1 #ifdef PETSC_RCS_HEADER 2 static char vcid[] = "$Id: aij.c,v 1.253 1998/03/16 18:43:04 bsmith Exp balay $"; 3 #endif 4 5 /* 6 Defines the basic matrix operations for the AIJ (compressed row) 7 matrix storage format. 8 */ 9 10 #include "pinclude/pviewer.h" 11 #include "sys.h" 12 #include "src/mat/impls/aij/seq/aij.h" 13 #include "src/vec/vecimpl.h" 14 #include "src/inline/spops.h" 15 #include "src/inline/dot.h" 16 #include "src/inline/bitarray.h" 17 18 /* 19 Basic AIJ format ILU based on drop tolerance 20 */ 21 #undef __FUNC__ 22 #define __FUNC__ "MatILUDTFactor_SeqAIJ" 23 int MatILUDTFactor_SeqAIJ(Mat A,double dt,int maxnz,IS row,IS col,Mat *fact) 24 { 25 /* Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; */ 26 int ierr = 1; 27 28 PetscFunctionBegin; 29 SETERRQ(ierr,0,"Not implemented"); 30 #if !defined(USE_PETSC_DEBUG) 31 PetscFunctionReturn(0); 32 #endif 33 } 34 35 extern int MatToSymmetricIJ_SeqAIJ(int,int*,int*,int,int,int**,int**); 36 37 #undef __FUNC__ 38 #define __FUNC__ "MatGetRowIJ_SeqAIJ" 39 int MatGetRowIJ_SeqAIJ(Mat A,int oshift,PetscTruth symmetric,int *m,int **ia,int **ja, 40 PetscTruth *done) 41 { 42 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 43 int ierr,i,ishift; 44 45 PetscFunctionBegin; 46 *m = A->m; 47 if (!ia) PetscFunctionReturn(0); 48 ishift = a->indexshift; 49 if (symmetric) { 50 ierr = MatToSymmetricIJ_SeqAIJ(a->m,a->i,a->j,ishift,oshift,ia,ja); CHKERRQ(ierr); 51 } else if (oshift == 0 && ishift == -1) { 52 int nz = a->i[a->m]; 53 /* malloc space and subtract 1 from i and j indices */ 54 *ia = (int *) PetscMalloc( (a->m+1)*sizeof(int) ); CHKPTRQ(*ia); 55 *ja = (int *) PetscMalloc( (nz+1)*sizeof(int) ); CHKPTRQ(*ja); 56 for ( i=0; i<nz; i++ ) (*ja)[i] = a->j[i] - 1; 57 for ( i=0; i<a->m+1; i++ ) (*ia)[i] = a->i[i] - 1; 58 } else if (oshift == 1 && ishift == 0) { 59 int nz = a->i[a->m] + 1; 60 /* malloc space and add 1 to i and j indices */ 61 *ia = (int *) PetscMalloc( (a->m+1)*sizeof(int) ); CHKPTRQ(*ia); 62 *ja = (int *) PetscMalloc( (nz+1)*sizeof(int) ); CHKPTRQ(*ja); 63 for ( i=0; i<nz; i++ ) (*ja)[i] = a->j[i] + 1; 64 for ( i=0; i<a->m+1; i++ ) (*ia)[i] = a->i[i] + 1; 65 } else { 66 *ia = a->i; *ja = a->j; 67 } 68 69 PetscFunctionReturn(0); 70 } 71 72 #undef __FUNC__ 73 #define __FUNC__ "MatRestoreRowIJ_SeqAIJ" 74 int MatRestoreRowIJ_SeqAIJ(Mat A,int oshift,PetscTruth symmetric,int *n,int **ia,int **ja, 75 PetscTruth *done) 76 { 77 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 78 int ishift = a->indexshift; 79 80 PetscFunctionBegin; 81 if (!ia) PetscFunctionReturn(0); 82 if (symmetric || (oshift == 0 && ishift == -1) || (oshift == 1 && ishift == 0)) { 83 PetscFree(*ia); 84 PetscFree(*ja); 85 } 86 PetscFunctionReturn(0); 87 } 88 89 #undef __FUNC__ 90 #define __FUNC__ "MatGetColumnIJ_SeqAIJ" 91 int MatGetColumnIJ_SeqAIJ(Mat A,int oshift,PetscTruth symmetric,int *nn,int **ia,int **ja, 92 PetscTruth *done) 93 { 94 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 95 int ierr,i,ishift = a->indexshift,*collengths,*cia,*cja,n = A->n,m = A->m; 96 int nz = a->i[m]+ishift,row,*jj,mr,col; 97 98 PetscFunctionBegin; 99 *nn = A->n; 100 if (!ia) PetscFunctionReturn(0); 101 if (symmetric) { 102 ierr = MatToSymmetricIJ_SeqAIJ(a->m,a->i,a->j,ishift,oshift,ia,ja); CHKERRQ(ierr); 103 } else { 104 collengths = (int *) PetscMalloc( (n+1)*sizeof(int) ); CHKPTRQ(collengths); 105 PetscMemzero(collengths,n*sizeof(int)); 106 cia = (int *) PetscMalloc( (n+1)*sizeof(int) ); CHKPTRQ(cia); 107 cja = (int *) PetscMalloc( (nz+1)*sizeof(int) ); CHKPTRQ(cja); 108 jj = a->j; 109 for ( i=0; i<nz; i++ ) { 110 collengths[jj[i] + ishift]++; 111 } 112 cia[0] = oshift; 113 for ( i=0; i<n; i++) { 114 cia[i+1] = cia[i] + collengths[i]; 115 } 116 PetscMemzero(collengths,n*sizeof(int)); 117 jj = a->j; 118 for ( row=0; row<m; row++ ) { 119 mr = a->i[row+1] - a->i[row]; 120 for ( i=0; i<mr; i++ ) { 121 col = *jj++ + ishift; 122 cja[cia[col] + collengths[col]++ - oshift] = row + oshift; 123 } 124 } 125 PetscFree(collengths); 126 *ia = cia; *ja = cja; 127 } 128 129 PetscFunctionReturn(0); 130 } 131 132 #undef __FUNC__ 133 #define __FUNC__ "MatRestoreColumnIJ_SeqAIJ" 134 int MatRestoreColumnIJ_SeqAIJ(Mat A,int oshift,PetscTruth symmetric,int *n,int **ia, 135 int **ja,PetscTruth *done) 136 { 137 PetscFunctionBegin; 138 if (!ia) PetscFunctionReturn(0); 139 140 PetscFree(*ia); 141 PetscFree(*ja); 142 143 PetscFunctionReturn(0); 144 } 145 146 #define CHUNKSIZE 15 147 148 #undef __FUNC__ 149 #define __FUNC__ "MatSetValues_SeqAIJ" 150 int MatSetValues_SeqAIJ(Mat A,int m,int *im,int n,int *in,Scalar *v,InsertMode is) 151 { 152 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 153 int *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax, N, sorted = a->sorted; 154 int *imax = a->imax, *ai = a->i, *ailen = a->ilen,roworiented = a->roworiented; 155 int *aj = a->j, nonew = a->nonew,shift = a->indexshift; 156 Scalar *ap,value, *aa = a->a; 157 158 PetscFunctionBegin; 159 for ( k=0; k<m; k++ ) { /* loop over added rows */ 160 row = im[k]; 161 #if defined(USE_PETSC_BOPT_g) 162 if (row < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Negative row"); 163 if (row >= a->m) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Row too large"); 164 #endif 165 rp = aj + ai[row] + shift; ap = aa + ai[row] + shift; 166 rmax = imax[row]; nrow = ailen[row]; 167 low = 0; 168 for ( l=0; l<n; l++ ) { /* loop over added columns */ 169 #if defined(USE_PETSC_BOPT_g) 170 if (in[l] < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Negative column"); 171 if (in[l] >= a->n) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Column too large"); 172 #endif 173 col = in[l] - shift; 174 if (roworiented) { 175 value = *v++; 176 } 177 else { 178 value = v[k + l*m]; 179 } 180 if (!sorted) low = 0; high = nrow; 181 while (high-low > 5) { 182 t = (low+high)/2; 183 if (rp[t] > col) high = t; 184 else low = t; 185 } 186 for ( i=low; i<high; i++ ) { 187 if (rp[i] > col) break; 188 if (rp[i] == col) { 189 if (is == ADD_VALUES) ap[i] += value; 190 else ap[i] = value; 191 goto noinsert; 192 } 193 } 194 if (nonew == 1) goto noinsert; 195 else if (nonew == -1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Inserting a new nonzero in the matrix"); 196 if (nrow >= rmax) { 197 /* there is no extra room in row, therefore enlarge */ 198 int new_nz = ai[a->m] + CHUNKSIZE,len,*new_i,*new_j; 199 Scalar *new_a; 200 201 if (nonew == -2) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Inserting a new nonzero in the matrix"); 202 203 /* malloc new storage space */ 204 len = new_nz*(sizeof(int)+sizeof(Scalar))+(a->m+1)*sizeof(int); 205 new_a = (Scalar *) PetscMalloc( len ); CHKPTRQ(new_a); 206 new_j = (int *) (new_a + new_nz); 207 new_i = new_j + new_nz; 208 209 /* copy over old data into new slots */ 210 for ( ii=0; ii<row+1; ii++ ) {new_i[ii] = ai[ii];} 211 for ( ii=row+1; ii<a->m+1; ii++ ) {new_i[ii] = ai[ii]+CHUNKSIZE;} 212 PetscMemcpy(new_j,aj,(ai[row]+nrow+shift)*sizeof(int)); 213 len = (new_nz - CHUNKSIZE - ai[row] - nrow - shift); 214 PetscMemcpy(new_j+ai[row]+shift+nrow+CHUNKSIZE,aj+ai[row]+shift+nrow, 215 len*sizeof(int)); 216 PetscMemcpy(new_a,aa,(ai[row]+nrow+shift)*sizeof(Scalar)); 217 PetscMemcpy(new_a+ai[row]+shift+nrow+CHUNKSIZE,aa+ai[row]+shift+nrow, 218 len*sizeof(Scalar)); 219 /* free up old matrix storage */ 220 PetscFree(a->a); 221 if (!a->singlemalloc) {PetscFree(a->i);PetscFree(a->j);} 222 aa = a->a = new_a; ai = a->i = new_i; aj = a->j = new_j; 223 a->singlemalloc = 1; 224 225 rp = aj + ai[row] + shift; ap = aa + ai[row] + shift; 226 rmax = imax[row] = imax[row] + CHUNKSIZE; 227 PLogObjectMemory(A,CHUNKSIZE*(sizeof(int) + sizeof(Scalar))); 228 a->maxnz += CHUNKSIZE; 229 a->reallocs++; 230 } 231 N = nrow++ - 1; a->nz++; 232 /* shift up all the later entries in this row */ 233 for ( ii=N; ii>=i; ii-- ) { 234 rp[ii+1] = rp[ii]; 235 ap[ii+1] = ap[ii]; 236 } 237 rp[i] = col; 238 ap[i] = value; 239 noinsert:; 240 low = i + 1; 241 } 242 ailen[row] = nrow; 243 } 244 PetscFunctionReturn(0); 245 } 246 247 #undef __FUNC__ 248 #define __FUNC__ "MatGetValues_SeqAIJ" 249 int MatGetValues_SeqAIJ(Mat A,int m,int *im,int n,int *in,Scalar *v) 250 { 251 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 252 int *rp, k, low, high, t, row, nrow, i, col, l, *aj = a->j; 253 int *ai = a->i, *ailen = a->ilen, shift = a->indexshift; 254 Scalar *ap, *aa = a->a, zero = 0.0; 255 256 PetscFunctionBegin; 257 for ( k=0; k<m; k++ ) { /* loop over rows */ 258 row = im[k]; 259 if (row < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Negative row"); 260 if (row >= a->m) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Row too large"); 261 rp = aj + ai[row] + shift; ap = aa + ai[row] + shift; 262 nrow = ailen[row]; 263 for ( l=0; l<n; l++ ) { /* loop over columns */ 264 if (in[l] < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Negative column"); 265 if (in[l] >= a->n) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Column too large"); 266 col = in[l] - shift; 267 high = nrow; low = 0; /* assume unsorted */ 268 while (high-low > 5) { 269 t = (low+high)/2; 270 if (rp[t] > col) high = t; 271 else low = t; 272 } 273 for ( i=low; i<high; i++ ) { 274 if (rp[i] > col) break; 275 if (rp[i] == col) { 276 *v++ = ap[i]; 277 goto finished; 278 } 279 } 280 *v++ = zero; 281 finished:; 282 } 283 } 284 PetscFunctionReturn(0); 285 } 286 287 288 #undef __FUNC__ 289 #define __FUNC__ "MatView_SeqAIJ_Binary" 290 extern int MatView_SeqAIJ_Binary(Mat A,Viewer viewer) 291 { 292 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 293 int i, fd, *col_lens, ierr; 294 295 PetscFunctionBegin; 296 ierr = ViewerBinaryGetDescriptor(viewer,&fd); CHKERRQ(ierr); 297 col_lens = (int *) PetscMalloc( (4+a->m)*sizeof(int) ); CHKPTRQ(col_lens); 298 col_lens[0] = MAT_COOKIE; 299 col_lens[1] = a->m; 300 col_lens[2] = a->n; 301 col_lens[3] = a->nz; 302 303 /* store lengths of each row and write (including header) to file */ 304 for ( i=0; i<a->m; i++ ) { 305 col_lens[4+i] = a->i[i+1] - a->i[i]; 306 } 307 ierr = PetscBinaryWrite(fd,col_lens,4+a->m,PETSC_INT,1); CHKERRQ(ierr); 308 PetscFree(col_lens); 309 310 /* store column indices (zero start index) */ 311 if (a->indexshift) { 312 for ( i=0; i<a->nz; i++ ) a->j[i]--; 313 } 314 ierr = PetscBinaryWrite(fd,a->j,a->nz,PETSC_INT,0); CHKERRQ(ierr); 315 if (a->indexshift) { 316 for ( i=0; i<a->nz; i++ ) a->j[i]++; 317 } 318 319 /* store nonzero values */ 320 ierr = PetscBinaryWrite(fd,a->a,a->nz,PETSC_SCALAR,0); CHKERRQ(ierr); 321 PetscFunctionReturn(0); 322 } 323 324 #undef __FUNC__ 325 #define __FUNC__ "MatView_SeqAIJ_ASCII" 326 extern int MatView_SeqAIJ_ASCII(Mat A,Viewer viewer) 327 { 328 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 329 int ierr, i,j, m = a->m, shift = a->indexshift, format, flg1,flg2; 330 FILE *fd; 331 char *outputname; 332 333 PetscFunctionBegin; 334 ierr = ViewerASCIIGetPointer(viewer,&fd); CHKERRQ(ierr); 335 ierr = ViewerFileGetOutputname_Private(viewer,&outputname); CHKERRQ(ierr); 336 ierr = ViewerGetFormat(viewer,&format); 337 if (format == VIEWER_FORMAT_ASCII_INFO) { 338 PetscFunctionReturn(0); 339 } else if (format == VIEWER_FORMAT_ASCII_INFO_LONG) { 340 ierr = OptionsHasName(PETSC_NULL,"-mat_aij_no_inode",&flg1); CHKERRQ(ierr); 341 ierr = OptionsHasName(PETSC_NULL,"-mat_no_unroll",&flg2); CHKERRQ(ierr); 342 if (flg1 || flg2) fprintf(fd," not using I-node routines\n"); 343 else fprintf(fd," using I-node routines: found %d nodes, limit used is %d\n", 344 a->inode.node_count,a->inode.limit); 345 } else if (format == VIEWER_FORMAT_ASCII_MATLAB) { 346 int nofinalvalue = 0; 347 if ((a->i[m] == a->i[m-1]) || (a->j[a->nz-1] != a->n-!shift)) { 348 nofinalvalue = 1; 349 } 350 fprintf(fd,"%% Size = %d %d \n",m,a->n); 351 fprintf(fd,"%% Nonzeros = %d \n",a->nz); 352 fprintf(fd,"zzz = zeros(%d,3);\n",a->nz+nofinalvalue); 353 fprintf(fd,"zzz = [\n"); 354 355 for (i=0; i<m; i++) { 356 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 357 #if defined(USE_PETSC_COMPLEX) 358 fprintf(fd,"%d %d %18.16e + %18.16e i \n",i+1,a->j[j]+!shift,real(a->a[j]),imag(a->a[j])); 359 #else 360 fprintf(fd,"%d %d %18.16e\n", i+1, a->j[j]+!shift, a->a[j]); 361 #endif 362 } 363 } 364 if (nofinalvalue) { 365 fprintf(fd,"%d %d %18.16e\n", m, a->n, 0.0); 366 } 367 fprintf(fd,"];\n %s = spconvert(zzz);\n",outputname); 368 } else if (format == VIEWER_FORMAT_ASCII_COMMON) { 369 for ( i=0; i<m; i++ ) { 370 fprintf(fd,"row %d:",i); 371 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 372 #if defined(USE_PETSC_COMPLEX) 373 if (imag(a->a[j]) > 0.0 && real(a->a[j]) != 0.0) 374 fprintf(fd," %d %g + %g i",a->j[j]+shift,real(a->a[j]),imag(a->a[j])); 375 else if (imag(a->a[j]) < 0.0 && real(a->a[j]) != 0.0) 376 fprintf(fd," %d %g - %g i",a->j[j]+shift,real(a->a[j]),-imag(a->a[j])); 377 else if (real(a->a[j]) != 0.0) 378 fprintf(fd," %d %g ",a->j[j]+shift,real(a->a[j])); 379 #else 380 if (a->a[j] != 0.0) fprintf(fd," %d %g ",a->j[j]+shift,a->a[j]); 381 #endif 382 } 383 fprintf(fd,"\n"); 384 } 385 } 386 else if (format == VIEWER_FORMAT_ASCII_SYMMODU) { 387 int nzd=0, fshift=1, *sptr; 388 sptr = (int *) PetscMalloc( (m+1)*sizeof(int) ); CHKPTRQ(sptr); 389 for ( i=0; i<m; i++ ) { 390 sptr[i] = nzd+1; 391 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 392 if (a->j[j] >= i) { 393 #if defined(USE_PETSC_COMPLEX) 394 if (imag(a->a[j]) != 0.0 || real(a->a[j]) != 0.0) nzd++; 395 #else 396 if (a->a[j] != 0.0) nzd++; 397 #endif 398 } 399 } 400 } 401 sptr[m] = nzd+1; 402 fprintf(fd," %d %d\n\n",m,nzd); 403 for ( i=0; i<m+1; i+=6 ) { 404 if (i+4<m) fprintf(fd," %d %d %d %d %d %d\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3],sptr[i+4],sptr[i+5]); 405 else if (i+3<m) fprintf(fd," %d %d %d %d %d\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3],sptr[i+4]); 406 else if (i+2<m) fprintf(fd," %d %d %d %d\n",sptr[i],sptr[i+1],sptr[i+2],sptr[i+3]); 407 else if (i+1<m) fprintf(fd," %d %d %d\n",sptr[i],sptr[i+1],sptr[i+2]); 408 else if (i<m) fprintf(fd," %d %d\n",sptr[i],sptr[i+1]); 409 else fprintf(fd," %d\n",sptr[i]); 410 } 411 fprintf(fd,"\n"); 412 PetscFree(sptr); 413 for ( i=0; i<m; i++ ) { 414 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 415 if (a->j[j] >= i) fprintf(fd," %d ",a->j[j]+fshift); 416 } 417 fprintf(fd,"\n"); 418 } 419 fprintf(fd,"\n"); 420 for ( i=0; i<m; i++ ) { 421 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 422 if (a->j[j] >= i) { 423 #if defined(USE_PETSC_COMPLEX) 424 if (imag(a->a[j]) != 0.0 || real(a->a[j]) != 0.0) 425 fprintf(fd," %18.16e %18.16e ",real(a->a[j]),imag(a->a[j])); 426 #else 427 if (a->a[j] != 0.0) fprintf(fd," %18.16e ",a->a[j]); 428 #endif 429 } 430 } 431 fprintf(fd,"\n"); 432 } 433 } else { 434 for ( i=0; i<m; i++ ) { 435 fprintf(fd,"row %d:",i); 436 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 437 #if defined(USE_PETSC_COMPLEX) 438 if (imag(a->a[j]) > 0.0) { 439 fprintf(fd," %d %g + %g i",a->j[j]+shift,real(a->a[j]),imag(a->a[j])); 440 } else if (imag(a->a[j]) < 0.0) { 441 fprintf(fd," %d %g - %g i",a->j[j]+shift,real(a->a[j]),-imag(a->a[j])); 442 } else { 443 fprintf(fd," %d %g ",a->j[j]+shift,real(a->a[j])); 444 } 445 #else 446 fprintf(fd," %d %g ",a->j[j]+shift,a->a[j]); 447 #endif 448 } 449 fprintf(fd,"\n"); 450 } 451 } 452 fflush(fd); 453 PetscFunctionReturn(0); 454 } 455 456 #undef __FUNC__ 457 #define __FUNC__ "MatView_SeqAIJ_Draw" 458 extern int MatView_SeqAIJ_Draw(Mat A,Viewer viewer) 459 { 460 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 461 int ierr, i,j, m = a->m, shift = a->indexshift,pause,color; 462 int format; 463 double xl,yl,xr,yr,w,h,xc,yc,scale = 1.0,x_l,x_r,y_l,y_r,maxv = 0.0; 464 Draw draw; 465 DrawButton button; 466 PetscTruth isnull; 467 468 PetscFunctionBegin; 469 ierr = ViewerDrawGetDraw(viewer,&draw); CHKERRQ(ierr); 470 ierr = DrawSynchronizedClear(draw); CHKERRQ(ierr); 471 ierr = ViewerGetFormat(viewer,&format); CHKERRQ(ierr); 472 ierr = DrawIsNull(draw,&isnull); CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 473 474 xr = a->n; yr = a->m; h = yr/10.0; w = xr/10.0; 475 xr += w; yr += h; xl = -w; yl = -h; 476 ierr = DrawSetCoordinates(draw,xl,yl,xr,yr); CHKERRQ(ierr); 477 /* loop over matrix elements drawing boxes */ 478 479 if (format != VIEWER_FORMAT_DRAW_CONTOUR) { 480 /* Blue for negative, Cyan for zero and Red for positive */ 481 color = DRAW_BLUE; 482 for ( i=0; i<m; i++ ) { 483 y_l = m - i - 1.0; y_r = y_l + 1.0; 484 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 485 x_l = a->j[j] + shift; x_r = x_l + 1.0; 486 #if defined(USE_PETSC_COMPLEX) 487 if (real(a->a[j]) >= 0.) continue; 488 #else 489 if (a->a[j] >= 0.) continue; 490 #endif 