Actual source code: blockmat.c
petsc-3.6.1 2015-08-06
2: /*
3: This provides a matrix that consists of Mats
4: */
6: #include <petsc/private/matimpl.h> /*I "petscmat.h" I*/
7: #include <../src/mat/impls/baij/seq/baij.h> /* use the common AIJ data-structure */
9: typedef struct {
10: SEQAIJHEADER(Mat);
11: SEQBAIJHEADER;
12: Mat *diags;
14: Vec left,right,middle,workb; /* dummy vectors to perform local parts of product */
15: } Mat_BlockMat;
17: extern PetscErrorCode MatBlockMatSetPreallocation(Mat,PetscInt,PetscInt,const PetscInt*);
21: PetscErrorCode MatSOR_BlockMat_Symmetric(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
22: {
23: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
24: PetscScalar *x;
25: const Mat *v;
26: const PetscScalar *b;
27: PetscErrorCode ierr;
28: PetscInt n = A->cmap->n,i,mbs = n/A->rmap->bs,j,bs = A->rmap->bs;
29: const PetscInt *idx;
30: IS row,col;
31: MatFactorInfo info;
32: Vec left = a->left,right = a->right, middle = a->middle;
33: Mat *diag;
36: its = its*lits;
37: if (its <= 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits);
38: if (flag & SOR_EISENSTAT) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for Eisenstat");
39: if (omega != 1.0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for omega not equal to 1.0");
40: if (fshift) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for fshift");
41: if ((flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) && !(flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP)) {
42: SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Cannot do backward sweep without forward sweep");
43: }
45: if (!a->diags) {
46: PetscMalloc1(mbs,&a->diags);
47: MatFactorInfoInitialize(&info);
48: for (i=0; i<mbs; i++) {
49: MatGetOrdering(a->a[a->diag[i]], MATORDERINGND,&row,&col);
50: MatCholeskyFactorSymbolic(a->diags[i],a->a[a->diag[i]],row,&info);
51: MatCholeskyFactorNumeric(a->diags[i],a->a[a->diag[i]],&info);
52: ISDestroy(&row);
53: ISDestroy(&col);
54: }
55: VecDuplicate(bb,&a->workb);
56: }
57: diag = a->diags;
59: VecSet(xx,0.0);
60: VecGetArray(xx,&x);
61: /* copy right hand side because it must be modified during iteration */
62: VecCopy(bb,a->workb);
63: VecGetArrayRead(a->workb,&b);
65: /* need to add code for when initial guess is zero, see MatSOR_SeqAIJ */
66: while (its--) {
67: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) {
69: for (i=0; i<mbs; i++) {
70: n = a->i[i+1] - a->i[i] - 1;
71: idx = a->j + a->i[i] + 1;
72: v = a->a + a->i[i] + 1;
74: VecSet(left,0.0);
75: for (j=0; j<n; j++) {
76: VecPlaceArray(right,x + idx[j]*bs);
77: MatMultAdd(v[j],right,left,left);
78: VecResetArray(right);
79: }
80: VecPlaceArray(right,b + i*bs);
81: VecAYPX(left,-1.0,right);
82: VecResetArray(right);
84: VecPlaceArray(right,x + i*bs);
85: MatSolve(diag[i],left,right);
87: /* now adjust right hand side, see MatSOR_SeqSBAIJ */
88: for (j=0; j<n; j++) {
89: MatMultTranspose(v[j],right,left);
90: VecPlaceArray(middle,b + idx[j]*bs);
91: VecAXPY(middle,-1.0,left);
92: VecResetArray(middle);
93: }
94: VecResetArray(right);
96: }
97: }
98: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) {
100: for (i=mbs-1; i>=0; i--) {
101: n = a->i[i+1] - a->i[i] - 1;
102: idx = a->j + a->i[i] + 1;
103: v = a->a + a->i[i] + 1;
105: VecSet(left,0.0);
106: for (j=0; j<n; j++) {
107: VecPlaceArray(right,x + idx[j]*bs);
108: MatMultAdd(v[j],right,left,left);
109: VecResetArray(right);
110: }
111: VecPlaceArray(right,b + i*bs);
112: VecAYPX(left,-1.0,right);
113: VecResetArray(right);
