Actual source code: aij.c
1: #define PETSCMAT_DLL
3: /*
4: Defines the basic matrix operations for the AIJ (compressed row)
5: matrix storage format.
6: */
8: #include src/mat/impls/aij/seq/aij.h
9: #include src/inline/spops.h
10: #include src/inline/dot.h
11: #include petscbt.h
15: PetscErrorCode MatDiagonalSet_SeqAIJ(Mat Y,Vec D,InsertMode is)
16: {
18: Mat_SeqAIJ *aij = (Mat_SeqAIJ*) Y->data;
19: PetscInt i,*diag, m = Y->rmap.n;
20: PetscScalar *v,*aa = aij->a;
21: PetscTruth missing;
24: if (Y->assembled) {
25: MatMissingDiagonal_SeqAIJ(Y,&missing,PETSC_NULL);
26: if (!missing) {
27: diag = aij->diag;
28: VecGetArray(D,&v);
29: if (is == INSERT_VALUES) {
30: for (i=0; i<m; i++) {
31: aa[diag[i]] = v[i];
32: }
33: } else {
34: for (i=0; i<m; i++) {
35: aa[diag[i]] += v[i];
36: }
37: }
38: VecRestoreArray(D,&v);
39: return(0);
40: }
41: }
42: MatDiagonalSet_Default(Y,D,is);
43: return(0);
44: }
48: PetscErrorCode MatGetRowIJ_SeqAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *m,PetscInt *ia[],PetscInt *ja[],PetscTruth *done)
49: {
50: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
52: PetscInt i,ishift;
53:
55: *m = A->rmap.n;
56: if (!ia) return(0);
57: ishift = 0;
58: if (symmetric && !A->structurally_symmetric) {
59: MatToSymmetricIJ_SeqAIJ(A->rmap.n,a->i,a->j,ishift,oshift,ia,ja);
60: } else if (oshift == 1) {
61: PetscInt nz = a->i[A->rmap.n];
62: /* malloc space and add 1 to i and j indices */
63: PetscMalloc((A->rmap.n+1)*sizeof(PetscInt),ia);
64: PetscMalloc((nz+1)*sizeof(PetscInt),ja);
65: for (i=0; i<nz; i++) (*ja)[i] = a->j[i] + 1;
66: for (i=0; i<A->rmap.n+1; i++) (*ia)[i] = a->i[i] + 1;
67: } else {
68: *ia = a->i; *ja = a->j;
69: }
70: return(0);
71: }
75: PetscErrorCode MatRestoreRowIJ_SeqAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *n,PetscInt *ia[],PetscInt *ja[],PetscTruth *done)
76: {
78:
80: if (!ia) return(0);
81: if ((symmetric && !A->structurally_symmetric) || oshift == 1) {
82: PetscFree(*ia);
83: PetscFree(*ja);
84: }
85: return(0);
86: }
90: PetscErrorCode MatGetColumnIJ_SeqAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *nn,PetscInt *ia[],PetscInt *ja[],PetscTruth *done)
91: {
92: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
94: PetscInt i,*collengths,*cia,*cja,n = A->cmap.n,m = A->rmap.n;
95: PetscInt nz = a->i[m],row,*jj,mr,col;
96:
98: *nn = n;
99: if (!ia) return(0);
100: if (symmetric) {
101: MatToSymmetricIJ_SeqAIJ(A->rmap.n,a->i,a->j,0,oshift,ia,ja);
102: } else {
103: PetscMalloc((n+1)*sizeof(PetscInt),&collengths);
104: PetscMemzero(collengths,n*sizeof(PetscInt));
105: PetscMalloc((n+1)*sizeof(PetscInt),&cia);
106: PetscMalloc((nz+1)*sizeof(PetscInt),&cja);
107: jj = a->j;
108: for (i=0; i<nz; i++) {
109: collengths[jj[i]]++;
110: }
111: cia[0] = oshift;
112: for (i=0; i<n; i++) {
113: cia[i+1] = cia[i] + collengths[i];
114: }
115: PetscMemzero(collengths,n*sizeof(PetscInt));
116: jj = a->j;
117: for (row=0; row<m; row++) {
118: mr = a->i[row+1] - a->i[row];
119: for (i=0; i<mr; i++) {
120: col = *jj++;
121: cja[cia[col] + collengths[col]++ - oshift] = row + oshift;
122: }
123: }
124: PetscFree(collengths);
125: *ia = cia; *ja = cja;
126: }
127: return(0);
128: }
132: PetscErrorCode MatRestoreColumnIJ_SeqAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *n,PetscInt *ia[],PetscInt *ja[],PetscTruth *done)
133: {
137: if (!ia) return(0);
139: PetscFree(*ia);
140: PetscFree(*ja);
141:
142: return(0);
143: }
147: PetscErrorCode MatSetValuesRow_SeqAIJ(Mat A,PetscInt row,const PetscScalar v[])
148: {
149: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
150: PetscInt *ai = a->i;
154: PetscMemcpy(a->a+ai[row],v,(ai[row+1]-ai[row])*sizeof(PetscScalar));
155: return(0);
156: }
158: #define CHUNKSIZE 15
162: PetscErrorCode MatSetValues_SeqAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode is)
163: {
164: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
165: PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N;
166: PetscInt *imax = a->imax,*ai = a->i,*ailen = a->ilen;
168: PetscInt *aj = a->j,nonew = a->nonew,lastcol = -1;
169: PetscScalar *ap,value,*aa = a->a;
170: PetscTruth ignorezeroentries = ((a->ignorezeroentries && is == ADD_VALUES) ? PETSC_TRUE:PETSC_FALSE);
171: PetscTruth roworiented = a->roworiented;
174: for (k=0; k<m; k++) { /* loop over added rows */
175: row = im[k];
176: if (row < 0) continue;
177: #if defined(PETSC_USE_DEBUG)
178: if (row >= A->rmap.n) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",row,A->rmap.n-1);
179: #endif
180: rp = aj + ai[row]; ap = aa + ai[row];
181: rmax = imax[row]; nrow = ailen[row];
182: low = 0;
183: high = nrow;
184: for (l=0; l<n; l++) { /* loop over added columns */
185: if (in[l] < 0) continue;
186: #if defined(PETSC_USE_DEBUG)
187: if (in[l] >= A->cmap.n) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[l],A->cmap.n-1);
188: #endif
189: col = in[l];
190: if (roworiented) {
191: value = v[l + k*n];
192: } else {
193: value = v[k + l*m];
194: }
195: if (value == 0.0 && ignorezeroentries && (is == ADD_VALUES)) continue;
197: if (col <= lastcol) low = 0; else high = nrow;
198: lastcol = col;
199: while (high-low > 5) {
200: t = (low+high)/2;
201: if (rp[t] > col) high = t;
202: else low = t;
203: }
204: for (i=low; i<high; i++) {
205: if (rp[i] > col) break;
206: if (rp[i] == col) {
207: if (is == ADD_VALUES) ap[i] += value;
208: else ap[i] = value;
209: goto noinsert;
210: }
211: }
212: if (value == 0.0 && ignorezeroentries) goto noinsert;
213: if (nonew == 1) goto noinsert;
214: if (nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero at (%D,%D) in the matrix",row,col);
215: MatSeqXAIJReallocateAIJ(A,A->rmap.n,1,nrow,row,col,rmax,aa,ai,aj,rp,ap,imax,nonew);
216: N = nrow++ - 1; a->nz++; high++;
217: /* shift up all the later entries in this row */
218: for (ii=N; ii>=i; ii--) {
219: rp[ii+1] = rp[ii];
220: ap[ii+1] = ap[ii];
221: }
222: rp[i] = col;
223: ap[i] = value;
224: noinsert:;
225: low = i + 1;
226: }
227: ailen[row] = nrow;
228: }
229: A->same_nonzero = PETSC_FALSE;
230: return(0);
231: }
236: PetscErrorCode MatGetValues_SeqAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],PetscScalar v[])
237: {
238: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
239: PetscInt *rp,k,low,high,t,row,nrow,i,col,l,*aj = a->j;
240: PetscInt *ai = a->i,*ailen = a->ilen;
241: PetscScalar *ap,*aa = a->a,zero = 0.0;
244: for (k=0; k<m; k++) { /* loop over rows */
245: row = im[k];
246: if (row < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",row);
247: if (row >= A->rmap.n) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",row,A->rmap.n-1);
248: rp = aj + ai[row]; ap = aa + ai[row];
249: nrow = ailen[row];
250: for (l=0; l<n; l++) { /* loop over columns */
251: if (in[l] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative column: %D",in[l]);
252: if (in[l] >= A->cmap.n) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[l],A->cmap.n-1);
253: col = in[l] ;
254: high = nrow; low = 0; /* assume unsorted */
255: while (high-low > 5) {
256: t = (low+high)/2;
257: if (rp[t] > col) high = t;
258: else low = t;
259: }
260: for (i=low; i<high; i++) {
261: if (rp[i] > col) break;
262: if (rp[i] == col) {
263: *v++ = ap[i];
264: goto finished;
265: }
266: }
267: *v++ = zero;
268: finished:;
269: }
270: }
271: return(0);
272: }
277: PetscErrorCode MatView_SeqAIJ_Binary(Mat A,PetscViewer viewer)
278: {
279: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
281: PetscInt i,*col_lens;
282: int fd;
285: PetscViewerBinaryGetDescriptor(viewer,&fd);
286: PetscMalloc((4+A->rmap.n)*sizeof(PetscInt),&col_lens);
287: col_lens[0] = MAT_FILE_COOKIE;
288: col_lens[1] = A->rmap.n;
289: col_lens[2] = A->cmap.n;
290: col_lens[3] = a->nz;
292: /* store lengths of each row and write (including header) to file */
293: for (i=0; i<A->rmap.n; i++) {
294: col_lens[4+i] = a->i[i+1] - a->i[i];
295: }
296: PetscBinaryWrite(fd,col_lens,4+A->rmap.n,PETSC_INT,PETSC_TRUE);
297: PetscFree(col_lens);
299: /* store column indices (zero start index) */
300: PetscBinaryWrite(fd,a->j,a->nz,PETSC_INT,PETSC_FALSE);
302: /* store nonzero values */
303: PetscBinaryWrite(fd,a->a,a->nz,PETSC_SCALAR,PETSC_FALSE);
304: return(0);
305: }
307: EXTERN PetscErrorCode MatSeqAIJFactorInfo_Matlab(Mat,PetscViewer);
311: PetscErrorCode MatView_SeqAIJ_ASCII(Mat A,PetscViewer viewer)
312: {
313: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
314: PetscErrorCode ierr;
315: PetscInt i,j,m = A->rmap.n,shift=0;
316: const char *name;
317: PetscViewerFormat format;
320: PetscObjectGetName((PetscObject)A,&name);
321: PetscViewerGetFormat(viewer,&format);
322: if (format == PETSC_VIEWER_ASCII_MATLAB) {
323: PetscInt nofinalvalue = 0;
324: if ((a->i[m] == a->i[m-1]) || (a->j[a->nz-1] != A->cmap.n-!shift)) {
325: nofinalvalue = 1;
326: }
327: PetscViewerASCIIUseTabs(viewer,PETSC_NO);
328: PetscViewerASCIIPrintf(viewer,"%% Size = %D %D \n",m,A->cmap.n);
329: PetscViewerASCIIPrintf(viewer,"%% Nonzeros = %D \n",a->nz);
330: PetscViewerASCIIPrintf(viewer,"zzz = zeros(%D,3);\n",a->nz+nofinalvalue);
331: PetscViewerASCIIPrintf(viewer,"zzz = [\n");
333: for (i=0; i<m; i++) {
334: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
335: #if defined(PETSC_USE_COMPLEX)
336: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e + %18.16ei \n",i+1,a->j[j]+!shift,PetscRealPart(a->a[j]),PetscImaginaryPart(a->a[j]));
337: #else
338: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e\n",i+1,a->j[j]+!shift,a->a[j]);
339: #endif
340: }
341: }
342: if (nofinalvalue) {
343: PetscViewerASCIIPrintf(viewer,"%D %D %18.16e\n",m,A->cmap.n,0.0);
344: }
345: PetscViewerASCIIPrintf(viewer,"];\n %s = spconvert(zzz);\n",name);
346: PetscViewerASCIIUseTabs(viewer,PETSC_YES);
347: } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO || format == PETSC_VIEWER_ASCII_INFO) {
348: return(0);
349: } else if (format == PETSC_VIEWER_ASCII_COMMON) {
350: PetscViewerASCIIUseTabs(viewer,PETSC_NO);
351: for (i=0; i<m; i++) {
352: PetscViewerASCIIPrintf(viewer,"row %D:",i);
353: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
354: #if defined(PETSC_USE_COMPLEX)
355: if (PetscImaginaryPart(a->a[j]) > 0.0 && PetscRealPart(a->a[j]) != 0.0) {
356: PetscViewerASCIIPrintf(viewer," (%D, %G + %G i)",a->j[j]+shift,PetscRealPart(a->a[j]),PetscImaginaryPart(a->a[j]));
357: } else if (PetscImaginaryPart(a->a[j]) < 0.0 && PetscRealPart(a->a[j]) != 0.0) {
358: PetscViewerASCIIPrintf(viewer," (%D, %G - %G i)",a->j[j]+shift,PetscRealPart(a->a[j]),-PetscImaginaryPart(a->a[j]));
359: } else if (PetscRealPart(a->a[j]) != 0.0) {
360: PetscViewerASCIIPrintf(viewer," (%D, %G) ",a->j[j]+shift,PetscRealPart(a->a[j]));
361: }
362: #else
363: if (a->a[j] != 0.0) {PetscViewerASCIIPrintf(viewer," (%D, %G) ",a->j[j]+shift,a->a[j]);}
364: #endif
365: }
366: PetscViewerASCIIPrintf(viewer,"\n");
367: }
368: PetscViewerASCIIUseTabs(viewer,PETSC_YES);
369: } else if (format == PETSC_VIEWER_ASCII_SYMMODU) {
370: PetscInt nzd=0,fshift=1,*sptr;
371: PetscViewerASCIIUseTabs(viewer,PETSC_NO);
372: PetscMalloc((m+1)*sizeof(PetscInt),&sptr);