491 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 492 } 493 } 494 color = DRAW_CYAN; 495 for ( i=0; i<m; i++ ) { 496 y_l = m - i - 1.0; y_r = y_l + 1.0; 497 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 498 x_l = a->j[j] + shift; x_r = x_l + 1.0; 499 if (a->a[j] != 0.) continue; 500 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 501 } 502 } 503 color = DRAW_RED; 504 for ( i=0; i<m; i++ ) { 505 y_l = m - i - 1.0; y_r = y_l + 1.0; 506 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 507 x_l = a->j[j] + shift; x_r = x_l + 1.0; 508 #if defined(USE_PETSC_COMPLEX) 509 if (real(a->a[j]) <= 0.) continue; 510 #else 511 if (a->a[j] <= 0.) continue; 512 #endif 513 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 514 } 515 } 516 } else { 517 /* use contour shading to indicate magnitude of values */ 518 /* first determine max of all nonzero values */ 519 int nz = a->nz,count; 520 Draw popup; 521 522 for ( i=0; i<nz; i++ ) { 523 if (PetscAbsScalar(a->a[i]) > maxv) maxv = PetscAbsScalar(a->a[i]); 524 } 525 ierr = DrawGetPopup(draw,&popup); CHKERRQ(ierr); 526 ierr = DrawScalePopup(popup,0.0,maxv); CHKERRQ(ierr); 527 count = 0; 528 for ( i=0; i<m; i++ ) { 529 y_l = m - i - 1.0; y_r = y_l + 1.0; 530 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 531 x_l = a->j[j] + shift; x_r = x_l + 1.0; 532 color = 32 + (int) ((200.0 - 32.0)*PetscAbsScalar(a->a[count])/maxv); 533 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 534 count++; 535 } 536 } 537 } 538 DrawSynchronizedFlush(draw); 539 DrawGetPause(draw,&pause); 540 if (pause >= 0) { PetscSleep(pause); PetscFunctionReturn(0);} 541 542 /* allow the matrix to zoom or shrink */ 543 ierr = DrawCheckResizedWindow(draw); 544 ierr = DrawSynchronizedGetMouseButton(draw,&button,&xc,&yc,0,0); 545 while (button != BUTTON_RIGHT) { 546 DrawSynchronizedClear(draw); 547 if (button == BUTTON_LEFT) scale = .5; 548 else if (button == BUTTON_CENTER) scale = 2.; 549 xl = scale*(xl + w - xc) + xc - w*scale; 550 xr = scale*(xr - w - xc) + xc + w*scale; 551 yl = scale*(yl + h - yc) + yc - h*scale; 552 yr = scale*(yr - h - yc) + yc + h*scale; 553 w *= scale; h *= scale; 554 ierr = DrawSetCoordinates(draw,xl,yl,xr,yr); CHKERRQ(ierr); 555 if (format != VIEWER_FORMAT_DRAW_CONTOUR) { 556 /* Blue for negative, Cyan for zero and Red for positive */ 557 color = DRAW_BLUE; 558 for ( i=0; i<m; i++ ) { 559 y_l = m - i - 1.0; y_r = y_l + 1.0; 560 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 561 x_l = a->j[j] + shift; x_r = x_l + 1.0; 562 #if defined(USE_PETSC_COMPLEX) 563 if (real(a->a[j]) >= 0.) continue; 564 #else 565 if (a->a[j] >= 0.) continue; 566 #endif 567 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 568 } 569 } 570 color = DRAW_CYAN; 571 for ( i=0; i<m; i++ ) { 572 y_l = m - i - 1.0; y_r = y_l + 1.0; 573 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 574 x_l = a->j[j] + shift; x_r = x_l + 1.0; 575 if (a->a[j] != 0.) continue; 576 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 577 } 578 } 579 color = DRAW_RED; 580 for ( i=0; i<m; i++ ) { 581 y_l = m - i - 1.0; y_r = y_l + 1.0; 582 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 583 x_l = a->j[j] + shift; x_r = x_l + 1.0; 584 #if defined(USE_PETSC_COMPLEX) 585 if (real(a->a[j]) <= 0.) continue; 586 #else 587 if (a->a[j] <= 0.) continue; 588 #endif 589 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); 590 } 591 } 592 } else { 593 /* use contour shading to indicate magnitude of values */ 594 int count = 0; 595 for ( i=0; i<m; i++ ) { 596 y_l = m - i - 1.0; y_r = y_l + 1.0; 597 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 598 x_l = a->j[j] + shift; x_r = x_l + 1.0; 599 color = 32 + (int) ((200.0 - 32.0)*PetscAbsScalar(a->a[count])/maxv); 600 DrawRectangle(draw,x_l,y_l,x_r,y_r,color,color,color,color); CHKERRQ(ierr); 601 count++; 602 } 603 } 604 } 605 606 ierr = DrawCheckResizedWindow(draw); CHKERRQ(ierr); 607 ierr = DrawSynchronizedGetMouseButton(draw,&button,&xc,&yc,0,0); CHKERRQ(ierr); 608 } 609 PetscFunctionReturn(0); 610 } 611 612 #undef __FUNC__ 613 #define __FUNC__ "MatView_SeqAIJ" 614 int MatView_SeqAIJ(PetscObject obj,Viewer viewer) 615 { 616 Mat A = (Mat) obj; 617 Mat_SeqAIJ *a = (Mat_SeqAIJ*) A->data; 618 ViewerType vtype; 619 int ierr; 620 621 PetscFunctionBegin; 622 ierr = ViewerGetType(viewer,&vtype); CHKERRQ(ierr); 623 if (vtype == MATLAB_VIEWER) { 624 ierr = ViewerMatlabPutSparse_Private(viewer,a->m,a->n,a->nz,a->a,a->i,a->j); CHKERRQ(ierr); 625 } 626 else if (vtype == ASCII_FILE_VIEWER || vtype == ASCII_FILES_VIEWER){ 627 ierr = MatView_SeqAIJ_ASCII(A,viewer); CHKERRQ(ierr); 628 } 629 else if (vtype == BINARY_FILE_VIEWER) { 630 ierr = MatView_SeqAIJ_Binary(A,viewer); CHKERRQ(ierr); 631 } 632 else if (vtype == DRAW_VIEWER) { 633 ierr = MatView_SeqAIJ_Draw(A,viewer); CHKERRQ(ierr); 634 } 635 PetscFunctionReturn(0); 636 } 637 638 extern int Mat_AIJ_CheckInode(Mat); 639 #undef __FUNC__ 640 #define __FUNC__ "MatAssemblyEnd_SeqAIJ" 641 int MatAssemblyEnd_SeqAIJ(Mat A,MatAssemblyType mode) 642 { 643 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 644 int fshift = 0,i,j,*ai = a->i, *aj = a->j, *imax = a->imax,ierr; 645 int m = a->m, *ip, N, *ailen = a->ilen,shift = a->indexshift,rmax = 0; 646 Scalar *aa = a->a, *ap; 647 648 PetscFunctionBegin; 649 if (mode == MAT_FLUSH_ASSEMBLY) PetscFunctionReturn(0); 650 651 if (m) rmax = ailen[0]; /* determine row with most nonzeros */ 652 for ( i=1; i<m; i++ ) { 653 /* move each row back by the amount of empty slots (fshift) before it*/ 654 fshift += imax[i-1] - ailen[i-1]; 655 rmax = PetscMax(rmax,ailen[i]); 656 if (fshift) { 657 ip = aj + ai[i] + shift; ap = aa + ai[i] + shift; 658 N = ailen[i]; 659 for ( j=0; j<N; j++ ) { 660 ip[j-fshift] = ip[j]; 661 ap[j-fshift] = ap[j]; 662 } 663 } 664 ai[i] = ai[i-1] + ailen[i-1]; 665 } 666 if (m) { 667 fshift += imax[m-1] - ailen[m-1]; 668 ai[m] = ai[m-1] + ailen[m-1]; 669 } 670 /* reset ilen and imax for each row */ 671 for ( i=0; i<m; i++ ) { 672 ailen[i] = imax[i] = ai[i+1] - ai[i]; 673 } 674 a->nz = ai[m] + shift; 675 676 /* diagonals may have moved, so kill the diagonal pointers */ 677 if (fshift && a->diag) { 678 PetscFree(a->diag); 679 PLogObjectMemory(A,-(m+1)*sizeof(int)); 680 a->diag = 0; 681 } 682 PLogInfo(A,"MatAssemblyEnd_SeqAIJ:Matrix size: %d X %d; storage space: %d unneeded, %d used\n", 683 m,a->n,fshift,a->nz); 684 PLogInfo(A,"MatAssemblyEnd_SeqAIJ:Number of mallocs during MatSetValues is %d\n", 685 a->reallocs); 686 PLogInfo(A,"MatAssemblyEnd_SeqAIJ:Most nonzeros in any row is %d\n",rmax); 687 a->reallocs = 0; 688 A->info.nz_unneeded = (double)fshift; 689 690 /* check out for identical nodes. If found, use inode functions */ 691 ierr = Mat_AIJ_CheckInode(A); CHKERRQ(ierr); 692 PetscFunctionReturn(0); 693 } 694 695 #undef __FUNC__ 696 #define __FUNC__ "MatZeroEntries_SeqAIJ" 697 int MatZeroEntries_SeqAIJ(Mat A) 698 { 699 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 700 701 PetscFunctionBegin; 702 PetscMemzero(a->a,(a->i[a->m]+a->indexshift)*sizeof(Scalar)); 703 PetscFunctionReturn(0); 704 } 705 706 #undef __FUNC__ 707 #define __FUNC__ "MatDestroy_SeqAIJ" 708 int MatDestroy_SeqAIJ(PetscObject obj) 709 { 710 Mat A = (Mat) obj; 711 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 712 int ierr; 713 714 PetscFunctionBegin; 715 #if defined(USE_PETSC_LOG) 716 PLogObjectState(obj,"Rows=%d, Cols=%d, NZ=%d",a->m,a->n,a->nz); 717 #endif 718 PetscFree(a->a); 719 if (!a->singlemalloc) { PetscFree(a->i); PetscFree(a->j);} 720 if (a->diag) PetscFree(a->diag); 721 if (a->ilen) PetscFree(a->ilen); 722 if (a->imax) PetscFree(a->imax); 723 if (a->solve_work) PetscFree(a->solve_work); 724 if (a->inode.size) PetscFree(a->inode.size); 725 if (a->icol) {ierr = ISDestroy(a->icol);CHKERRQ(ierr);} 726 PetscFree(a); 727 728 PLogObjectDestroy(A); 729 PetscHeaderDestroy(A); 730 PetscFunctionReturn(0); 731 } 732 733 #undef __FUNC__ 734 #define __FUNC__ "MatCompress_SeqAIJ" 735 int MatCompress_SeqAIJ(Mat A) 736 { 737 PetscFunctionBegin; 738 PetscFunctionReturn(0); 739 } 740 741 #undef __FUNC__ 742 #define __FUNC__ "MatSetOption_SeqAIJ" 743 int MatSetOption_SeqAIJ(Mat A,MatOption op) 744 { 745 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 746 747 PetscFunctionBegin; 748 if (op == MAT_ROW_ORIENTED) a->roworiented = 1; 749 else