115: VecPlaceArray(right,x + i*bs);
116: MatSolve(diag[i],left,right);
117: VecResetArray(right);
119: }
120: }
121: }
122: VecRestoreArray(xx,&x);
123: VecRestoreArrayRead(a->workb,&b);
124: return(0);
125: }
129: PetscErrorCode MatSOR_BlockMat(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
130: {
131: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
132: PetscScalar *x;
133: const Mat *v;
134: const PetscScalar *b;
135: PetscErrorCode ierr;
136: PetscInt n = A->cmap->n,i,mbs = n/A->rmap->bs,j,bs = A->rmap->bs;
137: const PetscInt *idx;
138: IS row,col;
139: MatFactorInfo info;
140: Vec left = a->left,right = a->right;
141: Mat *diag;
144: its = its*lits;
145: if (its <= 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits);
146: if (flag & SOR_EISENSTAT) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for Eisenstat");
147: if (omega != 1.0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for omega not equal to 1.0");
148: if (fshift) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support yet for fshift");
150: if (!a->diags) {
151: PetscMalloc1(mbs,&a->diags);
152: MatFactorInfoInitialize(&info);
153: for (i=0; i<mbs; i++) {
154: MatGetOrdering(a->a[a->diag[i]], MATORDERINGND,&row,&col);
155: MatLUFactorSymbolic(a->diags[i],a->a[a->diag[i]],row,col,&info);
156: MatLUFactorNumeric(a->diags[i],a->a[a->diag[i]],&info);
157: ISDestroy(&row);
158: ISDestroy(&col);
159: }
160: }
161: diag = a->diags;
163: VecSet(xx,0.0);
164: VecGetArray(xx,&x);
165: VecGetArrayRead(bb,&b);
167: /* need to add code for when initial guess is zero, see MatSOR_SeqAIJ */
168: while (its--) {
169: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) {
171: for (i=0; i<mbs; i++) {
172: n = a->i[i+1] - a->i[i];
173: idx = a->j + a->i[i];
174: v = a->a + a->i[i];
176: VecSet(left,0.0);
177: for (j=0; j<n; j++) {
178: if (idx[j] != i) {
179: VecPlaceArray(right,x + idx[j]*bs);
180: MatMultAdd(v[j],right,left,left);
181: VecResetArray(right);
182: }
183: }
184: VecPlaceArray(right,b + i*bs);
185: VecAYPX(left,-1.0,right);
186: VecResetArray(right);
188: VecPlaceArray(right,x + i*bs);
189: MatSolve(diag[i],left,right);
190: VecResetArray(right);
191: }
192: }
193: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) {
195: for (i=mbs-1; i>=0; i--) {
196: n = a->i[i+1] - a->i[i];
197: idx = a->j + a->i[i];
198: v = a->a + a->i[i];
200: VecSet(left,0.0);
201: for (j=0; j<n; j++) {
202: if (idx[j] != i) {
203: VecPlaceArray(right,x + idx[j]*bs);
204: MatMultAdd(v[j],right,left,left);
205: VecResetArray(right);
206: }
207: }
208: VecPlaceArray(right,b + i*bs);
209: VecAYPX(left,-1.0,right);
210: VecResetArray(right);
212: VecPlaceArray(right,x + i*bs);
213: MatSolve(diag[i],left,right);
214: VecResetArray(right);
216: }
217: }
218: }
219: VecRestoreArray(xx,&x);
220: VecRestoreArrayRead(bb,&b);
221: return(0);
222: }
226: PetscErrorCode MatSetValues_BlockMat(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode is)
227: {
228: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
229: PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N,lastcol = -1;
230: PetscInt *imax=a->imax,*ai=a->i,*ailen=a->ilen;
231: PetscInt *aj =a->j,nonew=a->nonew,bs=A->rmap->bs,brow,bcol;
233: PetscInt ridx,cidx;
234: PetscBool roworiented=a->roworiented;
235: MatScalar value;
236: Mat *ap,*aa = a->a;
239: for (k=0; k<m; k++) { /* loop over added rows */
240: row = im[k];
241: brow = row/bs;
242: if (row < 0) continue;
243: #if defined(PETSC_USE_DEBUG)
