MFEM  v4.6.0
Finite element discretization library
ex1.cpp
Go to the documentation of this file.
1 // MFEM Example 1
2 // GINKGO Modification
3 //
4 // Compile with: make ex1
5 //
6 // Sample runs: ex1 -m ../data/square-disc.mesh
7 // ex1 -m ../data/star.mesh
8 // ex1 -m ../data/star-mixed.mesh
9 // ex1 -m ../data/escher.mesh
10 // ex1 -m ../data/fichera.mesh
11 // ex1 -m ../data/fichera-mixed.mesh
12 // ex1 -m ../data/toroid-wedge.mesh
13 // ex1 -m ../data/square-disc-p2.vtk -o 2
14 // ex1 -m ../data/square-disc-p3.mesh -o 3
15 // ex1 -m ../data/square-disc-nurbs.mesh -o -1
16 // ex1 -m ../data/star-mixed-p2.mesh -o 2
17 // ex1 -m ../data/disc-nurbs.mesh -o -1
18 // ex1 -m ../data/pipe-nurbs.mesh -o -1
19 // ex1 -m ../data/fichera-mixed-p2.mesh -o 2
20 // ex1 -m ../data/star-surf.mesh
21 // ex1 -m ../data/square-disc-surf.mesh
22 // ex1 -m ../data/inline-segment.mesh
23 // ex1 -m ../data/amr-quad.mesh
24 // ex1 -m ../data/amr-hex.mesh
25 // ex1 -m ../data/fichera-amr.mesh
26 // ex1 -m ../data/mobius-strip.mesh
27 // ex1 -m ../data/mobius-strip.mesh -o -1 -sc
28 //
29 // Device sample runs:
30 // ex1 -pa -d cuda
31 // ex1 -pa -d raja-cuda
32 // ex1 -pa -d occa-cuda
33 // ex1 -pa -d raja-omp
34 // ex1 -pa -d occa-omp
35 // ex1 -m ../data/beam-hex.mesh -pa -d cuda
36 //
37 // Description: This example code demonstrates the use of MFEM to define a
38 // simple finite element discretization of the Laplace problem
39 // -Delta u = 1 with homogeneous Dirichlet boundary conditions.
40 // Specifically, we discretize using a FE space of the specified
41 // order, or if order < 1 using an isoparametric/isogeometric
42 // space (i.e. quadratic for quadratic curvilinear mesh, NURBS for
43 // NURBS mesh, etc.)
44 //
45 // The example highlights the use of mesh refinement, finite
46 // element grid functions, as well as linear and bilinear forms
47 // corresponding to the left-hand side and right-hand side of the
48 // discrete linear system. We also cover the explicit elimination
49 // of essential boundary conditions, static condensation, and the
50 // optional connection to the GLVis tool for visualization.
51 
52 #include "mfem.hpp"
53 #include <fstream>
54 #include <iostream>
55 
56 #ifndef MFEM_USE_GINKGO
57 #error This example requires that MFEM is built with MFEM_USE_GINKGO=YES
58 #endif
59 
60 using namespace std;
61 using namespace mfem;
62 
63 int main(int argc, char *argv[])
64 {
65  // 1. Parse command-line options.
