Actual source code: aij.c
petsc-3.7.7 2017-09-25
2: /*
3: Defines the basic matrix operations for the AIJ (compressed row)
4: matrix storage format.
5: */
8: #include <../src/mat/impls/aij/seq/aij.h> /*I "petscmat.h" I*/
9: #include <petscblaslapack.h>
10: #include <petscbt.h>
11: #include <petsc/private/kernels/blocktranspose.h>
15: PetscErrorCode MatGetColumnNorms_SeqAIJ(Mat A,NormType type,PetscReal *norms)
16: {
18: PetscInt i,m,n;
19: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)A->data;
22: MatGetSize(A,&m,&n);
23: PetscMemzero(norms,n*sizeof(PetscReal));
24: if (type == NORM_2) {
25: for (i=0; i<aij->i[m]; i++) {
26: norms[aij->j[i]] += PetscAbsScalar(aij->a[i]*aij->a[i]);
27: }
28: } else if (type == NORM_1) {
29: for (i=0; i<aij->i[m]; i++) {
30: norms[aij->j[i]] += PetscAbsScalar(aij->a[i]);
31: }
32: } else if (type == NORM_INFINITY) {
33: for (i=0; i<aij->i[m]; i++) {
34: norms[aij->j[i]] = PetscMax(PetscAbsScalar(aij->a[i]),norms[aij->j[i]]);
35: }
36: } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Unknown NormType");
38: if (type == NORM_2) {
39: for (i=0; i<n; i++) norms[i] = PetscSqrtReal(norms[i]);
40: }
41: return(0);
42: }
46: PetscErrorCode MatFindOffBlockDiagonalEntries_SeqAIJ(Mat A,IS *is)
47: {
48: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
49: PetscInt i,m=A->rmap->n,cnt = 0, bs = A->rmap->bs;
50: const PetscInt *jj = a->j,*ii = a->i;
51: PetscInt *rows;
52: PetscErrorCode ierr;
55: for (i=0; i<m; i++) {
56: if ((ii[i] != ii[i+1]) && ((jj[ii[i]] < bs*(i/bs)) || (jj[ii[i+1]-1] > bs*((i+bs)/bs)-1))) {
57: cnt++;
58: }
59: }
60: PetscMalloc1(cnt,&rows);
61: cnt = 0;
62: for (i=0; i<m; i++) {
63: if ((ii[i] != ii[i+1]) && ((jj[ii[i]] < bs*(i/bs)) || (jj[ii[i+1]-1] > bs*((i+bs)/bs)-1))) {
64: rows[cnt] = i;
65: cnt++;
66: }
67: }
68: ISCreateGeneral(PETSC_COMM_SELF,cnt,rows,PETSC_OWN_POINTER,is);
69: return(0);
70: }
74: PetscErrorCode MatFindZeroDiagonals_SeqAIJ_Private(Mat A,PetscInt *nrows,PetscInt **zrows)
75: {
76: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
77: const MatScalar *aa = a->a;
78: PetscInt i,m=A->rmap->n,cnt = 0;
79: const PetscInt *ii = a->i,*jj = a->j,*diag;
80: PetscInt *rows;
81: PetscErrorCode ierr;
84: MatMarkDiagonal_SeqAIJ(A);
85: diag = a->diag;
86: for (i=0; i<m; i++) {
87: if ((diag[i] >= ii[i+1]) || (jj[diag[i]] != i) || (aa[diag[i]] == 0.0)) {
88: cnt++;
89: }
90: }
91: PetscMalloc1(cnt,&rows);
92: cnt = 0;
93: for (i=0; i<m; i++) {
94: if ((diag[i] >= ii[i+1]) || (jj[diag[i]] != i) || (aa[diag[i]] == 0.0)) {
95: rows[cnt++] = i;
96: }
97: }
98: *nrows = cnt;
99: *zrows = rows;
100: return(0);
101: }
105: PetscErrorCode MatFindZeroDiagonals_SeqAIJ(Mat A,IS *zrows)
106: {
107: PetscInt nrows,*rows;
111: *zrows = NULL;
112: MatFindZeroDiagonals_SeqAIJ_Private(A,&nrows,&rows);
113: ISCreateGeneral(PetscObjectComm((PetscObject)A),nrows,rows,PETSC_OWN_POINTER,zrows);
114: return(0);
115: }
119: PetscErrorCode MatFindNonzeroRows_SeqAIJ(Mat A,IS *keptrows)
120: {
121: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
122: const MatScalar *aa;
123: PetscInt m=A->rmap->n,cnt = 0;
124: const PetscInt *ii;
125: PetscInt n,i,j,*rows;
126: PetscErrorCode ierr;
129: *keptrows = 0;
130: ii = a->i;
131: for (i=0; i<m; i++) {
132: n = ii[i+1] - ii[i];
133: if (!n) {
134: cnt++;
135: goto ok1;
136: }
137: aa = a->a + ii[i];
138: for (j=0; j<n; j++) {
139: if (aa[j] != 0.0) goto ok1;
140: }
141: cnt++;
142: ok1:;
143: }
144: if (!cnt) return(0);
145: PetscMalloc1(A->rmap->n-cnt,&rows);
146: cnt = 0;
147: for (i=0; i<m; i++) {
148: n = ii[i+1] - ii[i];
149: if (!n) continue;
150: aa = a->a + ii[i];
151: for (j=0; j<n; j++) {
152: if (aa[j] != 0.0) {
153: rows[cnt++] = i;
154: break;
155: }
156: }
157: }
158: ISCreateGeneral(PETSC_COMM_SELF,cnt,rows,PETSC_OWN_POINTER,keptrows);
159: return(0);
160: }
164: PetscErrorCode MatDiagonalSet_SeqAIJ(Mat Y,Vec D,InsertMode is)
165: {
166: PetscErrorCode ierr;
167: Mat_SeqAIJ *aij = (Mat_SeqAIJ*) Y->data;
168: PetscInt i,m = Y->rmap->n;
169: const PetscInt *diag;
170: MatScalar *aa = aij->a;
171: const PetscScalar *v;
172: PetscBool missing;
175: if (Y->assembled) {
176: MatMissingDiagonal_SeqAIJ(Y,&missing,NULL);
177: if (!missing) {
178: diag = aij->diag;
179: VecGetArrayRead(D,&v);
180: if (is == INSERT_VALUES) {
181: for (i=0; i<m; i++) {
182: aa[diag[i]] = v[i];
183: }
184: } else {
185: for (i=0; i<m; i++) {
186: aa[diag[i]] += v[i];
187: }
188: }
189: VecRestoreArrayRead(D,&v);
190: return(0);
191: }
192: MatSeqAIJInvalidateDiagonal(Y);
193: }
194: MatDiagonalSet_Default(Y,D,is);
195: return(0);
196: }
200: PetscErrorCode MatGetRowIJ_SeqAIJ(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *m,const PetscInt *ia[],const PetscInt *ja[],PetscBool *done)
201: {
202: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
204: PetscInt i,ishift;
207: *m = A->rmap->n;
208: if (!ia) return(0);
209: ishift = 0;
210: if (symmetric && !A->structurally_symmetric) {
211: MatToSymmetricIJ_SeqAIJ(A->rmap->n,a->i,a->j,PETSC_TRUE,ishift,oshift,(PetscInt**)ia,(PetscInt**)ja);
212: } else if (oshift == 1) {
213: PetscInt *tia;
214: PetscInt nz = a->i[A->rmap->n];
215: /* malloc space and add 1 to i and j indices */
216: PetscMalloc1(A->rmap->n+1,&tia);
217: for (i=0; i<A->rmap->n+1; i++) tia[i] = a->i[i] + 1;
218: *ia = tia;
219: if (ja) {
220: PetscInt *tja;
221: PetscMalloc1(nz+1,&tja);
222: for (i=0; i<nz; i++) tja[i] = a->j[i] + 1;
223: *ja = tja;
224: }
225: } else {
226: *ia = a->i;
227: if (ja) *ja = a->j;
228: }
229: return(0);
230: }
234: PetscErrorCode MatRestoreRowIJ_SeqAIJ(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,const PetscInt *ia[],const PetscInt *ja[],PetscBool *done)
235: {
239: if (!ia) return(0);
240: if ((symmetric && !A->structurally_symmetric) || oshift == 1) {
241: PetscFree(*ia);
242: if (ja) {PetscFree(*ja);}
243: }
244: return(0);
245: }
249: PetscErrorCode MatGetColumnIJ_SeqAIJ(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *nn,const PetscInt *ia[],const PetscInt *ja[],PetscBool *done)
250: {
251: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
253: PetscInt i,*collengths,*cia,*cja,n = A->cmap->n,m = A->rmap->n;
254: PetscInt nz = a->i[m],row,*jj,mr,col;
257: *nn = n;
258: if (!ia) return(0);
259: if (symmetric) {
260: MatToSymmetricIJ_SeqAIJ(A->rmap->n,a->i,a->j,PETSC_TRUE,0,oshift,(PetscInt**)ia,(PetscInt**)ja);
261: } else {
262: PetscCalloc1(n+1,&collengths);
263: PetscMalloc1(n+1,&cia);
264: PetscMalloc1(nz+1,&cja);
265: jj = a->j;
266: for (i=0; i<nz; i++) {
267: collengths[jj[i]]++;
268: }
269: cia[0] = oshift;
270: for (i=0; i<n; i++) {
271: cia[i+1] = cia[i] + collengths[i];
272: }
273: PetscMemzero(collengths,n*sizeof(PetscInt));
274: jj = a->j;
275: for (row=0; row<m; row++) {
276: mr = a->i[row+1] - a->i[row];
277: for (i=0; i<mr; i++) {
278: col = *jj++;
280: cja[cia[col] + collengths[col]++ - oshift] = row + oshift;
281: }
282: }
283: PetscFree(collengths);
284: *ia = cia; *ja = cja;
285: }
286: return(0);
287: }
291: PetscErrorCode MatRestoreColumnIJ_SeqAIJ(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,const PetscInt *ia[],const PetscInt *ja[],PetscBool *done)
292: {
296: if (!ia) return(0);
298: PetscFree(*ia);
299: PetscFree(*ja);
300: return(0);
301: }
303: /*
304: MatGetColumnIJ_SeqAIJ_Color() and MatRestoreColumnIJ_SeqAIJ_Color() are customized from
305: MatGetColumnIJ_SeqAIJ() and MatRestoreColumnIJ_SeqAIJ() by adding an output
306: spidx[], index of a->a, to be used in MatTransposeColoringCreate_SeqAIJ() and MatFDColoringCreate_SeqXAIJ()
307: */
310: PetscErrorCode MatGetColumnIJ_SeqAIJ_Color(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *nn,const PetscInt *ia[],const PetscInt *ja[],PetscInt *spidx[],PetscBool *done)
311: {
312: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
314: PetscInt i,*collengths,*cia,*cja,n = A->cmap->n,m = A->rmap->n;
315: PetscInt nz = a->i[m],row,*jj,mr,col;
316: PetscInt *cspidx;
319: *nn = n;
320: if (!ia) return(0);
322: PetscCalloc1(n+1,&collengths);
323: PetscMalloc1(n+1,&cia);
324: PetscMalloc1(nz+1,&cja);
325: PetscMalloc1(nz+1,&cspidx);
326: jj = a->j;
327: for (i=0; i<nz; i++) {
328: collengths[jj[i]]++;
329: }
330: cia[0] = oshift;
331: for (i=0; i<n; i++) {
332: cia[i+1] = cia[i] + collengths[i];
333: }
334: PetscMemzero(collengths,n*sizeof(PetscInt));
335: jj = a->j;
336: for (row=0; row<m; row++) {
337: mr = a->i[row+1] - a->i[row];
338: for (i=0; i<mr; i++) {
339: col = *jj++;
340: cspidx[cia[col] + collengths[col] - oshift] = a->i[row] + i; /* index of a->j */
341: cja[cia[col] + collengths[col]++ - oshift] = row + oshift;
342: }
343: }
344: PetscFree(collengths);
345: *ia = cia; *ja = cja;
346: *spidx = cspidx;
347: return(0);
348: }
352: PetscErrorCode MatRestoreColumnIJ_SeqAIJ_Color(Mat A,PetscInt oshift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,const PetscInt *ia[],const PetscInt *ja[],PetscInt *spidx[],PetscBool *done)
353: {
357: MatRestoreColumnIJ_SeqAIJ(A,oshift,symmetric,inodecompressed,n,ia,ja,done);
358: PetscFree(*spidx);
359: return(0);
360: }
364: PetscErrorCode MatSetValuesRow_SeqAIJ(Mat A,PetscInt row,const PetscScalar v[])
365: {
366: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
367: PetscInt *ai = a->i;
371: PetscMemcpy(a->a+ai[row],v,(ai[row+1]-ai[row])*sizeof(PetscScalar));
372: return(0);
373: }
375: /*
376: MatSeqAIJSetValuesLocalFast - An optimized version of MatSetValuesLocal() for SeqAIJ matrices with several assumptions
378: - a single row of values is set with each call
379: - no row or column indices are negative or (in error) larger than the number of rows or columns
380: - the values are always added to the matrix, not set
381: - no new locations are introduced in the nonzero structure of the matrix
383: This does NOT assume the global column indices are sorted
385: */
387: #include <petsc/private/isimpl.h>
390: PetscErrorCode MatSeqAIJSetValuesLocalFast(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode is)
391: {
392: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
393: PetscInt low,high,t,row,nrow,i,col,l;
394: const PetscInt *rp,*ai = a->i,*ailen = a->ilen,*aj = a->j;
395: PetscInt lastcol = -1;
396: MatScalar *ap,value,*aa = a->a;
397: const PetscInt *ridx = A->rmap->mapping->indices,*cidx = A->cmap->mapping->indices;
399: row = ridx[im[0]];
400: rp = aj + ai[row];
401: ap = aa + ai[row];
402: nrow = ailen[row];
403: low = 0;
404: high = nrow;
405: for (l=0; l<n; l++) { /* loop over added columns */
406: col = cidx[in[l]];
407: value = v[l];
409: if (col <= lastcol) low = 0;
410: else high = nrow;
411: lastcol = col;
412: while (high-low > 5) {
413: t = (low+high)/2;
414: if (rp[t] > col) high = t;
415: else low = t;
416: }
417: for (i=low; i<high; i++) {
418: if (rp[i] == col) {
419: ap[i] += value;
420: low = i + 1;
421: break;
422: }
423: }
424: }
425: return 0;
426: }
430: PetscErrorCode MatSetValues_SeqAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode is)
431: {
432: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
433: PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N;
434: PetscInt *imax = a->imax,*ai = a->i,*ailen = a->ilen;
436: PetscInt *aj = a->j,nonew = a->nonew,lastcol = -1;
437: MatScalar *ap,value,*aa = a->a;
438: PetscBool ignorezeroentries = a->ignorezeroentries;
439: PetscBool roworiented = a->roworiented;
442: for (k=0; k<m; k++) { /* loop over added rows */
443: row = im[k];
444: if (row < 0) continue;
445: #if defined(PETSC_USE_DEBUG)
446: 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);
447: #endif
448: rp = aj + ai[row]; ap = aa + ai[row];
449: rmax = imax[row]; nrow = ailen[row];
450: low = 0;
451: high = nrow;
452: for (l=0; l<n; l++) { /* loop over added columns */
453: if (in[l] < 0) continue;
454: #if defined(PETSC_USE_DEBUG)
455: 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);
456: #endif
457: col = in[l];
458: if (roworiented) {
459: value = v[l + k*n];
460: } else {
461: value = v[k + l*m];
462: }
463: if ((value == 0.0 && ignorezeroentries) && (is == ADD_VALUES)) continue;
465: if (col <= lastcol) low = 0;
466: else high = nrow;
467: lastcol = col;
468: while (high-low > 5) {
469: t = (low+high)/2;
470: if (rp[t] > col) high = t;
471: else low = t;
472: }
473: for (i=low; i<high; i++) {
474: if (rp[i] > col) break;
475: if (rp[i] == col) {
476: if (is == ADD_VALUES) ap[i] += value;
477: else ap[i] = value;
478: low = i + 1;
479: goto noinsert;
480: }
481: }
482: if (value == 0.0 && ignorezeroentries) goto noinsert;
483: if (nonew == 1) goto noinsert;
484: if (nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero at (%D,%D) in the matrix",row,col);
485: MatSeqXAIJReallocateAIJ(A,A->rmap->n,1,nrow,row,col,rmax,aa,ai,aj,rp,ap,imax,nonew,MatScalar);
486: N = nrow++ - 1; a->nz++; high++;
487: /* shift up all the later entries in this row */
488: for (ii=N; ii>=i; ii--) {
489: rp[ii+1] = rp[ii];
490: ap[ii+1] = ap[ii];
491: }
492: rp[i] = col;
493: ap[i] = value;
494: low = i + 1;
495: A->nonzerostate++;
496: noinsert:;
497: }
498: ailen[row] = nrow;
499: }
500: return(0);
501: }
506: PetscErrorCode MatGetValues_SeqAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],PetscScalar v[])
507: {
508: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
509: PetscInt *rp,k,low,high,t,row,nrow,i,col,l,*aj = a->j;
510: PetscInt *ai = a->i,*ailen = a->ilen;
511: MatScalar *ap,*aa = a->a;
514: for (k=0; k<m; k++) { /* loop over rows */
515: row = im[k];
516: if (row < 0) {v += n; continue;} /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",row); */
517: 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);
518: rp = aj + ai[row]; ap = aa + ai[row];
519: nrow = ailen[row];
520: for (l=0; l<n; l++) { /* loop over columns */
521: if (in[l] < 0) {v++; continue;} /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative column: %D",in[l]); */
522: 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);
523: col = in[l];
524: high = nrow; low = 0; /* assume unsorted */
525: while (high-low > 5) {
526: t = (low+high)/2;
527: if (rp[t] > col) high = t;
528: else low = t;
529: }
530: for (i=low; i<high; i++) {
531: if (rp[i] > col) break;
532: if (rp[i] == col) {
533: *v++ = ap[i];
534: goto finished;
535: }
536: }
537: *v++ = 0.0;
538: finished:;
539: }
540: }
541: return(0);
542: }
547: PetscErrorCode MatView_SeqAIJ_Binary(Mat A,PetscViewer viewer)
548: {
549: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
551: PetscInt i,*col_lens;
552: int fd;
553: FILE *file;
556: PetscViewerBinaryGetDescriptor(viewer,&fd);
557: PetscMalloc1(4+A->rmap->n,&col_lens);
559: col_lens[0] = MAT_FILE_CLASSID;
560: col_lens[1] = A->rmap->n;
561: col_lens[2] = A->cmap->n;
562: col_lens[3] = a->nz;
564: /* store lengths of each row and write (including header) to file */
565: for (i=0; i<A->rmap->n; i++) {
566: col_lens[4+i] = a->i[i+1] - a->i[i];
567: }
568: PetscBinaryWrite(fd,col_lens,4+A->rmap->n,PETSC_INT,PETSC_TRUE);
569: PetscFree(col_lens);
571: /* store column indices (zero start index) */
572: PetscBinaryWrite(fd,a->j,a->nz,PETSC_INT,PETSC_FALSE);
574: /* store nonzero values */
575: PetscBinaryWrite(fd,a->a,a->nz,PETSC_SCALAR,PETSC_FALSE);
577: PetscViewerBinaryGetInfoPointer(viewer,&file);
578: if (file) {
579: fprintf(file,"-matload_block_size %d\n",(int)PetscAbs(A->rmap->bs));
580: }
581: return(0);
582: }
584: extern PetscErrorCode MatSeqAIJFactorInfo_Matlab(Mat,PetscViewer);
588: PetscErrorCode MatView_SeqAIJ_ASCII(Mat A,PetscViewer viewer)
589: {
590: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
591: PetscErrorCode ierr;
592: PetscInt i,j,m = A->rmap->n;
593: const char *name;
594: PetscViewerFormat format;
597: if (!a->a) return(0);
599: PetscViewerGetFormat(viewer,&format);
600: if (format == PETSC_VIEWER_ASCII_MATLAB) {
601: PetscInt nofinalvalue = 0;
602: if (m && ((a->i[m] == a->i[m-1]) || (a->j[a->nz-1] != A->cmap->n-1))) {
603: /* Need a dummy value to ensure the dimension of the matrix. */
604: nofinalvalue = 1;
605: }
606: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
607: PetscViewerASCIIPrintf(viewer,"%% Size = %D %D \n",m,A->cmap->n);