244: if (row >= A->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",row,A->rmap->N-1);
245: #endif
246: rp = aj + ai[brow];
247: ap = aa + ai[brow];
248: rmax = imax[brow];
249: nrow = ailen[brow];
250: low = 0;
251: high = nrow;
252: for (l=0; l<n; l++) { /* loop over added columns */
253: if (in[l] < 0) continue;
254: #if defined(PETSC_USE_DEBUG)
255: if (in[l] >= A->cmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[l],A->cmap->n-1);
256: #endif
257: col = in[l]; bcol = col/bs;
258: if (A->symmetric && brow > bcol) continue;
259: ridx = row % bs; cidx = col % bs;
260: if (roworiented) value = v[l + k*n];
261: else value = v[k + l*m];
263: if (col <= lastcol) low = 0;
264: else high = nrow;
265: lastcol = col;
266: while (high-low > 7) {
267: t = (low+high)/2;
268: if (rp[t] > bcol) high = t;
269: else low = t;
270: }
271: for (i=low; i<high; i++) {
272: if (rp[i] > bcol) break;
273: if (rp[i] == bcol) goto noinsert1;
274: }
275: if (nonew == 1) goto noinsert1;
276: if (nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) in the matrix", row, col);
277: MatSeqXAIJReallocateAIJ(A,a->mbs,1,nrow,brow,bcol,rmax,aa,ai,aj,rp,ap,imax,nonew,Mat);
278: N = nrow++ - 1; high++;
279: /* shift up all the later entries in this row */
280: for (ii=N; ii>=i; ii--) {
281: rp[ii+1] = rp[ii];
282: ap[ii+1] = ap[ii];
283: }
284: if (N>=i) ap[i] = 0;
285: rp[i] = bcol;
286: a->nz++;
287: A->nonzerostate++;
288: noinsert1:;
289: if (!*(ap+i)) {
290: MatCreateSeqAIJ(PETSC_COMM_SELF,bs,bs,0,0,ap+i);
291: }
292: MatSetValues(ap[i],1,&ridx,1,&cidx,&value,is);
293: low = i;
294: }
295: ailen[brow] = nrow;
296: }
297: return(0);
298: }
302: PetscErrorCode MatLoad_BlockMat(Mat newmat, PetscViewer viewer)
303: {
304: PetscErrorCode ierr;
305: Mat tmpA;
306: PetscInt i,j,m,n,bs = 1,ncols,*lens,currentcol,mbs,**ii,*ilens,nextcol,*llens,cnt = 0;
307: const PetscInt *cols;
308: const PetscScalar *values;
309: PetscBool flg = PETSC_FALSE,notdone;
310: Mat_SeqAIJ *a;
311: Mat_BlockMat *amat;
314: /* force binary viewer to load .info file if it has not yet done so */
315: PetscViewerSetUp(viewer);
316: MatCreate(PETSC_COMM_SELF,&tmpA);
317: MatSetType(tmpA,MATSEQAIJ);
318: MatLoad_SeqAIJ(tmpA,viewer);
320: MatGetLocalSize(tmpA,&m,&n);
321: PetscOptionsBegin(PETSC_COMM_SELF,NULL,"Options for loading BlockMat matrix 1","Mat");
322: PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,NULL);
323: PetscOptionsBool("-matload_symmetric","Store the matrix as symmetric","MatLoad",flg,&flg,NULL);
324: PetscOptionsEnd();
326: /* Determine number of nonzero blocks for each block row */
327: a = (Mat_SeqAIJ*) tmpA->data;
328: mbs = m/bs;
329: PetscMalloc3(mbs,&lens,bs,&ii,bs,&ilens);
330: PetscMemzero(lens,mbs*sizeof(PetscInt));
332: for (i=0; i<mbs; i++) {
333: for (j=0; j<bs; j++) {
334: ii[j] = a->j + a->i[i*bs + j];
335: ilens[j] = a->i[i*bs + j + 1] - a->i[i*bs + j];
336: }
338: currentcol = -1;
339: notdone = PETSC_TRUE;
340: while (PETSC_TRUE) {
341: notdone = PETSC_FALSE;
342: nextcol = 1000000000;
343: for (j=0; j<bs; j++) {
344: while ((ilens[j] > 0 && ii[j][0]/bs <= currentcol)) {
345: ii[j]++;
346: ilens[j]--;
347: }
348: if (ilens[j] > 0) {
349: notdone = PETSC_TRUE;
350: nextcol = PetscMin(nextcol,ii[j][0]/bs);
351: }
352: }
353: if (!notdone) break;
354: if (!flg || (nextcol >= i)) lens[i]++;
355: currentcol = nextcol;
356: }
357: }
359: if (newmat->rmap->n < 0 && newmat->rmap->N < 0 && newmat->cmap->n < 0 && newmat->cmap->N < 0) {
360: MatSetSizes(newmat,m,n,PETSC_DETERMINE,PETSC_DETERMINE);
361: }
