Actual source code: ex96.c

petsc-3.7.3 2016-08-01
Report Typos and Errors
  2: static char help[] ="Tests sequential and parallel DMCreateMatrix(), MatMatMult() and MatPtAP()\n\
  3:   -Mx <xg>, where <xg> = number of coarse grid points in the x-direction\n\
  4:   -My <yg>, where <yg> = number of coarse grid points in the y-direction\n\
  5:   -Mz <zg>, where <zg> = number of coarse grid points in the z-direction\n\
  6:   -Npx <npx>, where <npx> = number of processors in the x-direction\n\
  7:   -Npy <npy>, where <npy> = number of processors in the y-direction\n\
  8:   -Npz <npz>, where <npz> = number of processors in the z-direction\n\n";

 10: /*
 11:     This test is modified from ~src/ksp/examples/tests/ex19.c.
 12:     Example of usage: mpiexec -n 3 ./ex96 -Mx 10 -My 10 -Mz 10
 13: */

 15: #include <petscdm.h>
 16: #include <petscdmda.h>
 17: #include <../src/mat/impls/aij/seq/aij.h>
 18: #include <../src/mat/impls/aij/mpi/mpiaij.h>

 20: /* User-defined application contexts */
 21: typedef struct {
 22:   PetscInt mx,my,mz;            /* number grid points in x, y and z direction */
 23:   Vec      localX,localF;       /* local vectors with ghost region */
 24:   DM       da;
 25:   Vec      x,b,r;               /* global vectors */
 26:   Mat      J;                   /* Jacobian on grid */
 27: } GridCtx;
 28: typedef struct {
 29:   GridCtx  fine;
 30:   GridCtx  coarse;
 31:   PetscInt ratio;
 32:   Mat      Ii;                  /* interpolation from coarse to fine */
 33: } AppCtx;

 35: #define COARSE_LEVEL 0
 36: #define FINE_LEVEL   1

 38: /*
 39:       Mm_ratio - ration of grid lines between fine and coarse grids.
 40: */
 43: int main(int argc,char **argv)
 44: {
 46:   AppCtx         user;
 47:   PetscInt       Npx=PETSC_DECIDE,Npy=PETSC_DECIDE,Npz=PETSC_DECIDE;
 48:   PetscMPIInt    size,rank;
 49:   PetscInt       m,n,M,N,i,nrows;
 50:   PetscScalar    one = 1.0;
 51:   PetscReal      fill=2.0;
 52:   Mat            A,A_tmp,P,C,C1,C2;
 53:   PetscScalar    *array,none = -1.0,alpha;
 54:   Vec            x,v1,v2,v3,v4;
 55:   PetscReal      norm,norm_tmp,norm_tmp1,tol=100.*PETSC_MACHINE_EPSILON;
 56:   PetscRandom    rdm;
 57:   PetscBool      Test_MatMatMult=PETSC_TRUE,Test_MatPtAP=PETSC_TRUE,Test_3D=PETSC_FALSE,flg;

 59:   PetscInitialize(&argc,&argv,NULL,help);
 60:   PetscOptionsGetReal(NULL,NULL,"-tol",&tol,NULL);

 62:   user.ratio     = 2;
 63:   user.coarse.mx = 2; user.coarse.my = 2; user.coarse.mz = 0;

 65:   PetscOptionsGetInt(NULL,NULL,"-Mx",&user.coarse.mx,NULL);
 66:   PetscOptionsGetInt(NULL,NULL,"-My",&user.coarse.my,NULL);
 67:   PetscOptionsGetInt(NULL,NULL,"-Mz",&user.coarse.mz,NULL);
 68:   PetscOptionsGetInt(NULL,NULL,"-ratio",&user.ratio,NULL);

 70:   if (user.coarse.mz) Test_3D = PETSC_TRUE;

 72:   user.fine.mx = user.ratio*(user.coarse.mx-1)+1;
 73:   user.fine.my = user.ratio*(user.coarse.my-1)+1;
 74:   user.fine.mz = user.ratio*(user.coarse.mz-1)+1;

 76:   MPI_Comm_size(PETSC_COMM_WORLD,&size);
 77:   MPI_Comm_rank(PETSC_COMM_WORLD,&rank);
 78:   PetscOptionsGetInt(NULL,NULL,"-Npx",&Npx,NULL);
 79:   PetscOptionsGetInt(NULL,NULL,"-Npy",&Npy,NULL);
 80:   PetscOptionsGetInt(NULL,NULL,"-Npz",&Npz,NULL);