608: PetscViewerASCIIPrintf(viewer,"%% Nonzeros = %D \n",a->nz);
609: #if defined(PETSC_USE_COMPLEX)
610: PetscViewerASCIIPrintf(viewer,"zzz = zeros(%D,4);\n",a->nz+nofinalvalue);
611: #else
612: PetscViewerASCIIPrintf(viewer,"zzz = zeros(%D,3);\n",a->nz+nofinalvalue);
613: #endif
614: PetscViewerASCIIPrintf(viewer,"zzz = [\n");
616: for (i=0; i<m; i++) {
617: for (j=a->i[i]; j<a->i[i+1]; j++) {
618: #if defined(PETSC_USE_COMPLEX)
619: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e %18.16e\n",i+1,a->j[j]+1,(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
620: #else
621: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e\n",i+1,a->j[j]+1,(double)a->a[j]);
622: #endif
623: }
624: }
625: if (nofinalvalue) {
626: #if defined(PETSC_USE_COMPLEX)
627: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e %18.16e\n",m,A->cmap->n,0.,0.);
628: #else
629: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e\n",m,A->cmap->n,0.0);
630: #endif
631: }
632: PetscObjectGetName((PetscObject)A,&name);
633: PetscViewerASCIIPrintf(viewer,"];\n %s = spconvert(zzz);\n",name);
634: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
635: } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO || format == PETSC_VIEWER_ASCII_INFO) {
636: return(0);
637: } else if (format == PETSC_VIEWER_ASCII_COMMON) {
638: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
639: for (i=0; i<m; i++) {
640: PetscViewerASCIIPrintf(viewer,"row %D:",i);
641: for (j=a->i[i]; j<a->i[i+1]; j++) {
642: #if defined(PETSC_USE_COMPLEX)
643: if (PetscImaginaryPart(a->a[j]) > 0.0 && PetscRealPart(a->a[j]) != 0.0) {
644: PetscViewerASCIIPrintf(viewer," (%D, %g + %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
645: } else if (PetscImaginaryPart(a->a[j]) < 0.0 && PetscRealPart(a->a[j]) != 0.0) {
646: PetscViewerASCIIPrintf(viewer," (%D, %g - %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)-PetscImaginaryPart(a->a[j]));
647: } else if (PetscRealPart(a->a[j]) != 0.0) {
648: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)PetscRealPart(a->a[j]));
649: }
650: #else
651: if (a->a[j] != 0.0) {PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)a->a[j]);}
652: #endif
653: }
654: PetscViewerASCIIPrintf(viewer,"\n");
655: }
656: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
657: } else if (format == PETSC_VIEWER_ASCII_SYMMODU) {
658: PetscInt nzd=0,fshift=1,*sptr;
659: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
660: PetscMalloc1(m+1,&sptr);
661: for (i=0; i<m; i++) {
662: sptr[i] = nzd+1;
663: for (j=a->i[i]; j<a->i[i+1]; j++) {
664: if (a->j[j] >= i) {
665: #if defined(PETSC_USE_COMPLEX)
666: if (PetscImaginaryPart(a->a[j]) != 0.0 || PetscRealPart(a->a[j]) != 0.0) nzd++;
667: #else
668: if (a->a[j] != 0.0) nzd++;
669: #endif
670: }
671: }
672: }
673: sptr[m] = nzd+1;
674: PetscViewerASCIIPrintf(viewer," %D %D\n\n",m,nzd);
675: for (i=0; i<m+1; i+=6) {
676: if (i+4<m) {
677: PetscViewerASCIIPrintf(viewer," %D %D %D %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3],sptr[i+4],sptr[i+5]);
678: } else if (i+3<m) {
679: PetscViewerASCIIPrintf(viewer," %D %D %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3],sptr[i+4]);
680: } else if (i+2<m) {
681: PetscViewerASCIIPrintf(viewer," %D %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3]);
682: } else if (i+1<m) {
683: PetscViewerASCIIPrintf(viewer," %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2]);
684: } else if (i<m) {
685: PetscViewerASCIIPrintf(viewer," %D %D\n",sptr[i],sptr[i+1]);
686: } else {
687: PetscViewerASCIIPrintf(viewer," %D\n",sptr[i]);
688: }
689: }
690: PetscViewerASCIIPrintf(viewer,"\n");
691: PetscFree(sptr);
692: for (i=0; i<m; i++) {
693: for (j=a->i[i]; j<a->i[i+1]; j++) {
694: if (a->j[j] >= i) {PetscViewerASCIIPrintf(viewer," %D ",a->j[j]+fshift);}
695: }
696: PetscViewerASCIIPrintf(viewer,"\n");
697: }
698: PetscViewerASCIIPrintf(viewer,"\n");
699: for (i=0; i<m; i++) {
700: for (j=a->i[i]; j<a->i[i+1]; j++) {
701: if (a->j[j] >= i) {
702: #if defined(PETSC_USE_COMPLEX)
703: if (PetscImaginaryPart(a->a[j]) != 0.0 || PetscRealPart(a->a[j]) != 0.0) {
704: PetscViewerASCIIPrintf(viewer," %18.16e %18.16e ",(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
705: }
706: #else
707: if (a->a[j] != 0.0) {PetscViewerASCIIPrintf(viewer," %18.16e ",(double)a->a[j]);}
708: #endif
709: }
710: }
711: PetscViewerASCIIPrintf(viewer,"\n");
712: }
713: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
714: } else if (format == PETSC_VIEWER_ASCII_DENSE) {
715: PetscInt cnt = 0,jcnt;
716: PetscScalar value;
717: #if defined(PETSC_USE_COMPLEX)
718: PetscBool realonly = PETSC_TRUE;
720: for (i=0; i<a->i[m]; i++) {
721: if (PetscImaginaryPart(a->a[i]) != 0.0) {
722: realonly = PETSC_FALSE;
723: break;
724: }
725: }
726: #endif
728: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
729: for (i=0; i<m; i++) {
730: jcnt = 0;
731: for (j=0; j<A->cmap->n; j++) {
732: if (jcnt < a->i[i+1]-a->i[i] && j == a->j[cnt]) {
733: value = a->a[cnt++];
734: jcnt++;
735: } else {
736: value = 0.0;
737: }
738: #if defined(PETSC_USE_COMPLEX)
739: if (realonly) {
740: PetscViewerASCIIPrintf(viewer," %7.5e ",(double)PetscRealPart(value));
741: } else {
742: PetscViewerASCIIPrintf(viewer," %7.5e+%7.5e i ",(double)PetscRealPart(value),(double)PetscImaginaryPart(value));
743: }
744: #else
745: PetscViewerASCIIPrintf(viewer," %7.5e ",(double)value);
746: #endif
747: }
748: PetscViewerASCIIPrintf(viewer,"\n");
749: }
750: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
751: } else if (format == PETSC_VIEWER_ASCII_MATRIXMARKET) {
752: PetscInt fshift=1;
753: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
754: #if defined(PETSC_USE_COMPLEX)
755: PetscViewerASCIIPrintf(viewer,"%%%%MatrixMarket matrix coordinate complex general\n");
756: #else
757: PetscViewerASCIIPrintf(viewer,"%%%%MatrixMarket matrix coordinate real general\n");
758: #endif
759: PetscViewerASCIIPrintf(viewer,"%D %D %D\n", m, A->cmap->n, a->nz);
760: for (i=0; i<m; i++) {
761: for (j=a->i[i]; j<a->i[i+1]; j++) {
762: #if defined(PETSC_USE_COMPLEX)
763: PetscViewerASCIIPrintf(viewer,"%D %D %g %g\n", i+fshift,a->j[j]+fshift,(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
764: #else
765: PetscViewerASCIIPrintf(viewer,"%D %D %g\n", i+fshift, a->j[j]+fshift, (double)a->a[j]);
766: #endif
767: }
768: }
769: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
770: } else {
771: PetscViewerASCIIUseTabs(viewer,PETSC_FALSE);
772: if (A->factortype) {
773: for (i=0; i<m; i++) {
774: PetscViewerASCIIPrintf(viewer,"row %D:",i);
775: /* L part */
776: for (j=a->i[i]; j<a->i[i+1]; j++) {
777: #if defined(PETSC_USE_COMPLEX)
778: if (PetscImaginaryPart(a->a[j]) > 0.0) {
779: PetscViewerASCIIPrintf(viewer," (%D, %g + %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
780: } else if (PetscImaginaryPart(a->a[j]) < 0.0) {
781: PetscViewerASCIIPrintf(viewer," (%D, %g - %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)(-PetscImaginaryPart(a->a[j])));
782: } else {
783: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)PetscRealPart(a->a[j]));
784: }
785: #else
786: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)a->a[j]);
787: #endif
788: }
789: /* diagonal */
790: j = a->diag[i];
791: #if defined(PETSC_USE_COMPLEX)
792: if (PetscImaginaryPart(a->a[j]) > 0.0) {
793: PetscViewerASCIIPrintf(viewer," (%D, %g + %g i)",a->j[j],(double)PetscRealPart(1.0/a->a[j]),(double)PetscImaginaryPart(1.0/a->a[j]));
794: } else if (PetscImaginaryPart(a->a[j]) < 0.0) {
795: PetscViewerASCIIPrintf(viewer," (%D, %g - %g i)",a->j[j],(double)PetscRealPart(1.0/a->a[j]),(double)(-PetscImaginaryPart(1.0/a->a[j])));
796: } else {
797: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)PetscRealPart(1.0/a->a[j]));
798: }
799: #else
800: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)(1.0/a->a[j]));
801: #endif
803: /* U part */
804: for (j=a->diag[i+1]+1; j<a->diag[i]; j++) {
805: #if defined(PETSC_USE_COMPLEX)
806: if (PetscImaginaryPart(a->a[j]) > 0.0) {
807: PetscViewerASCIIPrintf(viewer," (%D, %g + %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
808: } else if (PetscImaginaryPart(a->a[j]) < 0.0) {
809: PetscViewerASCIIPrintf(viewer," (%D, %g - %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)(-PetscImaginaryPart(a->a[j])));
810: } else {
811: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)PetscRealPart(a->a[j]));
812: }
813: #else
814: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)a->a[j]);
815: #endif
816: }
817: PetscViewerASCIIPrintf(viewer,"\n");
818: }
819: } else {
820: for (i=0; i<m; i++) {
821: PetscViewerASCIIPrintf(viewer,"row %D:",i);
822: for (j=a->i[i]; j<a->i[i+1]; j++) {
823: #if defined(PETSC_USE_COMPLEX)
824: if (PetscImaginaryPart(a->a[j]) > 0.0) {
825: PetscViewerASCIIPrintf(viewer," (%D, %g + %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)PetscImaginaryPart(a->a[j]));
826: } else if (PetscImaginaryPart(a->a[j]) < 0.0) {
827: PetscViewerASCIIPrintf(viewer," (%D, %g - %g i)",a->j[j],(double)PetscRealPart(a->a[j]),(double)-PetscImaginaryPart(a->a[j]));
828: } else {
829: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)PetscRealPart(a->a[j]));
830: }
831: #else
832: PetscViewerASCIIPrintf(viewer," (%D, %g) ",a->j[j],(double)a->a[j]);
833: #endif
834: }
835: PetscViewerASCIIPrintf(viewer,"\n");
836: }
837: }
838: PetscViewerASCIIUseTabs(viewer,PETSC_TRUE);
839: }
840: PetscViewerFlush(viewer);
841: return(0);
842: }
844: #include <petscdraw.h>
847: PetscErrorCode MatView_SeqAIJ_Draw_Zoom(PetscDraw draw,void *Aa)
848: {
849: Mat A = (Mat) Aa;
850: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
851: PetscErrorCode ierr;
852: PetscInt i,j,m = A->rmap->n;
853: int color;
854: PetscReal xl,yl,xr,yr,x_l,x_r,y_l,y_r;
855: PetscViewer viewer;
856: PetscViewerFormat format;
859: PetscObjectQuery((PetscObject)A,"Zoomviewer",(PetscObject*)&viewer);
860: PetscViewerGetFormat(viewer,&format);
861: PetscDrawGetCoordinates(draw,&xl,&yl,&xr,&yr);
863: /* loop over matrix elements drawing boxes */
865: if (format != PETSC_VIEWER_DRAW_CONTOUR) {
866: PetscDrawCollectiveBegin(draw);
867: /* Blue for negative, Cyan for zero and Red for positive */
868: color = PETSC_DRAW_BLUE;
869: for (i=0; i<m; i++) {
870: y_l = m - i - 1.0; y_r = y_l + 1.0;
871: for (j=a->i[i]; j<a->i[i+1]; j++) {
872: x_l = a->j[j]; x_r = x_l + 1.0;
873: if (PetscRealPart(a->a[j]) >= 0.) continue;
874: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
875: }
876: }
877: color = PETSC_DRAW_CYAN;
878: for (i=0; i<m; i++) {
879: y_l = m - i - 1.0; y_r = y_l + 1.0;
880: for (j=a->i[i]; j<a->i[i+1]; j++) {
881: x_l = a->j[j]; x_r = x_l + 1.0;
882: if (a->a[j] != 0.) continue;
883: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
884: }
885: }
886: color = PETSC_DRAW_RED;
887: for (i=0; i<m; i++) {
888: y_l = m - i - 1.0; y_r = y_l + 1.0;
889: for (j=a->i[i]; j<a->i[i+1]; j++) {
890: x_l = a->j[j]; x_r = x_l + 1.0;
891: if (PetscRealPart(a->a[j]) <= 0.) continue;
892: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
893: }
894: }
895: PetscDrawCollectiveEnd(draw);
896: } else {
897: /* use contour shading to indicate magnitude of values */
898: /* first determine max of all nonzero values */
899: PetscReal minv = 0.0, maxv = 0.0;
900: PetscInt nz = a->nz, count = 0;
901: PetscDraw popup;
903: for (i=0; i<nz; i++) {
904: if (PetscAbsScalar(a->a[i]) > maxv) maxv = PetscAbsScalar(a->a[i]);
905: }
906: if (minv >= maxv) maxv = minv + PETSC_SMALL;
907: PetscDrawGetPopup(draw,&popup);
908: PetscDrawScalePopup(popup,minv,maxv);
910: PetscDrawCollectiveBegin(draw);
911: for (i=0; i<m; i++) {
912: y_l = m - i - 1.0;
913: y_r = y_l + 1.0;
914: for (j=a->i[i]; j<a->i[i+1]; j++) {
915: x_l = a->j[j];
916: x_r = x_l + 1.0;
917: color = PetscDrawRealToColor(PetscAbsScalar(a->a[count]),minv,maxv);
918: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
919: count++;
920: }
921: }
922: PetscDrawCollectiveEnd(draw);
923: }
924: return(0);
925: }
927: #include <petscdraw.h>
930: PetscErrorCode MatView_SeqAIJ_Draw(Mat A,PetscViewer viewer)
931: {
933: PetscDraw draw;
934: PetscReal xr,yr,xl,yl,h,w;
935: PetscBool isnull;
938: PetscViewerDrawGetDraw(viewer,0,&draw);
939: PetscDrawIsNull(draw,&isnull);
940: if (isnull) return(0);
942: xr = A->cmap->n; yr = A->rmap->n; h = yr/10.0; w = xr/10.0;
943: xr += w; yr += h; xl = -w; yl = -h;
944: PetscDrawSetCoordinates(draw,xl,yl,xr,yr);
945: PetscObjectCompose((PetscObject)A,"Zoomviewer",(PetscObject)viewer);
946: PetscDrawZoom(draw,MatView_SeqAIJ_Draw_Zoom,A);
947: PetscObjectCompose((PetscObject)A,"Zoomviewer",NULL);
948: PetscDrawSave(draw);
949: return(0);
950: }
954: PetscErrorCode MatView_SeqAIJ(Mat A,PetscViewer viewer)
955: {
957: PetscBool iascii,isbinary,isdraw;
960: PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);
961: PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERBINARY,&isbinary);
962: PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERDRAW,&isdraw);
963: if (iascii) {
964: MatView_SeqAIJ_ASCII(A,viewer);
965: } else if (isbinary) {
966: MatView_SeqAIJ_Binary(A,viewer);
967: } else if (isdraw) {
968: MatView_SeqAIJ_Draw(A,viewer);
969: }
970: MatView_SeqAIJ_Inode(A,viewer);
971: return(0);
972: }
976: PetscErrorCode MatAssemblyEnd_SeqAIJ(Mat A,MatAssemblyType mode)
977: {
978: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
980: PetscInt fshift = 0,i,j,*ai = a->i,*aj = a->j,*imax = a->imax;
981: PetscInt m = A->rmap->n,*ip,N,*ailen = a->ilen,rmax = 0;
982: MatScalar *aa = a->a,*ap;
983: PetscReal ratio = 0.6;
986: if (mode == MAT_FLUSH_ASSEMBLY) return(0);
988: if (m) rmax = ailen[0]; /* determine row with most nonzeros */
989: for (i=1; i<m; i++) {
990: /* move each row back by the amount of empty slots (fshift) before it*/
991: fshift += imax[i-1] - ailen[i-1];
992: rmax = PetscMax(rmax,ailen[i]);
993: if (fshift) {
994: ip = aj + ai[i];
995: ap = aa + ai[i];
996: N = ailen[i];
997: for (j=0; j<N; j++) {
998: ip[j-fshift] = ip[j];
999: ap[j-fshift] = ap[j];
1000: }
1001: }
1002: ai[i] = ai[i-1] + ailen[i-1];
1003: }
1004: if (m) {
1005: fshift += imax[m-1] - ailen[m-1];
1006: ai[m] = ai[m-1] + ailen[m-1];
1007: }
1009: /* reset ilen and imax for each row */
1010: a->nonzerorowcnt = 0;
1011: for (i=0; i<m; i++) {
1012: ailen[i] = imax[i] = ai[i+1] - ai[i];
1013: a->nonzerorowcnt += ((ai[i+1] - ai[i]) > 0);
1014: }
1015: a->nz = ai[m];
1016: if (fshift && a->nounused == -1) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB, "Unused space detected in matrix: %D X %D, %D unneeded", m, A->cmap->n, fshift);
1018: MatMarkDiagonal_SeqAIJ(A);
1019: PetscInfo4(A,"Matrix size: %D X %D; storage space: %D unneeded,%D used\n",m,A->cmap->n,fshift,a->nz);
1020: PetscInfo1(A,"Number of mallocs during MatSetValues() is %D\n",a->reallocs);
1021: PetscInfo1(A,"Maximum nonzeros in any row is %D\n",rmax);
1023: A->info.mallocs += a->reallocs;
1024: a->reallocs = 0;
1025: A->info.nz_unneeded = (PetscReal)fshift;
1026: a->rmax = rmax;
1028: MatCheckCompressedRow(A,a->nonzerorowcnt,&a->compressedrow,a->i,m,ratio);
1029: MatAssemblyEnd_SeqAIJ_Inode(A,mode);
1030: MatSeqAIJInvalidateDiagonal(A);
1031: return(0);
1032: }
1036: PetscErrorCode MatRealPart_SeqAIJ(Mat A)
1037: {
1038: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1039: PetscInt i,nz = a->nz;
1040: MatScalar *aa = a->a;
1044: for (i=0; i<nz; i++) aa[i] = PetscRealPart(aa[i]);
1045: MatSeqAIJInvalidateDiagonal(A);
1046: return(0);
1047: }
1051: PetscErrorCode MatImaginaryPart_SeqAIJ(Mat A)
1052: {
1053: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1054: PetscInt i,nz = a->nz;
1055: MatScalar *aa = a->a;
1059: for (i=0; i<nz; i++) aa[i] = PetscImaginaryPart(aa[i]);
1060: MatSeqAIJInvalidateDiagonal(A);
1061: return(0);
1062: }
1066: PetscErrorCode MatZeroEntries_SeqAIJ(Mat A)
1067: {
1068: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1072: PetscMemzero(a->a,(a->i[A->rmap->n])*sizeof(PetscScalar));
1073: MatSeqAIJInvalidateDiagonal(A);
1074: return(0);
1075: }
1079: PetscErrorCode MatDestroy_SeqAIJ(Mat A)
1080: {
