Actual source code: test21.c

  1: /*
  2:    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
  3:    SLEPc - Scalable Library for Eigenvalue Problem Computations
  4:    Copyright (c) 2002-, Universitat Politecnica de Valencia, Spain

  6:    This file is part of SLEPc.
  7:    SLEPc is distributed under a 2-clause BSD license (see LICENSE).
  8:    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
  9: */

 11: static char help[] = "Test DSGSVD.\n\n";

 13: #include <slepcds.h>

 15: int main(int argc,char **argv)
 16: {
 17:   DS             ds;
 18:   SlepcSC        sc;
 19:   Mat            X;
 20:   Vec            x0;
 21:   PetscReal      sigma,rnorm,cond;
 22:   PetscScalar    *A,*B,*w;
 23:   PetscInt       i,j,k,n=15,m=10,p=10,m1,p1,ld;
 24:   PetscViewer    viewer;
 25:   PetscBool      verbose;

 27:   PetscFunctionBeginUser;
 28:   PetscCall(SlepcInitialize(&argc,&argv,NULL,help));
 29:   PetscCall(PetscOptionsGetInt(NULL,NULL,"-n",&n,NULL));
 30:   PetscCall(PetscOptionsGetInt(NULL,NULL,"-m",&m,NULL));
 31:   PetscCall(PetscOptionsGetInt(NULL,NULL,"-p",&p,NULL));
 32:   PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Solve a Dense System of type GSVD - dimension (%" PetscInt_FMT "+%" PetscInt_FMT ")x%" PetscInt_FMT ".\n",n,p,m));
 33:   PetscCall(PetscOptionsHasName(NULL,NULL,"-verbose",&verbose));

 35:   /* Create DS object */
 36:   PetscCall(DSCreate(PETSC_COMM_WORLD,&ds));
 37:   PetscCall(DSSetType(ds,DSGSVD));
 38:   PetscCall(DSSetFromOptions(ds));
 39:   ld   = PetscMax(PetscMax(p,m),n)+2;  /* test leading dimension larger than n */
 40:   PetscCall(DSAllocate(ds,ld));
 41:   PetscCall(DSSetDimensions(ds,n,0,0));
 42:   PetscCall(DSGSVDSetDimensions(ds,m,p));
 43:   PetscCall(DSGSVDGetDimensions(ds,&m1,&p1));
 44:   PetscCheck(m1==m && p1==p,PETSC_COMM_WORLD,PETSC_ERR_PLIB,"Inconsistent dimension values");

 46:   /* Set up viewer */
 47:   PetscCall(PetscViewerASCIIGetStdout(PETSC_COMM_WORLD,&viewer));
 48:   PetscCall(PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO_DETAIL));
 49:   PetscCall(DSView(ds,viewer));
 50:   PetscCall(PetscViewerPopFormat(viewer));

 52:   k = PetscMin(n,m);
 53:   /* Fill A with a rectangular Toeplitz matrix */
 54:   PetscCall(DSGetArray(ds,DS_MAT_A,&A));
 55:   for (i=0;i<k;i++) A[i+i*ld]=1.0;
 56:   for (j=1;j<3;j++) {
 57:     for (i=0;i<n-j;i++) {
 58:       if ((i+j)<m) A[i+(i+j)*ld]=(PetscScalar)(j+1);
 59:     }
 60:   }
 61:   for (j=1;j<n/2;j++) {
 62:     for (i=0;i<n-j;i++) {
 63:       if ((i+j)<n && i<m) A[(i+j)+i*ld]=-1.0;
 64:     }
 65:   }
 66:   PetscCall(DSRestoreArray(ds,DS_MAT_A,&A));

 68:   k = PetscMin(p,m);
 69:   /* Fill B with a shifted bidiagonal matrix */
 70:   PetscCall(DSGetArray(ds,DS_MAT_B,&B));
 71:   for (i=m-k;i<m;i++) {
 72:     B[i-m+k+i*ld]=2.0-1.0/(PetscScalar)(i+1);
 73:     if (i) B[i-1-m+k+i*ld]=1.0;
 74:   }
 75:   PetscCall(DSRestoreArray(ds,DS_MAT_B,&B));