if (op == MAT_COLUMN_ORIENTED) a->roworiented = 0; 750 else if (op == MAT_COLUMNS_SORTED) a->sorted = 1; 751 else if (op == MAT_COLUMNS_UNSORTED) a->sorted = 0; 752 else if (op == MAT_NO_NEW_NONZERO_LOCATIONS) a->nonew = 1; 753 else if (op == MAT_NEW_NONZERO_LOCATION_ERROR) a->nonew = -1; 754 else if (op == MAT_NEW_NONZERO_ALLOCATION_ERROR) a->nonew = -2; 755 else if (op == MAT_YES_NEW_NONZERO_LOCATIONS) a->nonew = 0; 756 else if (op == MAT_ROWS_SORTED || 757 op == MAT_ROWS_UNSORTED || 758 op == MAT_SYMMETRIC || 759 op == MAT_STRUCTURALLY_SYMMETRIC || 760 op == MAT_YES_NEW_DIAGONALS || 761 op == MAT_IGNORE_OFF_PROC_ENTRIES|| 762 op == MAT_USE_HASH_TABLE) 763 PLogInfo(A,"MatSetOption_SeqAIJ:Option ignored\n"); 764 else if (op == MAT_NO_NEW_DIAGONALS) { 765 SETERRQ(PETSC_ERR_SUP,0,"MAT_NO_NEW_DIAGONALS"); 766 } else if (op == MAT_INODE_LIMIT_1) a->inode.limit = 1; 767 else if (op == MAT_INODE_LIMIT_2) a->inode.limit = 2; 768 else if (op == MAT_INODE_LIMIT_3) a->inode.limit = 3; 769 else if (op == MAT_INODE_LIMIT_4) a->inode.limit = 4; 770 else if (op == MAT_INODE_LIMIT_5) a->inode.limit = 5; 771 else SETERRQ(PETSC_ERR_SUP,0,"unknown option"); 772 PetscFunctionReturn(0); 773 } 774 775 #undef __FUNC__ 776 #define __FUNC__ "MatGetDiagonal_SeqAIJ" 777 int MatGetDiagonal_SeqAIJ(Mat A,Vec v) 778 { 779 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 780 int i,j, n,shift = a->indexshift,ierr; 781 Scalar *x, zero = 0.0; 782 783 PetscFunctionBegin; 784 ierr = VecSet(&zero,v);CHKERRQ(ierr); 785 VecGetArray_Fast(v,x); VecGetLocalSize(v,&n); 786 if (n != a->m) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Nonconforming matrix and vector"); 787 for ( i=0; i<a->m; i++ ) { 788 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 789 if (a->j[j]+shift == i) { 790 x[i] = a->a[j]; 791 break; 792 } 793 } 794 } 795 PetscFunctionReturn(0); 796 } 797 798 /* -------------------------------------------------------*/ 799 /* Should check that shapes of vectors and matrices match */ 800 /* -------------------------------------------------------*/ 801 #undef __FUNC__ 802 #define __FUNC__ "MatMultTrans_SeqAIJ" 803 int MatMultTrans_SeqAIJ(Mat A,Vec xx,Vec yy) 804 { 805 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 806 Scalar *x, *y, *v, alpha; 807 int m = a->m, n, i, *idx, shift = a->indexshift; 808 809 PetscFunctionBegin; 810 VecGetArray_Fast(xx,x); VecGetArray_Fast(yy,y); 811 PetscMemzero(y,a->n*sizeof(Scalar)); 812 y = y + shift; /* shift for Fortran start by 1 indexing */ 813 for ( i=0; i<m; i++ ) { 814 idx = a->j + a->i[i] + shift; 815 v = a->a + a->i[i] + shift; 816 n = a->i[i+1] - a->i[i]; 817 alpha = x[i]; 818 while (n-->0) {y[*idx++] += alpha * *v++;} 819 } 820 PLogFlops(2*a->nz - a->n); 821 PetscFunctionReturn(0); 822 } 823 824 #undef __FUNC__ 825 #define __FUNC__ "MatMultTransAdd_SeqAIJ" 826 int MatMultTransAdd_SeqAIJ(Mat A,Vec xx,Vec zz,Vec yy) 827 { 828 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 829 Scalar *x, *y, *v, alpha; 830 int m = a->m, n, i, *idx,shift = a->indexshift; 831 832 PetscFunctionBegin; 833 VecGetArray_Fast(xx,x); VecGetArray_Fast(yy,y); 834 if (zz != yy) VecCopy(zz,yy); 835 y = y + shift; /* shift for Fortran start by 1 indexing */ 836 for ( i=0; i<m; i++ ) { 837 idx = a->j + a->i[i] + shift; 838 v = a->a + a->i[i] + shift; 839 n = a->i[i+1] - a->i[i]; 840 alpha = x[i]; 841 while (n-->0) {y[*idx++] += alpha * *v++;} 842 } 843 PLogFlops(2*a->nz); 844 PetscFunctionReturn(0); 845 } 846 847 #undef __FUNC__ 848 #define __FUNC__ "MatMult_SeqAIJ" 849 int MatMult_SeqAIJ(Mat A,Vec xx,Vec yy) 850 { 851 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 852 Scalar *x, *y, *v, sum; 853 int m = a->m, *idx, shift = a->indexshift,*ii; 854 #if !defined(USE_FORTRAN_KERNELS) 855 int n, i, jrow,j; 856 #endif 857 858 PetscFunctionBegin; 859 VecGetArray_Fast(xx,x); VecGetArray_Fast(yy,y); 860 x = x + shift; /* shift for Fortran start by 1 indexing */ 861 idx = a->j; 862 v = a->a; 863 ii = a->i; 864 #if defined(USE_FORTRAN_KERNELS) 865 fortranmultaij_(&m,x,ii,idx+shift,v+shift,y); 866 #else 867 v += shift; /* shift for Fortran start by 1 indexing */ 868 idx += shift; 869 for ( i=0; i<m; i++ ) { 870 jrow = ii[i]; 871 n = ii[i+1] - jrow; 872 sum = 0.0; 873 for ( j=0; j<n; j++) { 874 sum += v[jrow]*x[idx[jrow]]; jrow++; 875 } 876 y[i] = sum; 877 } 878 #endif 879 PLogFlops(2*a->nz - m); 880 PetscFunctionReturn(0); 881 } 882 883 #undef __FUNC__ 884 #define __FUNC__ "MatMultAdd_SeqAIJ" 885 int MatMultAdd_SeqAIJ(Mat A,Vec xx,Vec yy,Vec zz) 886 { 887 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 888 Scalar *x, *y, *z, *v, sum; 889 int m = a->m, *idx, shift = a->indexshift,*ii; 890 #if !defined(USE_FORTRAN_KERNELS) 891 int n,i,jrow,j; 892 #endif 893 894 PetscFunctionBegin; 895 VecGetArray_Fast(xx,x); VecGetArray_Fast(yy,y); VecGetArray_Fast(zz,z); 896 x = x + shift; /* shift for Fortran start by 1 indexing */ 897 idx = a->j; 898 v = a->a; 899 ii = a->i; 900 #if defined(USE_FORTRAN_KERNELS) 901 fortranmultaddaij_(&m,x,ii,idx+shift,v+shift,y,z); 902 #else 903 v += shift; /* shift for Fortran start by 1 indexing */ 904 idx += shift; 905 for ( i=0; i<m; i++ ) { 906 jrow = ii[i]; 907 n = ii[i+1] - jrow; 908 sum = y[i]; 909 for ( j=0; j<n; j++) { 910 sum += v[jrow]*x[idx[jrow]]; jrow++; 911 } 912 z[i] = sum; 913 } 914 #endif 915 PLogFlops(2*a->nz); 916 PetscFunctionReturn(0); 917 } 918 919 /* 920 Adds diagonal pointers to sparse matrix structure. 921 */ 922 923 #undef __FUNC__ 924 #define __FUNC__ "MatMarkDiag_SeqAIJ" 925 int MatMarkDiag_SeqAIJ(Mat A) 926 { 927 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 928 int i,j, *diag, m = a->m,shift = a->indexshift; 929 930 PetscFunctionBegin; 931 diag = (int *) PetscMalloc( (m+1)*sizeof(int)); CHKPTRQ(diag); 932 PLogObjectMemory(A,(m+1)*sizeof(int)); 933 for ( i=0; i<a->m; i++ ) { 934 for ( j=a->i[i]+shift; j<a->i[i+1]+shift; j++ ) { 935 if (a->j[j]+shift == i) { 936 diag[i] = j - shift; 937 break; 938 } 939 } 940 } 941 a->diag = diag; 942 PetscFunctionReturn(0); 943 } 944 945 #undef __FUNC__ 946 #define __FUNC__ "MatRelax_SeqAIJ" 947 int MatRelax_SeqAIJ(Mat A,Vec bb,double omega,MatSORType flag, 948 double fshift,int its,Vec xx) 949 { 950 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 951 Scalar *x, *b, *bs, d, *xs, sum, *v = a->a,*t,scale,*ts, *xb; 952 int ierr, *idx, *diag,n = a->n, m = a->m, i, shift = a->indexshift; 953 954 PetscFunctionBegin; 955 VecGetArray_Fast(xx,x); VecGetArray_Fast(bb,b); 956 if (!a->diag) {ierr = MatMarkDiag_SeqAIJ(A);CHKERRQ(ierr);} 957 diag = a->diag; 958 xs = x + shift; /* shifted by one for index start of a or a->j*/ 959 if (flag == SOR_APPLY_UPPER) { 960 /* apply ( U + D/omega) to the vector */ 961 bs = b + shift; 962 for ( i=0; i<m; i++ ) { 963 d = fshift + a->a[diag[i] + shift]; 964 n = a->i[i+1] - diag[i] - 1; 965 idx = a->j + diag[i] + (!shift); 966 v = a->a + diag[i] + (!shift); 967 sum = b[i]*d/omega; 968 SPARSEDENSEDOT(sum,bs,v,idx,n); 969 x[i] = sum; 970 } 971 PetscFunctionReturn(0); 972 } 973 if (flag == SOR_APPLY_LOWER) { 974 SETERRQ(PETSC_ERR_SUP,0,"SOR_APPLY_LOWER is not done"); 975 } else if (flag & SOR_EISENSTAT) { 976 /* Let A = L + U + D; where L is lower trianglar, 977 U is upper triangular, E is diagonal; This routine applies 978 979 (L + E)^{-1} A (U + E)^{-1} 980 981 to a vector efficiently using Eisenstat's trick. This is for 982 the case of SSOR preconditioner, so E is D/omega where omega 983 is the relaxation factor. 