1081: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1085: #if defined(PETSC_USE_LOG)
1086: PetscLogObjectState((PetscObject)A,"Rows=%D, Cols=%D, NZ=%D",A->rmap->n,A->cmap->n,a->nz);
1087: #endif
1088: MatSeqXAIJFreeAIJ(A,&a->a,&a->j,&a->i);
1089: ISDestroy(&a->row);
1090: ISDestroy(&a->col);
1091: PetscFree(a->diag);
1092: PetscFree(a->ibdiag);
1093: PetscFree2(a->imax,a->ilen);
1094: PetscFree3(a->idiag,a->mdiag,a->ssor_work);
1095: PetscFree(a->solve_work);
1096: ISDestroy(&a->icol);
1097: PetscFree(a->saved_values);
1098: ISColoringDestroy(&a->coloring);
1099: PetscFree2(a->compressedrow.i,a->compressedrow.rindex);
1100: PetscFree(a->matmult_abdense);
1102: MatDestroy_SeqAIJ_Inode(A);
1103: PetscFree(A->data);
1105: PetscObjectChangeTypeName((PetscObject)A,0);
1106: PetscObjectComposeFunction((PetscObject)A,"MatSeqAIJSetColumnIndices_C",NULL);
1107: PetscObjectComposeFunction((PetscObject)A,"MatStoreValues_C",NULL);
1108: PetscObjectComposeFunction((PetscObject)A,"MatRetrieveValues_C",NULL);
1109: PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqaij_seqsbaij_C",NULL);
1110: PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqaij_seqbaij_C",NULL);
1111: PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqaij_seqaijperm_C",NULL);
1112: #if defined(PETSC_HAVE_ELEMENTAL)
1113: PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqaij_elemental_C",NULL);
1114: #endif
1115: PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqaij_seqdense_C",NULL);
1116: PetscObjectComposeFunction((PetscObject)A,"MatIsTranspose_C",NULL);
1117: PetscObjectComposeFunction((PetscObject)A,"MatSeqAIJSetPreallocation_C",NULL);
1118: PetscObjectComposeFunction((PetscObject)A,"MatSeqAIJSetPreallocationCSR_C",NULL);
1119: PetscObjectComposeFunction((PetscObject)A,"MatReorderForNonzeroDiagonal_C",NULL);
1120: return(0);
1121: }
1125: PetscErrorCode MatSetOption_SeqAIJ(Mat A,MatOption op,PetscBool flg)
1126: {
1127: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1131: switch (op) {
1132: case MAT_ROW_ORIENTED:
1133: a->roworiented = flg;
1134: break;
1135: case MAT_KEEP_NONZERO_PATTERN:
1136: a->keepnonzeropattern = flg;
1137: break;
1138: case MAT_NEW_NONZERO_LOCATIONS:
1139: a->nonew = (flg ? 0 : 1);
1140: break;
1141: case MAT_NEW_NONZERO_LOCATION_ERR:
1142: a->nonew = (flg ? -1 : 0);
1143: break;
1144: case MAT_NEW_NONZERO_ALLOCATION_ERR:
1145: a->nonew = (flg ? -2 : 0);
1146: break;
1147: case MAT_UNUSED_NONZERO_LOCATION_ERR:
1148: a->nounused = (flg ? -1 : 0);
1149: break;
1150: case MAT_IGNORE_ZERO_ENTRIES:
1151: a->ignorezeroentries = flg;
1152: break;
1153: case MAT_SPD:
1154: case MAT_SYMMETRIC:
1155: case MAT_STRUCTURALLY_SYMMETRIC:
1156: case MAT_HERMITIAN:
1157: case MAT_SYMMETRY_ETERNAL:
1158: /* These options are handled directly by MatSetOption() */
1159: break;
1160: case MAT_NEW_DIAGONALS:
1161: case MAT_IGNORE_OFF_PROC_ENTRIES:
1162: case MAT_USE_HASH_TABLE:
1163: PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);
1164: break;
1165: case MAT_USE_INODES:
1166: /* Not an error because MatSetOption_SeqAIJ_Inode handles this one */
1167: break;
1168: default:
1169: SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"unknown option %d",op);
1170: }
1171: MatSetOption_SeqAIJ_Inode(A,op,flg);
1172: return(0);
1173: }
1177: PetscErrorCode MatGetDiagonal_SeqAIJ(Mat A,Vec v)
1178: {
1179: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1181: PetscInt i,j,n,*ai=a->i,*aj=a->j,nz;
1182: PetscScalar *aa=a->a,*x,zero=0.0;
1185: VecGetLocalSize(v,&n);
1186: if (n != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
1188: if (A->factortype == MAT_FACTOR_ILU || A->factortype == MAT_FACTOR_LU) {
1189: PetscInt *diag=a->diag;
1190: VecGetArray(v,&x);
1191: for (i=0; i<n; i++) x[i] = 1.0/aa[diag[i]];
1192: VecRestoreArray(v,&x);
1193: return(0);
1194: }
1196: VecSet(v,zero);
1197: VecGetArray(v,&x);
1198: for (i=0; i<n; i++) {
1199: nz = ai[i+1] - ai[i];
1200: if (!nz) x[i] = 0.0;
1201: for (j=ai[i]; j<ai[i+1]; j++) {
1202: if (aj[j] == i) {
1203: x[i] = aa[j];
1204: break;
1205: }
1206: }
1207: }
1208: VecRestoreArray(v,&x);
1209: return(0);
1210: }
1212: #include <../src/mat/impls/aij/seq/ftn-kernels/fmult.h>
1215: PetscErrorCode MatMultTransposeAdd_SeqAIJ(Mat A,Vec xx,Vec zz,Vec yy)
1216: {
1217: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1218: PetscScalar *y;
1219: const PetscScalar *x;
1220: PetscErrorCode ierr;
1221: PetscInt m = A->rmap->n;
1222: #if !defined(PETSC_USE_FORTRAN_KERNEL_MULTTRANSPOSEAIJ)
1223: const MatScalar *v;
1224: PetscScalar alpha;
1225: PetscInt n,i,j;
1226: const PetscInt *idx,*ii,*ridx=NULL;
1227: Mat_CompressedRow cprow = a->compressedrow;
1228: PetscBool usecprow = cprow.use;
1229: #endif
1232: if (zz != yy) {VecCopy(zz,yy);}
1233: VecGetArrayRead(xx,&x);
1234: VecGetArray(yy,&y);
1236: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTTRANSPOSEAIJ)
1237: fortranmulttransposeaddaij_(&m,x,a->i,a->j,a->a,y);
1238: #else
1239: if (usecprow) {
1240: m = cprow.nrows;
1241: ii = cprow.i;
1242: ridx = cprow.rindex;
1243: } else {
1244: ii = a->i;
1245: }
1246: for (i=0; i<m; i++) {
1247: idx = a->j + ii[i];
1248: v = a->a + ii[i];
1249: n = ii[i+1] - ii[i];
1250: if (usecprow) {
1251: alpha = x[ridx[i]];
1252: } else {
1253: alpha = x[i];
1254: }
1255: for (j=0; j<n; j++) y[idx[j]] += alpha*v[j];
1256: }
1257: #endif
1258: PetscLogFlops(2.0*a->nz);
1259: VecRestoreArrayRead(xx,&x);
1260: VecRestoreArray(yy,&y);
1261: return(0);
1262: }
1266: PetscErrorCode MatMultTranspose_SeqAIJ(Mat A,Vec xx,Vec yy)
1267: {
1271: VecSet(yy,0.0);
1272: MatMultTransposeAdd_SeqAIJ(A,xx,yy,yy);
1273: return(0);
1274: }
1276: #include <../src/mat/impls/aij/seq/ftn-kernels/fmult.h>
1280: PetscErrorCode MatMult_SeqAIJ(Mat A,Vec xx,Vec yy)
1281: {
1282: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1283: PetscScalar *y;
1284: const PetscScalar *x;
1285: const MatScalar *aa;
1286: PetscErrorCode ierr;
1287: PetscInt m=A->rmap->n;
1288: const PetscInt *aj,*ii,*ridx=NULL;
1289: PetscInt n,i;
1290: PetscScalar sum;
1291: PetscBool usecprow=a->compressedrow.use;
1293: #if defined(PETSC_HAVE_PRAGMA_DISJOINT)
1294: #pragma disjoint(*x,*y,*aa)
1295: #endif
1298: VecGetArrayRead(xx,&x);
1299: VecGetArray(yy,&y);
1300: ii = a->i;
1301: if (usecprow) { /* use compressed row format */
1302: PetscMemzero(y,m*sizeof(PetscScalar));
1303: m = a->compressedrow.nrows;
1304: ii = a->compressedrow.i;
1305: ridx = a->compressedrow.rindex;
1306: for (i=0; i<m; i++) {
1307: n = ii[i+1] - ii[i];
1308: aj = a->j + ii[i];
1309: aa = a->a + ii[i];
1310: sum = 0.0;
1311: PetscSparseDensePlusDot(sum,x,aa,aj,n);
1312: /* for (j=0; j<n; j++) sum += (*aa++)*x[*aj++]; */
1313: y[*ridx++] = sum;
1314: }
1315: } else { /* do not use compressed row format */
1316: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTAIJ)
1317: aj = a->j;
1318: aa = a->a;
1319: fortranmultaij_(&m,x,ii,aj,aa,y);
1320: #else
1321: for (i=0; i<m; i++) {
1322: n = ii[i+1] - ii[i];
1323: aj = a->j + ii[i];
1324: aa = a->a + ii[i];
1325: sum = 0.0;
1326: PetscSparseDensePlusDot(sum,x,aa,aj,n);
1327: y[i] = sum;
1328: }
1329: #endif
1330: }
1331: PetscLogFlops(2.0*a->nz - a->nonzerorowcnt);
1332: VecRestoreArrayRead(xx,&x);
1333: VecRestoreArray(yy,&y);
1334: return(0);
1335: }
1339: PetscErrorCode MatMultMax_SeqAIJ(Mat A,Vec xx,Vec yy)
1340: {
1341: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1342: PetscScalar *y;
1343: const PetscScalar *x;
1344: const MatScalar *aa;
1345: PetscErrorCode ierr;
1346: PetscInt m=A->rmap->n;
1347: const PetscInt *aj,*ii,*ridx=NULL;
1348: PetscInt n,i,nonzerorow=0;
1349: PetscScalar sum;
1350: PetscBool usecprow=a->compressedrow.use;
1352: #if defined(PETSC_HAVE_PRAGMA_DISJOINT)
1353: #pragma disjoint(*x,*y,*aa)
1354: #endif
1357: VecGetArrayRead(xx,&x);
1358: VecGetArray(yy,&y);
1359: if (usecprow) { /* use compressed row format */
1360: m = a->compressedrow.nrows;
1361: ii = a->compressedrow.i;
1362: ridx = a->compressedrow.rindex;
1363: for (i=0; i<m; i++) {
1364: n = ii[i+1] - ii[i];
1365: aj = a->j + ii[i];
1366: aa = a->a + ii[i];
1367: sum = 0.0;
1368: nonzerorow += (n>0);
1369: PetscSparseDenseMaxDot(sum,x,aa,aj,n);
1370: /* for (j=0; j<n; j++) sum += (*aa++)*x[*aj++]; */
1371: y[*ridx++] = sum;
1372: }
1373: } else { /* do not use compressed row format */
1374: ii = a->i;
1375: for (i=0; i<m; i++) {
1376: n = ii[i+1] - ii[i];
1377: aj = a->j + ii[i];
1378: aa = a->a + ii[i];
1379: sum = 0.0;
1380: nonzerorow += (n>0);
1381: PetscSparseDenseMaxDot(sum,x,aa,aj,n);
1382: y[i] = sum;
1383: }
1384: }
1385: PetscLogFlops(2.0*a->nz - nonzerorow);
1386: VecRestoreArrayRead(xx,&x);
1387: VecRestoreArray(yy,&y);
1388: return(0);
1389: }
1393: PetscErrorCode MatMultAddMax_SeqAIJ(Mat A,Vec xx,Vec yy,Vec zz)
1394: {
1395: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1396: PetscScalar *y,*z;
1397: const PetscScalar *x;
1398: const MatScalar *aa;
1399: PetscErrorCode ierr;
1400: PetscInt m = A->rmap->n,*aj,*ii;
1401: PetscInt n,i,*ridx=NULL;
1402: PetscScalar sum;
1403: PetscBool usecprow=a->compressedrow.use;
1406: VecGetArrayRead(xx,&x);
1407: VecGetArrayPair(yy,zz,&y,&z);
1408: if (usecprow) { /* use compressed row format */
1409: if (zz != yy) {
1410: PetscMemcpy(z,y,m*sizeof(PetscScalar));
1411: }
1412: m = a->compressedrow.nrows;
1413: ii = a->compressedrow.i;
1414: ridx = a->compressedrow.rindex;
1415: for (i=0; i<m; i++) {
1416: n = ii[i+1] - ii[i];
1417: aj = a->j + ii[i];
1418: aa = a->a + ii[i];
1419: sum = y[*ridx];
1420: PetscSparseDenseMaxDot(sum,x,aa,aj,n);
1421: z[*ridx++] = sum;
1422: }
1423: } else { /* do not use compressed row format */
1424: ii = a->i;
1425: for (i=0; i<m; i++) {
1426: n = ii[i+1] - ii[i];
1427: aj = a->j + ii[i];
1428: aa = a->a + ii[i];
1429: sum = y[i];
1430: PetscSparseDenseMaxDot(sum,x,aa,aj,n);
1431: z[i] = sum;
1432: }
1433: }
1434: PetscLogFlops(2.0*a->nz);
1435: VecRestoreArrayRead(xx,&x);
1436: VecRestoreArrayPair(yy,zz,&y,&z);
1437: return(0);
1438: }
1440: #include <../src/mat/impls/aij/seq/ftn-kernels/fmultadd.h>
1443: PetscErrorCode MatMultAdd_SeqAIJ(Mat A,Vec xx,Vec yy,Vec zz)
1444: {
1445: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1446: PetscScalar *y,*z;
1447: const PetscScalar *x;
1448: const MatScalar *aa;
1449: PetscErrorCode ierr;
1450: const PetscInt *aj,*ii,*ridx=NULL;
1451: PetscInt m = A->rmap->n,n,i;
1452: PetscScalar sum;
1453: PetscBool usecprow=a->compressedrow.use;
1456: VecGetArrayRead(xx,&x);
1457: VecGetArrayPair(yy,zz,&y,&z);
1458: if (usecprow) { /* use compressed row format */
1459: if (zz != yy) {
1460: PetscMemcpy(z,y,m*sizeof(PetscScalar));
1461: }
1462: m = a->compressedrow.nrows;
1463: ii = a->compressedrow.i;
1464: ridx = a->compressedrow.rindex;
1465: for (i=0; i<m; i++) {
1466: n = ii[i+1] - ii[i];
1467: aj = a->j + ii[i];
1468: aa = a->a + ii[i];
1469: sum = y[*ridx];
1470: PetscSparseDensePlusDot(sum,x,aa,aj,n);
1471: z[*ridx++] = sum;
1472: }
1473: } else { /* do not use compressed row format */
1474: ii = a->i;
1475: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTADDAIJ)
1476: aj = a->j;
1477: aa = a->a;
1478: fortranmultaddaij_(&m,x,ii,aj,aa,y,z);
1479: #else
1480: for (i=0; i<m; i++) {
1481: n = ii[i+1] - ii[i];
1482: aj = a->j + ii[i];
1483: aa = a->a + ii[i];
1484: sum = y[i];
1485: PetscSparseDensePlusDot(sum,x,aa,aj,n);
1486: z[i] = sum;
1487: }
1488: #endif
1489: }
1490: PetscLogFlops(2.0*a->nz);
1491: VecRestoreArrayRead(xx,&x);
1492: VecRestoreArrayPair(yy,zz,&y,&z);
1493: #if defined(PETSC_HAVE_CUSP)
1494: /*
1495: VecView(xx,0);
1496: VecView(zz,0);
1497: MatView(A,0);
1498: */
1499: #endif
1500: return(0);
1501: }
1503: /*
1504: Adds diagonal pointers to sparse matrix structure.
1505: */
1508: PetscErrorCode MatMarkDiagonal_SeqAIJ(Mat A)
1509: {
1510: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1512: PetscInt i,j,m = A->rmap->n;
1515: if (!a->diag) {
1516: PetscMalloc1(m,&a->diag);
1517: PetscLogObjectMemory((PetscObject)A, m*sizeof(PetscInt));
1518: }
1519: for (i=0; i<A->rmap->n; i++) {
1520: a->diag[i] = a->i[i+1];
1521: for (j=a->i[i]; j<a->i[i+1]; j++) {
1522: if (a->j[j] == i) {
1523: a->diag[i] = j;
1524: break;
1525: }
1526: }
1527: }
1528: return(0);
1529: }
1531: /*
1532: Checks for missing diagonals
1533: */
1536: PetscErrorCode MatMissingDiagonal_SeqAIJ(Mat A,PetscBool *missing,PetscInt *d)
1537: {
1538: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1539: PetscInt *diag,*ii = a->i,i;
1542: *missing = PETSC_FALSE;
1543: if (A->rmap->n > 0 && !ii) {
1544: *missing = PETSC_TRUE;
1545: if (d) *d = 0;
1546: PetscInfo(A,"Matrix has no entries therefore is missing diagonal\n");
1547: } else {
1548: diag = a->diag;
1549: for (i=0; i<A->rmap->n; i++) {
1550: if (diag[i] >= ii[i+1]) {
1551: *missing = PETSC_TRUE;
1552: if (d) *d = i;
1553: PetscInfo1(A,"Matrix is missing diagonal number %D\n",i);
1554: break;
1555: }
1556: }
1557: }
1558: return(0);
1559: }
1563: /*
1564: Negative shift indicates do not generate an error if there is a zero diagonal, just invert it anyways
1565: */
1566: PetscErrorCode MatInvertDiagonal_SeqAIJ(Mat A,PetscScalar omega,PetscScalar fshift)
1567: {
1568: Mat_SeqAIJ *a = (Mat_SeqAIJ*) A->data;
1570: PetscInt i,*diag,m = A->rmap->n;
1571: MatScalar *v = a->a;
1572: PetscScalar *idiag,*mdiag;
1575: if (a->idiagvalid) return(0);
1576: MatMarkDiagonal_SeqAIJ(A);
1577: diag = a->diag;
1578: if (!a->idiag) {
1579: PetscMalloc3(m,&a->idiag,m,&a->mdiag,m,&a->ssor_work);
1580: PetscLogObjectMemory((PetscObject)A, 3*m*sizeof(PetscScalar));
1581: v = a->a;
1582: }
1583: mdiag = a->mdiag;
1584: idiag = a->idiag;
1586: if (omega == 1.0 && PetscRealPart(fshift) <= 0.0) {
1587: for (i=0; i<m; i++) {
1588: mdiag[i] = v[diag[i]];
1589: if (!PetscAbsScalar(mdiag[i])) { /* zero diagonal */
1590: if (PetscRealPart(fshift)) {
1591: PetscInfo1(A,"Zero diagonal on row %D\n",i);
1592: A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
1593: } else {
1594: SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_INCOMP,"Zero diagonal on row %D",i);
1595: }
1596: }
1597: idiag[i] = 1.0/v[diag[i]];
1598: }
1599: PetscLogFlops(m);
1600: } else {
1601: for (i=0; i<m; i++) {
1602: mdiag[i] = v[diag[i]];
1603: idiag[i] = omega/(fshift + v[diag[i]]);
1604: }
1605: PetscLogFlops(2.0*m);
1606: }
1607: a->idiagvalid = PETSC_TRUE;
1608: return(0);
1609: }
1611: #include <../src/mat/impls/aij/seq/ftn-kernels/frelax.h>
1614: PetscErrorCode MatSOR_SeqAIJ(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
1615: {
1616: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1617: PetscScalar *x,d,sum,*t,scale;
1618: const MatScalar *v,*idiag=0,*mdiag;
1619: const PetscScalar *b, *bs,*xb, *ts;
1620: PetscErrorCode ierr;
1621: PetscInt n,m = A->rmap->n,i;
1622: const PetscInt *idx,*diag;
1625: its = its*lits;
1627: if (fshift != a->fshift || omega != a->omega) a->idiagvalid = PETSC_FALSE; /* must recompute idiag[] */
1628: if (!a->idiagvalid) {MatInvertDiagonal_SeqAIJ(A,omega,fshift);}
1629: a->fshift = fshift;
1630: a->omega = omega;
1632: diag = a->diag;
1633: t = a->ssor_work;
1634: idiag = a->idiag;
1635: mdiag = a->mdiag;
1637: VecGetArray(xx,&x);
1638: VecGetArrayRead(bb,&b);
1639: /* We count flops by assuming the upper triangular and lower triangular parts have the same number of nonzeros */
1640: if (flag == SOR_APPLY_UPPER) {
1641: /* apply (U + D/omega) to the vector */
1642: bs = b;
1643: for (i=0; i<m; i++) {
1644: d = fshift + mdiag[i];
1645: n = a->i[i+1] - diag[i] - 1;
1646: idx = a->j + diag[i] + 1;
1647: v = a->a + diag[i] + 1;
1648: sum = b[i]*d/omega;
1649: PetscSparseDensePlusDot(sum,bs,v,idx,n);
1650: x[i] = sum;
1651: }
1652: VecRestoreArray(xx,&x);
1653: VecRestoreArrayRead(bb,&b);
1654: PetscLogFlops(a->nz);
1655: return(0);
1656: }
1658: if (flag == SOR_APPLY_LOWER) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"SOR_APPLY_LOWER is not implemented");
1659: else if (flag & SOR_EISENSTAT) {
1660: /* Let A = L + U + D; where L is lower trianglar,
1661: U is upper triangular, E = D/omega; This routine applies
1663: (L + E)^{-1} A (U + E)^{-1}
1665: to a vector efficiently using Eisenstat's trick.