 77:   PetscCall(DSSetState(ds,DS_STATE_RAW));
 78:   if (verbose) {
 79:     PetscCall(PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_MATLAB));
 80:     PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Initial - - - - - - - - -\n"));
 81:   }
 82:   PetscCall(DSView(ds,viewer));

 84:   /* Condition number */
 85:   PetscCall(DSCond(ds,&cond));
 86:   PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Condition number = %.3f\n",(double)cond));

 88:   /* Solve */
 89:   PetscCall(PetscMalloc1(m,&w));
 90:   PetscCall(DSGetSlepcSC(ds,&sc));
 91:   sc->comparison    = SlepcCompareLargestReal;
 92:   sc->comparisonctx = NULL;
 93:   sc->map           = NULL;
 94:   sc->mapobj        = NULL;
 95:   PetscCall(DSSolve(ds,w,NULL));
 96:   PetscCall(DSSort(ds,w,NULL,NULL,NULL,NULL));
 97:   PetscCall(DSSynchronize(ds,w,NULL));
 98:   if (verbose) {
 99:     PetscCall(PetscPrintf(PETSC_COMM_WORLD,"After solve - - - - - - - - -\n"));
100:     PetscCall(DSView(ds,viewer));
101:   }
102:   /* Print singular values */
103:   PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Computed singular values =\n"));
104:   PetscCall(DSGetDimensions(ds,NULL,NULL,NULL,&k));
105:   for (i=0;i<k;i++) {
106:     sigma = PetscRealPart(w[i]);
107:     PetscCall(PetscViewerASCIIPrintf(viewer,"  %g\n",(double)sigma));
108:   }

110:   /* Singular vectors */
111:   PetscCall(DSVectors(ds,DS_MAT_X,NULL,NULL));  /* all singular vectors */
112:   PetscCall(DSGetMat(ds,DS_MAT_X,&X));
113:   PetscCall(MatCreateVecs(X,NULL,&x0));
114:   PetscCall(MatGetColumnVector(X,x0,0));
115:   PetscCall(VecNorm(x0,NORM_2,&rnorm));
116:   PetscCall(DSRestoreMat(ds,DS_MAT_X,&X));
117:   PetscCall(VecDestroy(&x0));
118:   PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Norm of 1st X vector = %.3f\n",(double)rnorm));

120:   PetscCall(DSGetMat(ds,DS_MAT_U,&X));
121:   PetscCall(MatCreateVecs(X,NULL,&x0));
122:   PetscCall(MatGetColumnVector(X,x0,0));
123:   PetscCall(VecNorm(x0,NORM_2,&rnorm));
124:   PetscCall(DSRestoreMat(ds,DS_MAT_U,&X));
125:   PetscCall(VecDestroy(&x0));
126:   if (PetscAbs(rnorm-1.0)>10*PETSC_MACHINE_EPSILON) PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Warning: the 1st U vector has norm %g\n",(double)rnorm));

128:   PetscCall(DSGetMat(ds,DS_MAT_V,&X));
129:   PetscCall(MatCreateVecs(X,NULL,&x0));
130:   PetscCall(MatGetColumnVector(X,x0,0));
131:   PetscCall(VecNorm(x0,NORM_2,&rnorm));
132:   PetscCall(DSRestoreMat(ds,DS_MAT_V,&X));
133:   PetscCall(VecDestroy(&x0));
134:   if (PetscAbs(rnorm-1.0)>10*PETSC_MACHINE_EPSILON) PetscCall(PetscPrintf(PETSC_COMM_WORLD,"Warning: the 1st V vector has norm %g\n",(double)rnorm));

136:   PetscCall(PetscFree(w));
137:   PetscCall(DSDestroy(&ds));
138:   PetscCall(SlepcFinalize());
139:   return 0;
140: }

142: /*TEST

144:    testset:
145:       output_file: output/test21_1.out
146:       requires: !single
147:       nsize: {{1 2 3}}
148:       filter: grep -v "parallel operation mode" | grep -v " MPI process"
149:       test:
150:          suffix: 1
151:          args: -ds_parallel redundant
152:       test:
153:          suffix: 2
154:          args: -ds_parallel synchronized

156: TEST*/