984 */ 985 t = (Scalar *) PetscMalloc( m*sizeof(Scalar) ); CHKPTRQ(t); 986 scale = (2.0/omega) - 1.0; 987 988 /* x = (E + U)^{-1} b */ 989 for ( i=m-1; i>=0; i-- ) { 990 d = fshift + a->a[diag[i] + shift]; 991 n = a->i[i+1] - diag[i] - 1; 992 idx = a->j + diag[i] + (!shift); 993 v = a->a + diag[i] + (!shift); 994 sum = b[i]; 995 SPARSEDENSEMDOT(sum,xs,v,idx,n); 996 x[i] = omega*(sum/d); 997 } 998 999 /* t = b - (2*E - D)x */ 1000 v = a->a; 1001 for ( i=0; i<m; i++ ) { t[i] = b[i] - scale*(v[*diag++ + shift])*x[i]; } 1002 1003 /* t = (E + L)^{-1}t */ 1004 ts = t + shift; /* shifted by one for index start of a or a->j*/ 1005 diag = a->diag; 1006 for ( i=0; i<m; i++ ) { 1007 d = fshift + a->a[diag[i]+shift]; 1008 n = diag[i] - a->i[i]; 1009 idx = a->j + a->i[i] + shift; 1010 v = a->a + a->i[i] + shift; 1011 sum = t[i]; 1012 SPARSEDENSEMDOT(sum,ts,v,idx,n); 1013 t[i] = omega*(sum/d); 1014 } 1015 1016 /* x = x + t */ 1017 for ( i=0; i<m; i++ ) { x[i] += t[i]; } 1018 PetscFree(t); 1019 PetscFunctionReturn(0); 1020 } 1021 if (flag & SOR_ZERO_INITIAL_GUESS) { 1022 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 1023 for ( i=0; i<m; i++ ) { 1024 d = fshift + a->a[diag[i]+shift]; 1025 n = diag[i] - a->i[i]; 1026 idx = a->j + a->i[i] + shift; 1027 v = a->a + a->i[i] + shift; 1028 sum = b[i]; 1029 SPARSEDENSEMDOT(sum,xs,v,idx,n); 1030 x[i] = omega*(sum/d); 1031 } 1032 xb = x; 1033 } else xb = b; 1034 if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) && 1035 (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) { 1036 for ( i=0; i<m; i++ ) { 1037 x[i] *= a->a[diag[i]+shift]; 1038 } 1039 } 1040 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 1041 for ( i=m-1; i>=0; i-- ) { 1042 d = fshift + a->a[diag[i] + shift]; 1043 n = a->i[i+1] - diag[i] - 1; 1044 idx = a->j + diag[i] + (!shift); 1045 v = a->a + diag[i] + (!shift); 1046 sum = xb[i]; 1047 SPARSEDENSEMDOT(sum,xs,v,idx,n); 1048 x[i] = omega*(sum/d); 1049 } 1050 } 1051 its--; 1052 } 1053 while (its--) { 1054 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 1055 for ( i=0; i<m; i++ ) { 1056 d = fshift + a->a[diag[i]+shift]; 1057 n = a->i[i+1] - a->i[i]; 1058 idx = a->j + a->i[i] + shift; 1059 v = a->a + a->i[i] + shift; 1060 sum = b[i]; 1061 SPARSEDENSEMDOT(sum,xs,v,idx,n); 1062 x[i] = (1. - omega)*x[i] + omega*(sum + a->a[diag[i]+shift]*x[i])/d; 1063 } 1064 } 1065 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 1066 for ( i=m-1; i>=0; i-- ) { 1067 d = fshift + a->a[diag[i] + shift]; 1068 n = a->i[i+1] - a->i[i]; 1069 idx = a->j + a->i[i] + shift; 1070 v = a->a + a->i[i] + shift; 1071 sum = b[i]; 1072 SPARSEDENSEMDOT(sum,xs,v,idx,n); 1073 x[i] = (1. - omega)*x[i] + omega*(sum + a->a[diag[i]+shift]*x[i])/d; 1074 } 1075 } 1076 } 1077 PetscFunctionReturn(0); 1078 } 1079 1080 #undef __FUNC__ 1081 #define __FUNC__ "MatGetInfo_SeqAIJ" 1082 int MatGetInfo_SeqAIJ(Mat A,MatInfoType flag,MatInfo *info) 1083 { 1084 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1085 1086 PetscFunctionBegin; 1087 info->rows_global = (double)a->m; 1088 info->columns_global = (double)a->n; 1089 info->rows_local = (double)a->m; 1090 info->columns_local = (double)a->n; 1091 info->block_size = 1.0; 1092 info->nz_allocated = (double)a->maxnz; 1093 info->nz_used = (double)a->nz; 1094 info->nz_unneeded = (double)(a->maxnz - a->nz); 1095 /* if (info->nz_unneeded != A->info.nz_unneeded) 1096 printf("space descrepancy: maxnz-nz = %d, nz_unneeded = %d\n",(int)info->nz_unneeded,(int)A->info.nz_unneeded); */ 1097 info->assemblies = (double)A->num_ass; 1098 info->mallocs = (double)a->reallocs; 1099 info->memory = A->mem; 1100 if (A->factor) { 1101 info->fill_ratio_given = A->info.fill_ratio_given; 1102 info->fill_ratio_needed = A->info.fill_ratio_needed; 1103 info->factor_mallocs = A->info.factor_mallocs; 1104 } else { 1105 info->fill_ratio_given = 0; 1106 info->fill_ratio_needed = 0; 1107 info->factor_mallocs = 0; 1108 } 1109 PetscFunctionReturn(0); 1110 } 1111 1112 extern int MatLUFactorSymbolic_SeqAIJ(Mat,IS,IS,double,Mat*); 1113 extern int MatLUFactorNumeric_SeqAIJ(Mat,Mat*); 1114 extern int MatLUFactor_SeqAIJ(Mat,IS,IS,double); 1115 extern int MatSolve_SeqAIJ(Mat,Vec,Vec); 1116 extern int MatSolveAdd_SeqAIJ(Mat,Vec,Vec,Vec); 1117 extern int MatSolveTrans_SeqAIJ(Mat,Vec,Vec); 1118 extern int MatSolveTransAdd_SeqAIJ(Mat,Vec,Vec,Vec); 1119 1120 #undef __FUNC__ 1121 #define __FUNC__ "MatZeroRows_SeqAIJ" 1122 int MatZeroRows_SeqAIJ(Mat A,IS is,Scalar *diag) 1123 { 1124 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1125 int i,ierr,N, *rows,m = a->m - 1,shift = a->indexshift; 1126 1127 PetscFunctionBegin; 1128 ierr = ISGetSize(is,&N); CHKERRQ(ierr); 1129 ierr = ISGetIndices(is,&rows); CHKERRQ(ierr); 1130 if (diag) { 1131 for ( i=0; i<N; i++ ) { 1132 if (rows[i] < 0 || rows[i] > m) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"row out of range"); 1133 if (a->ilen[rows[i]] > 0) { /* in case row was completely empty */ 1134 a->ilen[rows[i]] = 1; 1135 a->a[a->i[rows[i]]+shift] = *diag; 1136 a->j[a->i[rows[i]]+shift] = rows[i]+shift; 1137 } else { 1138 ierr = MatSetValues_SeqAIJ(A,1,&rows[i],1,&rows[i],diag,INSERT_VALUES);CHKERRQ(ierr); 1139 } 1140 } 1141 } else { 1142 for ( i=0; i<N; i++ ) { 1143 if (rows[i] < 0 || rows[i] > m) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"row out of range"); 1144 a->ilen[rows[i]] = 0; 1145 } 1146 } 1147 ISRestoreIndices(is,&rows); 1148 ierr = MatAssemblyEnd_SeqAIJ(A,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1149 PetscFunctionReturn(0); 1150 } 1151 1152 #undef __FUNC__ 1153 #define __FUNC__ "MatGetSize_SeqAIJ" 1154 int MatGetSize_SeqAIJ(Mat A,int *m,int *n) 1155 { 1156 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1157 1158 PetscFunctionBegin; 1159 if (m) *m = a->m; 1160 if (n) *n = a->n; 1161 PetscFunctionReturn(0); 1162 } 1163 1164 #undef __FUNC__ 1165 #define __FUNC__ "MatGetOwnershipRange_SeqAIJ" 1166 int MatGetOwnershipRange_SeqAIJ(Mat A,int *m,int *n) 1167 { 1168 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1169 1170 PetscFunctionBegin; 1171 *m = 0; *n = a->m; 1172 PetscFunctionReturn(0); 1173 } 1174 1175 #undef __FUNC__ 1176 #define __FUNC__ "MatGetRow_SeqAIJ" 1177 int MatGetRow_SeqAIJ(Mat A,int row,int *nz,int **idx,Scalar **v) 1178 { 1179 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1180 int *itmp,i,shift = a->indexshift; 1181 1182 PetscFunctionBegin; 1183 if (row < 0 || row >= a->m) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Row out of range"); 1184 1185 *nz = a->i[row+1] - a->i[row]; 1186 if (v) *v = a->a + a->i[row] + shift; 1187 if (idx) { 1188 itmp = a->j + a->i[row] + shift; 1189 if (*nz && shift) { 1190 *idx = (int *) PetscMalloc( (*nz)*sizeof(int) ); CHKPTRQ(*idx); 1191 for ( i=0; i<(*nz); i++ ) {(*idx)[i] = itmp[i] + shift;} 1192 } else if (*nz) { 1193 *idx = itmp; 1194 } 1195 else *idx = 0; 1196 } 1197 PetscFunctionReturn(0); 1198 } 1199 1200 #undef __FUNC__ 1201 #define __FUNC__ "MatRestoreRow_SeqAIJ" 1202 int MatRestoreRow_SeqAIJ(Mat A,int row,int *nz,int **idx,Scalar **v) 1203 { 1204 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1205 1206 PetscFunctionBegin; 1207 if (idx) {if (*idx && a->indexshift) PetscFree(*idx);} 1208 PetscFunctionReturn(0); 1209 } 1210 1211 #undef __FUNC__ 1212 #define __FUNC__ "MatNorm_SeqAIJ" 1213 int MatNorm_SeqAIJ(Mat A,NormType type,double *norm) 1214 { 1215 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1216 Scalar *v = a->a; 1217 double sum = 0.0; 1218 int i, j,shift = a->indexshift; 1219 1220 PetscFunctionBegin; 1221 if (type == NORM_FROBENIUS) { 1222 for (i=0; i<a->nz; i++ ) { 1223 #if defined(USE_PETSC_COMPLEX) 1224 sum += real(conj(*v)*(*v)); v++; 1225 #else 1226 sum += (*v)*(*v); v++; 1227 #endif 1228 } 1229 *norm = sqrt(sum); 1230 } else if (type == NORM_1) { 1231 double *tmp; 1232 int *jj = a->j; 1233 tmp = (double *) PetscMalloc( (a->n+1)*sizeof(double) ); CHKPTRQ(tmp); 1234 PetscMemzero(tmp,a->n*sizeof(double)); 1235 *norm = 0.0; 1236 for ( j=0; j<a->nz; j++ ) { 1237 tmp[*jj++ + shift] += PetscAbsScalar(*v); v++; 1238 } 1239 for ( j=0; j<a->n; j++ ) { 1240 if (tmp[j] > *norm) *norm = tmp[j]; 1241 } 1242 PetscFree(tmp); 1243 } else if (type == NORM_INFINITY) { 1244 *norm = 0.0; 1245 for ( j=0; j<a->m; j++ ) { 1246 v = a->a + a->i[j] + shift; 1247 sum = 0.0; 1248 for ( i=0; i<a->i[j+1]-a->i[j]; i++ ) { 1249 sum += PetscAbsScalar(*v); v++; 1250 } 1251 if (sum > *norm) *norm = sum; 1252 } 1253 } else { 1254 SETERRQ(PETSC_ERR_SUP,0,"No support for two norm"); 1255 } 1256 PetscFunctionReturn(0); 1257 } 1258 1259 #undef __FUNC__ 1260 #define __FUNC__ "MatTranspose_SeqAIJ" 1261 int MatTranspose_SeqAIJ(Mat A,Mat *B) 1262 { 1263 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1264 Mat C; 1265 int i, ierr, *aj = a->j, *ai = a->i, m = a->m, len, *col; 1266 int shift = a->indexshift; 1267 Scalar *array = a->a; 1268 1269 PetscFunctionBegin; 1270 if (B == PETSC_NULL && m != a->n) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Square matrix only for in-place"); 1271 col = (int *) PetscMalloc((1+a->n)*sizeof(int)); CHKPTRQ(col); 1272 PetscMemzero(col,(1+a->n)*sizeof(int)); 1273 if (shift) { 1274 for ( i=0; i<ai[m]-1; i++ ) aj[i] -= 1; 1275 } 1276 for ( i=0; i<ai[m]+shift; i++ ) col[aj[i]] += 1; 1277 ierr = MatCreateSeqAIJ(A->comm,a->n,m,0,col,&C); CHKERRQ(ierr); 1278 PetscFree(col); 1279 for ( i=0; i<m; i++ ) { 1280 len = ai[i+1]-ai[i]; 1281 ierr = MatSetValues(C,len,aj,1,&i,array,INSERT_VALUES); CHKERRQ(ierr); 1282 array += len; aj += len; 1283 } 1284 if (shift) { 1285 for ( i=0; i<ai[m]-1; i++ ) aj[i] += 1; 1286 } 1287 1288 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1289 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1290 1291 if (B != PETSC_NULL) { 1292 *B = C; 1293 } else { 1294 PetscOps *Abops; 1295 struct _MatOps *Aops; 1296 1297 /* This isn't really an in-place transpose */ 1298 PetscFree(a->a); 1299 if (!a->singlemalloc) {PetscFree(a->i); PetscFree(a->j);} 1300 if (a->diag) PetscFree(a->diag); 1301 if (a->ilen) PetscFree(a->ilen); 1302 if (a->imax) PetscFree(a->imax); 1303 if (a->solve_work) PetscFree(a->solve_work); 1304 if (a->inode.size) PetscFree(a->inode.size); 1305 PetscFree(a); 1306 1307 /* 1308 This is horrible, horrible code. We need to keep the 1309 A pointers for the bops and ops but copy everything 1310 else from C. 1311 */ 1312 Abops = A->bops; 1313 Aops = A->ops; 1314 PetscMemcpy(A,C,sizeof(struct _p_Mat)); 1315 A->bops = Abops; 1316 A->ops = Aops; 1317 1318 PetscHeaderDestroy(C); 1319 } 1320 PetscFunctionReturn(0); 1321 } 1322 1323 #undef __FUNC__ 1324 #define __FUNC__ "MatDiagonalScale_SeqAIJ" 1325 int MatDiagonalScale_SeqAIJ(Mat A,Vec ll,Vec rr) 1326 { 1327 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1328 Scalar *l,*r,x,*v; 1329 int i,j,m = a->m, n = a->n, M, nz = a->nz, *jj,shift = a->indexshift; 1330 1331 PetscFunctionBegin; 1332 if (ll) { 1333 /* The local size is used so that VecMPI can be passed to this routine 1334 by MatDiagonalScale_MPIAIJ */ 1335 VecGetLocalSize_Fast(ll,m); 1336 if (m != a->m) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Left scaling vector wrong length"); 1337 VecGetArray_Fast(ll,l); 1338 v = a->a; 1339 for ( i=0; i<m; i++ ) { 1340 x = l[i]; 1341 M = a->i[i+1] - a->i[i]; 1342 for ( j=0; j<M; j++ ) { (*v++) *= x;} 1343 } 1344 PLogFlops(nz); 1345 } 1346 if (rr) { 1347 VecGetLocalSize_Fast(rr,n); 1348 if (n != a->n) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Right scaling vector wrong length"); 1349 VecGetArray_Fast(rr,r); 1350 v = a->a; jj = a->j; 1351 for ( i=0; i<nz; i++ ) { 1352 (*v++) *= r[*jj++ + shift]; 1353 } 1354 PLogFlops(nz); 1355 } 1356 PetscFunctionReturn(0); 1357 } 1358 1359 #undef __FUNC__ 1360 #define __FUNC__ "MatGetSubMatrix_SeqAIJ" 1361 int MatGetSubMatrix_SeqAIJ(Mat A,IS isrow,IS iscol,int csize,MatGetSubMatrixCall scall,Mat *B) 1362 { 1363 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data,*c; 1364 int *smap, i, k, kstart, kend, ierr, oldcols = a->n,*lens; 1365 int row,mat_i,*mat_j,tcol,first,step,*mat_ilen; 1366 register int sum,lensi; 1367 int *irow, *icol, nrows, ncols, shift = a->indexshift,*ssmap; 1368 int *starts,*j_new,*i_new,*aj = a->j, *ai = a->i,ii,*ailen = a->ilen; 1369 Scalar *a_new,*mat_a; 1370 Mat C; 1371 1372 PetscFunctionBegin; 1373 ierr = ISSorted(isrow,(PetscTruth*)&i);CHKERRQ(ierr); 1374 if (!i) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,0,"ISrow is not sorted"); 1375 ierr = ISSorted(iscol,(PetscTruth*)&i);CHKERRQ(ierr); 1376 if (!i) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,0,"IScol is not sorted"); 1377 1378 ierr = ISGetIndices(isrow,&irow); CHKERRQ(ierr); 1379 ierr = ISGetSize(isrow,&nrows); CHKERRQ(ierr); 1380 ierr = ISGetSize(iscol,&ncols); CHKERRQ(ierr); 1381 1382 if (ISStrideGetInfo(iscol,&first,&step) && step == 1) { /* no need to sort */ 1383 /* special case of contiguous rows */ 1384 lens = (int *) PetscMalloc((ncols+nrows+1)*sizeof(int)); CHKPTRQ(lens); 1385 starts = lens + ncols; 1386 /* loop over new rows determining lens and starting points */ 1387 for (i=0; i<nrows; i++) { 1388 kstart = ai[irow[i]]+shift; 1389 kend = kstart + ailen[irow[i]]; 1390 for ( k=kstart; k<kend; k++ ) { 1391 if (aj[k]+shift >= first) { 1392 starts[i] = k; 1393 break; 1394 } 1395 } 1396 sum = 0; 1397 while (k < kend) { 1398 if (aj[k++]+shift >= first+ncols) break; 1399 sum++; 1400 } 1401 lens[i] = sum; 1402 } 1403 /* create submatrix */ 1404 if (scall == MAT_REUSE_MATRIX) { 1405 int n_cols,n_rows; 1406 ierr = MatGetSize(*B,&n_rows,&n_cols); CHKERRQ(ierr); 1407 if (n_rows != nrows || n_cols != ncols) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Reused submatrix wrong size"); 1408 ierr = MatZeroEntries(*B); CHKERRQ(ierr); 1409 C = *B; 1410 } else { 1411 ierr = MatCreateSeqAIJ(A->comm,nrows,ncols,0,lens,&C);CHKERRQ(ierr); 1412 } 1413 c = (Mat_SeqAIJ*) C->data; 1414 1415 /* loop over rows inserting into submatrix */ 1416 a_new = c->a; 1417 j_new = c->j; 1418 i_new = c->i; 1419 i_new[0] = -shift; 1420 for (i=0; i<nrows; i++) { 1421 ii = starts[i]; 1422 lensi = lens[i]; 1423 for ( k=0; k<lensi; k++ ) { 1424 *j_new++ = aj[ii+k] - first; 1425 } 1426 PetscMemcpy(a_new,a->a + starts[i],lensi*sizeof(Scalar)); 1427 a_new += lensi; 1428 i_new[i+1] = i_new[i] + lensi; 1429 c->ilen[i] = lensi; 1430 } 1431 PetscFree(lens); 1432 } else { 1433 ierr = ISGetIndices(iscol,&icol); CHKERRQ(ierr); 1434 smap = (int *) PetscMalloc((1+oldcols)*sizeof(int)); CHKPTRQ(smap); 1435 ssmap = smap + shift; 1436 lens = (int *) PetscMalloc((1+nrows)*sizeof(int)); CHKPTRQ(lens); 1437 PetscMemzero(smap,oldcols*sizeof(int)); 1438 for ( i=0; i<ncols; i++ ) smap[icol[i]] = i+1; 1439 /* determine lens of each row */ 1440 for (i=0; i<nrows; i++) { 1441 kstart = ai[irow[i]]+shift; 1442 kend = kstart + a->ilen[irow[i]]; 1443 lens[i] = 0; 1444 for ( k=kstart; k<kend; k++ ) { 1445 if (ssmap[aj[k]]) { 1446 lens[i]++; 1447 } 1448 } 1449 } 1450 /* Create and fill new matrix */ 1451 if (scall == MAT_REUSE_MATRIX) { 1452 c = (Mat_SeqAIJ *)((*B)->data); 1453 1454 if (c->m != nrows || c->n != ncols) SETERRQ(PETSC_ERR_ARG_SIZ,0,"Cannot reuse matrix. wrong size"); 1455 if (PetscMemcmp(c->ilen,lens, c->m *sizeof(int))) { 1456 SETERRQ(PETSC_ERR_ARG_SIZ,0,"Cannot reuse matrix. wrong no of nonzeros"); 1457 } 1458 PetscMemzero(c->ilen,c->m*sizeof(int)); 1459 C = *B; 1460 } else { 1461 ierr = MatCreateSeqAIJ(A->comm,nrows,ncols,0,lens,&C);CHKERRQ(ierr); 1462 } 1463 c = (Mat_SeqAIJ *)(C->data); 1464 for (i=0; i<nrows; i++) { 1465 row = irow[i]; 1466 kstart = ai[row]+shift; 1467 kend = kstart + a->ilen[row]; 1468 mat_i = c->i[i]+shift; 1469 mat_j = c->j + mat_i; 1470 mat_a = c->a + mat_i; 1471 mat_ilen = c->ilen + i; 1472 for ( k=kstart; k<kend; k++ ) { 1473 if ((tcol=ssmap[a->j[k]])) { 1474 *mat_j++ = tcol - (!shift); 1475 *mat_a++ = a->a[k]; 1476 (*mat_ilen)++; 1477 1478 } 1479 } 1480 } 1481 /* Free work space */ 1482 ierr = ISRestoreIndices(iscol,&icol); CHKERRQ(ierr); 1483 PetscFree(smap); PetscFree(lens); 1484 } 1485 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1486 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1487 1488 ierr = ISRestoreIndices(isrow,&irow); CHKERRQ(ierr); 1489 *B = C; 1490 PetscFunctionReturn(0); 1491 } 1492 1493 /* 1494 note: This can only work for identity for row and col. It would 1495 be good to check this and otherwise generate an error. 