362: MatBlockMatSetPreallocation(newmat,bs,0,lens);
363: if (flg) {
364: MatSetOption(newmat,MAT_SYMMETRIC,PETSC_TRUE);
365: }
366: amat = (Mat_BlockMat*)(newmat)->data;
368: /* preallocate the submatrices */
369: PetscMalloc1(bs,&llens);
370: for (i=0; i<mbs; i++) { /* loops for block rows */
371: for (j=0; j<bs; j++) {
372: ii[j] = a->j + a->i[i*bs + j];
373: ilens[j] = a->i[i*bs + j + 1] - a->i[i*bs + j];
374: }
376: currentcol = 1000000000;
377: for (j=0; j<bs; j++) { /* loop over rows in block finding first nonzero block */
378: if (ilens[j] > 0) {
379: currentcol = PetscMin(currentcol,ii[j][0]/bs);
380: }
381: }
383: notdone = PETSC_TRUE;
384: while (PETSC_TRUE) { /* loops over blocks in block row */
386: notdone = PETSC_FALSE;
387: nextcol = 1000000000;
388: PetscMemzero(llens,bs*sizeof(PetscInt));
389: for (j=0; j<bs; j++) { /* loop over rows in block */
390: while ((ilens[j] > 0 && ii[j][0]/bs <= currentcol)) { /* loop over columns in row */
391: ii[j]++;
392: ilens[j]--;
393: llens[j]++;
394: }
395: if (ilens[j] > 0) {
396: notdone = PETSC_TRUE;
397: nextcol = PetscMin(nextcol,ii[j][0]/bs);
398: }
399: }
400: if (cnt >= amat->maxnz) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Number of blocks found greater than expected %D",cnt);
401: if (!flg || currentcol >= i) {
402: amat->j[cnt] = currentcol;
403: MatCreateSeqAIJ(PETSC_COMM_SELF,bs,bs,0,llens,amat->a+cnt++);
404: }
406: if (!notdone) break;
407: currentcol = nextcol;
408: }
409: amat->ilen[i] = lens[i];
410: }
412: PetscFree3(lens,ii,ilens);
413: PetscFree(llens);
415: /* copy over the matrix, one row at a time */
416: for (i=0; i<m; i++) {
417: MatGetRow(tmpA,i,&ncols,&cols,&values);
418: MatSetValues(newmat,1,&i,ncols,cols,values,INSERT_VALUES);
419: MatRestoreRow(tmpA,i,&ncols,&cols,&values);
420: }
421: MatAssemblyBegin(newmat,MAT_FINAL_ASSEMBLY);
422: MatAssemblyEnd(newmat,MAT_FINAL_ASSEMBLY);
423: return(0);
424: }
428: PetscErrorCode MatView_BlockMat(Mat A,PetscViewer viewer)
429: {
430: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
431: PetscErrorCode ierr;
432: const char *name;
433: PetscViewerFormat format;
436: PetscObjectGetName((PetscObject)A,&name);
437: PetscViewerGetFormat(viewer,&format);
438: if (format == PETSC_VIEWER_ASCII_FACTOR_INFO || format == PETSC_VIEWER_ASCII_INFO) {
439: PetscViewerASCIIPrintf(viewer,"Nonzero block matrices = %D \n",a->nz);
440: if (A->symmetric) {
441: PetscViewerASCIIPrintf(viewer,"Only upper triangular part of symmetric matrix is stored\n");
442: }
443: }
444: return(0);
445: }
449: PetscErrorCode MatDestroy_BlockMat(Mat mat)
450: {
452: Mat_BlockMat *bmat = (Mat_BlockMat*)mat->data;
453: PetscInt i;
456: VecDestroy(&bmat->right);
457: VecDestroy(&bmat->left);
458: VecDestroy(&bmat->middle);
459: VecDestroy(&bmat->workb);
460: if (bmat->diags) {
461: for (i=0; i<mat->rmap->n/mat->rmap->bs; i++) {
462: MatDestroy(&bmat->diags[i]);
463: }
464: }
465: if (bmat->a) {
466: for (i=0; i<bmat->nz; i++) {
467: MatDestroy(&bmat->a[i]);
468: }
469: }
470: MatSeqXAIJFreeAIJ(mat,(PetscScalar**)&bmat->a,&bmat->j,&bmat->i);
471: PetscFree(mat->data);
472: return(0);
473: }
477: PetscErrorCode MatMult_BlockMat(Mat A,Vec x,Vec y)
478: {
479: Mat_BlockMat *bmat = (Mat_BlockMat*)A->data;
481: PetscScalar *xx,*yy;
482: PetscInt *aj,i,*ii,jrow,m = A->rmap->n/A->rmap->bs,bs = A->rmap->bs,n,j;
483: Mat *aa;
486: /*
487: Standard CSR multiply except each entry is a Mat
488: */
489: VecGetArray(x,&xx);
491: VecSet(y,0.0);
492: VecGetArray(y,&yy);
493: aj = bmat->j;
494: aa = bmat->a;