66  const char *mesh_file = "../../data/star.mesh";
67  int order = 1;
68  bool static_cond = false;
69  bool pa = false;
70  const char *device_config = "cpu";
71  bool visualization = true;
72  int solver_config = 0;
73  int print_lvl = 1;
74 
75  OptionsParser args(argc, argv);
76  args.AddOption(&mesh_file, "-m", "--mesh",
77  "Mesh file to use.");
78  args.AddOption(&order, "-o", "--order",
79  "Finite element order (polynomial degree) or -1 for"
80  " isoparametric space.");
81  args.AddOption(&static_cond, "-sc", "--static-condensation", "-no-sc",
82  "--no-static-condensation", "Enable static condensation.");
83  args.AddOption(&pa, "-pa", "--partial-assembly", "-no-pa",
84  "--no-partial-assembly", "Enable Partial Assembly.");
85  args.AddOption(&device_config, "-d", "--device",
86  "Device configuration string, see Device::Configure().");
87  args.AddOption(&visualization, "-vis", "--visualization", "-no-vis",
88  "--no-visualization",
89  "Enable or disable GLVis visualization.");
90  args.AddOption(&solver_config, "-s", "--solver-config",
91  "Solver and preconditioner combination: \n\t"
92  " 0 - Ginkgo solver and Ginkgo preconditioner, \n\t"
93  " 1 - Ginkgo solver and MFEM preconditioner, \n\t"
94  " 2 - MFEM solver and Ginkgo preconditioner, \n\t"
95  " 3 - MFEM solver and MFEM preconditioner.");
96  args.AddOption(&print_lvl, "-pl", "--print-level",
97  "Print level for iterative solver (1 prints every iteration).");
98  args.Parse();
99  if (!args.Good())
100  {
101  args.PrintUsage(cout);
102  return 1;
103  }
104  args.PrintOptions(cout);
105 
106  // 2. Enable hardware devices such as GPUs, and programming models such as
107  // CUDA, OCCA, RAJA and OpenMP based on command line options.
108  Device device(device_config);
109  device.Print();
110 
111  // 3. Read the mesh from the given mesh file. We can handle triangular,
112  // quadrilateral, tetrahedral, hexahedral, surface and volume meshes with
113  // the same code.
114  Mesh *mesh = new Mesh(mesh_file, 1, 1);
115  int dim = mesh->Dimension();
116 
117  // 4. Refine the mesh to increase the resolution. In this example we do
118  // 'ref_levels' of uniform refinement. We choose 'ref_levels' to be the
119  // largest number that gives a final mesh with no more than 50,000
120  // elements.
121  {
122  int ref_levels =
123  (int)floor(log(50000./mesh->GetNE())/log(2.)/dim);
124  for (int l = 0; l < ref_levels; l++)
125  {
126  mesh->UniformRefinement();
127  }
128  }
129 
130  // 5. Define a finite element space on the mesh. Here we use continuous
131  // Lagrange finite elements of the specified order. If order < 1, we
132  // instead use an isoparametric/isogeometric space.
134  if (order > 0)
135  {
136  fec = new H1_FECollection(order, dim);
137  }
138  else if (mesh->GetNodes())
139  {
140  fec = mesh->GetNodes()->OwnFEC();
141  cout << "Using isoparametric FEs: " << fec->Name() << endl;
142  }
143  else
144  {
145  fec = new H1_FECollection(order = 1, dim);
146  }
147  FiniteElementSpace *fespace = new FiniteElementSpace(mesh, fec);
148  cout << "Number of finite element unknowns: "
149  << fespace->GetTrueVSize() << endl;
150 
151  // 6. Determine the list of true (i.e. conforming) essential boundary dofs.
152  // In this example, the boundary conditions are defined by marking all
153  // the boundary attributes from the mesh as essential (Dirichlet) and
154  // converting them to a list of true dofs.
155  Array<int> ess_tdof_list;
156  if (mesh->bdr_attributes.Size())
157  {
158  Array<int> ess_bdr(mesh->bdr_attributes.Max());
159  ess_bdr = 1;
160  fespace->GetEssentialTrueDofs(ess_bdr, ess_tdof_list);
161  }
162 
163  // 7. Set up the linear form b(.) which corresponds to the right-hand side of
164  // the FEM linear system, which in this case is (1,phi_i) where phi_i are
165  // the basis functions in the finite element fespace.
166  LinearForm *b = new LinearForm(fespace);
167  ConstantCoefficient one(1.0);
168  b->AddDomainIntegrator(new DomainLFIntegrator(one));
169  b->Assemble();
170 
171  // 8. Define the solution vector x as a finite element grid function
172  // corresponding to fespace. Initialize x with initial guess of zero,
173  // which satisfies the boundary conditions.
174  GridFunction x(fespace);
175  x = 0.0;
176 
177  // 9. Set up the bilinear form a(.,.) on the finite element space
178  // corresponding to the Laplacian operator -Delta, by adding the Diffusion
179  // domain integrator.