373: for (i=0; i<m; i++) {
374: sptr[i] = nzd+1;
375: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
376: if (a->j[j] >= i) {
377: #if defined(PETSC_USE_COMPLEX)
378: if (PetscImaginaryPart(a->a[j]) != 0.0 || PetscRealPart(a->a[j]) != 0.0) nzd++;
379: #else
380: if (a->a[j] != 0.0) nzd++;
381: #endif
382: }
383: }
384: }
385: sptr[m] = nzd+1;
386: PetscViewerASCIIPrintf(viewer," %D %D\n\n",m,nzd);
387: for (i=0; i<m+1; i+=6) {
388: if (i+4<m) {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]);}
389: else if (i+3<m) {PetscViewerASCIIPrintf(viewer," %D %D %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3],sptr[i+4]);}
390: else if (i+2<m) {PetscViewerASCIIPrintf(viewer," %D %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3]);}
391: else if (i+1<m) {PetscViewerASCIIPrintf(viewer," %D %D %D\n",sptr[i],sptr[i+1],sptr[i+2]);}
392: else if (i<m) {PetscViewerASCIIPrintf(viewer," %D %D\n",sptr[i],sptr[i+1]);}
393: else {PetscViewerASCIIPrintf(viewer," %D\n",sptr[i]);}
394: }
395: PetscViewerASCIIPrintf(viewer,"\n");
396: PetscFree(sptr);
397: for (i=0; i<m; i++) {
398: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
399: if (a->j[j] >= i) {PetscViewerASCIIPrintf(viewer," %D ",a->j[j]+fshift);}
400: }
401: PetscViewerASCIIPrintf(viewer,"\n");
402: }
403: PetscViewerASCIIPrintf(viewer,"\n");
404: for (i=0; i<m; i++) {
405: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
406: if (a->j[j] >= i) {
407: #if defined(PETSC_USE_COMPLEX)
408: if (PetscImaginaryPart(a->a[j]) != 0.0 || PetscRealPart(a->a[j]) != 0.0) {
409: PetscViewerASCIIPrintf(viewer," %18.16e %18.16e ",PetscRealPart(a->a[j]),PetscImaginaryPart(a->a[j]));
410: }
411: #else
412: if (a->a[j] != 0.0) {PetscViewerASCIIPrintf(viewer," %18.16e ",a->a[j]);}
413: #endif
414: }
415: }
416: PetscViewerASCIIPrintf(viewer,"\n");
417: }
418: PetscViewerASCIIUseTabs(viewer,PETSC_YES);
419: } else if (format == PETSC_VIEWER_ASCII_DENSE) {
420: PetscInt cnt = 0,jcnt;
421: PetscScalar value;
423: PetscViewerASCIIUseTabs(viewer,PETSC_NO);
424: for (i=0; i<m; i++) {
425: jcnt = 0;
426: for (j=0; j<A->cmap.n; j++) {
427: if (jcnt < a->i[i+1]-a->i[i] && j == a->j[cnt]) {
428: value = a->a[cnt++];
429: jcnt++;
430: } else {
431: value = 0.0;
432: }
433: #if defined(PETSC_USE_COMPLEX)
434: PetscViewerASCIIPrintf(viewer," %7.5e+%7.5e i ",PetscRealPart(value),PetscImaginaryPart(value));
435: #else
436: PetscViewerASCIIPrintf(viewer," %7.5e ",value);
437: #endif
438: }
439: PetscViewerASCIIPrintf(viewer,"\n");
440: }
441: PetscViewerASCIIUseTabs(viewer,PETSC_YES);
442: } else {
443: PetscViewerASCIIUseTabs(viewer,PETSC_NO);
444: for (i=0; i<m; i++) {
445: PetscViewerASCIIPrintf(viewer,"row %D:",i);
446: for (j=a->i[i]+shift; j<a->i[i+1]+shift; j++) {
447: #if defined(PETSC_USE_COMPLEX)
448: if (PetscImaginaryPart(a->a[j]) > 0.0) {
449: PetscViewerASCIIPrintf(viewer," (%D, %G + %G i)",a->j[j]+shift,PetscRealPart(a->a[j]),PetscImaginaryPart(a->a[j]));
450: } else if (PetscImaginaryPart(a->a[j]) < 0.0) {
451: PetscViewerASCIIPrintf(viewer," (%D, %G - %G i)",a->j[j]+shift,PetscRealPart(a->a[j]),-PetscImaginaryPart(a->a[j]));
452: } else {
453: PetscViewerASCIIPrintf(viewer," (%D, %G) ",a->j[j]+shift,PetscRealPart(a->a[j]));
454: }
455: #else
456: PetscViewerASCIIPrintf(viewer," (%D, %G) ",a->j[j]+shift,a->a[j]);
457: #endif
458: }
459: PetscViewerASCIIPrintf(viewer,"\n");
460: }
461: PetscViewerASCIIUseTabs(viewer,PETSC_YES);
462: }
463: PetscViewerFlush(viewer);
464: return(0);
465: }
469: PetscErrorCode MatView_SeqAIJ_Draw_Zoom(PetscDraw draw,void *Aa)
470: {
471: Mat A = (Mat) Aa;
472: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
473: PetscErrorCode ierr;
474: PetscInt i,j,m = A->rmap.n,color;
475: PetscReal xl,yl,xr,yr,x_l,x_r,y_l,y_r,maxv = 0.0;
476: PetscViewer viewer;
477: PetscViewerFormat format;
480: PetscObjectQuery((PetscObject)A,"Zoomviewer",(PetscObject*)&viewer);
481: PetscViewerGetFormat(viewer,&format);
483: PetscDrawGetCoordinates(draw,&xl,&yl,&xr,&yr);
484: /* loop over matrix elements drawing boxes */
486: if (format != PETSC_VIEWER_DRAW_CONTOUR) {
487: /* Blue for negative, Cyan for zero and Red for positive */
488: color = PETSC_DRAW_BLUE;
489: for (i=0; i<m; i++) {
490: y_l = m - i - 1.0; y_r = y_l + 1.0;
491: for (j=a->i[i]; j<a->i[i+1]; j++) {
492: x_l = a->j[j] ; x_r = x_l + 1.0;
493: #if defined(PETSC_USE_COMPLEX)
494: if (PetscRealPart(a->a[j]) >= 0.) continue;
495: #else
496: if (a->a[j] >= 0.) continue;
497: #endif
498: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
499: }
500: }
501: color = PETSC_DRAW_CYAN;
502: for (i=0; i<m; i++) {
503: y_l = m - i - 1.0; y_r = y_l + 1.0;
504: for (j=a->i[i]; j<a->i[i+1]; j++) {
505: x_l = a->j[j]; x_r = x_l + 1.0;
506: if (a->a[j] != 0.) continue;
507: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
508: }
509: }
510: color = PETSC_DRAW_RED;
511: for (i=0; i<m; i++) {
512: y_l = m - i - 1.0; y_r = y_l + 1.0;
513: for (j=a->i[i]; j<a->i[i+1]; j++) {
514: x_l = a->j[j]; x_r = x_l + 1.0;
515: #if defined(PETSC_USE_COMPLEX)
516: if (PetscRealPart(a->a[j]) <= 0.) continue;
517: #else
518: if (a->a[j] <= 0.) continue;
519: #endif
520: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
521: }
522: }
523: } else {
524: /* use contour shading to indicate magnitude of values */
525: /* first determine max of all nonzero values */
526: PetscInt nz = a->nz,count;
527: PetscDraw popup;
528: PetscReal scale;
530: for (i=0; i<nz; i++) {
531: if (PetscAbsScalar(a->a[i]) > maxv) maxv = PetscAbsScalar(a->a[i]);
532: }
533: scale = (245.0 - PETSC_DRAW_BASIC_COLORS)/maxv;
534: PetscDrawGetPopup(draw,&popup);
535: if (popup) {PetscDrawScalePopup(popup,0.0,maxv);}
536: count = 0;
537: for (i=0; i<m; i++) {
538: y_l = m - i - 1.0; y_r = y_l + 1.0;
539: for (j=a->i[i]; j<a->i[i+1]; j++) {
540: x_l = a->j[j]; x_r = x_l + 1.0;
541: color = PETSC_DRAW_BASIC_COLORS + (PetscInt)(scale*PetscAbsScalar(a->a[count]));
542: PetscDrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color);
543: count++;
544: }
545: }
546: }
547: return(0);
548: }
552: PetscErrorCode MatView_SeqAIJ_Draw(Mat A,PetscViewer viewer)
553: {
555: PetscDraw draw;
556: PetscReal xr,yr,xl,yl,h,w;
557: PetscTruth isnull;
560: PetscViewerDrawGetDraw(viewer,0,&draw);
561: PetscDrawIsNull(draw,&isnull);
562: if (isnull) return(0);
564: PetscObjectCompose((PetscObject)A,"Zoomviewer",(PetscObject)viewer);
565: xr = A->cmap.n; yr = A->rmap.n; h = yr/10.0; w = xr/10.0;
566: xr += w; yr += h; xl = -w; yl = -h;
567: PetscDrawSetCoordinates(draw,xl,yl,xr,yr);
568: PetscDrawZoom(draw,MatView_SeqAIJ_Draw_Zoom,A);
569: PetscObjectCompose((PetscObject)A,"Zoomviewer",PETSC_NULL);
570: return(0);
571: }
575: PetscErrorCode MatView_SeqAIJ(Mat A,PetscViewer viewer)
576: {
578: PetscTruth issocket,iascii,isbinary,isdraw;
581: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_SOCKET,&issocket);
582: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);
583: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);
584: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);
585: if (iascii) {
586: MatView_SeqAIJ_ASCII(A,viewer);
587: #if defined(PETSC_USE_SOCKET_VIEWER)
588: } else if (issocket) {
589: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
590: PetscViewerSocketPutSparse_Private(viewer,A->rmap.n,A->cmap.n,a->nz,a->a,a->i,a->j);
591: #endif
592: } else if (isbinary) {
593: MatView_SeqAIJ_Binary(A,viewer);
594: } else if (isdraw) {
595: MatView_SeqAIJ_Draw(A,viewer);
596: } else {
597: SETERRQ1(PETSC_ERR_SUP,"Viewer type %s not supported by SeqAIJ matrices",((PetscObject)viewer)->type_name);
598: }
599: MatView_Inode(A,viewer);
600: return(0);
601: }
605: PetscErrorCode MatAssemblyEnd_SeqAIJ(Mat A,MatAssemblyType mode)
606: {
607: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
609: PetscInt fshift = 0,i,j,*ai = a->i,*aj = a->j,*imax = a->imax;
610: PetscInt m = A->rmap.n,*ip,N,*ailen = a->ilen,rmax = 0;
611: PetscScalar *aa = a->a,*ap;
612: PetscReal ratio=0.6;
615: if (mode == MAT_FLUSH_ASSEMBLY) return(0);
617: if (m) rmax = ailen[0]; /* determine row with most nonzeros */
618: for (i=1; i<m; i++) {
619: /* move each row back by the amount of empty slots (fshift) before it*/
620: fshift += imax[i-1] - ailen[i-1];
621: rmax = PetscMax(rmax,ailen[i]);
622: if (fshift) {
623: ip = aj + ai[i] ;
624: ap = aa + ai[i] ;
625: N = ailen[i];
626: for (j=0; j<N; j++) {
627: ip[j-fshift] = ip[j];
628: ap[j-fshift] = ap[j];
629: }
630: }
631: ai[i] = ai[i-1] + ailen[i-1];
632: }
633: if (m) {
634: fshift += imax[m-1] - ailen[m-1];
635: ai[m] = ai[m-1] + ailen[m-1];
636: }
637: /* reset ilen and imax for each row */
638: for (i=0; i<m; i++) {
639: ailen[i] = imax[i] = ai[i+1] - ai[i];
640: }
641: a->nz = ai[m];
643: MatMarkDiagonal_SeqAIJ(A);
644: PetscInfo4(A,"Matrix size: %D X %D; storage space: %D unneeded,%D used\n",m,A->cmap.n,fshift,a->nz);
645: PetscInfo1(A,"Number of mallocs during MatSetValues() is %D\n",a->reallocs);
646: PetscInfo1(A,"Maximum nonzeros in any row is %D\n",rmax);
647: a->reallocs = 0;
648: A->info.nz_unneeded = (double)fshift;
649: a->rmax = rmax;
651: /* check for zero rows. If found a large number of zero rows, use CompressedRow functions */
652: Mat_CheckCompressedRow(A,&a->compressedrow,a->i,m,ratio);
653: A->same_nonzero = PETSC_TRUE;
655: MatAssemblyEnd_Inode(A,mode);
656: return(0);
657: }
661: PetscErrorCode MatRealPart_SeqAIJ(Mat A)
662: {
663: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
664: PetscInt i,nz = a->nz;
665: PetscScalar *aa = a->a;
668: for (i=0; i<nz; i++) aa[i] = PetscRealPart(aa[i]);
669: return(0);
670: }
674: PetscErrorCode MatImaginaryPart_SeqAIJ(Mat A)
675: {
676: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
677: PetscInt i,nz = a->nz;
678: PetscScalar *aa = a->a;
681: for (i=0; i<nz; i++) aa[i] = PetscImaginaryPart(aa[i]);
682: return(0);
683: }
687: PetscErrorCode MatZeroEntries_SeqAIJ(Mat A)
688: {
689: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
693: PetscMemzero(a->a,(a->i[A->rmap.n])*sizeof(PetscScalar));
694: return(0);
695: }
699: PetscErrorCode MatDestroy_SeqAIJ(Mat A)
700: {
701: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
705: #if defined(PETSC_USE_LOG)
706: PetscLogObjectState((PetscObject)A,"Rows=%D, Cols=%D, NZ=%D",A->rmap.n,A->cmap.n,a->nz);
707: #endif
708: MatSeqXAIJFreeAIJ(A,&a->a,&a->j,&a->i);
709: if (a->row) {
710: ISDestroy(a->row);
711: }
712: if (a->col) {
713: ISDestroy(a->col);
714: }
715: PetscFree(a->diag);
716: PetscFree2(a->imax,a->ilen);
717: PetscFree(a->idiag);
718: PetscFree(a->solve_work);
719: if (a->icol) {ISDestroy(a->icol);}
720: PetscFree(a->saved_values);
721: if (a->coloring) {ISColoringDestroy(a->coloring);}
722: PetscFree(a->xtoy);
723: if (a->compressedrow.use){PetscFree(a->compressedrow.i);}
725: MatDestroy_Inode(A);
727: PetscFree(a);
729: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatSeqAIJSetColumnIndices_C","",PETSC_NULL);
730: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatStoreValues_C","",PETSC_NULL);
