Actual source code: ex11.c

  1: /*$Id: ex11.c,v 1.34 2001/03/23 23:23:55 balay Exp $*/

  3: static char help[] = "Solves a linear system in parallel with SLES.nn";

  5: /*T
  6:    Concepts: SLES^solving a Helmholtz equation
  7:    Concepts: complex numbers;
  8:    Concepts: Helmholtz equation
  9:    Processors: n
 10: T*/

 12: /*
 13:    Description: Solves a complex linear system in parallel with SLES.

 15:    The model problem:
 16:       Solve Helmholtz equation on the unit square: (0,1) x (0,1)
 17:           -delta u - sigma1*u + i*sigma2*u = f, 
 18:            where delta = Laplace operator
 19:       Dirichlet b.c.'s on all sides
 20:       Use the 2-D, five-point finite difference stencil.

 22:    Compiling the code:
 23:       This code uses the complex numbers version of PETSc, so one of the
 24:       following values of BOPT must be used for compiling the PETSc libraries
 25:       and this example:
 26:          BOPT=g_complex   - debugging version
 27:          BOPT=O_complex   - optimized version
 28:          BOPT=Opg_complex - profiling version
 29: */

 31: /* 
 32:   Include "petscsles.h" so that we can use SLES solvers.  Note that this file
 33:   automatically includes:
 34:      petsc.h       - base PETSc routines   petscvec.h - vectors
 35:      petscsys.h    - system routines       petscmat.h - matrices
 36:      petscis.h     - index sets            petscksp.h - Krylov subspace methods
 37:      petscviewer.h - viewers               petscpc.h  - preconditioners
 38: */
 39:  #include petscsles.h

 41: int main(int argc,char **args)
 42: {
 43:   Vec         x,b,u;      /* approx solution, RHS, exact solution */
 44:   Mat         A;            /* linear system matrix */
 45:   SLES        sles;         /* linear solver context */
 46:   double      norm;         /* norm of solution error */
 47:   int         dim,i,j,I,J,Istart,Iend,ierr,n = 6,its,use_random;
 48:   Scalar      v,none = -1.0,sigma2,pfive = 0.5,*xa;
 49:   PetscRandom rctx;
 50:   double      h2,sigma1 = 100.0;
 51:   PetscTruth  flg;

 53:   PetscInitialize(&argc,&args,(char *)0,help);
 54: #if !defined(PETSC_USE_COMPLEX)
 55:   SETERRQ(1,"This example requires complex numbers");
 56: #endif

 58:   PetscOptionsGetDouble(PETSC_NULL,"-sigma1",&sigma1,PETSC_NULL);
 59:   PetscOptionsGetInt(PETSC_NULL,"-n",&n,PETSC_NULL);
 60:   dim = n*n;

 62:   /* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
 63:          Compute the matrix and right-hand-side vector that define
 64:          the linear system, Ax = b.
 65:      - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */
 66:   /* 
 67:      Create parallel matrix, specifying only its global dimensions.
 68:      When using MatCreate(), the matrix format can be specified at
 69:      runtime. Also, the parallel partitioning of the matrix is
 70:      determined by PETSc at runtime.
 71:   */
 72:   MatCreate(PETSC_COMM_WORLD,PETSC_DECIDE,PETSC_DECIDE,dim,dim,&A);
 73:   MatSetFromOptions(A);

 75:   /* 
 76:      Currently, all PETSc parallel matrix formats are partitioned by
 77:      contiguous chunks of rows across the processors.  Determine which
 78:      rows of the matrix are locally owned. 
 79:   */
 80:   MatGetOwnershipRange(A,&Istart,&Iend);

 82:   /* 
 83:      Set matrix elements in parallel.
 84:       - Each processor needs to insert only elements that it owns
 85:         locally (but any non-local elements will be sent to the
 86:         appropriate processor during matrix assembly). 
 87:       - Always specify global rows and columns of matrix entries.
 88:   */