180  BilinearForm *a = new BilinearForm(fespace);
181  if (pa) { a->SetAssemblyLevel(AssemblyLevel::PARTIAL); }
182  a->AddDomainIntegrator(new DiffusionIntegrator(one));
183 
184  // 10. Assemble the bilinear form and the corresponding linear system,
185  // applying any necessary transformations such as: eliminating boundary
186  // conditions, applying conforming constraints for non-conforming AMR,
187  // static condensation, etc.
188  if (static_cond) { a->EnableStaticCondensation(); }
189  a->Assemble();
190 
191  OperatorPtr A;
192  Vector B, X;
193  a->FormLinearSystem(ess_tdof_list, x, *b, A, X, B);
194 
195  cout << "Size of linear system: " << A->Height() << endl;
196 
197  // 11. Solve the linear system A X = B.
198  if (!pa)
199  {
200  switch (solver_config)
201  {
202  // Solve the linear system with CG + IC from Ginkgo
203  case 0:
204  {
205  cout << "Using Ginkgo solver + preconditioner...\n";
206  Ginkgo::GinkgoExecutor exec(device);
207  Ginkgo::IcPreconditioner ginkgo_precond(exec, "paric", 30);
208  Ginkgo::CGSolver ginkgo_solver(exec, ginkgo_precond);
209  ginkgo_solver.SetPrintLevel(print_lvl);
210  ginkgo_solver.SetRelTol(1e-12);
211  ginkgo_solver.SetAbsTol(0.0);
212  ginkgo_solver.SetMaxIter(400);
213  ginkgo_solver.SetOperator(*(A.Ptr()));
214  ginkgo_solver.Mult(B, X);
215  break;
216  }
217 
218  // Solve the linear system with CG from Ginkgo + MFEM preconditioner
219  case 1:
220  {
221  cout << "Using Ginkgo solver + MFEM preconditioner...\n";
222  Ginkgo::GinkgoExecutor exec(device);
223  //Create MFEM preconditioner and wrap it for Ginkgo's use.
224  DSmoother M((SparseMatrix&)(*A));
225  Ginkgo::MFEMPreconditioner gko_M(exec, M);
226  Ginkgo::CGSolver ginkgo_solver(exec, gko_M);
227  ginkgo_solver.SetPrintLevel(print_lvl);
228  ginkgo_solver.SetRelTol(1e-12);
229  ginkgo_solver.SetAbsTol(0.0);
230  ginkgo_solver.SetMaxIter(400);
231  ginkgo_solver.SetOperator(*(A.Ptr()));
232  ginkgo_solver.Mult(B, X);
233  break;
234  }
235 
236  // Ginkgo IC preconditioner + MFEM CG solver
237  case 2:
238  {
239  cout << "Using MFEM solver + Ginkgo preconditioner...\n";
240  Ginkgo::GinkgoExecutor exec(device);
241  Ginkgo::IcPreconditioner M(exec, "paric", 30);
242  M.SetOperator(*(A.Ptr())); // Generate the preconditioner for the matrix A.
243  PCG(*A, M, B, X, print_lvl, 400, 1e-12, 0.0);
244  break;
245  }
246 
247  // MFEM solver + MFEM preconditioner
248  case 3:
249  {
250  cout << "Using MFEM solver + MFEM preconditioner...\n";
251  // Use a simple Jacobi preconditioner with PCG.
252  DSmoother M((SparseMatrix&)(*A));
253  PCG(*A, M, B, X, print_lvl, 400, 1e-12, 0.0);
254  break;
255  }
256  } // End switch on solver_config
257  }
258  // Partial assembly mode. Cannot use Ginkgo preconditioners, but can use Ginkgo
259  // solvers.
260  else
261  {
262  if (UsesTensorBasis(*fespace))
263  {
264  // Use Jacobi preconditioning in partial assembly mode.