1666: */
1667: scale = (2.0/omega) - 1.0;
1669: /* x = (E + U)^{-1} b */
1670: for (i=m-1; i>=0; i--) {
1671: n = a->i[i+1] - diag[i] - 1;
1672: idx = a->j + diag[i] + 1;
1673: v = a->a + diag[i] + 1;
1674: sum = b[i];
1675: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1676: x[i] = sum*idiag[i];
1677: }
1679: /* t = b - (2*E - D)x */
1680: v = a->a;
1681: for (i=0; i<m; i++) t[i] = b[i] - scale*(v[*diag++])*x[i];
1683: /* t = (E + L)^{-1}t */
1684: ts = t;
1685: diag = a->diag;
1686: for (i=0; i<m; i++) {
1687: n = diag[i] - a->i[i];
1688: idx = a->j + a->i[i];
1689: v = a->a + a->i[i];
1690: sum = t[i];
1691: PetscSparseDenseMinusDot(sum,ts,v,idx,n);
1692: t[i] = sum*idiag[i];
1693: /* x = x + t */
1694: x[i] += t[i];
1695: }
1697: PetscLogFlops(6.0*m-1 + 2.0*a->nz);
1698: VecRestoreArray(xx,&x);
1699: VecRestoreArrayRead(bb,&b);
1700: return(0);
1701: }
1702: if (flag & SOR_ZERO_INITIAL_GUESS) {
1703: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) {
1704: for (i=0; i<m; i++) {
1705: n = diag[i] - a->i[i];
1706: idx = a->j + a->i[i];
1707: v = a->a + a->i[i];
1708: sum = b[i];
1709: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1710: t[i] = sum;
1711: x[i] = sum*idiag[i];
1712: }
1713: xb = t;
1714: PetscLogFlops(a->nz);
1715: } else xb = b;
1716: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) {
1717: for (i=m-1; i>=0; i--) {
1718: n = a->i[i+1] - diag[i] - 1;
1719: idx = a->j + diag[i] + 1;
1720: v = a->a + diag[i] + 1;
1721: sum = xb[i];
1722: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1723: if (xb == b) {
1724: x[i] = sum*idiag[i];
1725: } else {
1726: x[i] = (1-omega)*x[i] + sum*idiag[i]; /* omega in idiag */
1727: }
1728: }
1729: PetscLogFlops(a->nz); /* assumes 1/2 in upper */
1730: }
1731: its--;
1732: }
1733: while (its--) {
1734: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) {
1735: for (i=0; i<m; i++) {
1736: /* lower */
1737: n = diag[i] - a->i[i];
1738: idx = a->j + a->i[i];
1739: v = a->a + a->i[i];
1740: sum = b[i];
1741: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1742: t[i] = sum; /* save application of the lower-triangular part */
1743: /* upper */
1744: n = a->i[i+1] - diag[i] - 1;
1745: idx = a->j + diag[i] + 1;
1746: v = a->a + diag[i] + 1;
1747: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1748: x[i] = (1. - omega)*x[i] + sum*idiag[i]; /* omega in idiag */
1749: }
1750: xb = t;
1751: PetscLogFlops(2.0*a->nz);
1752: } else xb = b;
1753: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) {
1754: for (i=m-1; i>=0; i--) {
1755: sum = xb[i];
1756: if (xb == b) {
1757: /* whole matrix (no checkpointing available) */
1758: n = a->i[i+1] - a->i[i];
1759: idx = a->j + a->i[i];
1760: v = a->a + a->i[i];
1761: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1762: x[i] = (1. - omega)*x[i] + (sum + mdiag[i]*x[i])*idiag[i];
1763: } else { /* lower-triangular part has been saved, so only apply upper-triangular */
1764: n = a->i[i+1] - diag[i] - 1;
1765: idx = a->j + diag[i] + 1;
1766: v = a->a + diag[i] + 1;
1767: PetscSparseDenseMinusDot(sum,x,v,idx,n);
1768: x[i] = (1. - omega)*x[i] + sum*idiag[i]; /* omega in idiag */
1769: }
1770: }
1771: if (xb == b) {
1772: PetscLogFlops(2.0*a->nz);
1773: } else {
1774: PetscLogFlops(a->nz); /* assumes 1/2 in upper */
1775: }
1776: }
1777: }
1778: VecRestoreArray(xx,&x);
1779: VecRestoreArrayRead(bb,&b);
1780: return(0);
1781: }
1786: PetscErrorCode MatGetInfo_SeqAIJ(Mat A,MatInfoType flag,MatInfo *info)
1787: {
1788: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1791: info->block_size = 1.0;
1792: info->nz_allocated = (double)a->maxnz;
1793: info->nz_used = (double)a->nz;
1794: info->nz_unneeded = (double)(a->maxnz - a->nz);
1795: info->assemblies = (double)A->num_ass;
1796: info->mallocs = (double)A->info.mallocs;
1797: info->memory = ((PetscObject)A)->mem;
1798: if (A->factortype) {
1799: info->fill_ratio_given = A->info.fill_ratio_given;
1800: info->fill_ratio_needed = A->info.fill_ratio_needed;
1801: info->factor_mallocs = A->info.factor_mallocs;
1802: } else {
1803: info->fill_ratio_given = 0;
1804: info->fill_ratio_needed = 0;
1805: info->factor_mallocs = 0;
1806: }
1807: return(0);
1808: }
1812: PetscErrorCode MatZeroRows_SeqAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag,Vec x,Vec b)
1813: {
1814: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1815: PetscInt i,m = A->rmap->n - 1;
1816: PetscErrorCode ierr;
1817: const PetscScalar *xx;
1818: PetscScalar *bb;
1819: #if defined(PETSC_USE_DEBUG)
1820: PetscInt d = 0;
1821: #endif
1824: if (x && b) {
1825: VecGetArrayRead(x,&xx);
1826: VecGetArray(b,&bb);
1827: for (i=0; i<N; i++) {
1828: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1829: bb[rows[i]] = diag*xx[rows[i]];
1830: }
1831: VecRestoreArrayRead(x,&xx);
1832: VecRestoreArray(b,&bb);
1833: }
1835: if (a->keepnonzeropattern) {
1836: for (i=0; i<N; i++) {
1837: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1838: PetscMemzero(&a->a[a->i[rows[i]]],a->ilen[rows[i]]*sizeof(PetscScalar));
1839: }
1840: if (diag != 0.0) {
1841: #if defined(PETSC_USE_DEBUG)
1842: for (i=0; i<N; i++) {
1843: d = rows[i];
1844: if (a->diag[d] >= a->i[d+1]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry in the zeroed row %D",d);
1845: }
1846: #endif
1847: for (i=0; i<N; i++) {
1848: a->a[a->diag[rows[i]]] = diag;
1849: }
1850: }
1851: } else {
1852: if (diag != 0.0) {
1853: for (i=0; i<N; i++) {
1854: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1855: if (a->ilen[rows[i]] > 0) {
1856: a->ilen[rows[i]] = 1;
1857: a->a[a->i[rows[i]]] = diag;
1858: a->j[a->i[rows[i]]] = rows[i];
1859: } else { /* in case row was completely empty */
1860: MatSetValues_SeqAIJ(A,1,&rows[i],1,&rows[i],&diag,INSERT_VALUES);
1861: }
1862: }
1863: } else {
1864: for (i=0; i<N; i++) {
1865: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1866: a->ilen[rows[i]] = 0;
1867: }
1868: }
1869: A->nonzerostate++;
1870: }
1871: MatAssemblyEnd_SeqAIJ(A,MAT_FINAL_ASSEMBLY);
1872: return(0);
1873: }
1877: PetscErrorCode MatZeroRowsColumns_SeqAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag,Vec x,Vec b)
1878: {
1879: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1880: PetscInt i,j,m = A->rmap->n - 1,d = 0;
1881: PetscErrorCode ierr;
1882: PetscBool missing,*zeroed,vecs = PETSC_FALSE;
1883: const PetscScalar *xx;
1884: PetscScalar *bb;
1887: if (x && b) {
1888: VecGetArrayRead(x,&xx);
1889: VecGetArray(b,&bb);
1890: vecs = PETSC_TRUE;
1891: }
1892: PetscCalloc1(A->rmap->n,&zeroed);
1893: for (i=0; i<N; i++) {
1894: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1895: PetscMemzero(&a->a[a->i[rows[i]]],a->ilen[rows[i]]*sizeof(PetscScalar));
1897: zeroed[rows[i]] = PETSC_TRUE;
1898: }
1899: for (i=0; i<A->rmap->n; i++) {
1900: if (!zeroed[i]) {
1901: for (j=a->i[i]; j<a->i[i+1]; j++) {
1902: if (zeroed[a->j[j]]) {
1903: if (vecs) bb[i] -= a->a[j]*xx[a->j[j]];
1904: a->a[j] = 0.0;
1905: }
1906: }
1907: } else if (vecs) bb[i] = diag*xx[i];
1908: }
1909: if (x && b) {
1910: VecRestoreArrayRead(x,&xx);
1911: VecRestoreArray(b,&bb);
1912: }
1913: PetscFree(zeroed);
1914: if (diag != 0.0) {
1915: MatMissingDiagonal_SeqAIJ(A,&missing,&d);
1916: if (missing) {
1917: if (a->nonew) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry in row %D",d);
1918: else {
1919: for (i=0; i<N; i++) {
1920: MatSetValues_SeqAIJ(A,1,&rows[i],1,&rows[i],&diag,INSERT_VALUES);
1921: }
1922: }
1923: } else {
1924: for (i=0; i<N; i++) {
1925: a->a[a->diag[rows[i]]] = diag;
1926: }
1927: }
1928: }
1929: MatAssemblyEnd_SeqAIJ(A,MAT_FINAL_ASSEMBLY);
1930: return(0);
1931: }
1935: PetscErrorCode MatGetRow_SeqAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1936: {
1937: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1938: PetscInt *itmp;
1941: if (row < 0 || row >= A->rmap->n) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row %D out of range",row);
1943: *nz = a->i[row+1] - a->i[row];
1944: if (v) *v = a->a + a->i[row];
1945: if (idx) {
1946: itmp = a->j + a->i[row];
1947: if (*nz) *idx = itmp;
1948: else *idx = 0;
1949: }
1950: return(0);
1951: }
1953: /* remove this function? */
1956: PetscErrorCode MatRestoreRow_SeqAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1957: {
1959: return(0);
1960: }
1964: PetscErrorCode MatNorm_SeqAIJ(Mat A,NormType type,PetscReal *nrm)
1965: {
1966: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1967: MatScalar *v = a->a;
1968: PetscReal sum = 0.0;
1970: PetscInt i,j;
1973: if (type == NORM_FROBENIUS) {
1974: for (i=0; i<a->nz; i++) {
1975: sum += PetscRealPart(PetscConj(*v)*(*v)); v++;
1976: }
1977: *nrm = PetscSqrtReal(sum);
1978: PetscLogFlops(2*a->nz);
1979: } else if (type == NORM_1) {
1980: PetscReal *tmp;
1981: PetscInt *jj = a->j;
1982: PetscCalloc1(A->cmap->n+1,&tmp);
1983: *nrm = 0.0;
1984: for (j=0; j<a->nz; j++) {
1985: tmp[*jj++] += PetscAbsScalar(*v); v++;
1986: }
1987: for (j=0; j<A->cmap->n; j++) {
1988: if (tmp[j] > *nrm) *nrm = tmp[j];
1989: }
1990: PetscFree(tmp);
1991: PetscLogFlops(PetscMax(a->nz-1,0));
1992: } else if (type == NORM_INFINITY) {
1993: *nrm = 0.0;
1994: for (j=0; j<A->rmap->n; j++) {
1995: v = a->a + a->i[j];
1996: sum = 0.0;
1997: for (i=0; i<a->i[j+1]-a->i[j]; i++) {
1998: sum += PetscAbsScalar(*v); v++;
1999: }
2000: if (sum > *nrm) *nrm = sum;
2001: }
2002: PetscLogFlops(PetscMax(a->nz-1,0));
2003: } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"No support for two norm");
2004: return(0);
2005: }
2007: /* Merged from MatGetSymbolicTranspose_SeqAIJ() - replace MatGetSymbolicTranspose_SeqAIJ()? */
2010: PetscErrorCode MatTransposeSymbolic_SeqAIJ(Mat A,Mat *B)
2011: {
2013: PetscInt i,j,anzj;
2014: Mat_SeqAIJ *a=(Mat_SeqAIJ*)A->data,*b;
2015: PetscInt an=A->cmap->N,am=A->rmap->N;
2016: PetscInt *ati,*atj,*atfill,*ai=a->i,*aj=a->j;
2019: /* Allocate space for symbolic transpose info and work array */
2020: PetscCalloc1(an+1,&ati);
2021: PetscMalloc1(ai[am],&atj);
2022: PetscMalloc1(an,&atfill);
2024: /* Walk through aj and count ## of non-zeros in each row of A^T. */
2025: /* Note: offset by 1 for fast conversion into csr format. */
2026: for (i=0;i<ai[am];i++) ati[aj[i]+1] += 1;
2027: /* Form ati for csr format of A^T. */
2028: for (i=0;i<an;i++) ati[i+1] += ati[i];
2030: /* Copy ati into atfill so we have locations of the next free space in atj */
2031: PetscMemcpy(atfill,ati,an*sizeof(PetscInt));
2033: /* Walk through A row-wise and mark nonzero entries of A^T. */
2034: for (i=0;i<am;i++) {
2035: anzj = ai[i+1] - ai[i];
2036: for (j=0;j<anzj;j++) {
2037: atj[atfill[*aj]] = i;
2038: atfill[*aj++] += 1;
2039: }
2040: }
2042: /* Clean up temporary space and complete requests. */
2043: PetscFree(atfill);
2044: MatCreateSeqAIJWithArrays(PetscObjectComm((PetscObject)A),an,am,ati,atj,NULL,B);
2045: MatSetBlockSizes(*B,PetscAbs(A->cmap->bs),PetscAbs(A->rmap->bs));
2047: b = (Mat_SeqAIJ*)((*B)->data);
2048: b->free_a = PETSC_FALSE;
2049: b->free_ij = PETSC_TRUE;
2050: b->nonew = 0;
2051: return(0);
2052: }
2056: PetscErrorCode MatTranspose_SeqAIJ(Mat A,MatReuse reuse,Mat *B)
2057: {
2058: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2059: Mat C;
2061: PetscInt i,*aj = a->j,*ai = a->i,m = A->rmap->n,len,*col;
2062: MatScalar *array = a->a;
2065: if (reuse == MAT_REUSE_MATRIX && A == *B && m != A->cmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Square matrix only for in-place");
2067: if (reuse == MAT_INITIAL_MATRIX || *B == A) {
2068: PetscCalloc1(1+A->cmap->n,&col);
2070: for (i=0; i<ai[m]; i++) col[aj[i]] += 1;
2071: MatCreate(PetscObjectComm((PetscObject)A),&C);
2072: MatSetSizes(C,A->cmap->n,m,A->cmap->n,m);
2073: MatSetBlockSizes(C,PetscAbs(A->cmap->bs),PetscAbs(A->rmap->bs));
2074: MatSetType(C,((PetscObject)A)->type_name);
2075: MatSeqAIJSetPreallocation_SeqAIJ(C,0,col);
2076: PetscFree(col);
2077: } else {
2078: C = *B;
2079: }
2081: for (i=0; i<m; i++) {
2082: len = ai[i+1]-ai[i];
2083: MatSetValues_SeqAIJ(C,len,aj,1,&i,array,INSERT_VALUES);
2084: array += len;
2085: aj += len;
2086: }
2087: MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);
2088: MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);
2090: if (reuse == MAT_INITIAL_MATRIX || *B != A) {
2091: *B = C;
2092: } else {
2093: MatHeaderMerge(A,&C);
2094: }
2095: return(0);
2096: }
2100: PetscErrorCode MatIsTranspose_SeqAIJ(Mat A,Mat B,PetscReal tol,PetscBool *f)
2101: {
2102: Mat_SeqAIJ *aij = (Mat_SeqAIJ*) A->data,*bij = (Mat_SeqAIJ*) B->data;
2103: PetscInt *adx,*bdx,*aii,*bii,*aptr,*bptr;
2104: MatScalar *va,*vb;
2106: PetscInt ma,na,mb,nb, i;
2109: MatGetSize(A,&ma,&na);
2110: MatGetSize(B,&mb,&nb);
2111: if (ma!=nb || na!=mb) {
2112: *f = PETSC_FALSE;
2113: return(0);
2114: }
2115: aii = aij->i; bii = bij->i;
2116: adx = aij->j; bdx = bij->j;
2117: va = aij->a; vb = bij->a;
2118: PetscMalloc1(ma,&aptr);
2119: PetscMalloc1(mb,&bptr);
2120: for (i=0; i<ma; i++) aptr[i] = aii[i];
2121: for (i=0; i<mb; i++) bptr[i] = bii[i];
2123: *f = PETSC_TRUE;
2124: for (i=0; i<ma; i++) {
2125: while (aptr[i]<aii[i+1]) {
2126: PetscInt idc,idr;
2127: PetscScalar vc,vr;
2128: /* column/row index/value */
2129: idc = adx[aptr[i]];
2130: idr = bdx[bptr[idc]];
2131: vc = va[aptr[i]];
2132: vr = vb[bptr[idc]];
2133: if (i!=idr || PetscAbsScalar(vc-vr) > tol) {
2134: *f = PETSC_FALSE;
2135: goto done;
2136: } else {
2137: aptr[i]++;
2138: if (B || i!=idc) bptr[idc]++;
2139: }
2140: }
2141: }
2142: done:
2143: PetscFree(aptr);
2144: PetscFree(bptr);
2145: return(0);
2146: }
2150: PetscErrorCode MatIsHermitianTranspose_SeqAIJ(Mat A,Mat B,PetscReal tol,PetscBool *f)
2151: {
2152: Mat_SeqAIJ *aij = (Mat_SeqAIJ*) A->data,*bij = (Mat_SeqAIJ*) B->data;
2153: PetscInt *adx,*bdx,*aii,*bii,*aptr,*bptr;
2154: MatScalar *va,*vb;
2156: PetscInt ma,na,mb,nb, i;
2159: MatGetSize(A,&ma,&na);
2160: MatGetSize(B,&mb,&nb);
2161: if (ma!=nb || na!=mb) {
2162: *f = PETSC_FALSE;
2163: return(0);
2164: }
2165: aii = aij->i; bii = bij->i;
2166: adx = aij->j; bdx = bij->j;
2167: va = aij->a; vb = bij->a;
2168: PetscMalloc1(ma,&aptr);
2169: PetscMalloc1(mb,&bptr);
2170: for (i=0; i<ma; i++) aptr[i] = aii[i];
2171: for (i=0; i<mb; i++) bptr[i] = bii[i];
2173: *f = PETSC_TRUE;
2174: for (i=0; i<ma; i++) {
2175: while (aptr[i]<aii[i+1]) {
2176: PetscInt idc,idr;
2177: PetscScalar vc,vr;
2178: /* column/row index/value */
2179: idc = adx[aptr[i]];
2180: idr = bdx[bptr[idc]];
2181: vc = va[aptr[i]];
2182: vr = vb[bptr[idc]];
2183: if (i!=idr || PetscAbsScalar(vc-PetscConj(vr)) > tol) {
2184: *f = PETSC_FALSE;
2185: goto done;
2186: } else {
2187: aptr[i]++;
2188: if (B || i!=idc) bptr[idc]++;
2189: }
2190: }
2191: }
2192: done:
2193: PetscFree(aptr);
2194: PetscFree(bptr);
2195: return(0);
2196: }
2200: PetscErrorCode MatIsSymmetric_SeqAIJ(Mat A,PetscReal tol,PetscBool *f)
2201: {
2205: MatIsTranspose_SeqAIJ(A,A,tol,f);
2206: return(0);
2207: }
2211: PetscErrorCode MatIsHermitian_SeqAIJ(Mat A,PetscReal tol,PetscBool *f)
2212: {
2216: MatIsHermitianTranspose_SeqAIJ(A,A,tol,f);
2217: return(0);
2218: }
2222: PetscErrorCode MatDiagonalScale_SeqAIJ(Mat A,Vec ll,Vec rr)
2223: {
2224: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2225: PetscScalar *l,*r,x;
2226: MatScalar *v;
2228: PetscInt i,j,m = A->rmap->n,n = A->cmap->n,M,nz = a->nz,*jj;
2231: if (ll) {
2232: /* The local size is used so that VecMPI can be passed to this routine
2233: by MatDiagonalScale_MPIAIJ */
2234: VecGetLocalSize(ll,&m);
2235: if (m != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Left scaling vector wrong length");
2236: VecGetArray(ll,&l);
2237: v = a->a;
2238: for (i=0; i<m; i++) {
2239: x = l[i];
2240: M = a->i[i+1] - a->i[i];
2241: for (j=0; j<M; j++) (*v++) *= x;
2242: }
2243: VecRestoreArray(ll,&l);
2244: PetscLogFlops(nz);
2245: }
2246: if (rr) {
2247: VecGetLocalSize(rr,&n);
2248: if (n != A->cmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Right scaling vector wrong length");
2249: VecGetArray(rr,&r);
2250: v = a->a; jj = a->j;
2251: for (i=0; i<nz; i++) (*v++) *= r[*jj++];
2252: VecRestoreArray(rr,&r);
2253: PetscLogFlops(nz);
2254: }
2255: MatSeqAIJInvalidateDiagonal(A);
2256: return(0);
2257: }
2261: PetscErrorCode MatGetSubMatrix_SeqAIJ(Mat A,IS isrow,IS iscol,PetscInt csize,MatReuse scall,Mat *B)
2262: {
2263: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data,*c;
2265: PetscInt *smap,i,k,kstart,kend,oldcols = A->cmap->n,*lens;
2266: PetscInt row,mat_i,*mat_j,tcol,first,step,*mat_ilen,sum,lensi;
2267: const PetscInt *irow,*icol;
2268: PetscInt nrows,ncols;
2269: PetscInt *starts,*j_new,*i_new,*aj = a->j,*ai = a->i,ii,*ailen = a->ilen;
2270: MatScalar *a_new,*mat_a;
2271: Mat C;
2272: PetscBool stride;
2276: ISGetIndices(isrow,&irow);
2277: ISGetLocalSize(isrow,&nrows);
2278: ISGetLocalSize(iscol,&ncols);
2280: PetscObjectTypeCompare((PetscObject)iscol,ISSTRIDE,&stride);
2281: if (stride) {
2282: ISStrideGetInfo(iscol,&first,&step);
2283: } else {
2284: first = 0;
2285: step = 0;
2286: }
2287: if (stride && step == 1) {
2288: /* special case of contiguous rows */
2289: PetscMalloc2(nrows,&lens,nrows,&starts);
2290: /* loop over new rows determining lens and starting points */
2291: for (i=0; i<nrows; i++) {
2292: kstart = ai[irow[i]];
2293: kend = kstart + ailen[irow[i]];
2294: starts[i] = kstart;
2295: for (k=kstart; k<kend; k++) {
2296: if (aj[k] >= first) {
2297: starts[i] = k;
2298: break;
2299: }
2300: }
2301: sum = 0;
2302: while (k < kend) {
2303: if (aj[k++] >= first+ncols) break;
2304: sum++;
2305: }
2306: lens[i] = sum;
2307: }
2308: /* create submatrix */
2309: if (scall == MAT_REUSE_MATRIX) {
2310: PetscInt n_cols,n_rows;
2311: MatGetSize(*B,&n_rows,&n_cols);
2312: if (n_rows != nrows || n_cols != ncols) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Reused submatrix wrong size");
2313: MatZeroEntries(*B);
2314: C = *B;
2315: } else {
2316: PetscInt rbs,cbs;
2317: MatCreate(PetscObjectComm((PetscObject)A),&C);
2318: MatSetSizes(C,nrows,ncols,PETSC_DETERMINE,PETSC_DETERMINE);
2319: ISGetBlockSize(isrow,&rbs);
2320: ISGetBlockSize(iscol,&cbs);
2321: MatSetBlockSizes(C,rbs,cbs);