731: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatRetrieveValues_C","",PETSC_NULL);
732: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatConvert_seqaij_seqsbaij_C","",PETSC_NULL);
733: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatConvert_seqaij_seqbaij_C","",PETSC_NULL);
734: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatConvert_seqaij_seqcsrperm_C","",PETSC_NULL);
735: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatIsTranspose_C","",PETSC_NULL);
736: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatSeqAIJSetPreallocation_C","",PETSC_NULL);
737: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatSeqAIJSetPreallocationCSR_C","",PETSC_NULL);
738: PetscObjectComposeFunctionDynamic((PetscObject)A,"MatReorderForNonzeroDiagonal_C","",PETSC_NULL);
739: return(0);
740: }
744: PetscErrorCode MatCompress_SeqAIJ(Mat A)
745: {
747: return(0);
748: }
752: PetscErrorCode MatSetOption_SeqAIJ(Mat A,MatOption op)
753: {
754: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
758: switch (op) {
759: case MAT_ROW_ORIENTED:
760: a->roworiented = PETSC_TRUE;
761: break;
762: case MAT_KEEP_ZEROED_ROWS:
763: a->keepzeroedrows = PETSC_TRUE;
764: break;
765: case MAT_COLUMN_ORIENTED:
766: a->roworiented = PETSC_FALSE;
767: break;
768: case MAT_COLUMNS_SORTED:
769: a->sorted = PETSC_TRUE;
770: break;
771: case MAT_COLUMNS_UNSORTED:
772: a->sorted = PETSC_FALSE;
773: break;
774: case MAT_NO_NEW_NONZERO_LOCATIONS:
775: a->nonew = 1;
776: break;
777: case MAT_NEW_NONZERO_LOCATION_ERR:
778: a->nonew = -1;
779: break;
780: case MAT_NEW_NONZERO_ALLOCATION_ERR:
781: a->nonew = -2;
782: break;
783: case MAT_YES_NEW_NONZERO_LOCATIONS:
784: a->nonew = 0;
785: break;
786: case MAT_IGNORE_ZERO_ENTRIES:
787: a->ignorezeroentries = PETSC_TRUE;
788: break;
789: case MAT_USE_COMPRESSEDROW:
790: a->compressedrow.use = PETSC_TRUE;
791: break;
792: case MAT_DO_NOT_USE_COMPRESSEDROW:
793: a->compressedrow.use = PETSC_FALSE;
794: break;
795: case MAT_ROWS_SORTED:
796: case MAT_ROWS_UNSORTED:
797: case MAT_YES_NEW_DIAGONALS:
798: case MAT_IGNORE_OFF_PROC_ENTRIES:
799: case MAT_USE_HASH_TABLE:
800: PetscInfo(A,"Option ignored\n");
801: break;
802: case MAT_NO_NEW_DIAGONALS:
803: SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS");
804: default:
805: break;
806: }
807: MatSetOption_Inode(A,op);
808: return(0);
809: }
813: PetscErrorCode MatGetDiagonal_SeqAIJ(Mat A,Vec v)
814: {
815: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
817: PetscInt i,j,n;
818: PetscScalar *x,zero = 0.0;
821: VecSet(v,zero);
822: VecGetArray(v,&x);
823: VecGetLocalSize(v,&n);
824: if (n != A->rmap.n) SETERRQ(PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
825: for (i=0; i<A->rmap.n; i++) {
826: for (j=a->i[i]; j<a->i[i+1]; j++) {
827: if (a->j[j] == i) {
828: x[i] = a->a[j];
829: break;
830: }
831: }
832: }
833: VecRestoreArray(v,&x);
834: return(0);
835: }
839: PetscErrorCode MatMultTransposeAdd_SeqAIJ(Mat A,Vec xx,Vec zz,Vec yy)
840: {
841: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
842: PetscScalar *x,*y;
843: PetscErrorCode ierr;
844: PetscInt m = A->rmap.n;
845: #if !defined(PETSC_USE_FORTRAN_KERNEL_MULTTRANSPOSEAIJ)
846: PetscScalar *v,alpha;
847: PetscInt n,i,*idx,*ii,*ridx=PETSC_NULL;
848: Mat_CompressedRow cprow = a->compressedrow;
849: PetscTruth usecprow = cprow.use;
850: #endif
853: if (zz != yy) {VecCopy(zz,yy);}
854: VecGetArray(xx,&x);
855: VecGetArray(yy,&y);
857: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTTRANSPOSEAIJ)
858: fortranmulttransposeaddaij_(&m,x,a->i,a->j,a->a,y);
859: #else
860: if (usecprow){
861: m = cprow.nrows;
862: ii = cprow.i;
863: ridx = cprow.rindex;
864: } else {
865: ii = a->i;
866: }
867: for (i=0; i<m; i++) {
868: idx = a->j + ii[i] ;
869: v = a->a + ii[i] ;
870: n = ii[i+1] - ii[i];
871: if (usecprow){
872: alpha = x[ridx[i]];
873: } else {
874: alpha = x[i];
875: }
876: while (n-->0) {y[*idx++] += alpha * *v++;}
877: }
878: #endif
879: PetscLogFlops(2*a->nz);
880: VecRestoreArray(xx,&x);
881: VecRestoreArray(yy,&y);
882: return(0);
883: }
887: PetscErrorCode MatMultTranspose_SeqAIJ(Mat A,Vec xx,Vec yy)
888: {
889: PetscScalar zero = 0.0;
893: VecSet(yy,zero);
894: MatMultTransposeAdd_SeqAIJ(A,xx,yy,yy);
895: return(0);
896: }
901: PetscErrorCode MatMult_SeqAIJ(Mat A,Vec xx,Vec yy)
902: {
903: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
904: PetscScalar *x,*y,*aa;
906: PetscInt m=A->rmap.n,*aj,*ii;
907: PetscInt n,i,j,*ridx=PETSC_NULL;
908: PetscScalar sum;
909: PetscTruth usecprow=a->compressedrow.use;
910: #if !defined(PETSC_USE_FORTRAN_KERNEL_MULTAIJ)
911: PetscInt jrow;
912: #endif
914: #if defined(PETSC_HAVE_PRAGMA_DISJOINT)
915: #pragma disjoint(*x,*y,*aa)
916: #endif
919: VecGetArray(xx,&x);
920: VecGetArray(yy,&y);
921: aj = a->j;
922: aa = a->a;
923: ii = a->i;
924: if (usecprow){ /* use compressed row format */
925: m = a->compressedrow.nrows;
926: ii = a->compressedrow.i;
927: ridx = a->compressedrow.rindex;
928: for (i=0; i<m; i++){
929: n = ii[i+1] - ii[i];
930: aj = a->j + ii[i];
931: aa = a->a + ii[i];
932: sum = 0.0;
933: for (j=0; j<n; j++) sum += (*aa++)*x[*aj++];
934: y[*ridx++] = sum;
935: }
936: } else { /* do not use compressed row format */
937: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTAIJ)
938: fortranmultaij_(&m,x,ii,aj,aa,y);
939: #else
940: for (i=0; i<m; i++) {
941: jrow = ii[i];
942: n = ii[i+1] - jrow;
943: sum = 0.0;
944: for (j=0; j<n; j++) {
945: sum += aa[jrow]*x[aj[jrow]]; jrow++;
946: }
947: y[i] = sum;
948: }
949: #endif
950: }
951: PetscLogFlops(2*a->nz - m);
952: VecRestoreArray(xx,&x);
953: VecRestoreArray(yy,&y);
954: return(0);
955: }
959: PetscErrorCode MatMultAdd_SeqAIJ(Mat A,Vec xx,Vec yy,Vec zz)
960: {
961: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
962: PetscScalar *x,*y,*z,*aa;
964: PetscInt m = A->rmap.n,*aj,*ii;
965: #if !defined(PETSC_USE_FORTRAN_KERNEL_MULTADDAIJ)
966: PetscInt n,i,jrow,j,*ridx=PETSC_NULL;
967: PetscScalar sum;
968: PetscTruth usecprow=a->compressedrow.use;
969: #endif
972: VecGetArray(xx,&x);
973: VecGetArray(yy,&y);
974: if (zz != yy) {
975: VecGetArray(zz,&z);
976: } else {
977: z = y;
978: }
980: aj = a->j;
981: aa = a->a;
982: ii = a->i;
983: #if defined(PETSC_USE_FORTRAN_KERNEL_MULTADDAIJ)
984: fortranmultaddaij_(&m,x,ii,aj,aa,y,z);
985: #else
986: if (usecprow){ /* use compressed row format */
987: if (zz != yy){
988: PetscMemcpy(z,y,m*sizeof(PetscScalar));
989: }
990: m = a->compressedrow.nrows;
991: ii = a->compressedrow.i;
992: ridx = a->compressedrow.rindex;
993: for (i=0; i<m; i++){
994: n = ii[i+1] - ii[i];
995: aj = a->j + ii[i];
996: aa = a->a + ii[i];
997: sum = y[*ridx];
998: for (j=0; j<n; j++) sum += (*aa++)*x[*aj++];
999: z[*ridx++] = sum;
1000: }
1001: } else { /* do not use compressed row format */
1002: for (i=0; i<m; i++) {
1003: jrow = ii[i];
1004: n = ii[i+1] - jrow;
1005: sum = y[i];
1006: for (j=0; j<n; j++) {
1007: sum += aa[jrow]*x[aj[jrow]]; jrow++;
1008: }
1009: z[i] = sum;
1010: }
1011: }
1012: #endif
1013: PetscLogFlops(2*a->nz);
1014: VecRestoreArray(xx,&x);
1015: VecRestoreArray(yy,&y);
1016: if (zz != yy) {
1017: VecRestoreArray(zz,&z);
1018: }
1019: return(0);
1020: }
1022: /*
1023: Adds diagonal pointers to sparse matrix structure.
1024: */
1027: PetscErrorCode MatMarkDiagonal_SeqAIJ(Mat A)
1028: {
1029: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1031: PetscInt i,j,m = A->rmap.n;
1034: if (!a->diag) {
1035: PetscMalloc(m*sizeof(PetscInt),&a->diag);
1036: }
1037: for (i=0; i<A->rmap.n; i++) {
1038: a->diag[i] = a->i[i+1];
1039: for (j=a->i[i]; j<a->i[i+1]; j++) {
1040: if (a->j[j] == i) {
1041: a->diag[i] = j;
1042: break;
1043: }
1044: }
1045: }
1046: return(0);
1047: }
1049: /*
1050: Checks for missing diagonals
1051: */
1054: PetscErrorCode MatMissingDiagonal_SeqAIJ(Mat A,PetscTruth *missing,PetscInt *d)
1055: {
1056: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1057: PetscInt *diag,*jj = a->j,i;
1060: *missing = PETSC_FALSE;
1061: if (A->rmap.n > 0 && !jj) {
1062: *missing = PETSC_TRUE;
1063: if (d) *d = 0;
1064: PetscInfo(A,"Matrix has no entries therefor is missing diagonal");
1065: } else {
1066: diag = a->diag;
1067: for (i=0; i<A->rmap.n; i++) {
1068: if (jj[diag[i]] != i) {
1069: *missing = PETSC_TRUE;
1070: if (d) *d = i;
1071: PetscInfo1(A,"Matrix is missing diagonal number %D",i);
1072: }
1073: }
1074: }
1075: return(0);
1076: }
1080: PetscErrorCode MatRelax_SeqAIJ(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
1081: {
1082: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1083: PetscScalar *x,d,*xs,sum,*t,scale,*idiag=0,*mdiag;
1084: const PetscScalar *v = a->a, *b, *bs,*xb, *ts;
1085: PetscErrorCode ierr;
1086: PetscInt n = A->cmap.n,m = A->rmap.n,i;
1087: const PetscInt *idx,*diag;
1090: its = its*lits;
1091: if (its <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits);
1093: diag = a->diag;
1094: if (!a->idiag) {
1095: PetscMalloc(3*m*sizeof(PetscScalar),&a->idiag);
1096: a->ssor = a->idiag + m;
1097: mdiag = a->ssor + m;
1099: v = a->a;
1101: /* this is wrong when fshift omega changes each iteration */
1102: if (omega == 1.0 && !fshift) {
1103: for (i=0; i<m; i++) {
1104: mdiag[i] = v[diag[i]];
1105: a->idiag[i] = 1.0/v[diag[i]];
1106: }
1107: PetscLogFlops(m);
1108: } else {
1109: for (i=0; i<m; i++) {
1110: mdiag[i] = v[diag[i]];
1111: a->idiag[i] = omega/(fshift + v[diag[i]]);
1112: }
1113: PetscLogFlops(2*m);
1114: }
1115: }
1116: t = a->ssor;
1117: idiag = a->idiag;
1118: mdiag = a->idiag + 2*m;
1120: VecGetArray(xx,&x);
1121: if (xx != bb) {
1122: VecGetArray(bb,(PetscScalar**)&b);
1123: } else {
1124: b = x;
1125: }
1127: /* We count flops by assuming the upper triangular and lower triangular parts have the same number of nonzeros */
1128: xs = x;
1129: if (flag == SOR_APPLY_UPPER) {
1130: /* apply (U + D/omega) to the vector */
1131: bs = b;
1132: for (i=0; i<m; i++) {
1133: d = fshift + a->a[diag[i]];
1134: n = a->i[i+1] - diag[i] - 1;
1135: idx = a->j + diag[i] + 1;
1136: v = a->a + diag[i] + 1;
1137: sum = b[i]*d/omega;
1138: SPARSEDENSEDOT(sum,bs,v,idx,n);
1139: x[i] = sum;
1140: }
1141: VecRestoreArray(xx,&x);
1142: if (bb != xx) {VecRestoreArray(bb,(PetscScalar**)&b);}
1143: PetscLogFlops(a->nz);
1144: return(0);
1145: }
1148: /* Let A = L + U + D; where L is lower trianglar,
1149: U is upper triangular, E is diagonal; This routine applies
1151: (L + E)^{-1} A (U + E)^{-1}
1153: to a vector efficiently using Eisenstat's trick. This is for
1154: the case of SSOR preconditioner, so E is D/omega where omega
1155: is the relaxation factor.
1156: */
1158: if (flag == SOR_APPLY_LOWER) {
1159: SETERRQ(PETSC_ERR_SUP,"SOR_APPLY_LOWER is not implemented");
1160: } else if (flag & SOR_EISENSTAT) {
1161: /* Let A = L + U + D; where L is lower trianglar,
1162: U is upper triangular, E is diagonal; This routine applies
1164: (L + E)^{-1} A (U + E)^{-1}
1166: to a vector efficiently using Eisenstat's trick. This is for
1167: the case of SSOR preconditioner, so E is D/omega where omega
1168: is the relaxation factor.