1496 */ 1497 #undef __FUNC__ 1498 #define __FUNC__ "MatILUFactor_SeqAIJ" 1499 int MatILUFactor_SeqAIJ(Mat inA,IS row,IS col,double efill,int fill) 1500 { 1501 Mat_SeqAIJ *a = (Mat_SeqAIJ *) inA->data; 1502 int ierr; 1503 Mat outA; 1504 1505 PetscFunctionBegin; 1506 if (fill != 0) SETERRQ(PETSC_ERR_SUP,0,"Only fill=0 supported"); 1507 1508 outA = inA; 1509 inA->factor = FACTOR_LU; 1510 a->row = row; 1511 a->col = col; 1512 1513 /* Create the invert permutation so that it can be used in MatLUFactorNumeric() */ 1514 ierr = ISInvertPermutation(col,&(a->icol)); CHKERRQ(ierr); 1515 1516 if (!a->solve_work) { /* this matrix may have been factored before */ 1517 a->solve_work = (Scalar *) PetscMalloc( (a->m+1)*sizeof(Scalar)); CHKPTRQ(a->solve_work); 1518 } 1519 1520 if (!a->diag) { 1521 ierr = MatMarkDiag_SeqAIJ(inA); CHKERRQ(ierr); 1522 } 1523 ierr = MatLUFactorNumeric_SeqAIJ(inA,&outA); CHKERRQ(ierr); 1524 PetscFunctionReturn(0); 1525 } 1526 1527 #include "pinclude/plapack.h" 1528 #undef __FUNC__ 1529 #define __FUNC__ "MatScale_SeqAIJ" 1530 int MatScale_SeqAIJ(Scalar *alpha,Mat inA) 1531 { 1532 Mat_SeqAIJ *a = (Mat_SeqAIJ *) inA->data; 1533 int one = 1; 1534 1535 PetscFunctionBegin; 1536 BLscal_( &a->nz, alpha, a->a, &one ); 1537 PLogFlops(a->nz); 1538 PetscFunctionReturn(0); 1539 } 1540 1541 #undef __FUNC__ 1542 #define __FUNC__ "MatGetSubMatrices_SeqAIJ" 1543 int MatGetSubMatrices_SeqAIJ(Mat A,int n, IS *irow,IS *icol,MatGetSubMatrixCall scall,Mat **B) 1544 { 1545 int ierr,i; 1546 1547 PetscFunctionBegin; 1548 if (scall == MAT_INITIAL_MATRIX) { 1549 *B = (Mat *) PetscMalloc( (n+1)*sizeof(Mat) ); CHKPTRQ(*B); 1550 } 1551 1552 for ( i=0; i<n; i++ ) { 1553 ierr = MatGetSubMatrix_SeqAIJ(A,irow[i],icol[i],PETSC_DECIDE,scall,&(*B)[i]);CHKERRQ(ierr); 1554 } 1555 PetscFunctionReturn(0); 1556 } 1557 1558 #undef __FUNC__ 1559 #define __FUNC__ "MatGetBlockSize_SeqAIJ" 1560 int MatGetBlockSize_SeqAIJ(Mat A, int *bs) 1561 { 1562 PetscFunctionBegin; 1563 *bs = 1; 1564 PetscFunctionReturn(0); 1565 } 1566 1567 #undef __FUNC__ 1568 #define __FUNC__ "MatIncreaseOverlap_SeqAIJ" 1569 int MatIncreaseOverlap_SeqAIJ(Mat A, int is_max, IS *is, int ov) 1570 { 1571 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1572 int shift, row, i,j,k,l,m,n, *idx,ierr, *nidx, isz, val; 1573 int start, end, *ai, *aj; 1574 BT table; 1575 1576 PetscFunctionBegin; 1577 shift = a->indexshift; 1578 m = a->m; 1579 ai = a->i; 1580 aj = a->j+shift; 1581 1582 if (ov < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"illegal overlap value used"); 1583 1584 nidx = (int *) PetscMalloc((m+1)*sizeof(int)); CHKPTRQ(nidx); 1585 ierr = BTCreate(m,table); CHKERRQ(ierr); 1586 1587 for ( i=0; i<is_max; i++ ) { 1588 /* Initialize the two local arrays */ 1589 isz = 0; 1590 BTMemzero(m,table); 1591 1592 /* Extract the indices, assume there can be duplicate entries */ 1593 ierr = ISGetIndices(is[i],&idx); CHKERRQ(ierr); 1594 ierr = ISGetSize(is[i],&n); CHKERRQ(ierr); 1595 1596 /* Enter these into the temp arrays. I.e., mark table[row], enter row into new index */ 1597 for ( j=0; j<n ; ++j){ 1598 if(!BTLookupSet(table, idx[j])) { nidx[isz++] = idx[j];} 1599 } 1600 ierr = ISRestoreIndices(is[i],&idx); CHKERRQ(ierr); 1601 ierr = ISDestroy(is[i]); CHKERRQ(ierr); 1602 1603 k = 0; 1604 for ( j=0; j<ov; j++){ /* for each overlap */ 1605 n = isz; 1606 for ( ; k<n ; k++){ /* do only those rows in nidx[k], which are not done yet */ 1607 row = nidx[k]; 1608 start = ai[row]; 1609 end = ai[row+1]; 1610 for ( l = start; l<end ; l++){ 1611 val = aj[l] + shift; 1612 if (!BTLookupSet(table,val)) {nidx[isz++] = val;} 1613 } 1614 } 1615 } 1616 ierr = ISCreateGeneral(PETSC_COMM_SELF, isz, nidx, (is+i)); CHKERRQ(ierr); 1617 } 1618 BTDestroy(table); 1619 PetscFree(nidx); 1620 PetscFunctionReturn(0); 1621 } 1622 1623 /* -------------------------------------------------------------- */ 1624 #undef __FUNC__ 1625 #define __FUNC__ "MatPermute_SeqAIJ" 1626 int MatPermute_SeqAIJ(Mat A, IS rowp, IS colp, Mat *B) 1627 { 1628 Mat_SeqAIJ *a = (Mat_SeqAIJ *) A->data; 1629 Scalar *vwork; 1630 int i, ierr, nz, m = a->m, n = a->n, *cwork; 1631 int *row,*col,*cnew,j,*lens; 1632 IS icolp,irowp; 1633 1634 PetscFunctionBegin; 1635 ierr = ISInvertPermutation(rowp,&irowp); CHKERRQ(ierr); 1636 ierr = ISGetIndices(irowp,&row); CHKERRQ(ierr); 1637 ierr = ISInvertPermutation(colp,&icolp); CHKERRQ(ierr); 1638 ierr = ISGetIndices(icolp,&col); CHKERRQ(ierr); 1639 1640 /* determine lengths of permuted rows */ 1641 lens = (int *) PetscMalloc( (m+1)*sizeof(int) ); CHKPTRQ(lens); 1642 for (i=0; i<m; i++ ) { 1643 lens[row[i]] = a->i[i+1] - a->i[i]; 1644 } 1645 ierr = MatCreateSeqAIJ(A->comm,m,n,0,lens,B);CHKERRQ(ierr); 1646 PetscFree(lens); 1647 1648 cnew = (int *) PetscMalloc( n*sizeof(int) ); CHKPTRQ(cnew); 1649 for (i=0; i<m; i++) { 1650 ierr = MatGetRow(A,i,&nz,&cwork,&vwork); CHKERRQ(ierr); 1651 for (j=0; j<nz; j++ ) { cnew[j] = col[cwork[j]];} 1652 ierr = MatSetValues(*B,1,&row[i],nz,cnew,vwork,INSERT_VALUES); CHKERRQ(ierr); 1653 ierr = MatRestoreRow(A,i,&nz,&cwork,&vwork); CHKERRQ(ierr); 1654 } 1655 PetscFree(cnew); 1656 ierr = MatAssemblyBegin(*B,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1657 ierr = MatAssemblyEnd(*B,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 1658 ierr = ISRestoreIndices(irowp,&row); CHKERRQ(ierr); 1659 ierr = ISRestoreIndices(icolp,&col); CHKERRQ(ierr); 1660 ierr = ISDestroy(irowp); CHKERRQ(ierr); 1661 ierr = ISDestroy(icolp); CHKERRQ(ierr); 1662 PetscFunctionReturn(0); 1663 } 1664 1665 #undef __FUNC__ 1666 #define __FUNC__ "MatPrintHelp_SeqAIJ" 1667 int MatPrintHelp_SeqAIJ(Mat A) 1668 { 1669 static int called = 0; 1670 MPI_Comm comm = A->comm; 1671 1672 PetscFunctionBegin; 1673 if (called) {PetscFunctionReturn(0);} else called = 1; 1674 (*PetscHelpPrintf)(comm," Options for MATSEQAIJ and MATMPIAIJ matrix formats (the defaults):\n"); 1675 (*PetscHelpPrintf)(comm," -mat_lu_pivotthreshold <threshold>: Set pivoting threshold\n"); 1676 (*PetscHelpPrintf)(comm," -mat_aij_oneindex: internal indices begin at 1 instead of the default 0.\n"); 1677 (*PetscHelpPrintf)(comm," -mat_aij_no_inode: Do not use inodes\n"); 1678 (*PetscHelpPrintf)(comm," -mat_aij_inode_limit <limit>: Set inode limit (max limit=5)\n"); 1679 #if defined(HAVE_ESSL) 1680 (*PetscHelpPrintf)(comm," -mat_aij_essl: Use IBM sparse LU factorization and solve.\n"); 1681 #endif 1682 PetscFunctionReturn(0); 1683 } 1684 extern int MatEqual_SeqAIJ(Mat A,Mat B, PetscTruth* flg); 1685 extern int MatFDColoringCreate_SeqAIJ(Mat,ISColoring,MatFDColoring); 1686 extern int MatColoringPatch_SeqAIJ(Mat,int,int *,ISColoring *); 1687 1688 /* -------------------------------------------------------------------*/ 1689 static struct _MatOps MatOps = {MatSetValues_SeqAIJ, 1690 MatGetRow_SeqAIJ,MatRestoreRow_SeqAIJ, 1691 MatMult_SeqAIJ,MatMultAdd_SeqAIJ, 1692 MatMultTrans_SeqAIJ,MatMultTransAdd_SeqAIJ, 1693 MatSolve_SeqAIJ,MatSolveAdd_SeqAIJ, 1694 MatSolveTrans_SeqAIJ,MatSolveTransAdd_SeqAIJ, 1695 MatLUFactor_SeqAIJ,0, 1696 MatRelax_SeqAIJ, 1697 MatTranspose_SeqAIJ, 1698 MatGetInfo_SeqAIJ,MatEqual_SeqAIJ, 1699 MatGetDiagonal_SeqAIJ,MatDiagonalScale_SeqAIJ,MatNorm_SeqAIJ, 1700 0,MatAssemblyEnd_SeqAIJ, 1701 MatCompress_SeqAIJ, 1702 MatSetOption_SeqAIJ,MatZeroEntries_SeqAIJ,MatZeroRows_SeqAIJ, 1703 MatLUFactorSymbolic_SeqAIJ,MatLUFactorNumeric_SeqAIJ,0,0, 1704 MatGetSize_SeqAIJ,MatGetSize_SeqAIJ,MatGetOwnershipRange_SeqAIJ, 1705 MatILUFactorSymbolic_SeqAIJ,0, 1706 0,0, 1707 MatConvertSameType_SeqAIJ,0,0, 1708 MatILUFactor_SeqAIJ,0,0, 1709 MatGetSubMatrices_SeqAIJ,MatIncreaseOverlap_SeqAIJ, 1710 MatGetValues_SeqAIJ,0, 1711 MatPrintHelp_SeqAIJ, 1712 MatScale_SeqAIJ,0,0, 1713 MatILUDTFactor_SeqAIJ, 1714 MatGetBlockSize_SeqAIJ, 1715 MatGetRowIJ_SeqAIJ, 1716 MatRestoreRowIJ_SeqAIJ, 1717 MatGetColumnIJ_SeqAIJ, 1718 MatRestoreColumnIJ_SeqAIJ, 1719 MatFDColoringCreate_SeqAIJ, 1720 MatColoringPatch_SeqAIJ, 1721 0, 1722 MatPermute_SeqAIJ}; 1723 1724 extern int MatUseSuperLU_SeqAIJ(Mat); 1725 extern int MatUseEssl_SeqAIJ(Mat); 1726 extern int MatUseDXML_SeqAIJ(Mat); 1727 1728 #undef __FUNC__ 1729 #define __FUNC__ "MatCreateSeqAIJ" 1730 /*@C 1731 MatCreateSeqAIJ - Creates a sparse matrix in AIJ (compressed row) format 1732 (the default parallel PETSc format). For good matrix assembly performance 1733 the user should preallocate the matrix storage by setting the parameter nz 1734 (or the array nzz). By setting these parameters accurately, performance 1735 during matrix assembly can be increased by more than a factor of 50. 1736 1737 Input Parameters: 1738 . comm - MPI communicator, set to PETSC_COMM_SELF 1739 . m - number of rows 1740 . n - number of columns 1741 . nz - number of nonzeros per row (same for all rows) 1742 . nzz - array containing the number of nonzeros in the various rows 1743 (possibly different for each row) or PETSC_NULL 1744 1745 Output Parameter: 1746 . A - the matrix 1747 1748 Notes: 1749 The AIJ format (also called the Yale sparse matrix format or 1750 compressed row storage), is fully compatible with standard Fortran 77 1751 storage. That is, the stored row and column indices can begin at 1752 either one (as in Fortran) or zero. See the users' manual for details. 1753 1754 Specify the preallocated storage with either nz or nnz (not both). 1755 Set nz=PETSC_DEFAULT and nnz=PETSC_NULL for PETSc to control dynamic memory 1756 allocation. For large problems you MUST preallocate memory or you 1757 will get TERRIBLE performance, see the users' manual chapter on matrices. 