2322: MatSetType(C,((PetscObject)A)->type_name);
2323: MatSeqAIJSetPreallocation_SeqAIJ(C,0,lens);
2324: }
2325: c = (Mat_SeqAIJ*)C->data;
2327: /* loop over rows inserting into submatrix */
2328: a_new = c->a;
2329: j_new = c->j;
2330: i_new = c->i;
2332: for (i=0; i<nrows; i++) {
2333: ii = starts[i];
2334: lensi = lens[i];
2335: for (k=0; k<lensi; k++) {
2336: *j_new++ = aj[ii+k] - first;
2337: }
2338: PetscMemcpy(a_new,a->a + starts[i],lensi*sizeof(PetscScalar));
2339: a_new += lensi;
2340: i_new[i+1] = i_new[i] + lensi;
2341: c->ilen[i] = lensi;
2342: }
2343: PetscFree2(lens,starts);
2344: } else {
2345: ISGetIndices(iscol,&icol);
2346: PetscCalloc1(oldcols,&smap);
2347: PetscMalloc1(1+nrows,&lens);
2348: for (i=0; i<ncols; i++) {
2349: #if defined(PETSC_USE_DEBUG)
2350: if (icol[i] >= oldcols) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Requesting column beyond largest column icol[%D] %D <= A->cmap->n %D",i,icol[i],oldcols);
2351: #endif
2352: smap[icol[i]] = i+1;
2353: }
2355: /* determine lens of each row */
2356: for (i=0; i<nrows; i++) {
2357: kstart = ai[irow[i]];
2358: kend = kstart + a->ilen[irow[i]];
2359: lens[i] = 0;
2360: for (k=kstart; k<kend; k++) {
2361: if (smap[aj[k]]) {
2362: lens[i]++;
2363: }
2364: }
2365: }
2366: /* Create and fill new matrix */
2367: if (scall == MAT_REUSE_MATRIX) {
2368: PetscBool equal;
2370: c = (Mat_SeqAIJ*)((*B)->data);
2371: if ((*B)->rmap->n != nrows || (*B)->cmap->n != ncols) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Cannot reuse matrix. wrong size");
2372: PetscMemcmp(c->ilen,lens,(*B)->rmap->n*sizeof(PetscInt),&equal);
2373: if (!equal) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Cannot reuse matrix. wrong no of nonzeros");
2374: PetscMemzero(c->ilen,(*B)->rmap->n*sizeof(PetscInt));
2375: C = *B;
2376: } else {
2377: PetscInt rbs,cbs;
2378: MatCreate(PetscObjectComm((PetscObject)A),&C);
2379: MatSetSizes(C,nrows,ncols,PETSC_DETERMINE,PETSC_DETERMINE);
2380: ISGetBlockSize(isrow,&rbs);
2381: ISGetBlockSize(iscol,&cbs);
2382: MatSetBlockSizes(C,rbs,cbs);
2383: MatSetType(C,((PetscObject)A)->type_name);
2384: MatSeqAIJSetPreallocation_SeqAIJ(C,0,lens);
2385: }
2386: c = (Mat_SeqAIJ*)(C->data);
2387: for (i=0; i<nrows; i++) {
2388: row = irow[i];
2389: kstart = ai[row];
2390: kend = kstart + a->ilen[row];
2391: mat_i = c->i[i];
2392: mat_j = c->j + mat_i;
2393: mat_a = c->a + mat_i;
2394: mat_ilen = c->ilen + i;
2395: for (k=kstart; k<kend; k++) {
2396: if ((tcol=smap[a->j[k]])) {
2397: *mat_j++ = tcol - 1;
2398: *mat_a++ = a->a[k];
2399: (*mat_ilen)++;
2401: }
2402: }
2403: }
2404: /* Free work space */
2405: ISRestoreIndices(iscol,&icol);
2406: PetscFree(smap);
2407: PetscFree(lens);
2408: /* sort */
2409: for (i = 0; i < nrows; i++) {
2410: PetscInt ilen;
2412: mat_i = c->i[i];
2413: mat_j = c->j + mat_i;
2414: mat_a = c->a + mat_i;
2415: ilen = c->ilen[i];
2416: PetscSortIntWithMatScalarArray(ilen,mat_j,mat_a);
2417: }
2418: }
2419: MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);
2420: MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);
2422: ISRestoreIndices(isrow,&irow);
2423: *B = C;
2424: return(0);
2425: }
2429: PetscErrorCode MatGetMultiProcBlock_SeqAIJ(Mat mat,MPI_Comm subComm,MatReuse scall,Mat *subMat)
2430: {
2432: Mat B;
2435: if (scall == MAT_INITIAL_MATRIX) {
2436: MatCreate(subComm,&B);
2437: MatSetSizes(B,mat->rmap->n,mat->cmap->n,mat->rmap->n,mat->cmap->n);
2438: MatSetBlockSizesFromMats(B,mat,mat);
2439: MatSetType(B,MATSEQAIJ);
2440: MatDuplicateNoCreate_SeqAIJ(B,mat,MAT_COPY_VALUES,PETSC_TRUE);
2441: *subMat = B;
2442: } else {
2443: MatCopy_SeqAIJ(mat,*subMat,SAME_NONZERO_PATTERN);
2444: }
2445: return(0);
2446: }
2450: PetscErrorCode MatILUFactor_SeqAIJ(Mat inA,IS row,IS col,const MatFactorInfo *info)
2451: {
2452: Mat_SeqAIJ *a = (Mat_SeqAIJ*)inA->data;
2454: Mat outA;
2455: PetscBool row_identity,col_identity;
2458: if (info->levels != 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only levels=0 supported for in-place ilu");
2460: ISIdentity(row,&row_identity);
2461: ISIdentity(col,&col_identity);
2463: outA = inA;
2464: outA->factortype = MAT_FACTOR_LU;
2465: PetscFree(inA->solvertype);
2466: PetscStrallocpy(MATSOLVERPETSC,&inA->solvertype);
2468: PetscObjectReference((PetscObject)row);
2469: ISDestroy(&a->row);
2471: a->row = row;
2473: PetscObjectReference((PetscObject)col);
2474: ISDestroy(&a->col);
2476: a->col = col;
2478: /* Create the inverse permutation so that it can be used in MatLUFactorNumeric() */
2479: ISDestroy(&a->icol);
2480: ISInvertPermutation(col,PETSC_DECIDE,&a->icol);
2481: PetscLogObjectParent((PetscObject)inA,(PetscObject)a->icol);
2483: if (!a->solve_work) { /* this matrix may have been factored before */
2484: PetscMalloc1(inA->rmap->n+1,&a->solve_work);
2485: PetscLogObjectMemory((PetscObject)inA, (inA->rmap->n+1)*sizeof(PetscScalar));
2486: }
2488: MatMarkDiagonal_SeqAIJ(inA);
2489: if (row_identity && col_identity) {
2490: MatLUFactorNumeric_SeqAIJ_inplace(outA,inA,info);
2491: } else {
2492: MatLUFactorNumeric_SeqAIJ_InplaceWithPerm(outA,inA,info);
2493: }
2494: return(0);
2495: }
2499: PetscErrorCode MatScale_SeqAIJ(Mat inA,PetscScalar alpha)
2500: {
2501: Mat_SeqAIJ *a = (Mat_SeqAIJ*)inA->data;
2502: PetscScalar oalpha = alpha;
2504: PetscBLASInt one = 1,bnz;
2507: PetscBLASIntCast(a->nz,&bnz);
2508: PetscStackCallBLAS("BLASscal",BLASscal_(&bnz,&oalpha,a->a,&one));
2509: PetscLogFlops(a->nz);
2510: MatSeqAIJInvalidateDiagonal(inA);
2511: return(0);
2512: }
2516: PetscErrorCode MatGetSubMatrices_SeqAIJ(Mat A,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *B[])
2517: {
2519: PetscInt i;
2522: if (scall == MAT_INITIAL_MATRIX) {
2523: PetscMalloc1(n+1,B);
2524: }
2526: for (i=0; i<n; i++) {
2527: MatGetSubMatrix_SeqAIJ(A,irow[i],icol[i],PETSC_DECIDE,scall,&(*B)[i]);
2528: }
2529: return(0);
2530: }
2534: PetscErrorCode MatIncreaseOverlap_SeqAIJ(Mat A,PetscInt is_max,IS is[],PetscInt ov)
2535: {
2536: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2538: PetscInt row,i,j,k,l,m,n,*nidx,isz,val;
2539: const PetscInt *idx;
2540: PetscInt start,end,*ai,*aj;
2541: PetscBT table;
2544: m = A->rmap->n;
2545: ai = a->i;
2546: aj = a->j;
2548: if (ov < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"illegal negative overlap value used");
2550: PetscMalloc1(m+1,&nidx);
2551: PetscBTCreate(m,&table);
2553: for (i=0; i<is_max; i++) {
2554: /* Initialize the two local arrays */
2555: isz = 0;
2556: PetscBTMemzero(m,table);
2558: /* Extract the indices, assume there can be duplicate entries */
2559: ISGetIndices(is[i],&idx);
2560: ISGetLocalSize(is[i],&n);
2562: /* Enter these into the temp arrays. I.e., mark table[row], enter row into new index */
2563: for (j=0; j<n; ++j) {
2564: if (!PetscBTLookupSet(table,idx[j])) nidx[isz++] = idx[j];
2565: }
2566: ISRestoreIndices(is[i],&idx);
2567: ISDestroy(&is[i]);
2569: k = 0;
2570: for (j=0; j<ov; j++) { /* for each overlap */
2571: n = isz;
2572: for (; k<n; k++) { /* do only those rows in nidx[k], which are not done yet */
2573: row = nidx[k];
2574: start = ai[row];
2575: end = ai[row+1];
2576: for (l = start; l<end; l++) {
2577: val = aj[l];
2578: if (!PetscBTLookupSet(table,val)) nidx[isz++] = val;
2579: }
2580: }
2581: }
2582: ISCreateGeneral(PETSC_COMM_SELF,isz,nidx,PETSC_COPY_VALUES,(is+i));
2583: }
2584: PetscBTDestroy(&table);
2585: PetscFree(nidx);
2586: return(0);
2587: }
2589: /* -------------------------------------------------------------- */
2592: PetscErrorCode MatPermute_SeqAIJ(Mat A,IS rowp,IS colp,Mat *B)
2593: {
2594: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2596: PetscInt i,nz = 0,m = A->rmap->n,n = A->cmap->n;
2597: const PetscInt *row,*col;
2598: PetscInt *cnew,j,*lens;
2599: IS icolp,irowp;
2600: PetscInt *cwork = NULL;
2601: PetscScalar *vwork = NULL;
2604: ISInvertPermutation(rowp,PETSC_DECIDE,&irowp);
2605: ISGetIndices(irowp,&row);
2606: ISInvertPermutation(colp,PETSC_DECIDE,&icolp);
2607: ISGetIndices(icolp,&col);
2609: /* determine lengths of permuted rows */
2610: PetscMalloc1(m+1,&lens);
2611: for (i=0; i<m; i++) lens[row[i]] = a->i[i+1] - a->i[i];
2612: MatCreate(PetscObjectComm((PetscObject)A),B);
2613: MatSetSizes(*B,m,n,m,n);
2614: MatSetBlockSizesFromMats(*B,A,A);
2615: MatSetType(*B,((PetscObject)A)->type_name);
2616: MatSeqAIJSetPreallocation_SeqAIJ(*B,0,lens);
2617: PetscFree(lens);
2619: PetscMalloc1(n,&cnew);
2620: for (i=0; i<m; i++) {
2621: MatGetRow_SeqAIJ(A,i,&nz,&cwork,&vwork);
2622: for (j=0; j<nz; j++) cnew[j] = col[cwork[j]];
2623: MatSetValues_SeqAIJ(*B,1,&row[i],nz,cnew,vwork,INSERT_VALUES);
2624: MatRestoreRow_SeqAIJ(A,i,&nz,&cwork,&vwork);
2625: }
2626: PetscFree(cnew);
2628: (*B)->assembled = PETSC_FALSE;
2630: MatAssemblyBegin(*B,MAT_FINAL_ASSEMBLY);
2631: MatAssemblyEnd(*B,MAT_FINAL_ASSEMBLY);
2632: ISRestoreIndices(irowp,&row);
2633: ISRestoreIndices(icolp,&col);
2634: ISDestroy(&irowp);
2635: ISDestroy(&icolp);
2636: return(0);
2637: }
2641: PetscErrorCode MatCopy_SeqAIJ(Mat A,Mat B,MatStructure str)
2642: {
2646: /* If the two matrices have the same copy implementation, use fast copy. */
2647: if (str == SAME_NONZERO_PATTERN && (A->ops->copy == B->ops->copy)) {
2648: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2649: Mat_SeqAIJ *b = (Mat_SeqAIJ*)B->data;
2651: if (a->i[A->rmap->n] != b->i[B->rmap->n]) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_INCOMP,"Number of nonzeros in two matrices are different");
2652: PetscMemcpy(b->a,a->a,(a->i[A->rmap->n])*sizeof(PetscScalar));
2653: } else {
2654: MatCopy_Basic(A,B,str);
2655: }
2656: return(0);
2657: }
2661: PetscErrorCode MatSetUp_SeqAIJ(Mat A)
2662: {
2666: MatSeqAIJSetPreallocation_SeqAIJ(A,PETSC_DEFAULT,0);
2667: return(0);
2668: }
2672: PetscErrorCode MatSeqAIJGetArray_SeqAIJ(Mat A,PetscScalar *array[])
2673: {
2674: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2677: *array = a->a;
2678: return(0);
2679: }
2683: PetscErrorCode MatSeqAIJRestoreArray_SeqAIJ(Mat A,PetscScalar *array[])
2684: {
2686: return(0);
2687: }
2689: /*
2690: Computes the number of nonzeros per row needed for preallocation when X and Y
2691: have different nonzero structure.
2692: */
2695: PetscErrorCode MatAXPYGetPreallocation_SeqX_private(PetscInt m,const PetscInt *xi,const PetscInt *xj,const PetscInt *yi,const PetscInt *yj,PetscInt *nnz)
2696: {
2697: PetscInt i,j,k,nzx,nzy;
2700: /* Set the number of nonzeros in the new matrix */
2701: for (i=0; i<m; i++) {
2702: const PetscInt *xjj = xj+xi[i],*yjj = yj+yi[i];
2703: nzx = xi[i+1] - xi[i];
2704: nzy = yi[i+1] - yi[i];
2705: nnz[i] = 0;
2706: for (j=0,k=0; j<nzx; j++) { /* Point in X */
2707: for (; k<nzy && yjj[k]<xjj[j]; k++) nnz[i]++; /* Catch up to X */
2708: if (k<nzy && yjj[k]==xjj[j]) k++; /* Skip duplicate */
2709: nnz[i]++;
2710: }
2711: for (; k<nzy; k++) nnz[i]++;
2712: }
2713: return(0);
2714: }
2718: PetscErrorCode MatAXPYGetPreallocation_SeqAIJ(Mat Y,Mat X,PetscInt *nnz)
2719: {
2720: PetscInt m = Y->rmap->N;
2721: Mat_SeqAIJ *x = (Mat_SeqAIJ*)X->data;
2722: Mat_SeqAIJ *y = (Mat_SeqAIJ*)Y->data;
2726: /* Set the number of nonzeros in the new matrix */
2727: MatAXPYGetPreallocation_SeqX_private(m,x->i,x->j,y->i,y->j,nnz);
2728: return(0);
2729: }
2733: PetscErrorCode MatAXPY_SeqAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str)
2734: {
2736: Mat_SeqAIJ *x = (Mat_SeqAIJ*)X->data,*y = (Mat_SeqAIJ*)Y->data;
2737: PetscBLASInt one=1,bnz;
2740: PetscBLASIntCast(x->nz,&bnz);
2741: if (str == SAME_NONZERO_PATTERN) {
2742: PetscScalar alpha = a;
2743: PetscStackCallBLAS("BLASaxpy",BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one));
2744: MatSeqAIJInvalidateDiagonal(Y);
2745: PetscObjectStateIncrease((PetscObject)Y);
2746: } else if (str == SUBSET_NONZERO_PATTERN) { /* nonzeros of X is a subset of Y's */
2747: MatAXPY_Basic(Y,a,X,str);
2748: } else {
2749: Mat B;
2750: PetscInt *nnz;
2751: PetscMalloc1(Y->rmap->N,&nnz);
2752: MatCreate(PetscObjectComm((PetscObject)Y),&B);
2753: PetscObjectSetName((PetscObject)B,((PetscObject)Y)->name);
2754: MatSetSizes(B,Y->rmap->n,Y->cmap->n,Y->rmap->N,Y->cmap->N);
2755: MatSetBlockSizesFromMats(B,Y,Y);
2756: MatSetType(B,(MatType) ((PetscObject)Y)->type_name);
2757: MatAXPYGetPreallocation_SeqAIJ(Y,X,nnz);
2758: MatSeqAIJSetPreallocation(B,0,nnz);
2759: MatAXPY_BasicWithPreallocation(B,Y,a,X,str);
2760: MatHeaderReplace(Y,&B);
2761: PetscFree(nnz);
2762: }
2763: return(0);
2764: }
2768: PetscErrorCode MatConjugate_SeqAIJ(Mat mat)
2769: {
2770: #if defined(PETSC_USE_COMPLEX)
2771: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)mat->data;
2772: PetscInt i,nz;
2773: PetscScalar *a;
2776: nz = aij->nz;
2777: a = aij->a;
2778: for (i=0; i<nz; i++) a[i] = PetscConj(a[i]);
2779: #else
2781: #endif
2782: return(0);
2783: }
2787: PetscErrorCode MatGetRowMaxAbs_SeqAIJ(Mat A,Vec v,PetscInt idx[])
2788: {
2789: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2791: PetscInt i,j,m = A->rmap->n,*ai,*aj,ncols,n;
2792: PetscReal atmp;
2793: PetscScalar *x;
2794: MatScalar *aa;
2797: if (A->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2798: aa = a->a;
2799: ai = a->i;
2800: aj = a->j;
2802: VecSet(v,0.0);
2803: VecGetArray(v,&x);
2804: VecGetLocalSize(v,&n);
2805: if (n != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
2806: for (i=0; i<m; i++) {
2807: ncols = ai[1] - ai[0]; ai++;
2808: x[i] = 0.0;
2809: for (j=0; j<ncols; j++) {
2810: atmp = PetscAbsScalar(*aa);
2811: if (PetscAbsScalar(x[i]) < atmp) {x[i] = atmp; if (idx) idx[i] = *aj;}
2812: aa++; aj++;
2813: }
2814: }
2815: VecRestoreArray(v,&x);
2816: return(0);
2817: }
2821: PetscErrorCode MatGetRowMax_SeqAIJ(Mat A,Vec v,PetscInt idx[])
2822: {
2823: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2825: PetscInt i,j,m = A->rmap->n,*ai,*aj,ncols,n;
2826: PetscScalar *x;
2827: MatScalar *aa;
2830: if (A->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2831: aa = a->a;
2832: ai = a->i;
2833: aj = a->j;
2835: VecSet(v,0.0);
2836: VecGetArray(v,&x);
2837: VecGetLocalSize(v,&n);
2838: if (n != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
2839: for (i=0; i<m; i++) {
2840: ncols = ai[1] - ai[0]; ai++;
2841: if (ncols == A->cmap->n) { /* row is dense */
2842: x[i] = *aa; if (idx) idx[i] = 0;
2843: } else { /* row is sparse so already KNOW maximum is 0.0 or higher */
2844: x[i] = 0.0;
2845: if (idx) {
2846: idx[i] = 0; /* in case ncols is zero */
2847: for (j=0;j<ncols;j++) { /* find first implicit 0.0 in the row */
2848: if (aj[j] > j) {
2849: idx[i] = j;
2850: break;
2851: }
2852: }
2853: }
2854: }
2855: for (j=0; j<ncols; j++) {
2856: if (PetscRealPart(x[i]) < PetscRealPart(*aa)) {x[i] = *aa; if (idx) idx[i] = *aj;}
2857: aa++; aj++;
2858: }
2859: }
2860: VecRestoreArray(v,&x);
2861: return(0);
2862: }
2866: PetscErrorCode MatGetRowMinAbs_SeqAIJ(Mat A,Vec v,PetscInt idx[])
2867: {
2868: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2870: PetscInt i,j,m = A->rmap->n,*ai,*aj,ncols,n;
2871: PetscReal atmp;
2872: PetscScalar *x;
2873: MatScalar *aa;
2876: if (A->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2877: aa = a->a;
2878: ai = a->i;
2879: aj = a->j;
2881: VecSet(v,0.0);
2882: VecGetArray(v,&x);
2883: VecGetLocalSize(v,&n);
2884: if (n != A->rmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector, %D vs. %D rows", A->rmap->n, n);
2885: for (i=0; i<m; i++) {
2886: ncols = ai[1] - ai[0]; ai++;
2887: if (ncols) {
2888: /* Get first nonzero */
2889: for (j = 0; j < ncols; j++) {
2890: atmp = PetscAbsScalar(aa[j]);
2891: if (atmp > 1.0e-12) {
2892: x[i] = atmp;
2893: if (idx) idx[i] = aj[j];
2894: break;
2895: }
2896: }
2897: if (j == ncols) {x[i] = PetscAbsScalar(*aa); if (idx) idx[i] = *aj;}
2898: } else {
2899: x[i] = 0.0; if (idx) idx[i] = 0;
2900: }
2901: for (j = 0; j < ncols; j++) {
2902: atmp = PetscAbsScalar(*aa);
2903: if (atmp > 1.0e-12 && PetscAbsScalar(x[i]) > atmp) {x[i] = atmp; if (idx) idx[i] = *aj;}
2904: aa++; aj++;
2905: }
2906: }
2907: VecRestoreArray(v,&x);
2908: return(0);
2909: }
2913: PetscErrorCode MatGetRowMin_SeqAIJ(Mat A,Vec v,PetscInt idx[])
2914: {
2915: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2916: PetscErrorCode ierr;
2917: PetscInt i,j,m = A->rmap->n,ncols,n;
2918: const PetscInt *ai,*aj;
2919: PetscScalar *x;
2920: const MatScalar *aa;
2923: if (A->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2924: aa = a->a;
2925: ai = a->i;
2926: aj = a->j;
2928: VecSet(v,0.0);
2929: VecGetArray(v,&x);
2930: VecGetLocalSize(v,&n);
2931: if (n != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
2932: for (i=0; i<m; i++) {
2933: ncols = ai[1] - ai[0]; ai++;
2934: if (ncols == A->cmap->n) { /* row is dense */
2935: x[i] = *aa; if (idx) idx[i] = 0;
2936: } else { /* row is sparse so already KNOW minimum is 0.0 or lower */
2937: x[i] = 0.0;
2938: if (idx) { /* find first implicit 0.0 in the row */
2939: idx[i] = 0; /* in case ncols is zero */
2940: for (j=0; j<ncols; j++) {
2941: if (aj[j] > j) {
2942: idx[i] = j;
2943: break;
2944: }
2945: }
2946: }
2947: }
2948: for (j=0; j<ncols; j++) {
2949: if (PetscRealPart(x[i]) > PetscRealPart(*aa)) {x[i] = *aa; if (idx) idx[i] = *aj;}
2950: aa++; aj++;
2951: }
2952: }
2953: VecRestoreArray(v,&x);
2954: return(0);
2955: }
2957: #include <petscblaslapack.h>
2958: #include <petsc/private/kernels/blockinvert.h>
2962: PetscErrorCode MatInvertBlockDiagonal_SeqAIJ(Mat A,const PetscScalar **values)
2963: {
2964: Mat_SeqAIJ *a = (Mat_SeqAIJ*) A->data;
2966: PetscInt i,bs = PetscAbs(A->rmap->bs),mbs = A->rmap->n/bs,ipvt[5],bs2 = bs*bs,*v_pivots,ij[7],*IJ,j;
2967: MatScalar *diag,work[25],*v_work;
2968: PetscReal shift = 0.0;
2969: PetscBool allowzeropivot,zeropivotdetected=PETSC_FALSE;
2972: allowzeropivot = PetscNot(A->erroriffailure);
2973: if (a->ibdiagvalid) {
2974: if (values) *values = a->ibdiag;
2975: return(0);
2976: }
2977: MatMarkDiagonal_SeqAIJ(A);
2978: if (!a->ibdiag) {
2979: PetscMalloc1(bs2*mbs,&a->ibdiag);
2980: PetscLogObjectMemory((PetscObject)A,bs2*mbs*sizeof(PetscScalar));
2981: }
2982: diag = a->ibdiag;
2983: if (values) *values = a->ibdiag;
2984: /* factor and invert each block */
