583  const auto& matrix = std::get<1>(m);
1584  op(std::span(v_size_t),
1585  mdspan_t<const F, 2>(matrix.first.data(), matrix.second), data,
1586  dofstart, block_size);
1587  }
1588 
1589  // Reflect a face (post rotation)
1590  if (post and cell_info >> (3 * f) & 1)
1591  {
1592  const auto& m = trans[1];
1593  const auto& v_size_t = std::get<0>(m);
1594  const auto& matrix = std::get<1>(m);
1595  op(std::span(v_size_t),
1596  mdspan_t<const F, 2>(matrix.first.data(), matrix.second), data,
1597  dofstart, block_size);
1598  }
1599 
1600  dofstart += _edofs[2][f].size();
1601  }
1602  }
1603  }
1604 }
1605 //-----------------------------------------------------------------------------
1606 template <std::floating_point F>
1607 template <typename T>
1608 void FiniteElement<F>::T_apply(std::span<T> u, int n,
1609  std::uint32_t cell_info) const
1610 {
1611  if (_dof_transformations_are_identity)
1612  return;
1613 
1614  if (_dof_transformations_are_permutations)
1615  permute_data<T, false>(u, n, cell_info, _eperm);
1616  else
1617  {
1618  transform_data<T, false>(u, n, cell_info, _etrans,
1619  precompute::apply_matrix<F, T>);
1620  }
1621 }
1622 //-----------------------------------------------------------------------------
1623 template <std::floating_point F>
1624 template <typename T>
1625 void FiniteElement<F>::Tt_apply(std::span<T> u, int n,
1626  std::uint32_t cell_info) const
1627 {
1628  if (_dof_transformations_are_identity)
1629  return;
1630  else if (_dof_transformations_are_permutations)
1631  permute_data<T, true>(u, n, cell_info, _eperm_rev);
1632  else
1633  {
1634  transform_data<T, true>(u, n, cell_info, _etransT,
1635  precompute::apply_matrix<F, T>);
1636  }
1637 }
1638 //-----------------------------------------------------------------------------
1639 template <std::floating_point F>
1640 template <typename T>
1641 void FiniteElement<F>::Tt_inv_apply(std::span<T> u, int n,
1642  std::uint32_t cell_info) const
1643 {
1644  if (_dof_transformations_are_identity)
1645  return;
1646  else if (_dof_transformations_are_permutations)
1647  permute_data<T, false>(u, n, cell_info, _eperm);
1648  else
1649  {
1650  transform_data<T, false>(u, n, cell_info, _etrans_invT,
1651  precompute::apply_matrix<F, T>);
1652  }
1653 }
1654 //-----------------------------------------------------------------------------
1655 template <std::floating_point F>
1656 template <typename T>
1657 void FiniteElement<F>::Tinv_apply(std::span<T> u, int n,
1658  std::uint32_t cell_info) const
1659 {
1660  if (_dof_transformations_are_identity)
1661  return;
1662  else if (_dof_transformations_are_permutations)
1663  permute_data<T, true>(u, n, cell_info, _eperm_rev);
1664  else
1665  {
1666  transform_data<T, true>(u, n, cell_info, _etrans_inv,
1667  precompute::apply_matrix<F, T>);
1668  }
1669 }
1670 //-----------------------------------------------------------------------------
1671 template <std::floating_point F>
1672 template <typename T>
1673 void FiniteElement<F>::Tt_apply_right(std::span<T> u, int n,
1674  std::uint32_t cell_info) const
1675 {
1676  if (_dof_transformations_are_identity)
1677  return;
1678  else if (_dof_transformations_are_permutations)
1679  {
1680  assert(u.size() % n == 0);
1681  const int step = u.size() / n;
1682  for (int i = 0; i < n; ++i)
1683  {
1684  std::span<T> dblock(u.data() + i * step, step);
1685  permute_data<T, false>(dblock, 1, cell_info, _eperm);
1686  }
1687  }
1688  else
1689  {
1690  transform_data<T, false>(u, n, cell_info, _etrans,
1691  precompute::apply_tranpose_matrix_right<F, T>);
1692  }
1693 }
1694 //-----------------------------------------------------------------------------
1695 template <std::floating_point F>
1696 template <typename T>
1697 void FiniteElement<F>::Tinv_apply_right(std::span<T> u, int n,
1698  std::uint32_t cell_info) const
1699 {
1700  if (_dof_transformations_are_identity)
1701  return;
1702  else if (_dof_transformations_are_permutations)
1703  {
1704  assert(u.size() % n == 0);
1705  const int step = u.size() / n;
1706  for (int i = 0; i < n; ++i)
1707  {
1708  std::span<T> dblock(u.data() + i * step, step);
1709  permute_data<T, false>(dblock, 1, cell_info, _eperm);
1710  }
1711  }
1712  else
1713  {
1714  transform_data<T, false>(u, n, cell_info, _etrans_invT,
1715  precompute::apply_tranpose_matrix_right<F, T>);
1716  }
1717 }