1169: */
1170: scale = (2.0/omega) - 1.0;
1172: /* x = (E + U)^{-1} b */
1173: for (i=m-1; i>=0; i--) {
1174: n = a->i[i+1] - diag[i] - 1;
1175: idx = a->j + diag[i] + 1;
1176: v = a->a + diag[i] + 1;
1177: sum = b[i];
1178: SPARSEDENSEMDOT(sum,xs,v,idx,n);
1179: x[i] = sum*idiag[i];
1180: }
1182: /* t = b - (2*E - D)x */
1183: v = a->a;
1184: for (i=0; i<m; i++) { t[i] = b[i] - scale*(v[*diag++])*x[i]; }
1186: /* t = (E + L)^{-1}t */
1187: ts = t;
1188: diag = a->diag;
1189: for (i=0; i<m; i++) {
1190: n = diag[i] - a->i[i];
1191: idx = a->j + a->i[i];
1192: v = a->a + a->i[i];
1193: sum = t[i];
1194: SPARSEDENSEMDOT(sum,ts,v,idx,n);
1195: t[i] = sum*idiag[i];
1196: /* x = x + t */
1197: x[i] += t[i];
1198: }
1200: PetscLogFlops(6*m-1 + 2*a->nz);
1201: VecRestoreArray(xx,&x);
1202: if (bb != xx) {VecRestoreArray(bb,(PetscScalar**)&b);}
1203: return(0);
1204: }
1205: if (flag & SOR_ZERO_INITIAL_GUESS) {
1206: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){
1207: #if defined(PETSC_USE_FORTRAN_KERNEL_RELAXAIJ)
1208: fortranrelaxaijforwardzero_(&m,&omega,x,a->i,a->j,(PetscInt*)diag,idiag,a->a,(void*)b);
1209: #else
1210: for (i=0; i<m; i++) {
1211: n = diag[i] - a->i[i];
1212: idx = a->j + a->i[i];
1213: v = a->a + a->i[i];
1214: sum = b[i];
1215: SPARSEDENSEMDOT(sum,xs,v,idx,n);
1216: x[i] = sum*idiag[i];
1217: }
1218: #endif
1219: xb = x;
1220: PetscLogFlops(a->nz);
1221: } else xb = b;
1222: if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) &&
1223: (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) {
1224: for (i=0; i<m; i++) {
1225: x[i] *= mdiag[i];
1226: }
1227: PetscLogFlops(m);
1228: }
1229: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){
1230: #if defined(PETSC_USE_FORTRAN_KERNEL_RELAXAIJ)
1231: fortranrelaxaijbackwardzero_(&m,&omega,x,a->i,a->j,(PetscInt*)diag,idiag,a->a,(void*)xb);
1232: #else
1233: for (i=m-1; i>=0; i--) {
1234: n = a->i[i+1] - diag[i] - 1;
1235: idx = a->j + diag[i] + 1;
1236: v = a->a + diag[i] + 1;
1237: sum = xb[i];
1238: SPARSEDENSEMDOT(sum,xs,v,idx,n);
1239: x[i] = sum*idiag[i];
1240: }
1241: #endif
1242: PetscLogFlops(a->nz);
1243: }
1244: its--;
1245: }
1246: while (its--) {
1247: if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){
1248: #if defined(PETSC_USE_FORTRAN_KERNEL_RELAXAIJ)
1249: fortranrelaxaijforward_(&m,&omega,x,a->i,a->j,(PetscInt*)diag,a->a,(void*)b);
1250: #else
1251: for (i=0; i<m; i++) {
1252: d = fshift + a->a[diag[i]];
1253: n = a->i[i+1] - a->i[i];
1254: idx = a->j + a->i[i];
1255: v = a->a + a->i[i];
1256: sum = b[i];
1257: SPARSEDENSEMDOT(sum,xs,v,idx,n);
1258: x[i] = (1. - omega)*x[i] + (sum + mdiag[i]*x[i])*idiag[i];
1259: }
1260: #endif
1261: PetscLogFlops(a->nz);
1262: }
1263: if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){
1264: #if defined(PETSC_USE_FORTRAN_KERNEL_RELAXAIJ)
1265: fortranrelaxaijbackward_(&m,&omega,x,a->i,a->j,(PetscInt*)diag,a->a,(void*)b);
1266: #else
1267: for (i=m-1; i>=0; i--) {
1268: d = fshift + a->a[diag[i]];
1269: n = a->i[i+1] - a->i[i];
1270: idx = a->j + a->i[i];
1271: v = a->a + a->i[i];
1272: sum = b[i];
1273: SPARSEDENSEMDOT(sum,xs,v,idx,n);
1274: x[i] = (1. - omega)*x[i] + (sum + mdiag[i]*x[i])*idiag[i];
1275: }
1276: #endif
1277: PetscLogFlops(a->nz);
1278: }
1279: }
1280: VecRestoreArray(xx,&x);
1281: if (bb != xx) {VecRestoreArray(bb,(PetscScalar**)&b);}
1282: return(0);
1283: }
1287: PetscErrorCode MatGetInfo_SeqAIJ(Mat A,MatInfoType flag,MatInfo *info)
1288: {
1289: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1292: info->rows_global = (double)A->rmap.n;
1293: info->columns_global = (double)A->cmap.n;
1294: info->rows_local = (double)A->rmap.n;
1295: info->columns_local = (double)A->cmap.n;
1296: info->block_size = 1.0;
1297: info->nz_allocated = (double)a->maxnz;
1298: info->nz_used = (double)a->nz;
1299: info->nz_unneeded = (double)(a->maxnz - a->nz);
1300: info->assemblies = (double)A->num_ass;
1301: info->mallocs = (double)a->reallocs;
1302: info->memory = A->mem;
1303: if (A->factor) {
1304: info->fill_ratio_given = A->info.fill_ratio_given;
1305: info->fill_ratio_needed = A->info.fill_ratio_needed;
1306: info->factor_mallocs = A->info.factor_mallocs;
1307: } else {
1308: info->fill_ratio_given = 0;
1309: info->fill_ratio_needed = 0;
1310: info->factor_mallocs = 0;
1311: }
1312: return(0);
1313: }
1317: PetscErrorCode MatZeroRows_SeqAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag)
1318: {
1319: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1320: PetscInt i,m = A->rmap.n - 1,d;
1322: PetscTruth missing;
1325: if (a->keepzeroedrows) {
1326: for (i=0; i<N; i++) {
1327: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1328: PetscMemzero(&a->a[a->i[rows[i]]],a->ilen[rows[i]]*sizeof(PetscScalar));
1329: }
1330: if (diag != 0.0) {
1331: MatMissingDiagonal_SeqAIJ(A,&missing,&d);
1332: if (missing) SETERRQ1(PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry in row %D",d);
1333: for (i=0; i<N; i++) {
1334: a->a[a->diag[rows[i]]] = diag;
1335: }
1336: }
1337: A->same_nonzero = PETSC_TRUE;
1338: } else {
1339: if (diag != 0.0) {
1340: for (i=0; i<N; i++) {
1341: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1342: if (a->ilen[rows[i]] > 0) {
1343: a->ilen[rows[i]] = 1;
1344: a->a[a->i[rows[i]]] = diag;
1345: a->j[a->i[rows[i]]] = rows[i];
1346: } else { /* in case row was completely empty */
1347: MatSetValues_SeqAIJ(A,1,&rows[i],1,&rows[i],&diag,INSERT_VALUES);
1348: }
1349: }
1350: } else {
1351: for (i=0; i<N; i++) {
1352: if (rows[i] < 0 || rows[i] > m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range", rows[i]);
1353: a->ilen[rows[i]] = 0;
1354: }
1355: }
1356: A->same_nonzero = PETSC_FALSE;
1357: }
1358: MatAssemblyEnd_SeqAIJ(A,MAT_FINAL_ASSEMBLY);
1359: return(0);
1360: }
1364: PetscErrorCode MatGetRow_SeqAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1365: {
1366: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1367: PetscInt *itmp;
1370: if (row < 0 || row >= A->rmap.n) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Row %D out of range",row);
1372: *nz = a->i[row+1] - a->i[row];
1373: if (v) *v = a->a + a->i[row];
1374: if (idx) {
1375: itmp = a->j + a->i[row];
1376: if (*nz) {
1377: *idx = itmp;
1378: }
1379: else *idx = 0;
1380: }
1381: return(0);
1382: }
1384: /* remove this function? */
1387: PetscErrorCode MatRestoreRow_SeqAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1388: {
1390: return(0);
1391: }
1395: PetscErrorCode MatNorm_SeqAIJ(Mat A,NormType type,PetscReal *nrm)
1396: {
1397: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1398: PetscScalar *v = a->a;
1399: PetscReal sum = 0.0;
1401: PetscInt i,j;
1404: if (type == NORM_FROBENIUS) {
1405: for (i=0; i<a->nz; i++) {
1406: #if defined(PETSC_USE_COMPLEX)
1407: sum += PetscRealPart(PetscConj(*v)*(*v)); v++;
1408: #else
1409: sum += (*v)*(*v); v++;
1410: #endif
1411: }
1412: *nrm = sqrt(sum);
1413: } else if (type == NORM_1) {
1414: PetscReal *tmp;
1415: PetscInt *jj = a->j;
1416: PetscMalloc((A->cmap.n+1)*sizeof(PetscReal),&tmp);
1417: PetscMemzero(tmp,A->cmap.n*sizeof(PetscReal));
1418: *nrm = 0.0;
1419: for (j=0; j<a->nz; j++) {
1420: tmp[*jj++] += PetscAbsScalar(*v); v++;
1421: }
1422: for (j=0; j<A->cmap.n; j++) {
1423: if (tmp[j] > *nrm) *nrm = tmp[j];
1424: }
1425: PetscFree(tmp);
1426: } else if (type == NORM_INFINITY) {
1427: *nrm = 0.0;
1428: for (j=0; j<A->rmap.n; j++) {
1429: v = a->a + a->i[j];
1430: sum = 0.0;
1431: for (i=0; i<a->i[j+1]-a->i[j]; i++) {
1432: sum += PetscAbsScalar(*v); v++;
1433: }
1434: if (sum > *nrm) *nrm = sum;
1435: }
1436: } else {
1437: SETERRQ(PETSC_ERR_SUP,"No support for two norm");
1438: }
1439: return(0);
1440: }
1444: PetscErrorCode MatTranspose_SeqAIJ(Mat A,Mat *B)
1445: {
1446: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1447: Mat C;
1449: PetscInt i,*aj = a->j,*ai = a->i,m = A->rmap.n,len,*col;
1450: PetscScalar *array = a->a;
1453: if (!B && m != A->cmap.n) SETERRQ(PETSC_ERR_ARG_SIZ,"Square matrix only for in-place");
1454: PetscMalloc((1+A->cmap.n)*sizeof(PetscInt),&col);
1455: PetscMemzero(col,(1+A->cmap.n)*sizeof(PetscInt));
1456:
1457: for (i=0; i<ai[m]; i++) col[aj[i]] += 1;
1458: MatCreate(A->comm,&C);
1459: MatSetSizes(C,A->cmap.n,m,A->cmap.n,m);
1460: MatSetType(C,A->type_name);
1461: MatSeqAIJSetPreallocation_SeqAIJ(C,0,col);
1462: PetscFree(col);
1463: for (i=0; i<m; i++) {
1464: len = ai[i+1]-ai[i];
1465: MatSetValues_SeqAIJ(C,len,aj,1,&i,array,INSERT_VALUES);
1466: array += len;
1467: aj += len;
1468: }
1470: MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);
1471: MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);
1473: if (B) {
1474: *B = C;
1475: } else {
1476: MatHeaderCopy(A,C);
1477: }
1478: return(0);
1479: }
1484: PetscErrorCode MatIsTranspose_SeqAIJ(Mat A,Mat B,PetscReal tol,PetscTruth *f)
1485: {
1486: Mat_SeqAIJ *aij = (Mat_SeqAIJ *) A->data,*bij = (Mat_SeqAIJ*) A->data;
1487: PetscInt *adx,*bdx,*aii,*bii,*aptr,*bptr; PetscScalar *va,*vb;
1489: PetscInt ma,na,mb,nb, i;
1492: bij = (Mat_SeqAIJ *) B->data;
1493:
1494: MatGetSize(A,&ma,&na);
1495: MatGetSize(B,&mb,&nb);
1496: if (ma!=nb || na!=mb){
1497: *f = PETSC_FALSE;
1498: return(0);
1499: }
1500: aii = aij->i; bii = bij->i;
1501: adx = aij->j; bdx = bij->j;
1502: va = aij->a; vb = bij->a;
1503: PetscMalloc(ma*sizeof(PetscInt),&aptr);
1504: PetscMalloc(mb*sizeof(PetscInt),&bptr);
1505: for (i=0; i<ma; i++) aptr[i] = aii[i];
1506: for (i=0; i<mb; i++) bptr[i] = bii[i];
1508: *f = PETSC_TRUE;
1509: for (i=0; i<ma; i++) {
1510: while (aptr[i]<aii[i+1]) {
1511: PetscInt idc,idr;
1512: PetscScalar vc,vr;
1513: /* column/row index/value */
1514: idc = adx[aptr[i]];
1515: idr = bdx[bptr[idc]];
1516: vc = va[aptr[i]];
1517: vr = vb[bptr[idc]];
1518: if (i!=idr || PetscAbsScalar(vc-vr) > tol) {
1519: *f = PETSC_FALSE;
1520: goto done;
1521: } else {
1522: aptr[i]++;
1523: if (B || i!=idc) bptr[idc]++;
1524: }
1525: }
1526: }
1527: done:
1528: PetscFree(aptr);
1529: if (B) {
1530: PetscFree(bptr);
1531: }
1532: return(0);
1533: }
1538: PetscErrorCode MatIsSymmetric_SeqAIJ(Mat A,PetscReal tol,PetscTruth *f)
1539: {
1542: MatIsTranspose_SeqAIJ(A,A,tol,f);
1543: return(0);
1544: }
1548: PetscErrorCode MatDiagonalScale_SeqAIJ(Mat A,Vec ll,Vec rr)
1549: {
1550: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1551: PetscScalar *l,*r,x,*v;
1553: PetscInt i,j,m = A->rmap.n,n = A->cmap.n,M,nz = a->nz,*jj;
1556: if (ll) {
1557: /* The local size is used so that VecMPI can be passed to this routine
1558: by MatDiagonalScale_MPIAIJ */
1559: VecGetLocalSize(ll,&m);
1560: if (m != A->rmap.n) SETERRQ(PETSC_ERR_ARG_SIZ,"Left scaling vector wrong length");
1561: VecGetArray(ll,&l);
1562: v = a->a;
1563: for (i=0; i<m; i++) {
1564: x = l[i];
1565: M = a->i[i+1] - a->i[i];
1566: for (j=0; j<M; j++) { (*v++) *= x;}
1567: }
1568: VecRestoreArray(ll,&l);
1569: PetscLogFlops(nz);
1570: }
1571: if (rr) {
1572: VecGetLocalSize(rr,&n);
1573: if (n != A->cmap.n) SETERRQ(PETSC_ERR_ARG_SIZ,"Right scaling vector wrong length");
1574: VecGetArray(rr,&r);
1575: v = a->a; jj = a->j;
1576: for (i=0; i<nz; i++) {
1577: (*v++) *= r[*jj++];
1578: }
1579: VecRestoreArray(rr,&r);
1580: PetscLogFlops(nz);
1581: }
1582: return(0);
1583: }
1587: PetscErrorCode MatGetSubMatrix_SeqAIJ(Mat A,IS isrow,IS iscol,PetscInt csize,MatReuse scall,Mat *B)
1588: {
1589: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data,*c;
1591: PetscInt *smap,i,k,kstart,kend,oldcols = A->cmap.n,*lens;
1592: PetscInt row,mat_i,*mat_j,tcol,first,step,*mat_ilen,sum,lensi;
1593: PetscInt *irow,*icol,nrows,ncols;
1594: PetscInt *starts,*j_new,*i_new,*aj = a->j,*ai = a->i,ii,*ailen = a->ilen;
1595: PetscScalar *a_new,*mat_a;
1596: Mat C;
1597: PetscTruth stride;
1600: ISSorted(isrow,(PetscTruth*)&i);
1601: if (!i) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"ISrow is not sorted");
1602: ISSorted(iscol,(PetscTruth*)&i);
1603: if (!i) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"IScol is not sorted");
1605: ISGetIndices(isrow,&irow);
1606: ISGetLocalSize(isrow,&nrows);
1607: ISGetLocalSize(iscol,&ncols);
1609: ISStrideGetInfo(iscol,&first,&step);
1610: ISStride(iscol,&stride);
1611: if (stride && step == 1) {
1612: /* special case of contiguous rows */
1613: PetscMalloc((2*nrows+1)*sizeof(PetscInt),&lens);
1614: starts = lens + nrows;
1615: /* loop over new rows determining lens and starting points */
1616: for (i=0; i<nrows; i++) {
1617: kstart = ai[irow[i]];
1618: kend = kstart + ailen[irow[i]];
1619: for (k=kstart; k<kend; k++) {
1620: if (aj[k] >= first) {
1621: starts[i] = k;
1622: break;
1623: }
1624: }