2985: switch (bs) {
2986: case 1:
2987: for (i=0; i<mbs; i++) {
2988: MatGetValues(A,1,&i,1,&i,diag+i);
2989: if (PetscAbsScalar(diag[i] + shift) < PETSC_MACHINE_EPSILON) {
2990: if (allowzeropivot) {
2991: A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
2992: PetscInfo1(A,"Zero pivot, row %D\n",i);
2993: } else SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_MAT_LU_ZRPVT,"Zero pivot, row %D",i);
2994: }
2995: diag[i] = (PetscScalar)1.0 / (diag[i] + shift);
2996: }
2997: break;
2998: case 2:
2999: for (i=0; i<mbs; i++) {
3000: ij[0] = 2*i; ij[1] = 2*i + 1;
3001: MatGetValues(A,2,ij,2,ij,diag);
3002: PetscKernel_A_gets_inverse_A_2(diag,shift,allowzeropivot,&zeropivotdetected);
3003: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3004: PetscKernel_A_gets_transpose_A_2(diag);
3005: diag += 4;
3006: }
3007: break;
3008: case 3:
3009: for (i=0; i<mbs; i++) {
3010: ij[0] = 3*i; ij[1] = 3*i + 1; ij[2] = 3*i + 2;
3011: MatGetValues(A,3,ij,3,ij,diag);
3012: PetscKernel_A_gets_inverse_A_3(diag,shift,allowzeropivot,&zeropivotdetected);
3013: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3014: PetscKernel_A_gets_transpose_A_3(diag);
3015: diag += 9;
3016: }
3017: break;
3018: case 4:
3019: for (i=0; i<mbs; i++) {
3020: ij[0] = 4*i; ij[1] = 4*i + 1; ij[2] = 4*i + 2; ij[3] = 4*i + 3;
3021: MatGetValues(A,4,ij,4,ij,diag);
3022: PetscKernel_A_gets_inverse_A_4(diag,shift,allowzeropivot,&zeropivotdetected);
3023: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3024: PetscKernel_A_gets_transpose_A_4(diag);
3025: diag += 16;
3026: }
3027: break;
3028: case 5:
3029: for (i=0; i<mbs; i++) {
3030: ij[0] = 5*i; ij[1] = 5*i + 1; ij[2] = 5*i + 2; ij[3] = 5*i + 3; ij[4] = 5*i + 4;
3031: MatGetValues(A,5,ij,5,ij,diag);
3032: PetscKernel_A_gets_inverse_A_5(diag,ipvt,work,shift,allowzeropivot,&zeropivotdetected);
3033: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3034: PetscKernel_A_gets_transpose_A_5(diag);
3035: diag += 25;
3036: }
3037: break;
3038: case 6:
3039: for (i=0; i<mbs; i++) {
3040: ij[0] = 6*i; ij[1] = 6*i + 1; ij[2] = 6*i + 2; ij[3] = 6*i + 3; ij[4] = 6*i + 4; ij[5] = 6*i + 5;
3041: MatGetValues(A,6,ij,6,ij,diag);
3042: PetscKernel_A_gets_inverse_A_6(diag,shift,allowzeropivot,&zeropivotdetected);
3043: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3044: PetscKernel_A_gets_transpose_A_6(diag);
3045: diag += 36;
3046: }
3047: break;
3048: case 7:
3049: for (i=0; i<mbs; i++) {
3050: ij[0] = 7*i; ij[1] = 7*i + 1; ij[2] = 7*i + 2; ij[3] = 7*i + 3; ij[4] = 7*i + 4; ij[5] = 7*i + 5; ij[5] = 7*i + 6;
3051: MatGetValues(A,7,ij,7,ij,diag);
3052: PetscKernel_A_gets_inverse_A_7(diag,shift,allowzeropivot,&zeropivotdetected);
3053: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3054: PetscKernel_A_gets_transpose_A_7(diag);
3055: diag += 49;
3056: }
3057: break;
3058: default:
3059: PetscMalloc3(bs,&v_work,bs,&v_pivots,bs,&IJ);
3060: for (i=0; i<mbs; i++) {
3061: for (j=0; j<bs; j++) {
3062: IJ[j] = bs*i + j;
3063: }
3064: MatGetValues(A,bs,IJ,bs,IJ,diag);
3065: PetscKernel_A_gets_inverse_A(bs,diag,v_pivots,v_work,allowzeropivot,&zeropivotdetected);
3066: if (zeropivotdetected) A->errortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
3067: PetscKernel_A_gets_transpose_A_N(diag,bs);
3068: diag += bs2;
3069: }
3070: PetscFree3(v_work,v_pivots,IJ);
3071: }
3072: a->ibdiagvalid = PETSC_TRUE;
3073: return(0);
3074: }
3078: static PetscErrorCode MatSetRandom_SeqAIJ(Mat x,PetscRandom rctx)
3079: {
3081: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)x->data;
3082: PetscScalar a;
3083: PetscInt m,n,i,j,col;
3086: if (!x->assembled) {
3087: MatGetSize(x,&m,&n);
3088: for (i=0; i<m; i++) {
3089: for (j=0; j<aij->imax[i]; j++) {
3090: PetscRandomGetValue(rctx,&a);
3091: col = (PetscInt)(n*PetscRealPart(a));
3092: MatSetValues(x,1,&i,1,&col,&a,ADD_VALUES);
3093: }
3094: }
3095: } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not yet coded");
3096: MatAssemblyBegin(x,MAT_FINAL_ASSEMBLY);
3097: MatAssemblyEnd(x,MAT_FINAL_ASSEMBLY);
3098: return(0);
3099: }
3103: PetscErrorCode MatShift_SeqAIJ(Mat Y,PetscScalar a)
3104: {
3106: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)Y->data;
3109: if (!Y->preallocated || !aij->nz) {
3110: MatSeqAIJSetPreallocation(Y,1,NULL);
3111: }
3112: MatShift_Basic(Y,a);
3113: return(0);
3114: }
3116: /* -------------------------------------------------------------------*/
3117: static struct _MatOps MatOps_Values = { MatSetValues_SeqAIJ,
3118: MatGetRow_SeqAIJ,
3119: MatRestoreRow_SeqAIJ,
3120: MatMult_SeqAIJ,
3121: /* 4*/ MatMultAdd_SeqAIJ,
3122: MatMultTranspose_SeqAIJ,
3123: MatMultTransposeAdd_SeqAIJ,
3124: 0,
3125: 0,
3126: 0,
3127: /* 10*/ 0,
3128: MatLUFactor_SeqAIJ,
3129: 0,
3130: MatSOR_SeqAIJ,
3131: MatTranspose_SeqAIJ,
3132: /*1 5*/ MatGetInfo_SeqAIJ,
3133: MatEqual_SeqAIJ,
3134: MatGetDiagonal_SeqAIJ,
3135: MatDiagonalScale_SeqAIJ,
3136: MatNorm_SeqAIJ,
3137: /* 20*/ 0,
3138: MatAssemblyEnd_SeqAIJ,
3139: MatSetOption_SeqAIJ,
3140: MatZeroEntries_SeqAIJ,
3141: /* 24*/ MatZeroRows_SeqAIJ,
3142: 0,
3143: 0,
3144: 0,
3145: 0,
3146: /* 29*/ MatSetUp_SeqAIJ,
3147: 0,
3148: 0,
3149: 0,
3150: 0,
3151: /* 34*/ MatDuplicate_SeqAIJ,
3152: 0,
3153: 0,
3154: MatILUFactor_SeqAIJ,
3155: 0,
3156: /* 39*/ MatAXPY_SeqAIJ,
3157: MatGetSubMatrices_SeqAIJ,
3158: MatIncreaseOverlap_SeqAIJ,
3159: MatGetValues_SeqAIJ,
3160: MatCopy_SeqAIJ,
3161: /* 44*/ MatGetRowMax_SeqAIJ,
3162: MatScale_SeqAIJ,
3163: MatShift_SeqAIJ,
3164: MatDiagonalSet_SeqAIJ,
3165: MatZeroRowsColumns_SeqAIJ,
3166: /* 49*/ MatSetRandom_SeqAIJ,
3167: MatGetRowIJ_SeqAIJ,
3168: MatRestoreRowIJ_SeqAIJ,
3169: MatGetColumnIJ_SeqAIJ,
3170: MatRestoreColumnIJ_SeqAIJ,
3171: /* 54*/ MatFDColoringCreate_SeqXAIJ,
3172: 0,
3173: 0,
3174: MatPermute_SeqAIJ,
3175: 0,
3176: /* 59*/ 0,
3177: MatDestroy_SeqAIJ,
3178: MatView_SeqAIJ,
3179: 0,
3180: MatMatMatMult_SeqAIJ_SeqAIJ_SeqAIJ,
3181: /* 64*/ MatMatMatMultSymbolic_SeqAIJ_SeqAIJ_SeqAIJ,
3182: MatMatMatMultNumeric_SeqAIJ_SeqAIJ_SeqAIJ,
3183: 0,
3184: 0,
3185: 0,
3186: /* 69*/ MatGetRowMaxAbs_SeqAIJ,
3187: MatGetRowMinAbs_SeqAIJ,
3188: 0,
3189: MatSetColoring_SeqAIJ,
3190: 0,
3191: /* 74*/ MatSetValuesAdifor_SeqAIJ,
3192: MatFDColoringApply_AIJ,
3193: 0,
3194: 0,
3195: 0,
3196: /* 79*/ MatFindZeroDiagonals_SeqAIJ,
3197: 0,
3198: 0,
3199: 0,
3200: MatLoad_SeqAIJ,
3201: /* 84*/ MatIsSymmetric_SeqAIJ,
3202: MatIsHermitian_SeqAIJ,
3203: 0,
3204: 0,
3205: 0,
3206: /* 89*/ MatMatMult_SeqAIJ_SeqAIJ,
3207: MatMatMultSymbolic_SeqAIJ_SeqAIJ,
3208: MatMatMultNumeric_SeqAIJ_SeqAIJ,
3209: MatPtAP_SeqAIJ_SeqAIJ,
3210: MatPtAPSymbolic_SeqAIJ_SeqAIJ_DenseAxpy,
3211: /* 94*/ MatPtAPNumeric_SeqAIJ_SeqAIJ,
3212: MatMatTransposeMult_SeqAIJ_SeqAIJ,
3213: MatMatTransposeMultSymbolic_SeqAIJ_SeqAIJ,
3214: MatMatTransposeMultNumeric_SeqAIJ_SeqAIJ,
3215: 0,
3216: /* 99*/ 0,
3217: 0,
3218: 0,
3219: MatConjugate_SeqAIJ,
3220: 0,
3221: /*104*/ MatSetValuesRow_SeqAIJ,
3222: MatRealPart_SeqAIJ,
3223: MatImaginaryPart_SeqAIJ,
3224: 0,
3225: 0,
3226: /*109*/ MatMatSolve_SeqAIJ,
3227: 0,
3228: MatGetRowMin_SeqAIJ,
3229: 0,
3230: MatMissingDiagonal_SeqAIJ,
3231: /*114*/ 0,
3232: 0,
3233: 0,
3234: 0,
3235: 0,
3236: /*119*/ 0,
3237: 0,
3238: 0,
3239: 0,
3240: MatGetMultiProcBlock_SeqAIJ,
3241: /*124*/ MatFindNonzeroRows_SeqAIJ,
3242: MatGetColumnNorms_SeqAIJ,
3243: MatInvertBlockDiagonal_SeqAIJ,
3244: 0,
3245: 0,
3246: /*129*/ 0,
3247: MatTransposeMatMult_SeqAIJ_SeqAIJ,
3248: MatTransposeMatMultSymbolic_SeqAIJ_SeqAIJ,
3249: MatTransposeMatMultNumeric_SeqAIJ_SeqAIJ,
3250: MatTransposeColoringCreate_SeqAIJ,
3251: /*134*/ MatTransColoringApplySpToDen_SeqAIJ,
3252: MatTransColoringApplyDenToSp_SeqAIJ,
3253: MatRARt_SeqAIJ_SeqAIJ,
3254: MatRARtSymbolic_SeqAIJ_SeqAIJ,
3255: MatRARtNumeric_SeqAIJ_SeqAIJ,
3256: /*139*/0,
3257: 0,
3258: 0,
3259: MatFDColoringSetUp_SeqXAIJ,
3260: MatFindOffBlockDiagonalEntries_SeqAIJ,
3261: /*144*/MatCreateMPIMatConcatenateSeqMat_SeqAIJ
3262: };
3266: PetscErrorCode MatSeqAIJSetColumnIndices_SeqAIJ(Mat mat,PetscInt *indices)
3267: {
3268: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)mat->data;
3269: PetscInt i,nz,n;
3272: nz = aij->maxnz;
3273: n = mat->rmap->n;
3274: for (i=0; i<nz; i++) {
3275: aij->j[i] = indices[i];
3276: }
3277: aij->nz = nz;
3278: for (i=0; i<n; i++) {
3279: aij->ilen[i] = aij->imax[i];
3280: }
3281: return(0);
3282: }
3286: /*@
3287: MatSeqAIJSetColumnIndices - Set the column indices for all the rows
3288: in the matrix.
3290: Input Parameters:
3291: + mat - the SeqAIJ matrix
3292: - indices - the column indices
3294: Level: advanced
3296: Notes:
3297: This can be called if you have precomputed the nonzero structure of the
3298: matrix and want to provide it to the matrix object to improve the performance
3299: of the MatSetValues() operation.
3301: You MUST have set the correct numbers of nonzeros per row in the call to
3302: MatCreateSeqAIJ(), and the columns indices MUST be sorted.
3304: MUST be called before any calls to MatSetValues();
3306: The indices should start with zero, not one.
3308: @*/
3309: PetscErrorCode MatSeqAIJSetColumnIndices(Mat mat,PetscInt *indices)
3310: {
3316: PetscUseMethod(mat,"MatSeqAIJSetColumnIndices_C",(Mat,PetscInt*),(mat,indices));
3317: return(0);
3318: }
3320: /* ----------------------------------------------------------------------------------------*/
3324: PetscErrorCode MatStoreValues_SeqAIJ(Mat mat)
3325: {
3326: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)mat->data;
3328: size_t nz = aij->i[mat->rmap->n];
3331: if (!aij->nonew) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NEW_NONZERO_LOCATIONS,PETSC_FALSE);first");
3333: /* allocate space for values if not already there */
3334: if (!aij->saved_values) {
3335: PetscMalloc1(nz+1,&aij->saved_values);
3336: PetscLogObjectMemory((PetscObject)mat,(nz+1)*sizeof(PetscScalar));
3337: }
3339: /* copy values over */
3340: PetscMemcpy(aij->saved_values,aij->a,nz*sizeof(PetscScalar));
3341: return(0);
3342: }
3346: /*@
3347: MatStoreValues - Stashes a copy of the matrix values; this allows, for
3348: example, reuse of the linear part of a Jacobian, while recomputing the
3349: nonlinear portion.
3351: Collect on Mat
3353: Input Parameters:
3354: . mat - the matrix (currently only AIJ matrices support this option)
3356: Level: advanced
3358: Common Usage, with SNESSolve():
3359: $ Create Jacobian matrix
3360: $ Set linear terms into matrix
3361: $ Apply boundary conditions to matrix, at this time matrix must have
3362: $ final nonzero structure (i.e. setting the nonlinear terms and applying
3363: $ boundary conditions again will not change the nonzero structure
3364: $ MatSetOption(mat,MAT_NEW_NONZERO_LOCATIONS,PETSC_FALSE);
3365: $ MatStoreValues(mat);
3366: $ Call SNESSetJacobian() with matrix
3367: $ In your Jacobian routine
3368: $ MatRetrieveValues(mat);
3369: $ Set nonlinear terms in matrix
3371: Common Usage without SNESSolve(), i.e. when you handle nonlinear solve yourself:
3372: $ // build linear portion of Jacobian
3373: $ MatSetOption(mat,MAT_NEW_NONZERO_LOCATIONS,PETSC_FALSE);
3374: $ MatStoreValues(mat);
3375: $ loop over nonlinear iterations
3376: $ MatRetrieveValues(mat);
3377: $ // call MatSetValues(mat,...) to set nonliner portion of Jacobian
3378: $ // call MatAssemblyBegin/End() on matrix
3379: $ Solve linear system with Jacobian
3380: $ endloop
3382: Notes:
3383: Matrix must already be assemblied before calling this routine
3384: Must set the matrix option MatSetOption(mat,MAT_NEW_NONZERO_LOCATIONS,PETSC_FALSE); before
3385: calling this routine.
3387: When this is called multiple times it overwrites the previous set of stored values
3388: and does not allocated additional space.
3390: .seealso: MatRetrieveValues()
3392: @*/
3393: PetscErrorCode MatStoreValues(Mat mat)
3394: {
3399: if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3400: if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3401: PetscUseMethod(mat,"MatStoreValues_C",(Mat),(mat));
3402: return(0);
3403: }
3407: PetscErrorCode MatRetrieveValues_SeqAIJ(Mat mat)
3408: {
3409: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)mat->data;
3411: PetscInt nz = aij->i[mat->rmap->n];
3414: if (!aij->nonew) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NEW_NONZERO_LOCATIONS,PETSC_FALSE);first");
3415: if (!aij->saved_values) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ORDER,"Must call MatStoreValues(A);first");
3416: /* copy values over */
3417: PetscMemcpy(aij->a,aij->saved_values,nz*sizeof(PetscScalar));
3418: return(0);
3419: }
3423: /*@
3424: MatRetrieveValues - Retrieves the copy of the matrix values; this allows, for
3425: example, reuse of the linear part of a Jacobian, while recomputing the
3426: nonlinear portion.
3428: Collect on Mat
3430: Input Parameters:
3431: . mat - the matrix (currently on AIJ matrices support this option)
3433: Level: advanced
3435: .seealso: MatStoreValues()
3437: @*/
3438: PetscErrorCode MatRetrieveValues(Mat mat)
3439: {
3444: if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3445: if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3446: PetscUseMethod(mat,"MatRetrieveValues_C",(Mat),(mat));
3447: return(0);
3448: }
3451: /* --------------------------------------------------------------------------------*/
3454: /*@C
3455: MatCreateSeqAIJ - Creates a sparse matrix in AIJ (compressed row) format
3456: (the default parallel PETSc format). For good matrix assembly performance
3457: the user should preallocate the matrix storage by setting the parameter nz
3458: (or the array nnz). By setting these parameters accurately, performance
3459: during matrix assembly can be increased by more than a factor of 50.
3461: Collective on MPI_Comm
3463: Input Parameters:
3464: + comm - MPI communicator, set to PETSC_COMM_SELF
3465: . m - number of rows
3466: . n - number of columns
3467: . nz - number of nonzeros per row (same for all rows)
3468: - nnz - array containing the number of nonzeros in the various rows
3469: (possibly different for each row) or NULL
3471: Output Parameter:
3472: . A - the matrix
3474: It is recommended that one use the MatCreate(), MatSetType() and/or MatSetFromOptions(),
3475: MatXXXXSetPreallocation() paradgm instead of this routine directly.
3476: [MatXXXXSetPreallocation() is, for example, MatSeqAIJSetPreallocation]
3478: Notes:
3479: If nnz is given then nz is ignored
3481: The AIJ format (also called the Yale sparse matrix format or
3482: compressed row storage), is fully compatible with standard Fortran 77
3483: storage. That is, the stored row and column indices can begin at
3484: either one (as in Fortran) or zero. See the users' manual for details.
3486: Specify the preallocated storage with either nz or nnz (not both).
3487: Set nz=PETSC_DEFAULT and nnz=NULL for PETSc to control dynamic memory
3488: allocation. For large problems you MUST preallocate memory or you
3489: will get TERRIBLE performance, see the users' manual chapter on matrices.
3491: By default, this format uses inodes (identical nodes) when possible, to
3492: improve numerical efficiency of matrix-vector products and solves. We
3493: search for consecutive rows with the same nonzero structure, thereby
3494: reusing matrix information to achieve increased efficiency.
3496: Options Database Keys:
3497: + -mat_no_inode - Do not use inodes
3498: - -mat_inode_limit <limit> - Sets inode limit (max limit=5)
3500: Level: intermediate
3502: .seealso: MatCreate(), MatCreateAIJ(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays()
3504: @*/
3505: PetscErrorCode MatCreateSeqAIJ(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt nz,const PetscInt nnz[],Mat *A)
3506: {
3510: MatCreate(comm,A);
3511: MatSetSizes(*A,m,n,m,n);
3512: MatSetType(*A,MATSEQAIJ);
3513: MatSeqAIJSetPreallocation_SeqAIJ(*A,nz,nnz);
3514: return(0);
3515: }
3519: /*@C
3520: MatSeqAIJSetPreallocation - For good matrix assembly performance
3521: the user should preallocate the matrix storage by setting the parameter nz
3522: (or the array nnz). By setting these parameters accurately, performance
3523: during matrix assembly can be increased by more than a factor of 50.
3525: Collective on MPI_Comm
3527: Input Parameters:
3528: + B - The matrix
3529: . nz - number of nonzeros per row (same for all rows)
3530: - nnz - array containing the number of nonzeros in the various rows
3531: (possibly different for each row) or NULL
3533: Notes:
3534: If nnz is given then nz is ignored
3536: The AIJ format (also called the Yale sparse matrix format or
3537: compressed row storage), is fully compatible with standard Fortran 77
3538: storage. That is, the stored row and column indices can begin at
3539: either one (as in Fortran) or zero. See the users' manual for details.
3541: Specify the preallocated storage with either nz or nnz (not both).
3542: Set nz=PETSC_DEFAULT and nnz=NULL for PETSc to control dynamic memory
3543: allocation. For large problems you MUST preallocate memory or you
3544: will get TERRIBLE performance, see the users' manual chapter on matrices.
3546: You can call MatGetInfo() to get information on how effective the preallocation was;
3547: for example the fields mallocs,nz_allocated,nz_used,nz_unneeded;
3548: You can also run with the option -info and look for messages with the string
3549: malloc in them to see if additional memory allocation was needed.
3551: Developers: Use nz of MAT_SKIP_ALLOCATION to not allocate any space for the matrix
3552: entries or columns indices
3554: By default, this format uses inodes (identical nodes) when possible, to
3555: improve numerical efficiency of matrix-vector products and solves. We
3556: search for consecutive rows with the same nonzero structure, thereby
3557: reusing matrix information to achieve increased efficiency.
3559: Options Database Keys:
3560: + -mat_no_inode - Do not use inodes
3561: . -mat_inode_limit <limit> - Sets inode limit (max limit=5)
3562: - -mat_aij_oneindex - Internally use indexing starting at 1
3563: rather than 0. Note that when calling MatSetValues(),
3564: the user still MUST index entries starting at 0!