265  OperatorJacobiSmoother M(*a, ess_tdof_list);
266  switch (solver_config)
267  {
268  // No Ginkgo preconditioners work with matrix-free; error
269  case 0:
270  {
271  cout << "Using Ginkgo solver + preconditioner...\n";
272  MFEM_ABORT("Cannot use Ginkgo preconditioner in partial assembly mode.\n"
273  " Try -s 1 to test Ginkgo solver with an MFEM preconditioner.");
274  break;
275  }
276 
277  // Use Ginkgo solver with MFEM preconditioner
278  case 1:
279  {
280  cout << "Using Ginkgo solver + MFEM preconditioner...\n";
281  Ginkgo::GinkgoExecutor exec(device);
282  // Wrap MFEM preconditioner for Ginkgo's use.
283  Ginkgo::MFEMPreconditioner gko_M(exec, M);
284  Ginkgo::CGSolver ginkgo_solver(exec, gko_M);
285  ginkgo_solver.SetPrintLevel(print_lvl);
286  ginkgo_solver.SetRelTol(1e-12);
287  ginkgo_solver.SetAbsTol(0.0);
288  ginkgo_solver.SetMaxIter(400);
289  ginkgo_solver.SetOperator(*(A.Ptr()));
290  ginkgo_solver.Mult(B, X);
291  break;
292  }
293 
294  // No Ginkgo preconditioners work with matrix-free; error
295  case 2:
296  {
297  cout << "Using MFEM solver + Ginkgo preconditioner...\n";
298  MFEM_ABORT("Cannot use Ginkgo preconditioner in partial assembly mode.\n"
299  " Try -s 1 to test Ginkgo solver with an MFEM preconditioner.");
300  break;
301  }
302 
303  // Use MFEM solver and preconditioner
304  case 3:
305  {
306  cout << "Using MFEM solver + MFEM preconditioner...\n";
307  PCG(*A, M, B, X, print_lvl, 400, 1e-12, 0.0);
308  break;
309  }
310  } // End switch on solver_config
311  }
312  else // CG with no preconditioning
313  {
314  cout << "Using MFEM solver + no preconditioner...\n";
315  CG(*A, B, X, print_lvl, 400, 1e-12, 0.0);
316  }
317  }
318 
319  // 12. Recover the solution as a finite element grid function.
320  a->RecoverFEMSolution(X, *b, x);
321 
322  // 13. Save the refined mesh and the solution. This output can be viewed later
323  // using GLVis: "glvis -m refined.mesh -g sol.gf".
324  ofstream mesh_ofs("refined.mesh");
325  mesh_ofs.precision(8);
326  mesh->Print(mesh_ofs);
327  ofstream sol_ofs("sol.gf");
328  sol_ofs.precision(8);
329  x.Save(sol_ofs);
330 
331  // 14. Send the solution by socket to a GLVis server.
332  if (visualization)
333  {
334  char vishost[] = "localhost";
335  int visport = 19916;
336  socketstream sol_sock(vishost, visport);
337  sol_sock.precision(8);
338  sol_sock << "solution\n" << *mesh << x << flush;
339  }
340 
341  // 15. Free the used memory.
342  delete a;
343  delete b;
344  delete fespace;
345  if (order > 0) { delete fec; }
346  delete mesh;
347 
348  return 0;
349 }
Class for domain integration L(v) := (f, v)
Definition: lininteg.hpp:108
int visport
void SetRelTol(double rtol)
Definition: ginkgo.hpp:793
Class for grid function - Vector with associated FE space.
Definition: gridfunc.hpp:30
Data type for scaled Jacobi-type smoother of sparse matrix.
virtual void SetOperator(const Operator &op)
Definition: ginkgo.cpp:415
A coefficient that is constant across space and time.
Definition: coefficient.hpp:84
void PrintOptions(std::ostream &out) const
Print the options.
Definition: optparser.cpp:331
int Dimension() const
Dimension of the reference space used within the elements.
Definition: mesh.hpp:1020
void PrintUsage(std::ostream &out) const
Print the usage message.
Definition: optparser.cpp:462
Pointer to an Operator of a specified type.
Definition: handle.hpp:33
T Max() const
Find the maximal element in the array, using the comparison operator < for class T.
Definition: array.cpp:68
virtual void GetEssentialTrueDofs(const Array< int > &bdr_attr_is_ess, Array< int > &ess_tdof_list, int component=-1)
Get a list of essential true dofs, ess_tdof_list, corresponding to the boundary attributes marked in ...