1625: sum = 0;
1626: while (k < kend) {
1627: if (aj[k++] >= first+ncols) break;
1628: sum++;
1629: }
1630: lens[i] = sum;
1631: }
1632: /* create submatrix */
1633: if (scall == MAT_REUSE_MATRIX) {
1634: PetscInt n_cols,n_rows;
1635: MatGetSize(*B,&n_rows,&n_cols);
1636: if (n_rows != nrows || n_cols != ncols) SETERRQ(PETSC_ERR_ARG_SIZ,"Reused submatrix wrong size");
1637: MatZeroEntries(*B);
1638: C = *B;
1639: } else {
1640: MatCreate(A->comm,&C);
1641: MatSetSizes(C,nrows,ncols,PETSC_DETERMINE,PETSC_DETERMINE);
1642: MatSetType(C,A->type_name);
1643: MatSeqAIJSetPreallocation_SeqAIJ(C,0,lens);
1644: }
1645: c = (Mat_SeqAIJ*)C->data;
1647: /* loop over rows inserting into submatrix */
1648: a_new = c->a;
1649: j_new = c->j;
1650: i_new = c->i;
1652: for (i=0; i<nrows; i++) {
1653: ii = starts[i];
1654: lensi = lens[i];
1655: for (k=0; k<lensi; k++) {
1656: *j_new++ = aj[ii+k] - first;
1657: }
1658: PetscMemcpy(a_new,a->a + starts[i],lensi*sizeof(PetscScalar));
1659: a_new += lensi;
1660: i_new[i+1] = i_new[i] + lensi;
1661: c->ilen[i] = lensi;
1662: }
1663: PetscFree(lens);
1664: } else {
1665: ISGetIndices(iscol,&icol);
1666: PetscMalloc((1+oldcols)*sizeof(PetscInt),&smap);
1667:
1668: PetscMalloc((1+nrows)*sizeof(PetscInt),&lens);
1669: PetscMemzero(smap,oldcols*sizeof(PetscInt));
1670: for (i=0; i<ncols; i++) smap[icol[i]] = i+1;
1671: /* determine lens of each row */
1672: for (i=0; i<nrows; i++) {
1673: kstart = ai[irow[i]];
1674: kend = kstart + a->ilen[irow[i]];
1675: lens[i] = 0;
1676: for (k=kstart; k<kend; k++) {
1677: if (smap[aj[k]]) {
1678: lens[i]++;
1679: }
1680: }
1681: }
1682: /* Create and fill new matrix */
1683: if (scall == MAT_REUSE_MATRIX) {
1684: PetscTruth equal;
1686: c = (Mat_SeqAIJ *)((*B)->data);
1687: if ((*B)->rmap.n != nrows || (*B)->cmap.n != ncols) SETERRQ(PETSC_ERR_ARG_SIZ,"Cannot reuse matrix. wrong size");
1688: PetscMemcmp(c->ilen,lens,(*B)->rmap.n*sizeof(PetscInt),&equal);
1689: if (!equal) {
1690: SETERRQ(PETSC_ERR_ARG_SIZ,"Cannot reuse matrix. wrong no of nonzeros");
1691: }
1692: PetscMemzero(c->ilen,(*B)->rmap.n*sizeof(PetscInt));
1693: C = *B;
1694: } else {
1695: MatCreate(A->comm,&C);
1696: MatSetSizes(C,nrows,ncols,PETSC_DETERMINE,PETSC_DETERMINE);
1697: MatSetType(C,A->type_name);
1698: MatSeqAIJSetPreallocation_SeqAIJ(C,0,lens);
1699: }
1700: c = (Mat_SeqAIJ *)(C->data);
1701: for (i=0; i<nrows; i++) {
1702: row = irow[i];
1703: kstart = ai[row];
1704: kend = kstart + a->ilen[row];
1705: mat_i = c->i[i];
1706: mat_j = c->j + mat_i;
1707: mat_a = c->a + mat_i;
1708: mat_ilen = c->ilen + i;
1709: for (k=kstart; k<kend; k++) {
1710: if ((tcol=smap[a->j[k]])) {
1711: *mat_j++ = tcol - 1;
1712: *mat_a++ = a->a[k];
1713: (*mat_ilen)++;
1715: }
1716: }
1717: }
1718: /* Free work space */
1719: ISRestoreIndices(iscol,&icol);
1720: PetscFree(smap);
1721: PetscFree(lens);
1722: }
1723: MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);
1724: MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);
1726: ISRestoreIndices(isrow,&irow);
1727: *B = C;
1728: return(0);
1729: }
1731: /*
1732: */
1735: PetscErrorCode MatILUFactor_SeqAIJ(Mat inA,IS row,IS col,MatFactorInfo *info)
1736: {
1737: Mat_SeqAIJ *a = (Mat_SeqAIJ*)inA->data;
1739: Mat outA;
1740: PetscTruth row_identity,col_identity;
1743: if (info->levels != 0) SETERRQ(PETSC_ERR_SUP,"Only levels=0 supported for in-place ilu");
1744: ISIdentity(row,&row_identity);
1745: ISIdentity(col,&col_identity);
1746: if (!row_identity || !col_identity) {
1747: SETERRQ(PETSC_ERR_ARG_WRONG,"Row and column permutations must be identity for in-place ILU");
1748: }
1750: outA = inA;
1751: inA->factor = FACTOR_LU;
1752: a->row = row;
1753: a->col = col;
1754: PetscObjectReference((PetscObject)row);
1755: PetscObjectReference((PetscObject)col);
1757: /* Create the inverse permutation so that it can be used in MatLUFactorNumeric() */
1758: if (a->icol) {ISDestroy(a->icol);} /* need to remove old one */
1759: ISInvertPermutation(col,PETSC_DECIDE,&a->icol);
1760: PetscLogObjectParent(inA,a->icol);
1762: if (!a->solve_work) { /* this matrix may have been factored before */
1763: PetscMalloc((inA->rmap.n+1)*sizeof(PetscScalar),&a->solve_work);
1764: }
1766: MatMarkDiagonal_SeqAIJ(inA);
1767: MatLUFactorNumeric_SeqAIJ(inA,info,&outA);
1768: return(0);
1769: }
1771: #include petscblaslapack.h
1774: PetscErrorCode MatScale_SeqAIJ(Mat inA,PetscScalar alpha)
1775: {
1776: Mat_SeqAIJ *a = (Mat_SeqAIJ*)inA->data;
1777: PetscBLASInt bnz = (PetscBLASInt)a->nz,one = 1;
1778: PetscScalar oalpha = alpha;
1783: BLASscal_(&bnz,&oalpha,a->a,&one);
1784: PetscLogFlops(a->nz);
1785: return(0);
1786: }
1790: PetscErrorCode MatGetSubMatrices_SeqAIJ(Mat A,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *B[])
1791: {
1793: PetscInt i;
1796: if (scall == MAT_INITIAL_MATRIX) {
1797: PetscMalloc((n+1)*sizeof(Mat),B);
1798: }
1800: for (i=0; i<n; i++) {
1801: MatGetSubMatrix_SeqAIJ(A,irow[i],icol[i],PETSC_DECIDE,scall,&(*B)[i]);
1802: }
1803: return(0);
1804: }
1808: PetscErrorCode MatIncreaseOverlap_SeqAIJ(Mat A,PetscInt is_max,IS is[],PetscInt ov)
1809: {
1810: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1812: PetscInt row,i,j,k,l,m,n,*idx,*nidx,isz,val;
1813: PetscInt start,end,*ai,*aj;
1814: PetscBT table;
1817: m = A->rmap.n;
1818: ai = a->i;
1819: aj = a->j;
1821: if (ov < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"illegal negative overlap value used");
1823: PetscMalloc((m+1)*sizeof(PetscInt),&nidx);
1824: PetscBTCreate(m,table);
1826: for (i=0; i<is_max; i++) {
1827: /* Initialize the two local arrays */
1828: isz = 0;
1829: PetscBTMemzero(m,table);
1830:
1831: /* Extract the indices, assume there can be duplicate entries */
1832: ISGetIndices(is[i],&idx);
1833: ISGetLocalSize(is[i],&n);
1834:
1835: /* Enter these into the temp arrays. I.e., mark table[row], enter row into new index */
1836: for (j=0; j<n ; ++j){
1837: if(!PetscBTLookupSet(table,idx[j])) { nidx[isz++] = idx[j];}
1838: }
1839: ISRestoreIndices(is[i],&idx);
1840: ISDestroy(is[i]);
1841:
1842: k = 0;
1843: for (j=0; j<ov; j++){ /* for each overlap */
1844: n = isz;
1845: for (; k<n ; k++){ /* do only those rows in nidx[k], which are not done yet */
1846: row = nidx[k];
1847: start = ai[row];
1848: end = ai[row+1];
1849: for (l = start; l<end ; l++){
1850: val = aj[l] ;
1851: if (!PetscBTLookupSet(table,val)) {nidx[isz++] = val;}
1852: }
1853: }
1854: }
1855: ISCreateGeneral(PETSC_COMM_SELF,isz,nidx,(is+i));
1856: }
1857: PetscBTDestroy(table);
1858: PetscFree(nidx);
1859: return(0);
1860: }
1862: /* -------------------------------------------------------------- */
1865: PetscErrorCode MatPermute_SeqAIJ(Mat A,IS rowp,IS colp,Mat *B)
1866: {
1867: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1869: PetscInt i,nz,m = A->rmap.n,n = A->cmap.n,*col;
1870: PetscInt *row,*cnew,j,*lens;
1871: IS icolp,irowp;
1872: PetscInt *cwork;
1873: PetscScalar *vwork;
1876: ISInvertPermutation(rowp,PETSC_DECIDE,&irowp);
1877: ISGetIndices(irowp,&row);
1878: ISInvertPermutation(colp,PETSC_DECIDE,&icolp);
1879: ISGetIndices(icolp,&col);
1880:
1881: /* determine lengths of permuted rows */
1882: PetscMalloc((m+1)*sizeof(PetscInt),&lens);
1883: for (i=0; i<m; i++) {
1884: lens[row[i]] = a->i[i+1] - a->i[i];
1885: }
1886: MatCreate(A->comm,B);
1887: MatSetSizes(*B,m,n,m,n);
1888: MatSetType(*B,A->type_name);
1889: MatSeqAIJSetPreallocation_SeqAIJ(*B,0,lens);
1890: PetscFree(lens);
1892: PetscMalloc(n*sizeof(PetscInt),&cnew);
1893: for (i=0; i<m; i++) {
1894: MatGetRow_SeqAIJ(A,i,&nz,&cwork,&vwork);
1895: for (j=0; j<nz; j++) { cnew[j] = col[cwork[j]];}
1896: MatSetValues_SeqAIJ(*B,1,&row[i],nz,cnew,vwork,INSERT_VALUES);
1897: MatRestoreRow_SeqAIJ(A,i,&nz,&cwork,&vwork);
1898: }
1899: PetscFree(cnew);
1900: (*B)->assembled = PETSC_FALSE;
1901: MatAssemblyBegin(*B,MAT_FINAL_ASSEMBLY);
1902: MatAssemblyEnd(*B,MAT_FINAL_ASSEMBLY);
1903: ISRestoreIndices(irowp,&row);
1904: ISRestoreIndices(icolp,&col);
1905: ISDestroy(irowp);
1906: ISDestroy(icolp);
1907: return(0);
1908: }
1912: PetscErrorCode MatCopy_SeqAIJ(Mat A,Mat B,MatStructure str)
1913: {
1917: /* If the two matrices have the same copy implementation, use fast copy. */
1918: if (str == SAME_NONZERO_PATTERN && (A->ops->copy == B->ops->copy)) {
1919: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1920: Mat_SeqAIJ *b = (Mat_SeqAIJ*)B->data;
1922: if (a->i[A->rmap.n] != b->i[B->rmap.n]) {
1923: SETERRQ(PETSC_ERR_ARG_INCOMP,"Number of nonzeros in two matrices are different");
1924: }
1925: PetscMemcpy(b->a,a->a,(a->i[A->rmap.n])*sizeof(PetscScalar));
1926: } else {
1927: MatCopy_Basic(A,B,str);
1928: }
1929: return(0);
1930: }
1934: PetscErrorCode MatSetUpPreallocation_SeqAIJ(Mat A)
1935: {
1939: MatSeqAIJSetPreallocation_SeqAIJ(A,PETSC_DEFAULT,0);
1940: return(0);
1941: }
1945: PetscErrorCode MatGetArray_SeqAIJ(Mat A,PetscScalar *array[])
1946: {
1947: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
1949: *array = a->a;
1950: return(0);
1951: }
1955: PetscErrorCode MatRestoreArray_SeqAIJ(Mat A,PetscScalar *array[])
1956: {
1958: return(0);
1959: }
1963: PetscErrorCode MatFDColoringApply_SeqAIJ(Mat J,MatFDColoring coloring,Vec x1,MatStructure *flag,void *sctx)
1964: {
1965: PetscErrorCode (*f)(void*,Vec,Vec,void*) = (PetscErrorCode (*)(void*,Vec,Vec,void *))coloring->f;
1967: PetscInt k,N,start,end,l,row,col,srow,**vscaleforrow,m1,m2;
1968: PetscScalar dx,*y,*xx,*w3_array;
1969: PetscScalar *vscale_array;
1970: PetscReal epsilon = coloring->error_rel,umin = coloring->umin;
1971: Vec w1,w2,w3;
1972: void *fctx = coloring->fctx;
1973: PetscTruth flg;
1976: if (!coloring->w1) {
1977: VecDuplicate(x1,&coloring->w1);
1978: PetscLogObjectParent(coloring,coloring->w1);
1979: VecDuplicate(x1,&coloring->w2);
1980: PetscLogObjectParent(coloring,coloring->w2);
1981: VecDuplicate(x1,&coloring->w3);
1982: PetscLogObjectParent(coloring,coloring->w3);
1983: }
1984: w1 = coloring->w1; w2 = coloring->w2; w3 = coloring->w3;
1986: MatSetUnfactored(J);
1987: PetscOptionsHasName(coloring->prefix,"-mat_fd_coloring_dont_rezero",&flg);
1988: if (flg) {
1989: PetscInfo(coloring,"Not calling MatZeroEntries()\n");
1990: } else {
1991: PetscTruth assembled;
1992: MatAssembled(J,&assembled);
1993: if (assembled) {
1994: MatZeroEntries(J);
1995: }
1996: }
1998: VecGetOwnershipRange(x1,&start,&end);
1999: VecGetSize(x1,&N);
2001: /*
2002: This is a horrible, horrible, hack. See DMMGComputeJacobian_Multigrid() it inproperly sets
2003: coloring->F for the coarser grids from the finest
2004: */
2005: if (coloring->F) {
2006: VecGetLocalSize(coloring->F,&m1);
2007: VecGetLocalSize(w1,&m2);
2008: if (m1 != m2) {
2009: coloring->F = 0;
2010: }
2011: }
2013: if (coloring->F) {
2014: w1 = coloring->F;
2015: coloring->F = 0;
2016: } else {
2018: (*f)(sctx,x1,w1,fctx);
2020: }
2022: /*
2023: Compute all the scale factors and share with other processors
2024: */
2025: VecGetArray(x1,&xx);xx = xx - start;
2026: VecGetArray(coloring->vscale,&vscale_array);vscale_array = vscale_array - start;
2027: for (k=0; k<coloring->ncolors; k++) {
2028: /*
2029: Loop over each column associated with color adding the
2030: perturbation to the vector w3.
2031: */
2032: for (l=0; l<coloring->ncolumns[k]; l++) {
2033: col = coloring->columns[k][l]; /* column of the matrix we are probing for */
2034: dx = xx[col];
2035: if (dx == 0.0) dx = 1.0;
2036: #if !defined(PETSC_USE_COMPLEX)
2037: if (dx < umin && dx >= 0.0) dx = umin;
2038: else if (dx < 0.0 && dx > -umin) dx = -umin;
2039: #else
2040: if (PetscAbsScalar(dx) < umin && PetscRealPart(dx) >= 0.0) dx = umin;
2041: else if (PetscRealPart(dx) < 0.0 && PetscAbsScalar(dx) < umin) dx = -umin;
2042: #endif
2043: dx *= epsilon;
2044: vscale_array[col] = 1.0/dx;
2045: }
2046: }
2047: vscale_array = vscale_array + start;VecRestoreArray(coloring->vscale,&vscale_array);
2048: VecGhostUpdateBegin(coloring->vscale,INSERT_VALUES,SCATTER_FORWARD);
2049: VecGhostUpdateEnd(coloring->vscale,INSERT_VALUES,SCATTER_FORWARD);
2051: /* VecView(coloring->vscale,PETSC_VIEWER_STDOUT_WORLD);
2052: VecView(x1,PETSC_VIEWER_STDOUT_WORLD);*/
2054: if (coloring->vscaleforrow) vscaleforrow = coloring->vscaleforrow;
2055: else vscaleforrow = coloring->columnsforrow;
2057: VecGetArray(coloring->vscale,&vscale_array);
2058: /*
2059: Loop over each color
2060: */
2061: for (k=0; k<coloring->ncolors; k++) {
2062: coloring->currentcolor = k;
2063: VecCopy(x1,w3);
2064: VecGetArray(w3,&w3_array);w3_array = w3_array - start;
2065: /*
2066: Loop over each column associated with color adding the
2067: perturbation to the vector w3.