 82:   /* Set up distributed array for fine grid */
 83:   if (!Test_3D) {
 84:     DMDACreate2d(PETSC_COMM_WORLD, DM_BOUNDARY_NONE, DM_BOUNDARY_NONE,DMDA_STENCIL_STAR,user.fine.mx,
 85:                         user.fine.my,Npx,Npy,1,1,NULL,NULL,&user.fine.da);
 86:   } else {
 87:     DMDACreate3d(PETSC_COMM_WORLD,DM_BOUNDARY_NONE,DM_BOUNDARY_NONE,DM_BOUNDARY_NONE,DMDA_STENCIL_STAR,
 88:                         user.fine.mx,user.fine.my,user.fine.mz,Npx,Npy,Npz,
 89:                         1,1,NULL,NULL,NULL,&user.fine.da);
 90:   }

 92:   /* Test DMCreateMatrix()                                         */
 93:   /*------------------------------------------------------------*/
 94:   DMSetMatType(user.fine.da,MATAIJ);
 95:   DMCreateMatrix(user.fine.da,&A);
 96:   DMSetMatType(user.fine.da,MATBAIJ);
 97:   DMCreateMatrix(user.fine.da,&C);

 99:   MatConvert(C,MATAIJ,MAT_INITIAL_MATRIX,&A_tmp); /* not work for mpisbaij matrix! */
100:   MatEqual(A,A_tmp,&flg);
101:   if (!flg) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_NOTSAMETYPE,"A != C");
102:   MatDestroy(&C);
103:   MatDestroy(&A_tmp);

105:   /*------------------------------------------------------------*/

107:   MatGetLocalSize(A,&m,&n);
108:   MatGetSize(A,&M,&N);
109:   /* set val=one to A */
110:   if (size == 1) {
111:     const PetscInt *ia,*ja;
112:     MatGetRowIJ(A,0,PETSC_FALSE,PETSC_FALSE,&nrows,&ia,&ja,&flg);
113:     if (flg) {
114:       MatSeqAIJGetArray(A,&array);
115:       for (i=0; i<ia[nrows]; i++) array[i] = one;
116:       MatSeqAIJRestoreArray(A,&array);
117:     }
118:     MatRestoreRowIJ(A,0,PETSC_FALSE,PETSC_FALSE,&nrows,&ia,&ja,&flg);
119:   } else {
120:     Mat_MPIAIJ *aij = (Mat_MPIAIJ*)A->data;
121:     Mat_SeqAIJ *a   = (Mat_SeqAIJ*)(aij->A)->data, *b=(Mat_SeqAIJ*)(aij->B)->data;
122:     /* A_part */
123:     for (i=0; i<a->i[m]; i++) a->a[i] = one;
124:     /* B_part */
125:     for (i=0; i<b->i[m]; i++) b->a[i] = one;

127:   }
128:   /* MatView(A, PETSC_VIEWER_STDOUT_WORLD); */

130:   /* Set up distributed array for coarse grid */
131:   if (!Test_3D) {
132:     DMDACreate2d(PETSC_COMM_WORLD, DM_BOUNDARY_NONE, DM_BOUNDARY_NONE,DMDA_STENCIL_STAR,user.coarse.mx,
133:                         user.coarse.my,Npx,Npy,1,1,NULL,NULL,&user.coarse.da);
134:   } else {
135:     DMDACreate3d(PETSC_COMM_WORLD,DM_BOUNDARY_NONE,DM_BOUNDARY_NONE,DM_BOUNDARY_NONE,DMDA_STENCIL_STAR,
136:                         user.coarse.mx,user.coarse.my,user.coarse.mz,Npx,Npy,Npz,
137:                         1,1,NULL,NULL,NULL,&user.coarse.da);
138:   }

140:   /* Create interpolation between the levels */
141:   DMCreateInterpolation(user.coarse.da,user.fine.da,&P,NULL);

143:   MatGetLocalSize(P,&m,&n);
144:   MatGetSize(P,&M,&N);

146:   /* Create vectors v1 and v2 that are compatible with A */
147:   VecCreate(PETSC_COMM_WORLD,&v1);
148:   MatGetLocalSize(A,&m,NULL);
149:   VecSetSizes(v1,m,PETSC_DECIDE);
150:   VecSetFromOptions(v1);
151:   VecDuplicate(v1,&v2);
152:   PetscRandomCreate(PETSC_COMM_WORLD,&rdm);
153:   PetscRandomSetFromOptions(rdm);

155:   /* Test MatMatMult(): C = A*P */
156:   /*----------------------------*/
157:   if (Test_MatMatMult) {
158:     MatDuplicate(A,MAT_COPY_VALUES,&A_tmp);
159:     MatMatMult(A_tmp,P,MAT_INITIAL_MATRIX,fill,&C);