3566: Level: intermediate
3568: .seealso: MatCreate(), MatCreateAIJ(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays(), MatGetInfo()
3570: @*/
3571: PetscErrorCode MatSeqAIJSetPreallocation(Mat B,PetscInt nz,const PetscInt nnz[])
3572: {
3578: PetscTryMethod(B,"MatSeqAIJSetPreallocation_C",(Mat,PetscInt,const PetscInt[]),(B,nz,nnz));
3579: return(0);
3580: }
3584: PetscErrorCode MatSeqAIJSetPreallocation_SeqAIJ(Mat B,PetscInt nz,const PetscInt *nnz)
3585: {
3586: Mat_SeqAIJ *b;
3587: PetscBool skipallocation = PETSC_FALSE,realalloc = PETSC_FALSE;
3589: PetscInt i;
3592: if (nz >= 0 || nnz) realalloc = PETSC_TRUE;
3593: if (nz == MAT_SKIP_ALLOCATION) {
3594: skipallocation = PETSC_TRUE;
3595: nz = 0;
3596: }
3598: PetscLayoutSetUp(B->rmap);
3599: PetscLayoutSetUp(B->cmap);
3601: if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 5;
3602: if (nz < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"nz cannot be less than 0: value %D",nz);
3603: if (nnz) {
3604: for (i=0; i<B->rmap->n; i++) {
3605: 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]);
3606: if (nnz[i] > B->cmap->n) 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],B->cmap->n);
3607: }
3608: }
3610: B->preallocated = PETSC_TRUE;
3612: b = (Mat_SeqAIJ*)B->data;
3614: if (!skipallocation) {
3615: if (!b->imax) {
3616: PetscMalloc2(B->rmap->n,&b->imax,B->rmap->n,&b->ilen);
3617: PetscLogObjectMemory((PetscObject)B,2*B->rmap->n*sizeof(PetscInt));
3618: }
3619: if (!nnz) {
3620: if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 10;
3621: else if (nz < 0) nz = 1;
3622: for (i=0; i<B->rmap->n; i++) b->imax[i] = nz;
3623: nz = nz*B->rmap->n;
3624: } else {
3625: nz = 0;
3626: for (i=0; i<B->rmap->n; i++) {b->imax[i] = nnz[i]; nz += nnz[i];}
3627: }
3628: /* b->ilen will count nonzeros in each row so far. */
3629: for (i=0; i<B->rmap->n; i++) b->ilen[i] = 0;
3631: /* allocate the matrix space */
3632: /* FIXME: should B's old memory be unlogged? */
3633: MatSeqXAIJFreeAIJ(B,&b->a,&b->j,&b->i);
3634: PetscMalloc3(nz,&b->a,nz,&b->j,B->rmap->n+1,&b->i);
3635: PetscLogObjectMemory((PetscObject)B,(B->rmap->n+1)*sizeof(PetscInt)+nz*(sizeof(PetscScalar)+sizeof(PetscInt)));
3636: b->i[0] = 0;
3637: for (i=1; i<B->rmap->n+1; i++) {
3638: b->i[i] = b->i[i-1] + b->imax[i-1];
3639: }
3640: b->singlemalloc = PETSC_TRUE;
3641: b->free_a = PETSC_TRUE;
3642: b->free_ij = PETSC_TRUE;
3643: } else {
3644: b->free_a = PETSC_FALSE;
3645: b->free_ij = PETSC_FALSE;
3646: }
3648: b->nz = 0;
3649: b->maxnz = nz;
3650: B->info.nz_unneeded = (double)b->maxnz;
3651: if (realalloc) {
3652: MatSetOption(B,MAT_NEW_NONZERO_ALLOCATION_ERR,PETSC_TRUE);
3653: }
3654: return(0);
3655: }
3657: #undef __FUNCT__
3659: /*@
3660: MatSeqAIJSetPreallocationCSR - Allocates memory for a sparse sequential matrix in AIJ format.
3662: Input Parameters:
3663: + B - the matrix
3664: . i - the indices into j for the start of each row (starts with zero)
3665: . j - the column indices for each row (starts with zero) these must be sorted for each row
3666: - v - optional values in the matrix
3668: Level: developer
3670: The i,j,v values are COPIED with this routine; to avoid the copy use MatCreateSeqAIJWithArrays()
3672: .keywords: matrix, aij, compressed row, sparse, sequential
3674: .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatSeqAIJSetPreallocation(), MatCreateSeqAIJ(), SeqAIJ
3675: @*/
3676: PetscErrorCode MatSeqAIJSetPreallocationCSR(Mat B,const PetscInt i[],const PetscInt j[],const PetscScalar v[])
3677: {
3683: PetscTryMethod(B,"MatSeqAIJSetPreallocationCSR_C",(Mat,const PetscInt[],const PetscInt[],const PetscScalar[]),(B,i,j,v));
3684: return(0);
3685: }
3687: #undef __FUNCT__
3689: PetscErrorCode MatSeqAIJSetPreallocationCSR_SeqAIJ(Mat B,const PetscInt Ii[],const PetscInt J[],const PetscScalar v[])
3690: {
3691: PetscInt i;
3692: PetscInt m,n;
3693: PetscInt nz;
3694: PetscInt *nnz, nz_max = 0;
3695: PetscScalar *values;
3699: if (Ii[0]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE, "Ii[0] must be 0 it is %D", Ii[0]);
3701: PetscLayoutSetUp(B->rmap);
3702: PetscLayoutSetUp(B->cmap);
3704: MatGetSize(B, &m, &n);
3705: PetscMalloc1(m+1, &nnz);
3706: for (i = 0; i < m; i++) {
3707: nz = Ii[i+1]- Ii[i];
3708: nz_max = PetscMax(nz_max, nz);
3709: if (nz < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE, "Local row %D has a negative number of columns %D", i, nnz);
3710: nnz[i] = nz;
3711: }
3712: MatSeqAIJSetPreallocation(B, 0, nnz);
3713: PetscFree(nnz);
3715: if (v) {
3716: values = (PetscScalar*) v;
3717: } else {
3718: PetscCalloc1(nz_max, &values);
3719: }
3721: for (i = 0; i < m; i++) {
3722: nz = Ii[i+1] - Ii[i];
3723: MatSetValues_SeqAIJ(B, 1, &i, nz, J+Ii[i], values + (v ? Ii[i] : 0), INSERT_VALUES);
3724: }
3726: MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);
3727: MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);
3729: if (!v) {
3730: PetscFree(values);
3731: }
3732: MatSetOption(B,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_TRUE);
3733: return(0);
3734: }
3736: #include <../src/mat/impls/dense/seq/dense.h>
3737: #include <petsc/private/kernels/petscaxpy.h>
3741: /*
3742: Computes (B'*A')' since computing B*A directly is untenable
3744: n p p
3745: ( ) ( ) ( )
3746: m ( A ) * n ( B ) = m ( C )
3747: ( ) ( ) ( )
3749: */
3750: PetscErrorCode MatMatMultNumeric_SeqDense_SeqAIJ(Mat A,Mat B,Mat C)
3751: {
3752: PetscErrorCode ierr;
3753: Mat_SeqDense *sub_a = (Mat_SeqDense*)A->data;
3754: Mat_SeqAIJ *sub_b = (Mat_SeqAIJ*)B->data;
3755: Mat_SeqDense *sub_c = (Mat_SeqDense*)C->data;
3756: PetscInt i,n,m,q,p;
3757: const PetscInt *ii,*idx;
3758: const PetscScalar *b,*a,*a_q;
3759: PetscScalar *c,*c_q;
3762: m = A->rmap->n;
3763: n = A->cmap->n;
3764: p = B->cmap->n;
3765: a = sub_a->v;
3766: b = sub_b->a;
3767: c = sub_c->v;
3768: PetscMemzero(c,m*p*sizeof(PetscScalar));
3770: ii = sub_b->i;
3771: idx = sub_b->j;
3772: for (i=0; i<n; i++) {
3773: q = ii[i+1] - ii[i];
3774: while (q-->0) {
3775: c_q = c + m*(*idx);
3776: a_q = a + m*i;
3777: PetscKernelAXPY(c_q,*b,a_q,m);
3778: idx++;
3779: b++;
3780: }
3781: }
3782: return(0);
3783: }
3787: PetscErrorCode MatMatMultSymbolic_SeqDense_SeqAIJ(Mat A,Mat B,PetscReal fill,Mat *C)
3788: {
3790: PetscInt m=A->rmap->n,n=B->cmap->n;
3791: Mat Cmat;
3794: if (A->cmap->n != B->rmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"A->cmap->n %D != B->rmap->n %D\n",A->cmap->n,B->rmap->n);
3795: MatCreate(PetscObjectComm((PetscObject)A),&Cmat);
3796: MatSetSizes(Cmat,m,n,m,n);
3797: MatSetBlockSizesFromMats(Cmat,A,B);
3798: MatSetType(Cmat,MATSEQDENSE);
3799: MatSeqDenseSetPreallocation(Cmat,NULL);
3801: Cmat->ops->matmultnumeric = MatMatMultNumeric_SeqDense_SeqAIJ;
3803: *C = Cmat;
3804: return(0);
3805: }
3807: /* ----------------------------------------------------------------*/
3810: PETSC_INTERN PetscErrorCode MatMatMult_SeqDense_SeqAIJ(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C)
3811: {
3815: if (scall == MAT_INITIAL_MATRIX) {
3816: PetscLogEventBegin(MAT_MatMultSymbolic,A,B,0,0);
3817: MatMatMultSymbolic_SeqDense_SeqAIJ(A,B,fill,C);
3818: PetscLogEventEnd(MAT_MatMultSymbolic,A,B,0,0);
3819: }
3820: PetscLogEventBegin(MAT_MatMultNumeric,A,B,0,0);
3821: MatMatMultNumeric_SeqDense_SeqAIJ(A,B,*C);
3822: PetscLogEventEnd(MAT_MatMultNumeric,A,B,0,0);
3823: return(0);
3824: }
3827: /*MC
3828: MATSEQAIJ - MATSEQAIJ = "seqaij" - A matrix type to be used for sequential sparse matrices,
3829: based on compressed sparse row format.
3831: Options Database Keys:
3832: . -mat_type seqaij - sets the matrix type to "seqaij" during a call to MatSetFromOptions()
3834: Level: beginner
3836: .seealso: MatCreateSeqAIJ(), MatSetFromOptions(), MatSetType(), MatCreate(), MatType
3837: M*/
3839: /*MC
3840: MATAIJ - MATAIJ = "aij" - A matrix type to be used for sparse matrices.
3842: This matrix type is identical to MATSEQAIJ when constructed with a single process communicator,
3843: and MATMPIAIJ otherwise. As a result, for single process communicators,
3844: MatSeqAIJSetPreallocation is supported, and similarly MatMPIAIJSetPreallocation is supported
3845: for communicators controlling multiple processes. It is recommended that you call both of
3846: the above preallocation routines for simplicity.
3848: Options Database Keys:
3849: . -mat_type aij - sets the matrix type to "aij" during a call to MatSetFromOptions()
3851: Developer Notes: Subclasses include MATAIJCUSP, MATAIJPERM, MATAIJCRL, and also automatically switches over to use inodes when
3852: enough exist.
3854: Level: beginner
3856: .seealso: MatCreateAIJ(), MatCreateSeqAIJ(), MATSEQAIJ,MATMPIAIJ
3857: M*/
3859: /*MC
3860: MATAIJCRL - MATAIJCRL = "aijcrl" - A matrix type to be used for sparse matrices.
3862: This matrix type is identical to MATSEQAIJCRL when constructed with a single process communicator,
3863: and MATMPIAIJCRL otherwise. As a result, for single process communicators,
3864: MatSeqAIJSetPreallocation() is supported, and similarly MatMPIAIJSetPreallocation() is supported
3865: for communicators controlling multiple processes. It is recommended that you call both of
3866: the above preallocation routines for simplicity.
3868: Options Database Keys:
3869: . -mat_type aijcrl - sets the matrix type to "aijcrl" during a call to MatSetFromOptions()
3871: Level: beginner
3873: .seealso: MatCreateMPIAIJCRL,MATSEQAIJCRL,MATMPIAIJCRL, MATSEQAIJCRL, MATMPIAIJCRL
3874: M*/
3876: PETSC_INTERN PetscErrorCode MatConvert_SeqAIJ_SeqAIJCRL(Mat,MatType,MatReuse,Mat*);
3877: #if defined(PETSC_HAVE_ELEMENTAL)
3878: PETSC_INTERN PetscErrorCode MatConvert_SeqAIJ_Elemental(Mat,MatType,MatReuse,Mat*);
3879: #endif
3880: PETSC_INTERN PetscErrorCode MatConvert_SeqAIJ_SeqDense(Mat,MatType,MatReuse,Mat*);
3882: #if defined(PETSC_HAVE_MATLAB_ENGINE)
3883: PETSC_EXTERN PetscErrorCode MatlabEnginePut_SeqAIJ(PetscObject,void*);
3884: PETSC_EXTERN PetscErrorCode MatlabEngineGet_SeqAIJ(PetscObject,void*);
3885: #endif
3890: /*@C
3891: MatSeqAIJGetArray - gives access to the array where the data for a SeqSeqAIJ matrix is stored
3893: Not Collective
3895: Input Parameter:
3896: . mat - a MATSEQAIJ matrix
3898: Output Parameter:
3899: . array - pointer to the data
3901: Level: intermediate
3903: .seealso: MatSeqAIJRestoreArray(), MatSeqAIJGetArrayF90()
3904: @*/
3905: PetscErrorCode MatSeqAIJGetArray(Mat A,PetscScalar **array)
3906: {
3910: PetscUseMethod(A,"MatSeqAIJGetArray_C",(Mat,PetscScalar**),(A,array));
3911: return(0);
3912: }
3916: /*@C
3917: MatSeqAIJGetMaxRowNonzeros - returns the maximum number of nonzeros in any row
3919: Not Collective
3921: Input Parameter:
3922: . mat - a MATSEQAIJ matrix
3924: Output Parameter:
3925: . nz - the maximum number of nonzeros in any row
3927: Level: intermediate
3929: .seealso: MatSeqAIJRestoreArray(), MatSeqAIJGetArrayF90()
3930: @*/
3931: PetscErrorCode MatSeqAIJGetMaxRowNonzeros(Mat A,PetscInt *nz)
3932: {
3933: Mat_SeqAIJ *aij = (Mat_SeqAIJ*)A->data;
3936: *nz = aij->rmax;
3937: return(0);
3938: }
3942: /*@C
3943: MatSeqAIJRestoreArray - returns access to the array where the data for a MATSEQAIJ matrix is stored obtained by MatSeqAIJGetArray()
3945: Not Collective
3947: Input Parameters:
3948: . mat - a MATSEQAIJ matrix
3949: . array - pointer to the data
3951: Level: intermediate
3953: .seealso: MatSeqAIJGetArray(), MatSeqAIJRestoreArrayF90()
3954: @*/
3955: PetscErrorCode MatSeqAIJRestoreArray(Mat A,PetscScalar **array)
3956: {
3960: PetscUseMethod(A,"MatSeqAIJRestoreArray_C",(Mat,PetscScalar**),(A,array));
3961: return(0);
3962: }
3966: PETSC_EXTERN PetscErrorCode MatCreate_SeqAIJ(Mat B)
3967: {
3968: Mat_SeqAIJ *b;
3970: PetscMPIInt size;
3973: MPI_Comm_size(PetscObjectComm((PetscObject)B),&size);
3974: if (size > 1) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Comm must be of size 1");
3976: PetscNewLog(B,&b);
3978: B->data = (void*)b;
3980: PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));
3982: b->row = 0;
3983: b->col = 0;
3984: b->icol = 0;
3985: b->reallocs = 0;
3986: b->ignorezeroentries = PETSC_FALSE;
3987: b->roworiented = PETSC_TRUE;
3988: b->nonew = 0;
3989: b->diag = 0;
3990: b->solve_work = 0;
3991: B->spptr = 0;
3992: b->saved_values = 0;
3993: b->idiag = 0;
3994: b->mdiag = 0;
3995: b->ssor_work = 0;
3996: b->omega = 1.0;
3997: b->fshift = 0.0;
3998: b->idiagvalid = PETSC_FALSE;
3999: b->ibdiagvalid = PETSC_FALSE;
4000: b->keepnonzeropattern = PETSC_FALSE;
4002: PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJ);
4003: PetscObjectComposeFunction((PetscObject)B,"MatSeqAIJGetArray_C",MatSeqAIJGetArray_SeqAIJ);
4004: PetscObjectComposeFunction((PetscObject)B,"MatSeqAIJRestoreArray_C",MatSeqAIJRestoreArray_SeqAIJ);
4006: #if defined(PETSC_HAVE_MATLAB_ENGINE)
4007: PetscObjectComposeFunction((PetscObject)B,"PetscMatlabEnginePut_C",MatlabEnginePut_SeqAIJ);
4008: PetscObjectComposeFunction((PetscObject)B,"PetscMatlabEngineGet_C",MatlabEngineGet_SeqAIJ);
4009: #endif
4011: PetscObjectComposeFunction((PetscObject)B,"MatSeqAIJSetColumnIndices_C",MatSeqAIJSetColumnIndices_SeqAIJ);
4012: PetscObjectComposeFunction((PetscObject)B,"MatStoreValues_C",MatStoreValues_SeqAIJ);
4013: PetscObjectComposeFunction((PetscObject)B,"MatRetrieveValues_C",MatRetrieveValues_SeqAIJ);
4014: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_seqsbaij_C",MatConvert_SeqAIJ_SeqSBAIJ);
4015: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_seqbaij_C",MatConvert_SeqAIJ_SeqBAIJ);
4016: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_seqaijperm_C",MatConvert_SeqAIJ_SeqAIJPERM);
4017: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_seqaijcrl_C",MatConvert_SeqAIJ_SeqAIJCRL);
4018: #if defined(PETSC_HAVE_ELEMENTAL)
4019: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_elemental_C",MatConvert_SeqAIJ_Elemental);
4020: #endif
4021: PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_seqdense_C",MatConvert_SeqAIJ_SeqDense);
4022: PetscObjectComposeFunction((PetscObject)B,"MatIsTranspose_C",MatIsTranspose_SeqAIJ);
4023: PetscObjectComposeFunction((PetscObject)B,"MatIsHermitianTranspose_C",MatIsTranspose_SeqAIJ);
4024: PetscObjectComposeFunction((PetscObject)B,"MatSeqAIJSetPreallocation_C",MatSeqAIJSetPreallocation_SeqAIJ);
4025: PetscObjectComposeFunction((PetscObject)B,"MatSeqAIJSetPreallocationCSR_C",MatSeqAIJSetPreallocationCSR_SeqAIJ);
4026: PetscObjectComposeFunction((PetscObject)B,"MatReorderForNonzeroDiagonal_C",MatReorderForNonzeroDiagonal_SeqAIJ);
4027: PetscObjectComposeFunction((PetscObject)B,"MatMatMult_seqdense_seqaij_C",MatMatMult_SeqDense_SeqAIJ);
4028: PetscObjectComposeFunction((PetscObject)B,"MatMatMultSymbolic_seqdense_seqaij_C",MatMatMultSymbolic_SeqDense_SeqAIJ);
4029: PetscObjectComposeFunction((PetscObject)B,"MatMatMultNumeric_seqdense_seqaij_C",MatMatMultNumeric_SeqDense_SeqAIJ);
4030: MatCreate_SeqAIJ_Inode(B);
4031: PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJ);
4032: return(0);
4033: }
4037: /*
4038: Given a matrix generated with MatGetFactor() duplicates all the information in A into B
4039: */
4040: PetscErrorCode MatDuplicateNoCreate_SeqAIJ(Mat C,Mat A,MatDuplicateOption cpvalues,PetscBool mallocmatspace)
4041: {
4042: Mat_SeqAIJ *c,*a = (Mat_SeqAIJ*)A->data;
4044: PetscInt i,m = A->rmap->n;
4047: c = (Mat_SeqAIJ*)C->data;
4049: C->factortype = A->factortype;
4050: c->row = 0;
4051: c->col = 0;
4052: c->icol = 0;
4053: c->reallocs = 0;
4055: C->assembled = PETSC_TRUE;
4057: PetscLayoutReference(A->rmap,&C->rmap);
4058: PetscLayoutReference(A->cmap,&C->cmap);
4060: PetscMalloc2(m,&c->imax,m,&c->ilen);
4061: PetscLogObjectMemory((PetscObject)C, 2*m*sizeof(PetscInt));
4062: for (i=0; i<m; i++) {
4063: c->imax[i] = a->imax[i];
4064: c->ilen[i] = a->ilen[i];
4065: }
4067: /* allocate the matrix space */
4068: if (mallocmatspace) {
4069: PetscMalloc3(a->i[m],&c->a,a->i[m],&c->j,m+1,&c->i);
4070: PetscLogObjectMemory((PetscObject)C, a->i[m]*(sizeof(PetscScalar)+sizeof(PetscInt))+(m+1)*sizeof(PetscInt));
4072: c->singlemalloc = PETSC_TRUE;
4074: PetscMemcpy(c->i,a->i,(m+1)*sizeof(PetscInt));
4075: if (m > 0) {
4076: PetscMemcpy(c->j,a->j,(a->i[m])*sizeof(PetscInt));
4077: if (cpvalues == MAT_COPY_VALUES) {
4078: PetscMemcpy(c->a,a->a,(a->i[m])*sizeof(PetscScalar));
4079: } else {
4080: PetscMemzero(c->a,(a->i[m])*sizeof(PetscScalar));
4081: }
4082: }
4083: }
4085: c->ignorezeroentries = a->ignorezeroentries;
4086: c->roworiented = a->roworiented;
4087: c->nonew = a->nonew;
4088: if (a->diag) {
4089: PetscMalloc1(m+1,&c->diag);
4090: PetscLogObjectMemory((PetscObject)C,(m+1)*sizeof(PetscInt));
4091: for (i=0; i<m; i++) {
4092: c->diag[i] = a->diag[i];
4093: }
4094: } else c->diag = 0;
4096: c->solve_work = 0;
4097: c->saved_values = 0;
4098: c->idiag = 0;
4099: c->ssor_work = 0;
4100: c->keepnonzeropattern = a->keepnonzeropattern;
4101: c->free_a = PETSC_TRUE;
4102: c->free_ij = PETSC_TRUE;
4104: c->rmax = a->rmax;
4105: c->nz = a->nz;
4106: c->maxnz = a->nz; /* Since we allocate exactly the right amount */
4107: C->preallocated = PETSC_TRUE;
4109: c->compressedrow.use = a->compressedrow.use;
4110: c->compressedrow.nrows = a->compressedrow.nrows;
4111: if (a->compressedrow.use) {
4112: i = a->compressedrow.nrows;
4113: PetscMalloc2(i+1,&c->compressedrow.i,i,&c->compressedrow.rindex);
4114: PetscMemcpy(c->compressedrow.i,a->compressedrow.i,(i+1)*sizeof(PetscInt));
4115: PetscMemcpy(c->compressedrow.rindex,a->compressedrow.rindex,i*sizeof(PetscInt));
4116: } else {
4117: c->compressedrow.use = PETSC_FALSE;
4118: c->compressedrow.i = NULL;
4119: c->compressedrow.rindex = NULL;
4120: }
4121: c->nonzerorowcnt = a->nonzerorowcnt;
4122: C->nonzerostate = A->nonzerostate;
4124: MatDuplicate_SeqAIJ_Inode(A,cpvalues,&C);
4125: PetscFunctionListDuplicate(((PetscObject)A)->qlist,&((PetscObject)C)->qlist);
4126: return(0);
4127: }
4131: PetscErrorCode MatDuplicate_SeqAIJ(Mat A,MatDuplicateOption cpvalues,Mat *B)
4132: {
4136: MatCreate(PetscObjectComm((PetscObject)A),B);
4137: MatSetSizes(*B,A->rmap->n,A->cmap->n,A->rmap->n,A->cmap->n);
4138: if (!(A->rmap->n % A->rmap->bs) && !(A->cmap->n % A->cmap->bs)) {
4139: MatSetBlockSizesFromMats(*B,A,A);
4140: }
4141: MatSetType(*B,((PetscObject)A)->type_name);
4142: MatDuplicateNoCreate_SeqAIJ(*B,A,cpvalues,PETSC_TRUE);
4143: return(0);
4144: }
4148: PetscErrorCode MatLoad_SeqAIJ(Mat newMat, PetscViewer viewer)
4149: {
4150: Mat_SeqAIJ *a;
4152: PetscInt i,sum,nz,header[4],*rowlengths = 0,M,N,rows,cols;
4153: int fd;
4154: PetscMPIInt size;
4155: MPI_Comm comm;
4156: PetscInt bs = newMat->rmap->bs;
4159: /* force binary viewer to load .info file if it has not yet done so */
4160: PetscViewerSetUp(viewer);
4161: PetscObjectGetComm((PetscObject)viewer,&comm);
4162: MPI_Comm_size(comm,&size);
4163: if (size > 1) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"view must have one processor");
4165: PetscOptionsBegin(comm,NULL,"Options for loading SEQAIJ matrix","Mat");
4166: PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,NULL);
4167: PetscOptionsEnd();
4168: if (bs < 0) bs = 1;
4169: MatSetBlockSize(newMat,bs);
4171: PetscViewerBinaryGetDescriptor(viewer,&fd);
4172: PetscBinaryRead(fd,header,4,PETSC_INT);
4173: if (header[0] != MAT_FILE_CLASSID) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"not matrix object in file");
4174: M = header[1]; N = header[2]; nz = header[3];
4176: if (nz < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format on disk,cannot load as SeqAIJ");
4178: /* read in row lengths */
4179: PetscMalloc1(M,&rowlengths);
4180: PetscBinaryRead(fd,rowlengths,M,PETSC_INT);
4182: /* check if sum of rowlengths is same as nz */
4183: for (i=0,sum=0; i< M; i++) sum +=rowlengths[i];
4184: if (sum != nz) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_FILE_READ,"Inconsistant matrix data in file. no-nonzeros = %dD, sum-row-lengths = %D\n",nz,sum);
4186: /* set global size if not set already*/
4187: if (newMat->rmap->n < 0 && newMat->rmap->N < 0 && newMat->cmap->n < 0 && newMat->cmap->N < 0) {
4188: MatSetSizes(newMat,PETSC_DECIDE,PETSC_DECIDE,M,N);
4189: } else {
4190: /* if sizes and type are already set, check if the matrix global sizes are correct */
4191: MatGetSize(newMat,&rows,&cols);
4192: if (rows < 0 && cols < 0) { /* user might provide local size instead of global size */
4193: MatGetLocalSize(newMat,&rows,&cols);
4194: }
4195: if (M != rows || N != cols) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED, "Matrix in file of different length (%D, %D) than the input matrix (%D, %D)",M,N,rows,cols);
4196: }
4197: MatSeqAIJSetPreallocation_SeqAIJ(newMat,0,rowlengths);
4198: a = (Mat_SeqAIJ*)newMat->data;
4200: PetscBinaryRead(fd,a->j,nz,PETSC_INT);
4202: /* read in nonzero values */
4203: PetscBinaryRead(fd,a->a,nz,PETSC_SCALAR);
4205: /* set matrix "i" values */
4206: a->i[0] = 0;
4207: for (i=1; i<= M; i++) {
4208: a->i[i] = a->i[i-1] + rowlengths[i-1];
4209: a->ilen[i-1] = rowlengths[i-1];
4210: }
4211: PetscFree(rowlengths);
4213: MatAssemblyBegin(newMat,MAT_FINAL_ASSEMBLY);
4214: MatAssemblyEnd(newMat,MAT_FINAL_ASSEMBLY);
4215: return(0);
4216: }
4220: PetscErrorCode MatEqual_SeqAIJ(Mat A,Mat B,PetscBool * flg)
4221: {
4222: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data,*b = (Mat_SeqAIJ*)B->data;
4224: #if defined(PETSC_USE_COMPLEX)
4225: PetscInt k;
4226: #endif
4229: /* If the matrix dimensions are not equal,or no of nonzeros */
4230: if ((A->rmap->n != B->rmap->n) || (A->cmap->n != B->cmap->n) ||(a->nz != b->nz)) {
4231: *flg = PETSC_FALSE;
4232: return(0);
4233: }
4235: /* if the a->i are the same */
4236: PetscMemcmp(a->i,b->i,(A->rmap->n+1)*sizeof(PetscInt),flg);
4237: if (!*flg) return(0);
4239: /* if a->j are the same */
4240: PetscMemcmp(a->j,b->j,(a->nz)*sizeof(PetscInt),flg);
4241: if (!*flg) return(0);
4243: /* if a->a are the same */
4244: #if defined(PETSC_USE_COMPLEX)
4245: for (k=0; k<a->nz; k++) {
4246: if (PetscRealPart(a->a[k]) != PetscRealPart(b->a[k]) || PetscImaginaryPart(a->a[k]) != PetscImaginaryPart(b->a[k])) {
4247: *flg = PETSC_FALSE;
4248: return(0);
4249: }
4250: }
4251: #else
4252: PetscMemcmp(a->a,b->a,(a->nz)*sizeof(PetscScalar),flg);
4253: #endif
4254: return(0);
4255: }
4259: /*@
4260: MatCreateSeqAIJWithArrays - Creates an sequential AIJ matrix using matrix elements (in CSR format)
4261: provided by the user.