2068: */
2069: for (l=0; l<coloring->ncolumns[k]; l++) {
2070: col = coloring->columns[k][l]; /* column of the matrix we are probing for */
2071: dx = xx[col];
2072: if (dx == 0.0) dx = 1.0;
2073: #if !defined(PETSC_USE_COMPLEX)
2074: if (dx < umin && dx >= 0.0) dx = umin;
2075: else if (dx < 0.0 && dx > -umin) dx = -umin;
2076: #else
2077: if (PetscAbsScalar(dx) < umin && PetscRealPart(dx) >= 0.0) dx = umin;
2078: else if (PetscRealPart(dx) < 0.0 && PetscAbsScalar(dx) < umin) dx = -umin;
2079: #endif
2080: dx *= epsilon;
2081: if (!PetscAbsScalar(dx)) SETERRQ(PETSC_ERR_PLIB,"Computed 0 differencing parameter");
2082: w3_array[col] += dx;
2083: }
2084: w3_array = w3_array + start; VecRestoreArray(w3,&w3_array);
2086: /*
2087: Evaluate function at x1 + dx (here dx is a vector of perturbations)
2088: */
2091: (*f)(sctx,w3,w2,fctx);
2093: VecAXPY(w2,-1.0,w1);
2095: /*
2096: Loop over rows of vector, putting results into Jacobian matrix
2097: */
2098: VecGetArray(w2,&y);
2099: for (l=0; l<coloring->nrows[k]; l++) {
2100: row = coloring->rows[k][l];
2101: col = coloring->columnsforrow[k][l];
2102: y[row] *= vscale_array[vscaleforrow[k][l]];
2103: srow = row + start;
2104: MatSetValues_SeqAIJ(J,1,&srow,1,&col,y+row,INSERT_VALUES);
2105: }
2106: VecRestoreArray(w2,&y);
2107: }
2108: coloring->currentcolor = k;
2109: VecRestoreArray(coloring->vscale,&vscale_array);
2110: xx = xx + start; VecRestoreArray(x1,&xx);
2111: MatAssemblyBegin(J,MAT_FINAL_ASSEMBLY);
2112: MatAssemblyEnd(J,MAT_FINAL_ASSEMBLY);
2113: return(0);
2114: }
2116: #include petscblaslapack.h
2119: PetscErrorCode MatAXPY_SeqAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str)
2120: {
2122: PetscInt i;
2123: Mat_SeqAIJ *x = (Mat_SeqAIJ *)X->data,*y = (Mat_SeqAIJ *)Y->data;
2124: PetscBLASInt one=1,bnz = (PetscBLASInt)x->nz;
2127: if (str == SAME_NONZERO_PATTERN) {
2128: PetscScalar alpha = a;
2129: BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one);
2130: } else if (str == SUBSET_NONZERO_PATTERN) { /* nonzeros of X is a subset of Y's */
2131: if (y->xtoy && y->XtoY != X) {
2132: PetscFree(y->xtoy);
2133: MatDestroy(y->XtoY);
2134: }
2135: if (!y->xtoy) { /* get xtoy */
2136: MatAXPYGetxtoy_Private(X->rmap.n,x->i,x->j,PETSC_NULL, y->i,y->j,PETSC_NULL, &y->xtoy);
2137: y->XtoY = X;
2138: }
2139: for (i=0; i<x->nz; i++) y->a[y->xtoy[i]] += a*(x->a[i]);
2140: PetscInfo3(0,"ratio of nnz(X)/nnz(Y): %d/%d = %G\n",x->nz,y->nz,(PetscReal)(x->nz)/y->nz);
2141: } else {
2142: MatAXPY_Basic(Y,a,X,str);
2143: }
2144: return(0);
2145: }
2149: PetscErrorCode MatSetBlockSize_SeqAIJ(Mat A,PetscInt bs)
2150: {
2152: return(0);
2153: }
2157: PetscErrorCode MatConjugate_SeqAIJ(Mat mat)
2158: {
2159: #if defined(PETSC_USE_COMPLEX)
2160: Mat_SeqAIJ *aij = (Mat_SeqAIJ *)mat->data;
2161: PetscInt i,nz;
2162: PetscScalar *a;
2165: nz = aij->nz;
2166: a = aij->a;
2167: for (i=0; i<nz; i++) {
2168: a[i] = PetscConj(a[i]);
2169: }
2170: #else
2172: #endif
2173: return(0);
2174: }
2178: PetscErrorCode MatGetRowMax_SeqAIJ(Mat A,Vec v)
2179: {
2180: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
2182: PetscInt i,j,m = A->rmap.n,*ai,*aj,ncols,n;
2183: PetscReal atmp;
2184: PetscScalar *x,zero = 0.0;
2185: MatScalar *aa;
2188: if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2189: aa = a->a;
2190: ai = a->i;
2191: aj = a->j;
2193: VecSet(v,zero);
2194: VecGetArray(v,&x);
2195: VecGetLocalSize(v,&n);
2196: if (n != A->rmap.n) SETERRQ(PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector");
2197: for (i=0; i<m; i++) {
2198: ncols = ai[1] - ai[0]; ai++;
2199: for (j=0; j<ncols; j++){
2200: atmp = PetscAbsScalar(*aa); aa++;
2201: if (PetscAbsScalar(x[i]) < atmp) x[i] = atmp;
2202: aj++;
2203: }
2204: }
2205: VecRestoreArray(v,&x);
2206: return(0);
2207: }
2209: /* -------------------------------------------------------------------*/
2210: static struct _MatOps MatOps_Values = {MatSetValues_SeqAIJ,
2211: MatGetRow_SeqAIJ,
2212: MatRestoreRow_SeqAIJ,
2213: MatMult_SeqAIJ,
2214: /* 4*/ MatMultAdd_SeqAIJ,
2215: MatMultTranspose_SeqAIJ,
2216: MatMultTransposeAdd_SeqAIJ,
2217: MatSolve_SeqAIJ,
2218: MatSolveAdd_SeqAIJ,
2219: MatSolveTranspose_SeqAIJ,
2220: /*10*/ MatSolveTransposeAdd_SeqAIJ,
2221: MatLUFactor_SeqAIJ,
2222: 0,
2223: MatRelax_SeqAIJ,
2224: MatTranspose_SeqAIJ,
2225: /*15*/ MatGetInfo_SeqAIJ,
2226: MatEqual_SeqAIJ,
2227: MatGetDiagonal_SeqAIJ,
2228: MatDiagonalScale_SeqAIJ,
2229: MatNorm_SeqAIJ,
2230: /*20*/ 0,
2231: MatAssemblyEnd_SeqAIJ,
2232: MatCompress_SeqAIJ,
2233: MatSetOption_SeqAIJ,
2234: MatZeroEntries_SeqAIJ,
2235: /*25*/ MatZeroRows_SeqAIJ,
2236: MatLUFactorSymbolic_SeqAIJ,
2237: MatLUFactorNumeric_SeqAIJ,
2238: MatCholeskyFactorSymbolic_SeqAIJ,
2239: MatCholeskyFactorNumeric_SeqAIJ,
2240: /*30*/ MatSetUpPreallocation_SeqAIJ,
2241: MatILUFactorSymbolic_SeqAIJ,
2242: MatICCFactorSymbolic_SeqAIJ,
2243: MatGetArray_SeqAIJ,
2244: MatRestoreArray_SeqAIJ,
2245: /*35*/ MatDuplicate_SeqAIJ,
2246: 0,
2247: 0,
2248: MatILUFactor_SeqAIJ,
2249: 0,
2250: /*40*/ MatAXPY_SeqAIJ,
2251: MatGetSubMatrices_SeqAIJ,
2252: MatIncreaseOverlap_SeqAIJ,
2253: MatGetValues_SeqAIJ,
2254: MatCopy_SeqAIJ,
2255: /*45*/ 0,
2256: MatScale_SeqAIJ,
2257: 0,
2258: MatDiagonalSet_SeqAIJ,
2259: MatILUDTFactor_SeqAIJ,
2260: /*50*/ MatSetBlockSize_SeqAIJ,
2261: MatGetRowIJ_SeqAIJ,
2262: MatRestoreRowIJ_SeqAIJ,
2263: MatGetColumnIJ_SeqAIJ,
2264: MatRestoreColumnIJ_SeqAIJ,
2265: /*55*/ MatFDColoringCreate_SeqAIJ,
2266: 0,
2267: 0,
2268: MatPermute_SeqAIJ,
2269: 0,
2270: /*60*/ 0,
2271: MatDestroy_SeqAIJ,
2272: MatView_SeqAIJ,
2273: 0,
2274: 0,
2275: /*65*/ 0,
2276: 0,
2277: 0,
2278: 0,
2279: 0,
2280: /*70*/ MatGetRowMax_SeqAIJ,
2281: 0,
2282: MatSetColoring_SeqAIJ,
2283: #if defined(PETSC_HAVE_ADIC)
2284: MatSetValuesAdic_SeqAIJ,
2285: #else
2286: 0,
2287: #endif
2288: MatSetValuesAdifor_SeqAIJ,
2289: /*75*/ 0,
2290: 0,
2291: 0,
2292: 0,
2293: 0,
2294: /*80*/ 0,
2295: 0,
2296: 0,
2297: 0,
2298: MatLoad_SeqAIJ,
2299: /*85*/ MatIsSymmetric_SeqAIJ,
2300: 0,
2301: 0,
2302: 0,
2303: 0,
2304: /*90*/ MatMatMult_SeqAIJ_SeqAIJ,
2305: MatMatMultSymbolic_SeqAIJ_SeqAIJ,
2306: MatMatMultNumeric_SeqAIJ_SeqAIJ,
2307: MatPtAP_Basic,
2308: MatPtAPSymbolic_SeqAIJ,
2309: /*95*/ MatPtAPNumeric_SeqAIJ,
2310: MatMatMultTranspose_SeqAIJ_SeqAIJ,
2311: MatMatMultTransposeSymbolic_SeqAIJ_SeqAIJ,
2312: MatMatMultTransposeNumeric_SeqAIJ_SeqAIJ,
2313: MatPtAPSymbolic_SeqAIJ_SeqAIJ,
2314: /*100*/MatPtAPNumeric_SeqAIJ_SeqAIJ,
2315: 0,
2316: 0,
2317: MatConjugate_SeqAIJ,
2318: 0,
2319: /*105*/MatSetValuesRow_SeqAIJ,
2320: MatRealPart_SeqAIJ,
2321: MatImaginaryPart_SeqAIJ,
2322: 0,
2323: 0,
2324: /*110*/MatMatSolve_SeqAIJ
2325: };
2330: PetscErrorCode MatSeqAIJSetColumnIndices_SeqAIJ(Mat mat,PetscInt *indices)
2331: {
2332: Mat_SeqAIJ *aij = (Mat_SeqAIJ *)mat->data;
2333: PetscInt i,nz,n;
2337: nz = aij->maxnz;
2338: n = mat->cmap.n;
2339: for (i=0; i<nz; i++) {
2340: aij->j[i] = indices[i];
2341: }
2342: aij->nz = nz;
2343: for (i=0; i<n; i++) {
2344: aij->ilen[i] = aij->imax[i];
2345: }
2347: return(0);
2348: }
2353: /*@
2354: MatSeqAIJSetColumnIndices - Set the column indices for all the rows
2355: in the matrix.
2357: Input Parameters:
2358: + mat - the SeqAIJ matrix
2359: - indices - the column indices
2361: Level: advanced
2363: Notes:
2364: This can be called if you have precomputed the nonzero structure of the
2365: matrix and want to provide it to the matrix object to improve the performance
2366: of the MatSetValues() operation.
2368: You MUST have set the correct numbers of nonzeros per row in the call to
2369: MatCreateSeqAIJ(), and the columns indices MUST be sorted.
2371: MUST be called before any calls to MatSetValues();
2373: The indices should start with zero, not one.
2375: @*/
2376: PetscErrorCode MatSeqAIJSetColumnIndices(Mat mat,PetscInt *indices)
2377: {
2378: PetscErrorCode ierr,(*f)(Mat,PetscInt *);
2383: PetscObjectQueryFunction((PetscObject)mat,"MatSeqAIJSetColumnIndices_C",(void (**)(void))&f);
2384: if (f) {
2385: (*f)(mat,indices);
2386: } else {
2387: SETERRQ(PETSC_ERR_SUP,"Wrong type of matrix to set column indices");
2388: }
2389: return(0);
2390: }
2392: /* ----------------------------------------------------------------------------------------*/
2397: PetscErrorCode MatStoreValues_SeqAIJ(Mat mat)
2398: {
2399: Mat_SeqAIJ *aij = (Mat_SeqAIJ *)mat->data;
2401: size_t nz = aij->i[mat->rmap.n];
2404: if (aij->nonew != 1) {
2405: SETERRQ(PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NO_NEW_NONZERO_LOCATIONS);first");
2406: }
2408: /* allocate space for values if not already there */
2409: if (!aij->saved_values) {
2410: PetscMalloc((nz+1)*sizeof(PetscScalar),&aij->saved_values);
2411: }
2413: /* copy values over */
2414: PetscMemcpy(aij->saved_values,aij->a,nz*sizeof(PetscScalar));
2415: return(0);
2416: }
2421: /*@
2422: MatStoreValues - Stashes a copy of the matrix values; this allows, for
2423: example, reuse of the linear part of a Jacobian, while recomputing the
2424: nonlinear portion.
2426: Collect on Mat
2428: Input Parameters:
2429: . mat - the matrix (currently only AIJ matrices support this option)
2431: Level: advanced
2433: Common Usage, with SNESSolve():
2434: $ Create Jacobian matrix
2435: $ Set linear terms into matrix
2436: $ Apply boundary conditions to matrix, at this time matrix must have
2437: $ final nonzero structure (i.e. setting the nonlinear terms and applying
2438: $ boundary conditions again will not change the nonzero structure
2439: $ MatSetOption(mat,MAT_NO_NEW_NONZERO_LOCATIONS);
2440: $ MatStoreValues(mat);
2441: $ Call SNESSetJacobian() with matrix
2442: $ In your Jacobian routine
2443: $ MatRetrieveValues(mat);
2444: $ Set nonlinear terms in matrix
2445:
2446: Common Usage without SNESSolve(), i.e. when you handle nonlinear solve yourself:
2447: $ // build linear portion of Jacobian
2448: $ MatSetOption(mat,MAT_NO_NEW_NONZERO_LOCATIONS);
2449: $ MatStoreValues(mat);
2450: $ loop over nonlinear iterations
2451: $ MatRetrieveValues(mat);
2452: $ // call MatSetValues(mat,...) to set nonliner portion of Jacobian
2453: $ // call MatAssemblyBegin/End() on matrix
2454: $ Solve linear system with Jacobian
2455: $ endloop
2457: Notes:
2458: Matrix must already be assemblied before calling this routine
2459: Must set the matrix option MatSetOption(mat,MAT_NO_NEW_NONZERO_LOCATIONS); before
2460: calling this routine.
2462: When this is called multiple times it overwrites the previous set of stored values
2463: and does not allocated additional space.
2465: .seealso: MatRetrieveValues()
2467: @*/
2468: PetscErrorCode MatStoreValues(Mat mat)
2469: {
2470: PetscErrorCode ierr,(*f)(Mat);
2474: if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2475: if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2477: PetscObjectQueryFunction((PetscObject)mat,"MatStoreValues_C",(void (**)(void))&f);
2478: if (f) {
2479: (*f)(mat);
2480: } else {
2481: SETERRQ(PETSC_ERR_SUP,"Wrong type of matrix to store values");
2482: }
2483: return(0);
2484: }
2489: PetscErrorCode MatRetrieveValues_SeqAIJ(Mat mat)
2490: {
2491: Mat_SeqAIJ *aij = (Mat_SeqAIJ *)mat->data;
2493: PetscInt nz = aij->i[mat->rmap.n];
2496: if (aij->nonew != 1) {
2497: SETERRQ(PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NO_NEW_NONZERO_LOCATIONS);first");
2498: }
2499: if (!aij->saved_values) {
2500: SETERRQ(PETSC_ERR_ORDER,"Must call MatStoreValues(A);first");
2501: }
2502: /* copy values over */
2503: PetscMemcpy(aij->a,aij->saved_values,nz*sizeof(PetscScalar));
2504: return(0);
2505: }
2510: /*@
2511: MatRetrieveValues - Retrieves the copy of the matrix values; this allows, for
2512: example, reuse of the linear part of a Jacobian, while recomputing the
2513: nonlinear portion.
2515: Collect on Mat
2517: Input Parameters:
2518: . mat - the matrix (currently on AIJ matrices support this option)
2520: Level: advanced
2522: .seealso: MatStoreValues()
2524: @*/
2525: PetscErrorCode MatRetrieveValues(Mat mat)
2526: {
2527: PetscErrorCode ierr,(*f)(Mat);
2531: if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2532: if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2534: PetscObjectQueryFunction((PetscObject)mat,"MatRetrieveValues_C",(void (**)(void))&f);
2535: if (f) {
2536: (*f)(mat);
2537: } else {
2538: SETERRQ(PETSC_ERR_SUP,"Wrong type of matrix to retrieve values");
2539: }
2540: return(0);
2541: }
2544: /* --------------------------------------------------------------------------------*/
2547: /*@C
2548: MatCreateSeqAIJ - Creates a sparse matrix in AIJ (compressed row) format
2549: (the default parallel PETSc format). For good matrix assembly performance
2550: the user should preallocate the matrix storage by setting the parameter nz
2551: (or the array nnz). By setting these parameters accurately, performance
2552: during matrix assembly can be increased by more than a factor of 50.
2554: Collective on MPI_Comm
2556: Input Parameters:
2557: + comm - MPI communicator, set to PETSC_COMM_SELF
2558: . m - number of rows
2559: . n - number of columns
2560: . nz - number of nonzeros per row (same for all rows)
2561: - nnz - array containing the number of nonzeros in the various rows
2562: (possibly different for each row) or PETSC_NULL
2564: Output Parameter:
2565: . A - the matrix
2567: Notes:
2568: If nnz is given then nz is ignored
2570: The AIJ format (also called the Yale sparse matrix format or
2571: compressed row storage), is fully compatible with standard Fortran 77
2572: storage. That is, the stored row and column indices can begin at
2573: either one (as in Fortran) or zero. See the users' manual for details.