495: ii = bmat->i;
496: for (i=0; i<m; i++) {
497: jrow = ii[i];
498: VecPlaceArray(bmat->left,yy + bs*i);
499: n = ii[i+1] - jrow;
500: for (j=0; j<n; j++) {
501: VecPlaceArray(bmat->right,xx + bs*aj[jrow]);
502: MatMultAdd(aa[jrow],bmat->right,bmat->left,bmat->left);
503: VecResetArray(bmat->right);
504: jrow++;
505: }
506: VecResetArray(bmat->left);
507: }
508: VecRestoreArray(x,&xx);
509: VecRestoreArray(y,&yy);
510: return(0);
511: }
515: PetscErrorCode MatMult_BlockMat_Symmetric(Mat A,Vec x,Vec y)
516: {
517: Mat_BlockMat *bmat = (Mat_BlockMat*)A->data;
519: PetscScalar *xx,*yy;
520: PetscInt *aj,i,*ii,jrow,m = A->rmap->n/A->rmap->bs,bs = A->rmap->bs,n,j;
521: Mat *aa;
524: /*
525: Standard CSR multiply except each entry is a Mat
526: */
527: VecGetArray(x,&xx);
529: VecSet(y,0.0);
530: VecGetArray(y,&yy);
531: aj = bmat->j;
532: aa = bmat->a;
533: ii = bmat->i;
534: for (i=0; i<m; i++) {
535: jrow = ii[i];
536: n = ii[i+1] - jrow;
537: VecPlaceArray(bmat->left,yy + bs*i);
538: VecPlaceArray(bmat->middle,xx + bs*i);
539: /* if we ALWAYS required a diagonal entry then could remove this if test */
540: if (aj[jrow] == i) {
541: VecPlaceArray(bmat->right,xx + bs*aj[jrow]);
542: MatMultAdd(aa[jrow],bmat->right,bmat->left,bmat->left);
543: VecResetArray(bmat->right);
544: jrow++;
545: n--;
546: }
547: for (j=0; j<n; j++) {
548: VecPlaceArray(bmat->right,xx + bs*aj[jrow]); /* upper triangular part */
549: MatMultAdd(aa[jrow],bmat->right,bmat->left,bmat->left);
550: VecResetArray(bmat->right);
552: VecPlaceArray(bmat->right,yy + bs*aj[jrow]); /* lower triangular part */
553: MatMultTransposeAdd(aa[jrow],bmat->middle,bmat->right,bmat->right);
554: VecResetArray(bmat->right);
555: jrow++;
556: }
557: VecResetArray(bmat->left);
558: VecResetArray(bmat->middle);
559: }
560: VecRestoreArray(x,&xx);
561: VecRestoreArray(y,&yy);
562: return(0);
563: }
567: PetscErrorCode MatMultAdd_BlockMat(Mat A,Vec x,Vec y,Vec z)
568: {
570: return(0);
571: }
575: PetscErrorCode MatMultTranspose_BlockMat(Mat A,Vec x,Vec y)
576: {
578: return(0);
579: }
583: PetscErrorCode MatMultTransposeAdd_BlockMat(Mat A,Vec x,Vec y,Vec z)
584: {
586: return(0);
587: }
589: /*
590: Adds diagonal pointers to sparse matrix structure.
591: */
594: PetscErrorCode MatMarkDiagonal_BlockMat(Mat A)
595: {
596: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
598: PetscInt i,j,mbs = A->rmap->n/A->rmap->bs;
601: if (!a->diag) {
602: PetscMalloc1(mbs,&a->diag);
603: }
604: for (i=0; i<mbs; i++) {
605: a->diag[i] = a->i[i+1];
606: for (j=a->i[i]; j<a->i[i+1]; j++) {
607: if (a->j[j] == i) {
608: a->diag[i] = j;
609: break;
610: }
611: }
612: }
613: return(0);
614: }
618: PetscErrorCode MatGetSubMatrix_BlockMat(Mat A,IS isrow,IS iscol,MatReuse scall,Mat *B)
619: {
620: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
621: Mat_SeqAIJ *c;
623: PetscInt i,k,first,step,lensi,nrows,ncols;
624: PetscInt *j_new,*i_new,*aj = a->j,*ailen = a->ilen;
625: PetscScalar *a_new;
626: Mat C,*aa = a->a;
627: PetscBool stride,equal;
630: ISEqual(isrow,iscol,&equal);
631: if (!equal) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only for idential column and row indices");
632: PetscObjectTypeCompare((PetscObject)iscol,ISSTRIDE,&stride);
633: if (!stride) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only for stride indices");
634: ISStrideGetInfo(iscol,&first,&step);
635: if (step != A->rmap->bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Can only select one entry from each block");
637: ISGetLocalSize(isrow,&nrows);