Definition: fespace.cpp:587
void Print(std::ostream &out=mfem::out)
Print the configuration of the MFEM virtual device object.
Definition: device.cpp:279
bool Good() const
Return true if the command line options were parsed successfully.
Definition: optparser.hpp:159
STL namespace.
void SetMaxIter(int max_it)
Definition: ginkgo.hpp:817
bool UsesTensorBasis(const FiniteElementSpace &fes)
Return true if the mesh contains only one topology and the elements are tensor elements.
Definition: fespace.hpp:1306
int main(int argc, char *argv[])
Definition: ex1.cpp:74
virtual void Mult(const Vector &x, Vector &y) const
Definition: ginkgo.cpp:287
void Parse()
Parse the command-line options. Note that this function expects all the options provided through the ...
Definition: optparser.cpp:151
char vishost[]
Data type sparse matrix.
Definition: sparsemat.hpp:50
Jacobi smoothing for a given bilinear form (no matrix necessary).
Definition: solvers.hpp:302
double b
Definition: lissajous.cpp:42
void UniformRefinement(int i, const DSTable &, int *, int *, int *)
Definition: mesh.cpp:10232
virtual const char * Name() const
Definition: fe_coll.hpp:80
void CG(const Operator &A, const Vector &b, Vector &x, int print_iter, int max_num_iter, double RTOLERANCE, double ATOLERANCE)
Conjugate gradient method. (tolerances are squared)
Definition: solvers.cpp:898
void PCG(const Operator &A, Solver &B, const Vector &b, Vector &x, int print_iter, int max_num_iter, double RTOLERANCE, double ATOLERANCE)
Preconditioned conjugate gradient method. (tolerances are squared)
Definition: solvers.cpp:913
void SetPrintLevel(int print_lvl)
Definition: ginkgo.hpp:626
virtual int GetTrueVSize() const
Return the number of vector true (conforming) dofs.
Definition: fespace.hpp:712
Array< int > bdr_attributes
A list of all unique boundary attributes used by the Mesh.
Definition: mesh.hpp:275
Class FiniteElementSpace - responsible for providing FEM view of the mesh, mainly managing the set of...
Definition: fespace.hpp:219
Collection of finite elements from the same family in multiple dimensions. This class is used to matc...
Definition: fe_coll.hpp:26
void AddOption(bool *var, const char *enable_short_name, const char *enable_long_name, const char *disable_short_name, const char *disable_long_name, const char *description, bool required=false)
Add a boolean option and set &#39;var&#39; to receive the value. Enable/disable tags are used to set the bool...
Definition: optparser.hpp:82
int Height() const
Get the height (size of output) of the Operator. Synonym with NumRows().
Definition: operator.hpp:66
int GetNE() const
Returns number of elements.
Definition: mesh.hpp:1086
double a
Definition: lissajous.cpp:41
A "square matrix" operator for the associated FE space and BLFIntegrators The sum of all the BLFInteg...
int dim
Definition: ex24.cpp:53
void SetAbsTol(double atol)
Definition: ginkgo.hpp:805
Operator * Ptr() const
Access the underlying Operator pointer.
Definition: handle.hpp:87
int Size() const
Return the logical size of the array.
Definition: array.hpp:141
Vector data type.
Definition: vector.hpp:58
virtual void Print(std::ostream &os=mfem::out) const
Definition: mesh.hpp:2011
Arbitrary order H1-conforming (continuous) finite elements.
Definition: fe_coll.hpp:259
Vector with associated FE space and LinearFormIntegrators.
Definition: linearform.hpp:24
void GetNodes(Vector &node_coord) const
Definition: mesh.cpp:8302
virtual void SetOperator(const Operator &op)
Definition: ginkgo.cpp:884
virtual void Save(std::ostream &out) const
Save the GridFunction to an output stream.
Definition: gridfunc.cpp:3696
The MFEM Device class abstracts hardware devices such as GPUs, as well as programming models such as ...
Definition: device.hpp:121