2575: Specify the preallocated storage with either nz or nnz (not both).
2576: Set nz=PETSC_DEFAULT and nnz=PETSC_NULL for PETSc to control dynamic memory
2577: allocation. For large problems you MUST preallocate memory or you
2578: will get TERRIBLE performance, see the users' manual chapter on matrices.
2580: By default, this format uses inodes (identical nodes) when possible, to
2581: improve numerical efficiency of matrix-vector products and solves. We
2582: search for consecutive rows with the same nonzero structure, thereby
2583: reusing matrix information to achieve increased efficiency.
2585: Options Database Keys:
2586: + -mat_no_inode - Do not use inodes
2587: . -mat_inode_limit <limit> - Sets inode limit (max limit=5)
2588: - -mat_aij_oneindex - Internally use indexing starting at 1
2589: rather than 0. Note that when calling MatSetValues(),
2590: the user still MUST index entries starting at 0!
2592: Level: intermediate
2594: .seealso: MatCreate(), MatCreateMPIAIJ(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays()
2596: @*/
2597: PetscErrorCode MatCreateSeqAIJ(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt nz,const PetscInt nnz[],Mat *A)
2598: {
2602: MatCreate(comm,A);
2603: MatSetSizes(*A,m,n,m,n);
2604: MatSetType(*A,MATSEQAIJ);
2605: MatSeqAIJSetPreallocation_SeqAIJ(*A,nz,(PetscInt*)nnz);
2606: return(0);
2607: }
2611: /*@C
2612: MatSeqAIJSetPreallocation - For good matrix assembly performance
2613: the user should preallocate the matrix storage by setting the parameter nz
2614: (or the array nnz). By setting these parameters accurately, performance
2615: during matrix assembly can be increased by more than a factor of 50.
2617: Collective on MPI_Comm
2619: Input Parameters:
2620: + B - The matrix
2621: . nz - number of nonzeros per row (same for all rows)
2622: - nnz - array containing the number of nonzeros in the various rows
2623: (possibly different for each row) or PETSC_NULL
2625: Notes:
2626: If nnz is given then nz is ignored
2628: The AIJ format (also called the Yale sparse matrix format or
2629: compressed row storage), is fully compatible with standard Fortran 77
2630: storage. That is, the stored row and column indices can begin at
2631: either one (as in Fortran) or zero. See the users' manual for details.
2633: Specify the preallocated storage with either nz or nnz (not both).
2634: Set nz=PETSC_DEFAULT and nnz=PETSC_NULL for PETSc to control dynamic memory
2635: allocation. For large problems you MUST preallocate memory or you
2636: will get TERRIBLE performance, see the users' manual chapter on matrices.
2638: Developers: Use nz of MAT_SKIP_ALLOCATION to not allocate any space for the matrix
2639: entries or columns indices
2641: By default, this format uses inodes (identical nodes) when possible, to
2642: improve numerical efficiency of matrix-vector products and solves. We
2643: search for consecutive rows with the same nonzero structure, thereby
2644: reusing matrix information to achieve increased efficiency.
2646: Options Database Keys:
2647: + -mat_no_inode - Do not use inodes
2648: . -mat_inode_limit <limit> - Sets inode limit (max limit=5)
2649: - -mat_aij_oneindex - Internally use indexing starting at 1
2650: rather than 0. Note that when calling MatSetValues(),
2651: the user still MUST index entries starting at 0!
2653: Level: intermediate
2655: .seealso: MatCreate(), MatCreateMPIAIJ(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays()
2657: @*/
2658: PetscErrorCode MatSeqAIJSetPreallocation(Mat B,PetscInt nz,const PetscInt nnz[])
2659: {
2660: PetscErrorCode ierr,(*f)(Mat,PetscInt,const PetscInt[]);
2663: PetscObjectQueryFunction((PetscObject)B,"MatSeqAIJSetPreallocation_C",(void (**)(void))&f);
2664: if (f) {
2665: (*f)(B,nz,nnz);
2666: }
2667: return(0);
2668: }
2673: PetscErrorCode MatSeqAIJSetPreallocation_SeqAIJ(Mat B,PetscInt nz,PetscInt *nnz)
2674: {
2675: Mat_SeqAIJ *b;
2676: PetscTruth skipallocation = PETSC_FALSE;
2678: PetscInt i;
2681:
2682: if (nz == MAT_SKIP_ALLOCATION) {
2683: skipallocation = PETSC_TRUE;
2684: nz = 0;
2685: }
2687: B->rmap.bs = B->cmap.bs = 1;
2688: PetscMapInitialize(B->comm,&B->rmap);
2689: PetscMapInitialize(B->comm,&B->cmap);
2691: if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 5;
2692: if (nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"nz cannot be less than 0: value %d",nz);
2693: if (nnz) {
2694: for (i=0; i<B->rmap.n; i++) {
2695: if (nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be less than 0: local row %d value %d",i,nnz[i]);
2696: if (nnz[i] > B->cmap.n) SETERRQ3(PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be greater than row length: local row %d value %d rowlength %d",i,nnz[i],B->cmap.n);
2697: }
2698: }
2700: B->preallocated = PETSC_TRUE;
2701: b = (Mat_SeqAIJ*)B->data;
2703: if (!skipallocation) {
2704: PetscMalloc2(B->rmap.n,PetscInt,&b->imax,B->rmap.n,PetscInt,&b->ilen);
2705: if (!nnz) {
2706: if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 10;
2707: else if (nz <= 0) nz = 1;
2708: for (i=0; i<B->rmap.n; i++) b->imax[i] = nz;
2709: nz = nz*B->rmap.n;
2710: } else {
2711: nz = 0;
2712: for (i=0; i<B->rmap.n; i++) {b->imax[i] = nnz[i]; nz += nnz[i];}
2713: }
2715: /* b->ilen will count nonzeros in each row so far. */
2716: for (i=0; i<B->rmap.n; i++) { b->ilen[i] = 0;}
2718: /* allocate the matrix space */
2719: PetscMalloc3(nz,PetscScalar,&b->a,nz,PetscInt,&b->j,B->rmap.n+1,PetscInt,&b->i);
2720: b->i[0] = 0;
2721: for (i=1; i<B->rmap.n+1; i++) {
2722: b->i[i] = b->i[i-1] + b->imax[i-1];
2723: }
2724: b->singlemalloc = PETSC_TRUE;
2725: b->free_a = PETSC_TRUE;
2726: b->free_ij = PETSC_TRUE;
2727: } else {
2728: b->free_a = PETSC_FALSE;
2729: b->free_ij = PETSC_FALSE;
2730: }
2732: b->nz = 0;
2733: b->maxnz = nz;
2734: B->info.nz_unneeded = (double)b->maxnz;
2735: return(0);
2736: }
2739: #undef __FUNCT__
2741: /*@C
2742: MatSeqAIJSetPreallocationCSR - Allocates memory for a sparse sequential matrix in AIJ format.
2744: Input Parameters:
2745: + B - the matrix
2746: . i - the indices into j for the start of each row (starts with zero)
2747: . j - the column indices for each row (starts with zero) these must be sorted for each row
2748: - v - optional values in the matrix
2750: Contributed by: Lisandro Dalchin
2752: Level: developer
2754: .keywords: matrix, aij, compressed row, sparse, sequential
2756: .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatSeqAIJSetPreallocation(), MatCreateSeqAIJ(), SeqAIJ
2757: @*/
2758: PetscErrorCode MatSeqAIJSetPreallocationCSR(Mat B,const PetscInt i[],const PetscInt j[],const PetscScalar v[])
2759: {
2760: PetscErrorCode (*f)(Mat,const PetscInt[],const PetscInt[],const PetscScalar[]);
2765: PetscObjectQueryFunction((PetscObject)B,"MatSeqAIJSetPreallocationCSR_C",(void (**)(void))&f);
2766: if (f) {
2767: (*f)(B,i,j,v);
2768: }
2769: return(0);
2770: }
2773: #undef __FUNCT__
2775: PetscErrorCode MatSeqAIJSetPreallocationCSR_SeqAIJ(Mat B,const PetscInt Ii[],const PetscInt J[],const PetscScalar v[])
2776: {
2777: PetscInt i;
2778: PetscInt m,n;
2779: PetscInt nz;
2780: PetscInt *nnz, nz_max = 0;
2781: PetscScalar *values;
2785: MatGetSize(B, &m, &n);
2787: if (Ii[0]) {
2788: SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE, "Ii[0] must be 0 it is %D", Ii[0]);
2789: }
2790: PetscMalloc((m+1) * sizeof(PetscInt), &nnz);
2791: for(i = 0; i < m; i++) {
2792: nz = Ii[i+1]- Ii[i];
2793: nz_max = PetscMax(nz_max, nz);
2794: if (nz < 0) {
2795: SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE, "Local row %D has a negative number of columns %D", i, nnz);
2796: }
2797: nnz[i] = nz;
2798: }
2799: MatSeqAIJSetPreallocation(B, 0, nnz);
2800: PetscFree(nnz);
2802: if (v) {
2803: values = (PetscScalar*) v;
2804: } else {
2805: PetscMalloc((nz_max+1)*sizeof(PetscScalar), &values);
2806: PetscMemzero(values, nz_max*sizeof(PetscScalar));
2807: }
2809: MatSetOption(B,MAT_COLUMNS_SORTED);
2811: for(i = 0; i < m; i++) {
2812: nz = Ii[i+1] - Ii[i];
2813: MatSetValues_SeqAIJ(B, 1, &i, nz, J+Ii[i], values + (v ? Ii[i] : 0), INSERT_VALUES);
2814: }
2816: MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);
2817: MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);
2818: MatSetOption(B,MAT_COLUMNS_UNSORTED);
2820: if (!v) {
2821: PetscFree(values);
2822: }
2823: return(0);
2824: }
2827: /*MC
2828: MATSEQAIJ - MATSEQAIJ = "seqaij" - A matrix type to be used for sequential sparse matrices,
2829: based on compressed sparse row format.
2831: Options Database Keys:
2832: . -mat_type seqaij - sets the matrix type to "seqaij" during a call to MatSetFromOptions()
2834: Level: beginner
2836: .seealso: MatCreateSeqAIJ(), MatSetFromOptions(), MatSetType(), MatCreate(), MatType
2837: M*/
2846: PetscErrorCode MatCreate_SeqAIJ(Mat B)
2847: {
2848: Mat_SeqAIJ *b;
2850: PetscMPIInt size;
2853: MPI_Comm_size(B->comm,&size);
2854: if (size > 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Comm must be of size 1");
2856: PetscNew(Mat_SeqAIJ,&b);
2857: B->data = (void*)b;
2858: PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));
2859: B->factor = 0;
2860: B->mapping = 0;
2861: b->row = 0;
2862: b->col = 0;
2863: b->icol = 0;
2864: b->reallocs = 0;
2865: b->sorted = PETSC_FALSE;
2866: b->ignorezeroentries = PETSC_FALSE;
2867: b->roworiented = PETSC_TRUE;
2868: b->nonew = 0;
2869: b->diag = 0;
2870: b->solve_work = 0;
2871: B->spptr = 0;
2872: b->saved_values = 0;
2873: b->idiag = 0;
2874: b->ssor = 0;
2875: b->keepzeroedrows = PETSC_FALSE;
2876: b->xtoy = 0;
2877: b->XtoY = 0;
2878: b->compressedrow.use = PETSC_FALSE;
2879: b->compressedrow.nrows = B->rmap.n;
2880: b->compressedrow.i = PETSC_NULL;
2881: b->compressedrow.rindex = PETSC_NULL;
2882: b->compressedrow.checked = PETSC_FALSE;
2883: B->same_nonzero = PETSC_FALSE;
2885: PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJ);
2886: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSeqAIJSetColumnIndices_C",
2887: "MatSeqAIJSetColumnIndices_SeqAIJ",
2888: MatSeqAIJSetColumnIndices_SeqAIJ);
2889: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C",
2890: "MatStoreValues_SeqAIJ",
2891: MatStoreValues_SeqAIJ);
2892: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C",
2893: "MatRetrieveValues_SeqAIJ",
2894: MatRetrieveValues_SeqAIJ);
2895: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqaij_seqsbaij_C",
2896: "MatConvert_SeqAIJ_SeqSBAIJ",
2897: MatConvert_SeqAIJ_SeqSBAIJ);
2898: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqaij_seqbaij_C",
2899: "MatConvert_SeqAIJ_SeqBAIJ",
2900: MatConvert_SeqAIJ_SeqBAIJ);
2901: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqaij_seqcsrperm_C",
2902: "MatConvert_SeqAIJ_SeqCSRPERM",
2903: MatConvert_SeqAIJ_SeqCSRPERM);
2904: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqaij_seqcrl_C",
2905: "MatConvert_SeqAIJ_SeqCRL",
2906: MatConvert_SeqAIJ_SeqCRL);
2907: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatIsTranspose_C",
2908: "MatIsTranspose_SeqAIJ",
2909: MatIsTranspose_SeqAIJ);
2910: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSeqAIJSetPreallocation_C",
2911: "MatSeqAIJSetPreallocation_SeqAIJ",
2912: MatSeqAIJSetPreallocation_SeqAIJ);
2913: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSeqAIJSetPreallocationCSR_C",
2914: "MatSeqAIJSetPreallocationCSR_SeqAIJ",
2915: MatSeqAIJSetPreallocationCSR_SeqAIJ);
2916: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatReorderForNonzeroDiagonal_C",
2917: "MatReorderForNonzeroDiagonal_SeqAIJ",
2918: MatReorderForNonzeroDiagonal_SeqAIJ);
2919: MatCreate_Inode(B);
2920: PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJ);
2921: return(0);
2922: }
2927: PetscErrorCode MatDuplicate_SeqAIJ(Mat A,MatDuplicateOption cpvalues,Mat *B)
2928: {
2929: Mat C;
2930: Mat_SeqAIJ *c,*a = (Mat_SeqAIJ*)A->data;
2932: PetscInt i,m = A->rmap.n;
2935: *B = 0;
2936: MatCreate(A->comm,&C);
2937: MatSetSizes(C,A->rmap.n,A->cmap.n,A->rmap.n,A->cmap.n);
2938: MatSetType(C,A->type_name);
2939: PetscMemcpy(C->ops,A->ops,sizeof(struct _MatOps));
2940:
2941: PetscMapCopy(A->comm,&A->rmap,&C->rmap);
2942: PetscMapCopy(A->comm,&A->cmap,&C->cmap);
2944: c = (Mat_SeqAIJ*)C->data;
2946: C->factor = A->factor;
2948: c->row = 0;
2949: c->col = 0;
2950: c->icol = 0;
2951: c->reallocs = 0;
2953: C->assembled = PETSC_TRUE;
2954:
2955: PetscMalloc2(m,PetscInt,&c->imax,m,PetscInt,&c->ilen);
2956: for (i=0; i<m; i++) {
2957: c->imax[i] = a->imax[i];
2958: c->ilen[i] = a->ilen[i];
2959: }
2961: /* allocate the matrix space */
2962: PetscMalloc3(a->i[m],PetscScalar,&c->a,a->i[m],PetscInt,&c->j,m+1,PetscInt,&c->i);
2963: c->singlemalloc = PETSC_TRUE;
2964: PetscMemcpy(c->i,a->i,(m+1)*sizeof(PetscInt));
2965: if (m > 0) {
2966: PetscMemcpy(c->j,a->j,(a->i[m])*sizeof(PetscInt));