4263: Collective on MPI_Comm
4265: Input Parameters:
4266: + comm - must be an MPI communicator of size 1
4267: . m - number of rows
4268: . n - number of columns
4269: . i - row indices
4270: . j - column indices
4271: - a - matrix values
4273: Output Parameter:
4274: . mat - the matrix
4276: Level: intermediate
4278: Notes:
4279: The i, j, and a arrays are not copied by this routine, the user must free these arrays
4280: once the matrix is destroyed and not before
4282: You cannot set new nonzero locations into this matrix, that will generate an error.
4284: The i and j indices are 0 based
4286: The format which is used for the sparse matrix input, is equivalent to a
4287: row-major ordering.. i.e for the following matrix, the input data expected is
4288: as shown
4290: $ 1 0 0
4291: $ 2 0 3
4292: $ 4 5 6
4293: $
4294: $ i = {0,1,3,6} [size = nrow+1 = 3+1]
4295: $ j = {0,0,2,0,1,2} [size = 6]; values must be sorted for each row
4296: $ v = {1,2,3,4,5,6} [size = 6]
4299: .seealso: MatCreate(), MatCreateAIJ(), MatCreateSeqAIJ(), MatCreateMPIAIJWithArrays(), MatMPIAIJSetPreallocationCSR()
4301: @*/
4302: PetscErrorCode MatCreateSeqAIJWithArrays(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt *i,PetscInt *j,PetscScalar *a,Mat *mat)
4303: {
4305: PetscInt ii;
4306: Mat_SeqAIJ *aij;
4307: #if defined(PETSC_USE_DEBUG)
4308: PetscInt jj;
4309: #endif
4312: if (m > 0 && i[0]) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"i (row indices) must start with 0");
4313: MatCreate(comm,mat);
4314: MatSetSizes(*mat,m,n,m,n);
4315: /* MatSetBlockSizes(*mat,,); */
4316: MatSetType(*mat,MATSEQAIJ);
4317: MatSeqAIJSetPreallocation_SeqAIJ(*mat,MAT_SKIP_ALLOCATION,0);
4318: aij = (Mat_SeqAIJ*)(*mat)->data;
4319: PetscMalloc2(m,&aij->imax,m,&aij->ilen);
4321: aij->i = i;
4322: aij->j = j;
4323: aij->a = a;
4324: aij->singlemalloc = PETSC_FALSE;
4325: aij->nonew = -1; /*this indicates that inserting a new value in the matrix that generates a new nonzero is an error*/
4326: aij->free_a = PETSC_FALSE;
4327: aij->free_ij = PETSC_FALSE;
4329: for (ii=0; ii<m; ii++) {
4330: aij->ilen[ii] = aij->imax[ii] = i[ii+1] - i[ii];
4331: #if defined(PETSC_USE_DEBUG)
4332: if (i[ii+1] - i[ii] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row length in i (row indices) row = %D length = %D",ii,i[ii+1] - i[ii]);
4333: for (jj=i[ii]+1; jj<i[ii+1]; jj++) {
4334: if (j[jj] < j[jj-1]) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column entry number %D (actual colum %D) in row %D is not sorted",jj-i[ii],j[jj],ii);
4335: if (j[jj] == j[jj]-1) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column entry number %D (actual colum %D) in row %D is identical to previous entry",jj-i[ii],j[jj],ii);
4336: }
4337: #endif
4338: }
4339: #if defined(PETSC_USE_DEBUG)
4340: for (ii=0; ii<aij->i[m]; ii++) {
4341: if (j[ii] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative column index at location = %D index = %D",ii,j[ii]);
4342: if (j[ii] > n - 1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column index to large at location = %D index = %D",ii,j[ii]);
4343: }
4344: #endif
4346: MatAssemblyBegin(*mat,MAT_FINAL_ASSEMBLY);
4347: MatAssemblyEnd(*mat,MAT_FINAL_ASSEMBLY);
4348: return(0);
4349: }
4352: /*@C
4353: MatCreateSeqAIJFromTriple - Creates an sequential AIJ matrix using matrix elements (in COO format)
4354: provided by the user.
4356: Collective on MPI_Comm
4358: Input Parameters:
4359: + comm - must be an MPI communicator of size 1
4360: . m - number of rows
4361: . n - number of columns
4362: . i - row indices
4363: . j - column indices
4364: . a - matrix values
4365: . nz - number of nonzeros
4366: - idx - 0 or 1 based
4368: Output Parameter:
4369: . mat - the matrix
4371: Level: intermediate
4373: Notes:
4374: The i and j indices are 0 based
4376: The format which is used for the sparse matrix input, is equivalent to a
4377: row-major ordering.. i.e for the following matrix, the input data expected is
4378: as shown:
4380: 1 0 0
4381: 2 0 3
4382: 4 5 6
4384: i = {0,1,1,2,2,2}
4385: j = {0,0,2,0,1,2}
4386: v = {1,2,3,4,5,6}
4389: .seealso: MatCreate(), MatCreateAIJ(), MatCreateSeqAIJ(), MatCreateSeqAIJWithArrays(), MatMPIAIJSetPreallocationCSR()
4391: @*/
4392: PetscErrorCode MatCreateSeqAIJFromTriple(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt *i,PetscInt *j,PetscScalar *a,Mat *mat,PetscInt nz,PetscBool idx)
4393: {
4395: PetscInt ii, *nnz, one = 1,row,col;
4399: PetscCalloc1(m,&nnz);
4400: for (ii = 0; ii < nz; ii++) {
4401: nnz[i[ii] - !!idx] += 1;
4402: }
4403: MatCreate(comm,mat);
4404: MatSetSizes(*mat,m,n,m,n);
4405: MatSetType(*mat,MATSEQAIJ);
4406: MatSeqAIJSetPreallocation_SeqAIJ(*mat,0,nnz);
4407: for (ii = 0; ii < nz; ii++) {
4408: if (idx) {
4409: row = i[ii] - 1;
4410: col = j[ii] - 1;
4411: } else {
4412: row = i[ii];
4413: col = j[ii];
4414: }
4415: MatSetValues(*mat,one,&row,one,&col,&a[ii],ADD_VALUES);
4416: }
4417: MatAssemblyBegin(*mat,MAT_FINAL_ASSEMBLY);
4418: MatAssemblyEnd(*mat,MAT_FINAL_ASSEMBLY);
4419: PetscFree(nnz);
4420: return(0);
4421: }
4425: PetscErrorCode MatSetColoring_SeqAIJ(Mat A,ISColoring coloring)
4426: {
4428: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
4431: if (coloring->ctype == IS_COLORING_GLOBAL) {
4432: ISColoringReference(coloring);
4433: a->coloring = coloring;
4434: } else if (coloring->ctype == IS_COLORING_GHOSTED) {
4435: PetscInt i,*larray;
4436: ISColoring ocoloring;
4437: ISColoringValue *colors;
4439: /* set coloring for diagonal portion */
4440: PetscMalloc1(A->cmap->n,&larray);
4441: for (i=0; i<A->cmap->n; i++) larray[i] = i;
4442: ISGlobalToLocalMappingApply(A->cmap->mapping,IS_GTOLM_MASK,A->cmap->n,larray,NULL,larray);
4443: PetscMalloc1(A->cmap->n,&colors);
4444: for (i=0; i<A->cmap->n; i++) colors[i] = coloring->colors[larray[i]];
4445: PetscFree(larray);
4446: ISColoringCreate(PETSC_COMM_SELF,coloring->n,A->cmap->n,colors,PETSC_OWN_POINTER,&ocoloring);
4447: a->coloring = ocoloring;
4448: }
4449: return(0);
4450: }
4454: PetscErrorCode MatSetValuesAdifor_SeqAIJ(Mat A,PetscInt nl,void *advalues)
4455: {
4456: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
4457: PetscInt m = A->rmap->n,*ii = a->i,*jj = a->j,nz,i,j;
4458: MatScalar *v = a->a;
4459: PetscScalar *values = (PetscScalar*)advalues;
4460: ISColoringValue *color;
4463: if (!a->coloring) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Coloring not set for matrix");
4464: color = a->coloring->colors;
4465: /* loop over rows */
4466: for (i=0; i<m; i++) {
4467: nz = ii[i+1] - ii[i];
4468: /* loop over columns putting computed value into matrix */
4469: for (j=0; j<nz; j++) *v++ = values[color[*jj++]];
4470: values += nl; /* jump to next row of derivatives */
4471: }
4472: return(0);
4473: }
4477: PetscErrorCode MatSeqAIJInvalidateDiagonal(Mat A)
4478: {
4479: Mat_SeqAIJ *a=(Mat_SeqAIJ*)A->data;
4483: a->idiagvalid = PETSC_FALSE;
4484: a->ibdiagvalid = PETSC_FALSE;
4486: MatSeqAIJInvalidateDiagonal_Inode(A);
4487: return(0);
4488: }
4492: PetscErrorCode MatCreateMPIMatConcatenateSeqMat_SeqAIJ(MPI_Comm comm,Mat inmat,PetscInt n,MatReuse scall,Mat *outmat)
4493: {
4497: MatCreateMPIMatConcatenateSeqMat_MPIAIJ(comm,inmat,n,scall,outmat);
4498: return(0);
4499: }
4501: /*
4502: Permute A into C's *local* index space using rowemb,colemb.
4503: The embedding are supposed to be injections and the above implies that the range of rowemb is a subset
4504: of [0,m), colemb is in [0,n).
4505: If pattern == DIFFERENT_NONZERO_PATTERN, C is preallocated according to A.
4506: */
4509: PetscErrorCode MatSetSeqMat_SeqAIJ(Mat C,IS rowemb,IS colemb,MatStructure pattern,Mat B)
4510: {
4511: /* If making this function public, change the error returned in this function away from _PLIB. */
4513: Mat_SeqAIJ *Baij;
4514: PetscBool seqaij;
4515: PetscInt m,n,*nz,i,j,count;
4516: PetscScalar v;
4517: const PetscInt *rowindices,*colindices;
4520: if (!B) return(0);
4521: /* Check to make sure the target matrix (and embeddings) are compatible with C and each other. */
4522: PetscObjectTypeCompare((PetscObject)B,MATSEQAIJ,&seqaij);
4523: if (!seqaij) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Input matrix is of wrong type");
4524: if (rowemb) {
4525: ISGetLocalSize(rowemb,&m);
4526: if (m != B->rmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Row IS of size %D is incompatible with matrix row size %D",m,B->rmap->n);
4527: } else {
4528: if (C->rmap->n != B->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Input matrix is row-incompatible with the target matrix");
4529: }
4530: if (colemb) {
4531: ISGetLocalSize(colemb,&n);
4532: if (n != B->cmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Diag col IS of size %D is incompatible with input matrix col size %D",n,B->cmap->n);
4533: } else {
4534: if (C->cmap->n != B->cmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Input matrix is col-incompatible with the target matrix");
4535: }
4537: Baij = (Mat_SeqAIJ*)(B->data);
4538: if (pattern == DIFFERENT_NONZERO_PATTERN) {
4539: PetscMalloc1(B->rmap->n,&nz);
4540: for (i=0; i<B->rmap->n; i++) {
4541: nz[i] = Baij->i[i+1] - Baij->i[i];
4542: }
4543: MatSeqAIJSetPreallocation(C,0,nz);
4544: PetscFree(nz);
4545: }
4546: if (pattern == SUBSET_NONZERO_PATTERN) {
4547: MatZeroEntries(C);
4548: }
4549: count = 0;
4550: rowindices = NULL;
4551: colindices = NULL;
4552: if (rowemb) {
4553: ISGetIndices(rowemb,&rowindices);
4554: }
4555: if (colemb) {
4556: ISGetIndices(colemb,&colindices);
4557: }
4558: for (i=0; i<B->rmap->n; i++) {
4559: PetscInt row;
4560: row = i;
4561: if (rowindices) row = rowindices[i];
4562: for (j=Baij->i[i]; j<Baij->i[i+1]; j++) {
4563: PetscInt col;
4564: col = Baij->j[count];
4565: if (colindices) col = colindices[col];
4566: v = Baij->a[count];
4567: MatSetValues(C,1,&row,1,&col,&v,INSERT_VALUES);
4568: ++count;
4569: }
4570: }
4571: /* FIXME: set C's nonzerostate correctly. */
4572: /* Assembly for C is necessary. */
4573: C->preallocated = PETSC_TRUE;
4574: C->assembled = PETSC_TRUE;
4575: C->was_assembled = PETSC_FALSE;
4576: return(0);
4577: }
4580: /*
4581: Special version for direct calls from Fortran
4582: */
4583: #include <petsc/private/fortranimpl.h>
4584: #if defined(PETSC_HAVE_FORTRAN_CAPS)
4585: #define matsetvaluesseqaij_ MATSETVALUESSEQAIJ
4586: #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE)
4587: #define matsetvaluesseqaij_ matsetvaluesseqaij
4588: #endif
4590: /* Change these macros so can be used in void function */
4591: #undef CHKERRQ
4592: #define CHKERRQ(ierr) CHKERRABORT(PetscObjectComm((PetscObject)A),ierr)
4593: #undef SETERRQ2
4594: #define SETERRQ2(comm,ierr,b,c,d) CHKERRABORT(comm,ierr)
4595: #undef SETERRQ3
4596: #define SETERRQ3(comm,ierr,b,c,d,e) CHKERRABORT(comm,ierr)
4600: PETSC_EXTERN void PETSC_STDCALL matsetvaluesseqaij_(Mat *AA,PetscInt *mm,const PetscInt im[],PetscInt *nn,const PetscInt in[],const PetscScalar v[],InsertMode *isis, PetscErrorCode *_ierr)
4601: {
4602: Mat A = *AA;
4603: PetscInt m = *mm, n = *nn;
4604: InsertMode is = *isis;
4605: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
4606: PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N;
4607: PetscInt *imax,*ai,*ailen;
4609: PetscInt *aj,nonew = a->nonew,lastcol = -1;
4610: MatScalar *ap,value,*aa;
4611: PetscBool ignorezeroentries = a->ignorezeroentries;
4612: PetscBool roworiented = a->roworiented;
4615: MatCheckPreallocated(A,1);
4616: imax = a->imax;
4617: ai = a->i;
4618: ailen = a->ilen;
4619: aj = a->j;
4620: aa = a->a;
4622: for (k=0; k<m; k++) { /* loop over added rows */
4623: row = im[k];
4624: if (row < 0) continue;
4625: #if defined(PETSC_USE_DEBUG)
4626: if (row >= A->rmap->n) SETERRABORT(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_OUTOFRANGE,"Row too large");
4627: #endif
4628: rp = aj + ai[row]; ap = aa + ai[row];
4629: rmax = imax[row]; nrow = ailen[row];
4630: low = 0;
4631: high = nrow;
4632: for (l=0; l<n; l++) { /* loop over added columns */
4633: if (in[l] < 0) continue;
4634: #if defined(PETSC_USE_DEBUG)
4635: if (in[l] >= A->cmap->n) SETERRABORT(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_OUTOFRANGE,"Column too large");
4636: #endif
4637: col = in[l];
4638: if (roworiented) value = v[l + k*n];
4639: else value = v[k + l*m];
4641: if (value == 0.0 && ignorezeroentries && (is == ADD_VALUES)) continue;
4643: if (col <= lastcol) low = 0;
4644: else high = nrow;
4645: lastcol = col;
4646: while (high-low > 5) {
4647: t = (low+high)/2;
4648: if (rp[t] > col) high = t;
4649: else low = t;
4650: }
4651: for (i=low; i<high; i++) {
4652: if (rp[i] > col) break;
4653: if (rp[i] == col) {
4654: if (is == ADD_VALUES) ap[i] += value;
4655: else ap[i] = value;
4656: goto noinsert;
4657: }
4658: }
4659: if (value == 0.0 && ignorezeroentries) goto noinsert;
4660: if (nonew == 1) goto noinsert;
4661: if (nonew == -1) SETERRABORT(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero in the matrix");
4662: MatSeqXAIJReallocateAIJ(A,A->rmap->n,1,nrow,row,col,rmax,aa,ai,aj,rp,ap,imax,nonew,MatScalar);
4663: N = nrow++ - 1; a->nz++; high++;
4664: /* shift up all the later entries in this row */
4665: for (ii=N; ii>=i; ii--) {
4666: rp[ii+1] = rp[ii];
4667: ap[ii+1] = ap[ii];
4668: }
4669: rp[i] = col;
4670: ap[i] = value;
4671: A->nonzerostate++;
4672: noinsert:;
4673: low = i + 1;
4674: }
4675: ailen[row] = nrow;
4676: }
4677: PetscFunctionReturnVoid();
4678: }