1718 //-----------------------------------------------------------------------------
1719 template <std::floating_point F>
1720 template <typename T>
1721 void FiniteElement<F>::T_apply_right(std::span<T> u, int n,
1722  std::uint32_t cell_info) const
1723 {
1724  if (_dof_transformations_are_identity)
1725  return;
1726  else if (_dof_transformations_are_permutations)
1727  {
1728  assert(u.size() % n == 0);
1729  const int step = u.size() / n;
1730  for (int i = 0; i < n; ++i)
1731  {
1732  std::span<T> dblock(u.data() + i * step, step);
1733  permute_data<T, true>(dblock, 1, cell_info, _eperm_rev);
1734  }
1735  }
1736  else
1737  {
1738  transform_data<T, true>(u, n, cell_info, _etransT,
1739  precompute::apply_tranpose_matrix_right<F, T>);
1740  }
1741 }
1742 //-----------------------------------------------------------------------------
1743 template <std::floating_point F>
1744 template <typename T>
1745 void FiniteElement<F>::Tt_inv_apply_right(std::span<T> u, int n,
1746  std::uint32_t cell_info) const
1747 {
1748  if (_dof_transformations_are_identity)
1749  return;
1750  else if (_dof_transformations_are_permutations)
1751  {
1752  assert(u.size() % n == 0);
1753  const int step = u.size() / n;
1754  for (int i = 0; i < n; ++i)
1755  {
1756  std::span<T> dblock(u.data() + i * step, step);
1757  permute_data<T, true>(dblock, 1, cell_info, _eperm_rev);
1758  }
1759  }
1760  else
1761  {
1762  transform_data<T, true>(u, n, cell_info, _etrans_inv,
1763  precompute::apply_tranpose_matrix_right<F, T>);
1764  }
1765 }
1766 //-----------------------------------------------------------------------------
1767 
1768 } // namespace basix
A finite element.
Definition: finite-element.h:139
std::map< cell::type, std::pair< std::vector< F >, std::array< std::size_t, 3 > > > entity_transformations() const
Return the entity dof transformation matrices.
Definition: finite-element.h:768
void Tinv_apply_right(std::span< T > u, int n, std::uint32_t cell_info) const
Right(post)-apply the inverse of the operator applied by T_apply().
Definition: finite-element.h:1697
const std::vector< int > & dof_ordering() const
Get dof layout.
Definition: finite-element.h:1192
std::vector< std::vector< FiniteElement< F > > > get_tensor_product_representation() const
Get the tensor product representation of this element.
Definition: finite-element.h:1175
const std::pair< std::vector< F >, std::array< std::size_t, 2 > > & wcoeffs() const
Get the coefficients that define the polynomial set in terms of the orthonormal polynomials.
Definition: finite-element.h:1079
const std::array< std::vector< std::pair< std::vector< F >, std::array< std::size_t, 4 > > >, 4 > & M() const
Get the interpolation matrices for each subentity.
Definition: finite-element.h:1133
std::pair< std::vector< F >, std::array< std::size_t, 3 > > base_transformations() const
Get the base transformations.
Definition: finite-element.cpp:1339
void T_apply(std::span< T > u, int n, std::uint32_t cell_info) const
Transform basis functions from the reference element ordering and orientation to the globally consist...
Definition: finite-element.h:1608
void Tt_apply(std::span< T > u, int n, std::uint32_t cell_info) const
Apply the transpose of the operator applied by T_apply().
Definition: finite-element.h:1625
const std::pair< std::vector< F >, std::array< std::size_t, 2 > > & coefficient_matrix() const
Get the matrix of coefficients.
Definition: finite-element.h:1142
const std::pair< std::vector< F >, std::array< std::size_t, 2 > > & dual_matrix() const
Get the dual matrix.
Definition: finite-element.h:1038
bool dof_transformations_are_identity() const
Indicates is the dof transformations are all the identity.
Definition: finite-element.h:557
bool operator==(const FiniteElement &e) const
Check if two elements are the same.
Definition: finite-element.cpp:1157
int embedded_subdegree() const
Highest degree n such that a Lagrange (or vector Lagrange) element of degree n is a subspace of this ...
Definition: finite-element.h:505
void T_apply_right(std::span< T > u, int n, std::uint32_t cell_info) const
Right(post)-apply the operator applied by T_apply().
Definition: finite-element.h:1721
FiniteElement(FiniteElement &&element)=default
Move constructor.
bool has_tensor_product_factorisation() const
Indicates whether or not this element can be represented as a product of elements defined on lower-di...