2967: if (cpvalues == MAT_COPY_VALUES) {
2968: PetscMemcpy(c->a,a->a,(a->i[m])*sizeof(PetscScalar));
2969: } else {
2970: PetscMemzero(c->a,(a->i[m])*sizeof(PetscScalar));
2971: }
2972: }
2974: c->sorted = a->sorted;
2975: c->ignorezeroentries = a->ignorezeroentries;
2976: c->roworiented = a->roworiented;
2977: c->nonew = a->nonew;
2978: if (a->diag) {
2979: PetscMalloc((m+1)*sizeof(PetscInt),&c->diag);
2980: PetscLogObjectMemory(C,(m+1)*sizeof(PetscInt));
2981: for (i=0; i<m; i++) {
2982: c->diag[i] = a->diag[i];
2983: }
2984: } else c->diag = 0;
2985: c->solve_work = 0;
2986: c->saved_values = 0;
2987: c->idiag = 0;
2988: c->ssor = 0;
2989: c->keepzeroedrows = a->keepzeroedrows;
2990: c->free_a = PETSC_TRUE;
2991: c->free_ij = PETSC_TRUE;
2992: c->xtoy = 0;
2993: c->XtoY = 0;
2995: c->nz = a->nz;
2996: c->maxnz = a->maxnz;
2997: C->preallocated = PETSC_TRUE;
2999: c->compressedrow.use = a->compressedrow.use;
3000: c->compressedrow.nrows = a->compressedrow.nrows;
3001: c->compressedrow.checked = a->compressedrow.checked;
3002: if ( a->compressedrow.checked && a->compressedrow.use){
3003: i = a->compressedrow.nrows;
3004: PetscMalloc((2*i+1)*sizeof(PetscInt),&c->compressedrow.i);
3005: c->compressedrow.rindex = c->compressedrow.i + i + 1;
3006: PetscMemcpy(c->compressedrow.i,a->compressedrow.i,(i+1)*sizeof(PetscInt));
3007: PetscMemcpy(c->compressedrow.rindex,a->compressedrow.rindex,i*sizeof(PetscInt));
3008: } else {
3009: c->compressedrow.use = PETSC_FALSE;
3010: c->compressedrow.i = PETSC_NULL;
3011: c->compressedrow.rindex = PETSC_NULL;
3012: }
3013: C->same_nonzero = A->same_nonzero;
3014: MatDuplicate_Inode(A,cpvalues,&C);
3016: *B = C;
3017: PetscFListDuplicate(A->qlist,&C->qlist);
3018: return(0);
3019: }
3023: PetscErrorCode MatLoad_SeqAIJ(PetscViewer viewer, MatType type,Mat *A)
3024: {
3025: Mat_SeqAIJ *a;
3026: Mat B;
3028: PetscInt i,sum,nz,header[4],*rowlengths = 0,M,N;
3029: int fd;
3030: PetscMPIInt size;
3031: MPI_Comm comm;
3032:
3034: PetscObjectGetComm((PetscObject)viewer,&comm);
3035: MPI_Comm_size(comm,&size);
3036: if (size > 1) SETERRQ(PETSC_ERR_ARG_SIZ,"view must have one processor");
3037: PetscViewerBinaryGetDescriptor(viewer,&fd);
3038: PetscBinaryRead(fd,header,4,PETSC_INT);
3039: if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not matrix object in file");
3040: M = header[1]; N = header[2]; nz = header[3];
3042: if (nz < 0) {
3043: SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format on disk,cannot load as SeqAIJ");
3044: }
3046: /* read in row lengths */
3047: PetscMalloc(M*sizeof(PetscInt),&rowlengths);
3048: PetscBinaryRead(fd,rowlengths,M,PETSC_INT);
3050: /* check if sum of rowlengths is same as nz */
3051: for (i=0,sum=0; i< M; i++) sum +=rowlengths[i];
3052: if (sum != nz) SETERRQ2(PETSC_ERR_FILE_READ,"Inconsistant matrix data in file. no-nonzeros = %d, sum-row-lengths = %d\n",nz,sum);
3054: /* create our matrix */
3055: MatCreate(comm,&B);
3056: MatSetSizes(B,PETSC_DECIDE,PETSC_DECIDE,M,N);
3057: MatSetType(B,type);
3058: MatSeqAIJSetPreallocation_SeqAIJ(B,0,rowlengths);
3059: a = (Mat_SeqAIJ*)B->data;
3061: /* read in column indices and adjust for Fortran indexing*/
3062: PetscBinaryRead(fd,a->j,nz,PETSC_INT);
3064: /* read in nonzero values */
3065: PetscBinaryRead(fd,a->a,nz,PETSC_SCALAR);
3067: /* set matrix "i" values */
3068: a->i[0] = 0;
3069: for (i=1; i<= M; i++) {
3070: a->i[i] = a->i[i-1] + rowlengths[i-1];
3071: a->ilen[i-1] = rowlengths[i-1];
3072: }
3073: PetscFree(rowlengths);
3075: MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);
3076: MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);
3077: *A = B;
3078: return(0);
3079: }
3083: PetscErrorCode MatEqual_SeqAIJ(Mat A,Mat B,PetscTruth* flg)
3084: {
3085: Mat_SeqAIJ *a = (Mat_SeqAIJ *)A->data,*b = (Mat_SeqAIJ *)B->data;
3089: /* If the matrix dimensions are not equal,or no of nonzeros */
3090: if ((A->rmap.n != B->rmap.n) || (A->cmap.n != B->cmap.n) ||(a->nz != b->nz)) {
3091: *flg = PETSC_FALSE;
3092: return(0);
3093: }
3094:
3095: /* if the a->i are the same */
3096: PetscMemcmp(a->i,b->i,(A->rmap.n+1)*sizeof(PetscInt),flg);
3097: if (!*flg) return(0);
3098:
3099: /* if a->j are the same */
3100: PetscMemcmp(a->j,b->j,(a->nz)*sizeof(PetscInt),flg);
3101: if (!*flg) return(0);
3102:
3103: /* if a->a are the same */
3104: PetscMemcmp(a->a,b->a,(a->nz)*sizeof(PetscScalar),flg);
3106: return(0);
3107:
3108: }
3112: /*@
3113: MatCreateSeqAIJWithArrays - Creates an sequential AIJ matrix using matrix elements (in CSR format)
3114: provided by the user.
3116: Collective on MPI_Comm
3118: Input Parameters:
3119: + comm - must be an MPI communicator of size 1
3120: . m - number of rows
3121: . n - number of columns
3122: . i - row indices
3123: . j - column indices
3124: - a - matrix values
3126: Output Parameter:
3127: . mat - the matrix
3129: Level: intermediate
3131: Notes:
3132: The i, j, and a arrays are not copied by this routine, the user must free these arrays
3133: once the matrix is destroyed
3135: You cannot set new nonzero locations into this matrix, that will generate an error.
3137: The i and j indices are 0 based
3139: .seealso: MatCreate(), MatCreateMPIAIJ(), MatCreateSeqAIJ(), MatCreateMPIAIJWithArrays(), MatMPIAIJSetPreallocationCSR()
3141: @*/
3142: PetscErrorCode MatCreateSeqAIJWithArrays(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt* i,PetscInt*j,PetscScalar *a,Mat *mat)
3143: {
3145: PetscInt ii;
3146: Mat_SeqAIJ *aij;
3149: if (i[0]) {
3150: SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"i (row indices) must start with 0");
3151: }
3152: MatCreate(comm,mat);
3153: MatSetSizes(*mat,m,n,m,n);
3154: MatSetType(*mat,MATSEQAIJ);
3155: MatSeqAIJSetPreallocation_SeqAIJ(*mat,MAT_SKIP_ALLOCATION,0);
3156: aij = (Mat_SeqAIJ*)(*mat)->data;
3157: PetscMalloc2(m,PetscInt,&aij->imax,m,PetscInt,&aij->ilen);
3159: aij->i = i;
3160: aij->j = j;
3161: aij->a = a;
3162: aij->singlemalloc = PETSC_FALSE;
3163: aij->nonew = -1; /*this indicates that inserting a new value in the matrix that generates a new nonzero is an error*/
3164: aij->free_a = PETSC_FALSE;
3165: aij->free_ij = PETSC_FALSE;
3167: for (ii=0; ii<m; ii++) {
3168: aij->ilen[ii] = aij->imax[ii] = i[ii+1] - i[ii];
3169: #if defined(PETSC_USE_DEBUG)
3170: if (i[ii+1] - i[ii] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Negative row length in i (row indices) row = %d length = %d",ii,i[ii+1] - i[ii]);
3171: #endif
3172: }
3173: #if defined(PETSC_USE_DEBUG)
3174: for (ii=0; ii<aij->i[m]; ii++) {
3175: if (j[ii] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Negative column index at location = %d index = %d",ii,j[ii]);
3176: if (j[ii] > n - 1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column index to large at location = %d index = %d",ii,j[ii]);
3177: }
3178: #endif
3180: MatAssemblyBegin(*mat,MAT_FINAL_ASSEMBLY);
3181: MatAssemblyEnd(*mat,MAT_FINAL_ASSEMBLY);
3182: return(0);
3183: }
3187: PetscErrorCode MatSetColoring_SeqAIJ(Mat A,ISColoring coloring)
3188: {
3190: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
3193: if (coloring->ctype == IS_COLORING_LOCAL) {
3194: ISColoringReference(coloring);
3195: a->coloring = coloring;
3196: } else if (coloring->ctype == IS_COLORING_GHOSTED) {
3197: PetscInt i,*larray;
3198: ISColoring ocoloring;
3199: ISColoringValue *colors;
3201: /* set coloring for diagonal portion */
3202: PetscMalloc((A->cmap.n+1)*sizeof(PetscInt),&larray);
3203: for (i=0; i<A->cmap.n; i++) {
3204: larray[i] = i;
3205: }
3206: ISGlobalToLocalMappingApply(A->mapping,IS_GTOLM_MASK,A->cmap.n,larray,PETSC_NULL,larray);
3207: PetscMalloc((A->cmap.n+1)*sizeof(ISColoringValue),&colors);
3208: for (i=0; i<A->cmap.n; i++) {
3209: colors[i] = coloring->colors[larray[i]];
3210: }
3211: PetscFree(larray);
3212: ISColoringCreate(PETSC_COMM_SELF,coloring->n,A->cmap.n,colors,&ocoloring);
3213: a->coloring = ocoloring;
3214: }
3215: return(0);
3216: }
3218: #if defined(PETSC_HAVE_ADIC)
3220: #include "adic/ad_utils.h"
3225: PetscErrorCode MatSetValuesAdic_SeqAIJ(Mat A,void *advalues)
3226: {
3227: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
3228: PetscInt m = A->rmap.n,*ii = a->i,*jj = a->j,nz,i,j,nlen;
3229: PetscScalar *v = a->a,*values = ((PetscScalar*)advalues)+1;
3230: ISColoringValue *color;
3233: if (!a->coloring) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Coloring not set for matrix");
3234: nlen = PetscADGetDerivTypeSize()/sizeof(PetscScalar);
3235: color = a->coloring->colors;
3236: /* loop over rows */
3237: for (i=0; i<m; i++) {
3238: nz = ii[i+1] - ii[i];
3239: /* loop over columns putting computed value into matrix */
3240: for (j=0; j<nz; j++) {
3241: *v++ = values[color[*jj++]];
3242: }
3243: values += nlen; /* jump to next row of derivatives */
3244: }
3245: return(0);
3246: }
3247: #endif
3251: PetscErrorCode MatSetValuesAdifor_SeqAIJ(Mat A,PetscInt nl,void *advalues)
3252: {
3253: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
3254: PetscInt m = A->rmap.n,*ii = a->i,*jj = a->j,nz,i,j;
3255: PetscScalar *v = a->a,*values = (PetscScalar *)advalues;
3256: ISColoringValue *color;
3259: if (!a->coloring) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Coloring not set for matrix");
3260: color = a->coloring->colors;
3261: /* loop over rows */
3262: for (i=0; i<m; i++) {
3263: nz = ii[i+1] - ii[i];
3264: /* loop over columns putting computed value into matrix */
3265: for (j=0; j<nz; j++) {
3266: *v++ = values[color[*jj++]];
3267: }
3268: values += nl; /* jump to next row of derivatives */
3269: }
3270: return(0);
3271: }
3273: /*
3274: Special version for direct calls from Fortran
3275: */
3276: #if defined(PETSC_HAVE_FORTRAN_CAPS)
3277: #define matsetvaluesseqaij_ MATSETVALUESSEQAIJ
3278: #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE)
3279: #define matsetvaluesseqaij_ matsetvaluesseqaij
3280: #endif
3282: /* Change these macros so can be used in void function */
3283: #undef CHKERRQ
3284: #define CHKERRQ(ierr) CHKERRABORT(A->comm,ierr)
3285: #undef SETERRQ2
3286: #define SETERRQ2(ierr,b,c,d) CHKERRABORT(A->comm,ierr)
3291: void PETSC_STDCALL matsetvaluesseqaij_(Mat *AA,PetscInt *mm,const PetscInt im[],PetscInt *nn,const PetscInt in[],const PetscScalar v[],InsertMode *isis, PetscErrorCode *_ierr)
3292: {
3293: Mat A = *AA;
3294: PetscInt m = *mm, n = *nn;
3295: InsertMode is = *isis;
3296: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data;
3297: PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N;
3298: PetscInt *imax,*ai,*ailen;
3300: PetscInt *aj,nonew = a->nonew,lastcol = -1;
3301: PetscScalar *ap,value,*aa;
3302: PetscTruth ignorezeroentries = ((a->ignorezeroentries && is == ADD_VALUES) ? PETSC_TRUE:PETSC_FALSE);
3303: PetscTruth roworiented = a->roworiented;
3306: MatPreallocated(A);
3307: imax = a->imax;
3308: ai = a->i;
3309: ailen = a->ilen;
3310: aj = a->j;
3311: aa = a->a;
3313: for (k=0; k<m; k++) { /* loop over added rows */
3314: row = im[k];
3315: if (row < 0) continue;
3316: #if defined(PETSC_USE_DEBUG)
3317: if (row >= A->rmap.n) SETERRABORT(A->comm,PETSC_ERR_ARG_OUTOFRANGE,"Row too large");
3318: #endif
3319: rp = aj + ai[row]; ap = aa + ai[row];
3320: rmax = imax[row]; nrow = ailen[row];
3321: low = 0;
3322: high = nrow;
3323: for (l=0; l<n; l++) { /* loop over added columns */
3324: if (in[l] < 0) continue;
3325: #if defined(PETSC_USE_DEBUG)
3326: if (in[l] >= A->cmap.n) SETERRABORT(A->comm,PETSC_ERR_ARG_OUTOFRANGE,"Column too large");
3327: #endif
3328: col = in[l];
3329: if (roworiented) {
3330: value = v[l + k*n];
3331: } else {
3332: value = v[k + l*m];
3333: }
3334: if (value == 0.0 && ignorezeroentries && (is == ADD_VALUES)) continue;
3336: if (col <= lastcol) low = 0; else high = nrow;
3337: lastcol = col;
3338: while (high-low > 5) {
3339: t = (low+high)/2;
3340: if (rp[t] > col) high = t;
3341: else low = t;
3342: }
3343: for (i=low; i<high; i++) {
3344: if (rp[i] > col) break;
3345: if (rp[i] == col) {
3346: if (is == ADD_VALUES) ap[i] += value;
3347: else ap[i] = value;
3348: goto noinsert;
3349: }
3350: }
3351: if (value == 0.0 && ignorezeroentries) goto noinsert;
3352: if (nonew == 1) goto noinsert;
3353: if (nonew == -1) SETERRABORT(A->comm,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero in the matrix");
3354: MatSeqXAIJReallocateAIJ(A,A->rmap.n,1,nrow,row,col,rmax,aa,ai,aj,rp,ap,imax,nonew);
3355: N = nrow++ - 1; a->nz++; high++;
3356: /* shift up all the later entries in this row */
3357: for (ii=N; ii>=i; ii--) {
3358: rp[ii+1] = rp[ii];
3359: ap[ii+1] = ap[ii];
3360: }
3361: rp[i] = col;
3362: ap[i] = value;
3363: noinsert:;
3364: low = i + 1;
3365: }
3366: ailen[row] = nrow;
3367: }
3368: A->same_nonzero = PETSC_FALSE;
3369: PetscFunctionReturnVoid();
3370: }