638: ncols = nrows;
640: /* create submatrix */
641: if (scall == MAT_REUSE_MATRIX) {
642: PetscInt n_cols,n_rows;
643: C = *B;
644: MatGetSize(C,&n_rows,&n_cols);
645: if (n_rows != nrows || n_cols != ncols) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Reused submatrix wrong size");
646: MatZeroEntries(C);
647: } else {
648: MatCreate(PetscObjectComm((PetscObject)A),&C);
649: MatSetSizes(C,nrows,ncols,PETSC_DETERMINE,PETSC_DETERMINE);
650: if (A->symmetric) {
651: MatSetType(C,MATSEQSBAIJ);
652: } else {
653: MatSetType(C,MATSEQAIJ);
654: }
655: MatSeqAIJSetPreallocation(C,0,ailen);
656: MatSeqSBAIJSetPreallocation(C,1,0,ailen);
657: }
658: c = (Mat_SeqAIJ*)C->data;
660: /* loop over rows inserting into submatrix */
661: a_new = c->a;
662: j_new = c->j;
663: i_new = c->i;
665: for (i=0; i<nrows; i++) {
666: lensi = ailen[i];
667: for (k=0; k<lensi; k++) {
668: *j_new++ = *aj++;
669: MatGetValue(*aa++,first,first,a_new++);
670: }
671: i_new[i+1] = i_new[i] + lensi;
672: c->ilen[i] = lensi;
673: }
675: MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);
676: MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);
677: *B = C;
678: return(0);
679: }
683: PetscErrorCode MatAssemblyEnd_BlockMat(Mat A,MatAssemblyType mode)
684: {
685: Mat_BlockMat *a = (Mat_BlockMat*)A->data;
687: PetscInt fshift = 0,i,j,*ai = a->i,*aj = a->j,*imax = a->imax;
688: PetscInt m = a->mbs,*ip,N,*ailen = a->ilen,rmax = 0;
689: Mat *aa = a->a,*ap;
692: if (mode == MAT_FLUSH_ASSEMBLY) return(0);
694: if (m) rmax = ailen[0]; /* determine row with most nonzeros */
695: for (i=1; i<m; i++) {
696: /* move each row back by the amount of empty slots (fshift) before it*/
697: fshift += imax[i-1] - ailen[i-1];
698: rmax = PetscMax(rmax,ailen[i]);
699: if (fshift) {
700: ip = aj + ai[i];
701: ap = aa + ai[i];
702: N = ailen[i];
703: for (j=0; j<N; j++) {
704: ip[j-fshift] = ip[j];
705: ap[j-fshift] = ap[j];
706: }
707: }
708: ai[i] = ai[i-1] + ailen[i-1];
709: }
710: if (m) {
711: fshift += imax[m-1] - ailen[m-1];
712: ai[m] = ai[m-1] + ailen[m-1];
713: }
714: /* reset ilen and imax for each row */
715: for (i=0; i<m; i++) {
716: ailen[i] = imax[i] = ai[i+1] - ai[i];
717: }
718: a->nz = ai[m];
719: for (i=0; i<a->nz; i++) {
720: #if defined(PETSC_USE_DEBUG)
721: if (!aa[i]) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Null matrix at location %D column %D nz %D",i,aj[i],a->nz);
722: #endif
723: MatAssemblyBegin(aa[i],MAT_FINAL_ASSEMBLY);
724: MatAssemblyEnd(aa[i],MAT_FINAL_ASSEMBLY);
725: }
726: PetscInfo4(A,"Matrix size: %D X %D; storage space: %D unneeded,%D used\n",m,A->cmap->n/A->cmap->bs,fshift,a->nz);
727: PetscInfo1(A,"Number of mallocs during MatSetValues() is %D\n",a->reallocs);
728: PetscInfo1(A,"Maximum nonzeros in any row is %D\n",rmax);
730: A->info.mallocs += a->reallocs;
731: a->reallocs = 0;
732: A->info.nz_unneeded = (double)fshift;
733: a->rmax = rmax;
734: MatMarkDiagonal_BlockMat(A);
735: return(0);
736: }
740: PetscErrorCode MatSetOption_BlockMat(Mat A,MatOption opt,PetscBool flg)
741: {
743: if (opt == MAT_SYMMETRIC && flg) {
744: A->ops->sor = MatSOR_BlockMat_Symmetric;
745: A->ops->mult = MatMult_BlockMat_Symmetric;
746: } else {
747: PetscInfo1(A,"Unused matrix option %s\n",MatOptions[opt]);
748: }
749: return(0);
750: }
753: static struct _MatOps MatOps_Values = {MatSetValues_BlockMat,
754: 0,
755: 0,
756: MatMult_BlockMat,
757: /* 4*/ MatMultAdd_BlockMat,
758: MatMultTranspose_BlockMat,
759: MatMultTransposeAdd_BlockMat,
760: 0,
761: 0,
762: 0,