 90:   PetscOptionsHasName(PETSC_NULL,"-norandom",&flg);
 91:   if (flg) use_random = 0;
 92:   else     use_random = 1;
 93:   if (use_random) {
 94:     PetscRandomCreate(PETSC_COMM_WORLD,RANDOM_DEFAULT_IMAGINARY,&rctx);
 95:   } else {
 96:     sigma2 = 10.0*PETSC_i;
 97:   }
 98:   h2 = 1.0/((n+1)*(n+1));
 99:   for (I=Istart; I<Iend; I++) {
100:     v = -1.0; i = I/n; j = I - i*n;
101:     if (i>0) {
102:       J = I-n; MatSetValues(A,1,&I,1,&J,&v,ADD_VALUES);}
103:     if (i<n-1) {
104:       J = I+n; MatSetValues(A,1,&I,1,&J,&v,ADD_VALUES);}
105:     if (j>0) {
106:       J = I-1; MatSetValues(A,1,&I,1,&J,&v,ADD_VALUES);}
107:     if (j<n-1) {
108:       J = I+1; MatSetValues(A,1,&I,1,&J,&v,ADD_VALUES);}
109:     if (use_random) {PetscRandomGetValue(rctx,&sigma2);}
110:     v = 4.0 - sigma1*h2 + sigma2*h2;
111:     MatSetValues(A,1,&I,1,&I,&v,ADD_VALUES);
112:   }
113:   if (use_random) {PetscRandomDestroy(rctx);}

115:   /* 
116:      Assemble matrix, using the 2-step process:
117:        MatAssemblyBegin(), MatAssemblyEnd()
118:      Computations can be done while messages are in transition
119:      by placing code between these two statements.
120:   */
121:   MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);
122:   MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);

124:   /* 
125:      Create parallel vectors.
126:       - When using VecCreate() and VecSetFromOptions(), we specify only the vector's global
127:         dimension; the parallel partitioning is determined at runtime. 
128:       - Note: We form 1 vector from scratch and then duplicate as needed.
129:   */
130:   VecCreate(PETSC_COMM_WORLD,PETSC_DECIDE,dim,&u);
131:   VecSetFromOptions(u);
132:   VecDuplicate(u,&b);
133:   VecDuplicate(b,&x);

135:   /* 
136:      Set exact solution; then compute right-hand-side vector.
137:   */
138: 
139:   if (use_random) {
140:     PetscRandomCreate(PETSC_COMM_WORLD,RANDOM_DEFAULT,&rctx);
141:     VecSetRandom(rctx,u);
142:   } else {
143:     VecSet(&pfive,u);
144:   }
145:   MatMult(A,u,b);

147:   /* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
148:                 Create the linear solver and set various options
149:      - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */

151:   /* 
152:      Create linear solver context
153:   */
154:   SLESCreate(PETSC_COMM_WORLD,&sles);

156:   /* 
157:      Set operators. Here the matrix that defines the linear system
158:      also serves as the preconditioning matrix.
159:   */
160:   SLESSetOperators(sles,A,A,DIFFERENT_NONZERO_PATTERN);

162:   /* 
163:     Set runtime options, e.g.,
164:         -ksp_type <type> -pc_type <type> -ksp_monitor -ksp_rtol <rtol>
165:   */
166:   SLESSetFromOptions(sles);

168:   /* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
169:                       Solve the linear system
170:      - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */

172:   SLESSolve(sles,b,x,&its);

174:   /* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
175:                       Check solution and clean up
176:      - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */

178:   /*
179:       Print the first 3 entries of x; this demonstrates extraction of the
180:       real and imaginary components of the complex vector, x.
181:   */
182:   PetscOptionsHasName(PETSC_NULL,"-print_x3",&flg);
183:   if (flg) {
184:     VecGetArray(x,&xa);
185:     PetscPrintf(PETSC_COMM_WORLD,"The first three entries of x are:n");
186:     for (i=0; i<3; i++){
187:       PetscPrintf(PETSC_COMM_WORLD,"x[%d] = %g + %g in",i,PetscRealPart(xa[i]),PetscImaginaryPart(xa[i]));
188:   }
189:     VecRestoreArray(x,&xa);
190:   }

192:   /* 
193:      Check the error
194:   */
195:   VecAXPY(&none,u,x);
196:   VecNorm(x,NORM_2,&norm);
197:   PetscPrintf(PETSC_COMM_WORLD,"Norm of error %A iterations %dn",norm,its);

199:   /* 
200:      Free work space.  All PETSc objects should be destroyed when they
201:      are no longer needed.
202:   */
203:   SLESDestroy(sles);
204:   if (use_random) {PetscRandomDestroy(rctx);}
205:   VecDestroy(u); VecDestroy(x);
206:   VecDestroy(b); MatDestroy(A);
207:   PetscFinalize();
208:   return 0;
209: }