Definition: finite-element.h:1157
const std::vector< std::vector< std::vector< int > > > & entity_closure_dofs() const
Get the dofs on the closure of each topological entity: (vertices, edges, faces, cell) in that order.
Definition: finite-element.h:674
int interpolation_nderivs() const
The number of derivatives needed when interpolating.
Definition: finite-element.h:1189
std::pair< std::vector< F >, std::array< std::size_t, 4 > > tabulate(int nd, impl::mdspan_t< const F, 2 > x) const
Compute basis values and derivatives at set of points.
Definition: finite-element.cpp:1247
int embedded_superdegree() const
Lowest degree n such that the highest degree polynomial in this element is contained in a Lagrange (o...
Definition: finite-element.h:500
std::function< void(O &, const P &, const Q &, F, const R &)> map_fn() const
Return a function that performs the appropriate push-forward/pull-back for the element type.
Definition: finite-element.h:623
int dim() const
Dimension of the finite element space.
Definition: finite-element.h:518
FiniteElement(element::family family, cell::type cell_type, polyset::type poly_type, int degree, const std::vector< std::size_t > &value_shape, mdspan_t< const F, 2 > wcoeffs, const std::array< std::vector< mdspan_t< const F, 2 >>, 4 > &x, const std::array< std::vector< mdspan_t< const F, 4 >>, 4 > &M, int interpolation_nderivs, maps::type map_type, sobolev::space sobolev_space, bool discontinuous, int embedded_subdegree, int embedded_superdegree, element::lagrange_variant lvariant, element::dpc_variant dvariant, std::vector< int > dof_ordering={})
Construct a finite element.
void permute(std::span< std::int32_t > d, std::uint32_t cell_info) const
Permute indices associated with degree-of-freedoms on the reference element ordering to the globally ...
Definition: finite-element.h:795
std::size_t hash() const
Get a unique hash of this element.
Definition: finite-element.cpp:1197
void permute_inv(std::span< std::int32_t > d, std::uint32_t cell_info) const
Perform the inverse of the operation applied by permute().
Definition: finite-element.h:826
void Tt_apply_right(std::span< T > u, int n, std::uint32_t cell_info) const
Right(post)-apply the transpose of the operator applied by T_apply().
Definition: finite-element.h:1673
void Tt_inv_apply(std::span< T > u, int n, std::uint32_t cell_info) const
Apply the inverse transpose of the operator applied by T_apply().
Definition: finite-element.h:1641
element::lagrange_variant lagrange_variant() const
Lagrange variant of the element.
Definition: finite-element.h:526
void Tt_inv_apply_right(std::span< T > u, int n, std::uint32_t cell_info) const
Right(post)-apply the transpose inverse of the operator applied by T_apply().
Definition: finite-element.h:1745
const std::vector< std::size_t > & value_shape() const
Element value tensor shape.
Definition: finite-element.h:512
int degree() const
Get the element polynomial degree.
Definition: finite-element.h:494
const std::pair< std::vector< F >, std::array< std::size_t, 2 > > & points() const
Return the interpolation points.
Definition: finite-element.h:970
void Tinv_apply(std::span< T > u, int n, std::uint32_t cell_info) const
Apply the inverse of the operator applied by T_apply().
Definition: finite-element.h:1657
std::array< std::size_t, 4 > tabulate_shape(std::size_t nd, std::size_t num_points) const
Array shape for tabulate basis values and derivatives at set of points.
Definition: finite-element.h:363
FiniteElement & operator=(const FiniteElement &element)=default
Assignment operator.
sobolev::space sobolev_space() const
Underlying Sobolev space for this element.
Definition: finite-element.h:541
FiniteElement(const FiniteElement &element)=default
Copy constructor.
polyset::type polyset_type() const
Get the element polyset type.
Definition: finite-element.h:490
const std::pair< std::vector< F >, std::array< std::size_t, 2 > > & interpolation_matrix() const
Return a matrix of weights interpolation.
Definition: finite-element.h:1027
std::pair< std::vector< F >, std::array< std::size_t, 3 > > pull_back(impl::mdspan_t< const F, 3 > u, impl::mdspan_t< const F, 3 > J, std::span< const F > detJ, impl::mdspan_t< const F, 3 > K) const
Map function values from a physical cell to the reference.
Definition: finite-element.cpp:1448
bool dof_transformations_are_permutations() const
Indicates if the degree-of-freedom transformations are all permutations.
Definition: finite-element.h:551
std::pair< std::vector< F >, std::array< std::size_t, 3 > > push_forward(impl::mdspan_t< const F, 3 > U, impl::mdspan_t< const F, 3 > J, std::span< const F > detJ, impl::mdspan_t< const F, 3 > K) const
Map function values from the reference to a physical cell.