763: /* 10*/ 0,
764: 0,
765: 0,
766: MatSOR_BlockMat,
767: 0,
768: /* 15*/ 0,
769: 0,
770: 0,
771: 0,
772: 0,
773: /* 20*/ 0,
774: MatAssemblyEnd_BlockMat,
775: MatSetOption_BlockMat,
776: 0,
777: /* 24*/ 0,
778: 0,
779: 0,
780: 0,
781: 0,
782: /* 29*/ 0,
783: 0,
784: 0,
785: 0,
786: 0,
787: /* 34*/ 0,
788: 0,
789: 0,
790: 0,
791: 0,
792: /* 39*/ 0,
793: 0,
794: 0,
795: 0,
796: 0,
797: /* 44*/ 0,
798: 0,
799: MatShift_Basic,
800: 0,
801: 0,
802: /* 49*/ 0,
803: 0,
804: 0,
805: 0,
806: 0,
807: /* 54*/ 0,
808: 0,
809: 0,
810: 0,
811: 0,
812: /* 59*/ MatGetSubMatrix_BlockMat,
813: MatDestroy_BlockMat,
814: MatView_BlockMat,
815: 0,
816: 0,
817: /* 64*/ 0,
818: 0,
819: 0,
820: 0,
821: 0,
822: /* 69*/ 0,
823: 0,
824: 0,
825: 0,
826: 0,
827: /* 74*/ 0,
828: 0,
829: 0,
830: 0,
831: 0,
832: /* 79*/ 0,
833: 0,
834: 0,
835: 0,
836: MatLoad_BlockMat,
837: /* 84*/ 0,
838: 0,
839: 0,
840: 0,
841: 0,
842: /* 89*/ 0,
843: 0,
844: 0,
845: 0,
846: 0,
847: /* 94*/ 0,
848: 0,
849: 0,
850: 0,
851: 0,
852: /* 99*/ 0,
853: 0,
854: 0,
855: 0,
856: 0,
857: /*104*/ 0,
858: 0,
859: 0,
860: 0,
861: 0,
862: /*109*/ 0,
863: 0,
864: 0,
865: 0,
866: 0,
867: /*114*/ 0,
868: 0,
869: 0,
870: 0,
871: 0,
872: /*119*/ 0,
873: 0,
874: 0,
875: 0,
876: 0,
877: /*124*/ 0,
878: 0,
879: 0,
880: 0,
881: 0,
882: /*129*/ 0,
883: 0,
884: 0,
885: 0,
886: 0,
887: /*134*/ 0,
888: 0,
889: 0,
890: 0,
891: 0,
892: /*139*/ 0,
893: 0,
894: 0
895: };
899: /*@C
900: MatBlockMatSetPreallocation - For good matrix assembly performance
901: the user should preallocate the matrix storage by setting the parameter nz
902: (or the array nnz). By setting these parameters accurately, performance
903: during matrix assembly can be increased by more than a factor of 50.
905: Collective on MPI_Comm
907: Input Parameters:
908: + B - The matrix
909: . bs - size of each block in matrix
910: . nz - number of nonzeros per block row (same for all rows)
911: - nnz - array containing the number of nonzeros in the various block rows
912: (possibly different for each row) or NULL
914: Notes:
915: If nnz is given then nz is ignored
917: Specify the preallocated storage with either nz or nnz (not both).
918: Set nz=PETSC_DEFAULT and nnz=NULL for PETSc to control dynamic memory
919: allocation. For large problems you MUST preallocate memory or you
920: will get TERRIBLE performance, see the users' manual chapter on matrices.
922: Level: intermediate
924: .seealso: MatCreate(), MatCreateBlockMat(), MatSetValues()
926: @*/
927: PetscErrorCode MatBlockMatSetPreallocation(Mat B,PetscInt bs,PetscInt nz,const PetscInt nnz[])
928: {
932: PetscTryMethod(B,"MatBlockMatSetPreallocation_C",(Mat,PetscInt,PetscInt,const PetscInt[]),(B,bs,nz,nnz));
933: return(0);
934: }
938: PetscErrorCode MatBlockMatSetPreallocation_BlockMat(Mat A,PetscInt bs,PetscInt nz,PetscInt *nnz)
939: {
940: Mat_BlockMat *bmat = (Mat_BlockMat*)A->data;
942: PetscInt i;
945: PetscLayoutSetBlockSize(A->rmap,bs);
946: PetscLayoutSetBlockSize(A->cmap,bs);
947: PetscLayoutSetUp(A->rmap);
948: PetscLayoutSetUp(A->cmap);
949: PetscLayoutGetBlockSize(A->rmap,&bs);
951: if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 5;
952: if (nz < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"nz cannot be less than 0: value %d",nz);
953: if (nnz) {
954: for (i=0; i<A->rmap->n/bs; i++) {
955: if (nnz[i] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be less than 0: local row %d value %d",i,nnz[i]);
956: if (nnz[i] > A->cmap->n/bs) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be greater than row length: local row %d value %d rowlength %d",i,nnz[i],A->cmap->n/bs);