161:     /* Test MAT_REUSE_MATRIX - reuse symbolic C */
162:     alpha=1.0;
163:     for (i=0; i<2; i++) {
164:       alpha -=0.1;
165:       MatScale(A_tmp,alpha);
166:       MatMatMult(A_tmp,P,MAT_REUSE_MATRIX,fill,&C);
167:     }

169:     /* Test MatDuplicate()        */
170:     /*----------------------------*/
171:     MatDuplicate(C,MAT_COPY_VALUES,&C1);
172:     MatDuplicate(C1,MAT_COPY_VALUES,&C2);
173:     MatDestroy(&C1);
174:     MatDestroy(&C2);

176:     /* Create vector x that is compatible with P */
177:     VecCreate(PETSC_COMM_WORLD,&x);
178:     MatGetLocalSize(P,NULL,&n);
179:     VecSetSizes(x,n,PETSC_DECIDE);
180:     VecSetFromOptions(x);

182:     norm = 0.0;
183:     for (i=0; i<10; i++) {
184:       VecSetRandom(x,rdm);
185:       MatMult(P,x,v1);
186:       MatMult(A_tmp,v1,v2); /* v2 = A*P*x */
187:       MatMult(C,x,v1);  /* v1 = C*x   */
188:       VecAXPY(v1,none,v2);
189:       VecNorm(v1,NORM_1,&norm_tmp);
190:       VecNorm(v2,NORM_1,&norm_tmp1);
191:       norm_tmp /= norm_tmp1;
192:       if (norm_tmp > norm) norm = norm_tmp;
193:     }
194:     if (norm >= tol && !rank) {
195:       PetscPrintf(PETSC_COMM_SELF,"Error: MatMatMult(), |v1 - v2|/|v2|: %g\n",(double)norm);
196:     }

198:     VecDestroy(&x);
199:     MatDestroy(&C);
200:     MatDestroy(&A_tmp);
201:   }

203:   /* Test P^T * A * P - MatPtAP() */
204:   /*------------------------------*/
205:   if (Test_MatPtAP) {
206:     MatPtAP(A,P,MAT_INITIAL_MATRIX,fill,&C);
207:     MatGetLocalSize(C,&m,&n);

209:     /* Test MAT_REUSE_MATRIX - reuse symbolic C */
210:     alpha=1.0;
211:     for (i=0; i<1; i++) {
212:       alpha -=0.1;
213:       MatScale(A,alpha);
214:       MatPtAP(A,P,MAT_REUSE_MATRIX,fill,&C);
215:     }

217:     /* Test MatDuplicate()        */
218:     /*----------------------------*/
219:     MatDuplicate(C,MAT_COPY_VALUES,&C1);
220:     MatDuplicate(C1,MAT_COPY_VALUES,&C2);
221:     MatDestroy(&C1);
222:     MatDestroy(&C2);

224:     /* Create vector x that is compatible with P */
225:     VecCreate(PETSC_COMM_WORLD,&x);
226:     MatGetLocalSize(P,&m,&n);
227:     VecSetSizes(x,n,PETSC_DECIDE);
228:     VecSetFromOptions(x);

230:     VecCreate(PETSC_COMM_WORLD,&v3);
231:     VecSetSizes(v3,n,PETSC_DECIDE);
232:     VecSetFromOptions(v3);
233:     VecDuplicate(v3,&v4);

235:     norm = 0.0;
236:     for (i=0; i<10; i++) {
237:       VecSetRandom(x,rdm);
238:       MatMult(P,x,v1);
239:       MatMult(A,v1,v2);  /* v2 = A*P*x */

241:       MatMultTranspose(P,v2,v3); /* v3 = Pt*A*P*x */
242:       MatMult(C,x,v4);           /* v3 = C*x   */
243:       VecAXPY(v4,none,v3);
244:       VecNorm(v4,NORM_1,&norm_tmp);
245:       VecNorm(v3,NORM_1,&norm_tmp1);

247:       norm_tmp /= norm_tmp1;
248:       if (norm_tmp > norm) norm = norm_tmp;
249:     }
250:     if (norm >= tol && !rank) {
251:       PetscPrintf(PETSC_COMM_SELF,"Error: MatPtAP(), |v3 - v4|/|v3|: %g\n",(double)norm);
252:     }
253:     MatDestroy(&C);
254:     VecDestroy(&v3);
255:     VecDestroy(&v4);
256:     VecDestroy(&x);
257:   }

259:   /* Clean up */
260:   MatDestroy(&A);
261:   PetscRandomDestroy(&rdm);
262:   VecDestroy(&v1);
263:   VecDestroy(&v2);
264:   DMDestroy(&user.fine.da);
265:   DMDestroy(&user.coarse.da);
266:   MatDestroy(&P);
267:   PetscFinalize();
268:   return 0;
269: }