Definition: finite-element.cpp:1408
const std::vector< std::vector< std::vector< int > > > & entity_dofs() const
Get the dofs on each topological entity: (vertices, edges, faces, cell) in that order.
Definition: finite-element.h:660
cell::type cell_type() const
Get the element cell type.
Definition: finite-element.h:486
bool interpolation_is_identity() const
Indicates whether or not the interpolation matrix for this element is an identity matrix.
Definition: finite-element.h:1186
const std::array< std::vector< std::pair< std::vector< F >, std::array< std::size_t, 2 > > >, 4 > & x() const
Get the interpolation points for each subentity.
Definition: finite-element.h:1090
FiniteElement & operator=(FiniteElement &&element)=default
Move assignment operator.
bool discontinuous() const
Indicates whether this element is the discontinuous variant.
Definition: finite-element.h:547
maps::type map_type() const
Map type for the element.
Definition: finite-element.h:537
element::dpc_variant dpc_variant() const
DPC variant of the element.
Definition: finite-element.h:533
element::family family() const
The finite element family.
Definition: finite-element.h:522
~FiniteElement()=default
Destructor.
type
Cell type.
Definition: cell.h:21
lagrange_variant
Variants of a Lagrange space that can be created.
Definition: element-families.h:12
impl::mdspan_t< T, d > mdspan_t
Typedef for mdspan.
Definition: finite-element.h:104
std::tuple< std::array< std::vector< std::vector< T > >, 4 >, std::array< std::vector< std::array< std::size_t, 2 > >, 4 >, std::array< std::vector< std::vector< T > >, 4 >, std::array< std::vector< std::array< std::size_t, 4 > >, 4 > > make_discontinuous(const std::array< std::vector< mdspan_t< const T, 2 >>, 4 > &x, const std::array< std::vector< mdspan_t< const T, 4 >>, 4 > &M, std::size_t tdim, std::size_t value_size)
Definition: finite-element.cpp:520
dpc_variant
Definition: element-families.h:32
family
Available element families.
Definition: element-families.h:45
void covariant_piola(O &&r, const P &U, const Q &, double, const R &K)
Covariant Piola map.
Definition: maps.h:60
void contravariant_piola(O &&r, const P &U, const Q &J, double detJ, const R &)
Contravariant Piola map.
Definition: maps.h:80
void double_contravariant_piola(O &&r, const P &U, const Q &J, double detJ, const R &)
Double contravariant Piola map.
Definition: maps.h:134
void double_covariant_piola(O &&r, const P &U, const Q &J, double, const R &K)
Double covariant Piola map.
Definition: maps.h:102
type
Map type.
Definition: maps.h:38
type
Cell type.
Definition: polyset.h:136
void apply_permutation_mapped(std::span< const std::size_t > perm, std::span< E > data, std::span< const int > emap, std::size_t n=1)
Permutation of mapped data.
Definition: precompute.h:154
space
Sobolev space type.
Definition: sobolev-spaces.h:13
Basix: FEniCS runtime basis evaluation library.
Definition: cell.h:17
FiniteElement< T > create_element(element::family family, cell::type cell, int degree, element::lagrange_variant lvariant, element::dpc_variant dvariant, bool discontinuous, std::vector< int > dof_ordering={})
Definition: finite-element.cpp:193
std::vector< int > tp_dof_ordering(element::family family, cell::type cell, int degree, element::lagrange_variant lvariant, element::dpc_variant dvariant, bool discontinuous)
Definition: finite-element.cpp:399
std::vector< std::vector< FiniteElement< T > > > tp_factors(element::family family, cell::type cell, int degree, element::lagrange_variant lvariant, element::dpc_variant dvariant, bool discontinuous, std::vector< int > dof_ordering)
Definition: finite-element.cpp:349
std::string version()
Definition: finite-element.cpp:1486
FiniteElement< T > create_tp_element(element::family family, cell::type cell, int degree, element::lagrange_variant lvariant, element::dpc_variant dvariant, bool discontinuous)
Definition: finite-element.cpp:330
FiniteElement< T > create_custom_element(cell::type cell_type, const std::vector< std::size_t > &value_shape, impl::mdspan_t< const T, 2 > wcoeffs, const std::array< std::vector< impl::mdspan_t< const T, 2 >>, 4 > &x, const std::array< std::vector< impl::mdspan_t< const T, 4 >>, 4 > &M, int interpolation_nderivs, maps::type map_type, sobolev::space sobolev_space, bool discontinuous, int embedded_subdegree, int embedded_superdegree, polyset::type poly_type)
Definition: finite-element.cpp:609