957: }
958: }
959: bmat->mbs = A->rmap->n/bs;
961: VecCreateSeqWithArray(PETSC_COMM_SELF,1,bs,NULL,&bmat->right);
962: VecCreateSeqWithArray(PETSC_COMM_SELF,1,bs,NULL,&bmat->middle);
963: VecCreateSeq(PETSC_COMM_SELF,bs,&bmat->left);
965: if (!bmat->imax) {
966: PetscMalloc2(A->rmap->n,&bmat->imax,A->rmap->n,&bmat->ilen);
967: PetscLogObjectMemory((PetscObject)A,2*A->rmap->n*sizeof(PetscInt));
968: }
969: if (nnz) {
970: nz = 0;
971: for (i=0; i<A->rmap->n/A->rmap->bs; i++) {
972: bmat->imax[i] = nnz[i];
973: nz += nnz[i];
974: }
975: } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Currently requires block row by row preallocation");
977: /* bmat->ilen will count nonzeros in each row so far. */
978: for (i=0; i<bmat->mbs; i++) bmat->ilen[i] = 0;
980: /* allocate the matrix space */
981: MatSeqXAIJFreeAIJ(A,(PetscScalar**)&bmat->a,&bmat->j,&bmat->i);
982: PetscMalloc3(nz,&bmat->a,nz,&bmat->j,A->rmap->n+1,&bmat->i);
983: PetscLogObjectMemory((PetscObject)A,(A->rmap->n+1)*sizeof(PetscInt)+nz*(sizeof(PetscScalar)+sizeof(PetscInt)));
984: bmat->i[0] = 0;
985: for (i=1; i<bmat->mbs+1; i++) {
986: bmat->i[i] = bmat->i[i-1] + bmat->imax[i-1];
987: }
988: bmat->singlemalloc = PETSC_TRUE;
989: bmat->free_a = PETSC_TRUE;
990: bmat->free_ij = PETSC_TRUE;
992: bmat->nz = 0;
993: bmat->maxnz = nz;
994: A->info.nz_unneeded = (double)bmat->maxnz;
995: MatSetOption(A,MAT_NEW_NONZERO_ALLOCATION_ERR,PETSC_TRUE);
996: return(0);
997: }
999: /*MC
1000: MATBLOCKMAT - A matrix that is defined by a set of Mat's that represents a sparse block matrix
1001: consisting of (usually) sparse blocks.
1003: Level: advanced
1005: .seealso: MatCreateBlockMat()
1007: M*/
1011: PETSC_EXTERN PetscErrorCode MatCreate_BlockMat(Mat A)
1012: {
1013: Mat_BlockMat *b;
1017: PetscNewLog(A,&b);
1018: A->data = (void*)b;
1019: PetscMemcpy(A->ops,&MatOps_Values,sizeof(struct _MatOps));
1021: A->assembled = PETSC_TRUE;
1022: A->preallocated = PETSC_FALSE;
1023: PetscObjectChangeTypeName((PetscObject)A,MATBLOCKMAT);
1025: PetscObjectComposeFunction((PetscObject)A,"MatBlockMatSetPreallocation_C",MatBlockMatSetPreallocation_BlockMat);
1026: return(0);
1027: }
1031: /*@C
1032: MatCreateBlockMat - Creates a new matrix in which each block contains a uniform-size sequential Mat object
1034: Collective on MPI_Comm
1036: Input Parameters:
1037: + comm - MPI communicator
1038: . m - number of rows
1039: . n - number of columns
1040: . bs - size of each submatrix
1041: . nz - expected maximum number of nonzero blocks in row (use PETSC_DEFAULT if not known)
1042: - nnz - expected number of nonzers per block row if known (use NULL otherwise)
1045: Output Parameter:
1046: . A - the matrix
1048: Level: intermediate
1050: Notes: Matrices of this type are nominally-sparse matrices in which each "entry" is a Mat object. Each Mat must
1051: have the same size and be sequential. The local and global sizes must be compatible with this decomposition.
1053: For matrices containing parallel submatrices and variable block sizes, see MATNEST.
1055: .keywords: matrix, bmat, create
1057: .seealso: MATBLOCKMAT, MatCreateNest()
1058: @*/
1059: PetscErrorCode MatCreateBlockMat(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt bs,PetscInt nz,PetscInt *nnz, Mat *A)
1060: {
1064: MatCreate(comm,A);
1065: MatSetSizes(*A,m,n,PETSC_DETERMINE,PETSC_DETERMINE);
1066: MatSetType(*A,MATBLOCKMAT);
1067: MatBlockMatSetPreallocation(*A,bs,nz,nnz);
1068: return(0);
1069: }