1758 1759 By default, this format uses inodes (identical nodes) when possible, to 1760 improve numerical efficiency of matrix-vector products and solves. We 1761 search for consecutive rows with the same nonzero structure, thereby 1762 reusing matrix information to achieve increased efficiency. 1763 1764 Options Database Keys: 1765 $ -mat_aij_no_inode - Do not use inodes 1766 $ -mat_aij_inode_limit <limit> - Set inode limit. 1767 $ (max limit=5) 1768 $ -mat_aij_oneindex - Internally use indexing starting at 1 1769 $ rather than 0. Note: When calling MatSetValues(), 1770 $ the user still MUST index entries starting at 0! 1771 1772 .seealso: MatCreate(), MatCreateMPIAIJ(), MatSetValues() 1773 @*/ 1774 int MatCreateSeqAIJ(MPI_Comm comm,int m,int n,int nz,int *nnz, Mat *A) 1775 { 1776 Mat B; 1777 Mat_SeqAIJ *b; 1778 int i, len, ierr, flg,size; 1779 1780 PetscFunctionBegin; 1781 MPI_Comm_size(comm,&size); 1782 if (size > 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,0,"Comm must be of size 1"); 1783 1784 *A = 0; 1785 PetscHeaderCreate(B,_p_Mat,struct _MatOps,MAT_COOKIE,MATSEQAIJ,comm,MatDestroy,MatView); 1786 PLogObjectCreate(B); 1787 B->data = (void *) (b = PetscNew(Mat_SeqAIJ)); CHKPTRQ(b); 1788 PetscMemzero(b,sizeof(Mat_SeqAIJ)); 1789 PetscMemcpy(B->ops,&MatOps,sizeof(struct _MatOps)); 1790 B->destroy = MatDestroy_SeqAIJ; 1791 B->view = MatView_SeqAIJ; 1792 B->factor = 0; 1793 B->lupivotthreshold = 1.0; 1794 B->mapping = 0; 1795 ierr = OptionsGetDouble(PETSC_NULL,"-mat_lu_pivotthreshold",&B->lupivotthreshold, 1796 &flg); CHKERRQ(ierr); 1797 b->ilu_preserve_row_sums = PETSC_FALSE; 1798 ierr = OptionsHasName(PETSC_NULL,"-pc_ilu_preserve_row_sums", 1799 (int*) &b->ilu_preserve_row_sums); CHKERRQ(ierr); 1800 b->row = 0; 1801 b->col = 0; 1802 b->icol = 0; 1803 b->indexshift = 0; 1804 b->reallocs = 0; 1805 ierr = OptionsHasName(PETSC_NULL,"-mat_aij_oneindex", &flg); CHKERRQ(ierr); 1806 if (flg) b->indexshift = -1; 1807 1808 b->m = m; B->m = m; B->M = m; 1809 b->n = n; B->n = n; B->N = n; 1810 b->imax = (int *) PetscMalloc( (m+1)*sizeof(int) ); CHKPTRQ(b->imax); 1811 if (nnz == PETSC_NULL) { 1812 if (nz == PETSC_DEFAULT) nz = 10; 1813 else if (nz <= 0) nz = 1; 1814 for ( i=0; i<m; i++ ) b->imax[i] = nz; 1815 nz = nz*m; 1816 } else { 1817 nz = 0; 1818 for ( i=0; i<m; i++ ) {b->imax[i] = nnz[i]; nz += nnz[i];} 1819 } 1820 1821 /* allocate the matrix space */ 1822 len = nz*(sizeof(int) + sizeof(Scalar)) + (b->m+1)*sizeof(int); 1823 b->a = (Scalar *) PetscMalloc( len ); CHKPTRQ(b->a); 1824 b->j = (int *) (b->a + nz); 1825 PetscMemzero(b->j,nz*sizeof(int)); 1826 b->i = b->j + nz; 1827 b->singlemalloc = 1; 1828 1829 b->i[0] = -b->indexshift; 1830 for (i=1; i<m+1; i++) { 1831 b->i[i] = b->i[i-1] + b->imax[i-1]; 1832 } 1833 1834 /* b->ilen will count nonzeros in each row so far. */ 1835 b->ilen = (int *) PetscMalloc((m+1)*sizeof(int)); 1836 PLogObjectMemory(B,len+2*(m+1)*sizeof(int)+sizeof(struct _p_Mat)+sizeof(Mat_SeqAIJ)); 1837 for ( i=0; i<b->m; i++ ) { b->ilen[i] = 0;} 1838 1839 b->nz = 0; 1840 b->maxnz = nz; 1841 b->sorted = 0; 1842 b->roworiented = 1; 1843 b->nonew = 0; 1844 b->diag = 0; 1845 b->solve_work = 0; 1846 b->spptr = 0; 1847 b->inode.node_count = 0; 1848 b->inode.size = 0; 1849 b->inode.limit = 5; 1850 b->inode.max_limit = 5; 1851 B->info.nz_unneeded = (double)b->maxnz; 1852 1853 *A = B; 1854 1855 /* SuperLU is not currently supported through PETSc */ 1856 #if defined(HAVE_SUPERLU) 1857 ierr = OptionsHasName(PETSC_NULL,"-mat_aij_superlu", &flg); CHKERRQ(ierr); 1858 if (flg) { ierr = MatUseSuperLU_SeqAIJ(B); CHKERRQ(ierr); } 1859 #endif 1860 ierr = OptionsHasName(PETSC_NULL,"-mat_aij_essl", &flg); CHKERRQ(ierr); 1861 if (flg) { ierr = MatUseEssl_SeqAIJ(B); CHKERRQ(ierr); } 1862 ierr = OptionsHasName(PETSC_NULL,"-mat_aij_dxml", &flg); CHKERRQ(ierr); 1863 if (flg) { 1864 if (!b->indexshift) SETERRQ( PETSC_ERR_LIB,0,"need -mat_aij_oneindex with -mat_aij_dxml"); 1865 ierr = MatUseDXML_SeqAIJ(B); CHKERRQ(ierr); 1866 } 1867 ierr = OptionsHasName(PETSC_NULL,"-help", &flg); CHKERRQ(ierr); 1868 if (flg) {ierr = MatPrintHelp(B); CHKERRQ(ierr); } 1869 PetscFunctionReturn(0); 1870 } 1871 1872 #undef __FUNC__ 1873 #define __FUNC__ "MatConvertSameType_SeqAIJ" 1874 int MatConvertSameType_SeqAIJ(Mat A,Mat *B,int cpvalues) 1875 { 1876 Mat C; 1877 Mat_SeqAIJ *c,*a = (Mat_SeqAIJ *) A->data; 1878 int i,len, m = a->m,shift = a->indexshift; 1879 1880 PetscFunctionBegin; 1881 *B = 0; 1882 PetscHeaderCreate(C,_p_Mat,struct _MatOps,MAT_COOKIE,MATSEQAIJ,A->comm,MatDestroy,MatView); 1883 PLogObjectCreate(C); 1884 C->data = (void *) (c = PetscNew(Mat_SeqAIJ)); CHKPTRQ(c); 1885 PetscMemcpy(C->ops,A->ops,sizeof(struct _MatOps)); 1886 C->destroy = MatDestroy_SeqAIJ; 1887 C->view = MatView_SeqAIJ; 1888 C->factor = A->factor; 1889 c->row = 0; 1890 c->col = 0; 1891 c->icol = 0; 1892 c->indexshift = shift; 1893 C->assembled = PETSC_TRUE; 1894 1895 c->m = C->m = a->m; 1896 c->n = C->n = a->n; 1897 C->M = a->m; 1898 C->N = a->n; 1899 1900 c->imax = (int *) PetscMalloc((m+1)*sizeof(int)); CHKPTRQ(c->imax); 1901 c->ilen = (int *) PetscMalloc((m+1)*sizeof(int)); CHKPTRQ(c->ilen); 1902 for ( i=0; i<m; i++ ) { 1903 c->imax[i] = a->imax[i]; 1904 c->ilen[i] = a->ilen[i]; 1905 } 1906 1907 /* allocate the matrix space */ 1908 c->singlemalloc = 1; 1909 len = (m+1)*sizeof(int)+(a->i[m])*(sizeof(Scalar)+sizeof(int)); 1910 c->a = (Scalar *) PetscMalloc( len ); CHKPTRQ(c->a); 1911 c->j = (int *) (c->a + a->i[m] + shift); 1912 c->i = c->j + a->i[m] + shift; 1913 PetscMemcpy(c->i,a->i,(m+1)*sizeof(int)); 1914 if (m > 0) { 1915 PetscMemcpy(c->j,a->j,(a->i[m]+shift)*sizeof(int)); 1916 if (cpvalues == COPY_VALUES) { 1917 PetscMemcpy(c->a,a->a,(a->i[m]+shift)*sizeof(Scalar)); 1918 } 1919 } 1920 1921 PLogObjectMemory(C,len+2*(m+1)*sizeof(int)+sizeof(struct _p_Mat)+sizeof(Mat_SeqAIJ)); 1922 c->sorted = a->sorted; 1923 c->roworiented = a->roworiented; 1924 c->nonew = a->nonew; 1925 c->ilu_preserve_row_sums = a->ilu_preserve_row_sums; 1926 1927 if (a->diag) { 1928 c->diag = (int *) PetscMalloc( (m+1)*sizeof(int) ); CHKPTRQ(c->diag); 1929 PLogObjectMemory(C,(m+1)*sizeof(int)); 1930 for ( i=0; i<m; i++ ) { 1931 c->diag[i] = a->diag[i]; 1932 } 1933 } else c->diag = 0; 1934 c->inode.limit = a->inode.limit; 1935 c->inode.max_limit = a->inode.max_limit; 1936 if (a->inode.size){ 1937 c->inode.size = (int *) PetscMalloc( (m+1)*sizeof(int) ); CHKPTRQ(c->inode.size); 1938 c->inode.node_count = a->inode.node_count; 1939 PetscMemcpy( c->inode.size, a->inode.size, (m+1)*sizeof(int)); 1940 } else { 1941 c->inode.size = 0; 1942 c->inode.node_count = 0; 1943 } 1944 c->nz = a->nz; 1945 c->maxnz = a->maxnz; 1946 c->solve_work = 0; 1947 c->spptr = 0; /* Dangerous -I'm throwing away a->spptr */ 1948 1949 *B = C; 1950 PetscFunctionReturn(0); 1951 } 1952 1953 #undef __FUNC__ 1954 #define __FUNC__ "MatLoad_SeqAIJ" 1955 int MatLoad_SeqAIJ(Viewer viewer,MatType type,Mat *A) 1956 { 1957 Mat_SeqAIJ *a; 1958 Mat B; 1959 int i, nz, ierr, fd, header[4],size,*rowlengths = 0,M,N,shift; 1960 MPI_Comm comm; 1961 1962 PetscFunctionBegin; 1963 PetscObjectGetComm((PetscObject) viewer,&comm); 1964 MPI_Comm_size(comm,&size); 1965 if (size > 1) SETERRQ(PETSC_ERR_ARG_SIZ,0,"view must have one processor"); 1966 ierr = ViewerBinaryGetDescriptor(viewer,&fd); CHKERRQ(ierr); 1967 ierr = PetscBinaryRead(fd,header,4,PETSC_INT); CHKERRQ(ierr); 1968 if (header[0] != MAT_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,0,"not matrix object in file"); 1969 M = header[1]; N = header[2]; nz = header[3]; 1970 1971 if (nz < 0) { 1972 SETERRQ(PETSC_ERR_FILE_UNEXPECTED,1,"Matrix stored in special format on disk, cannot load as SeqAIJ"); 1973 } 1974 1975 /* read in row lengths */ 1976 rowlengths = (int*) PetscMalloc( M*sizeof(int) ); CHKPTRQ(rowlengths); 1977 ierr = PetscBinaryRead(fd,rowlengths,M,PETSC_INT); CHKERRQ(ierr); 1978 1979 /* create our matrix */ 1980 ierr = MatCreateSeqAIJ(comm,M,N,0,rowlengths,A); CHKERRQ(ierr); 1981 B = *A; 1982 a = (Mat_SeqAIJ *) B->data; 1983 shift = a->indexshift; 1984 1985 /* read in column indices and adjust for Fortran indexing*/ 1986 ierr = PetscBinaryRead(fd,a->j,nz,PETSC_INT); CHKERRQ(ierr); 1987 if (shift) { 1988 for ( i=0; i<nz; i++ ) { 1989 a->j[i] += 1; 1990 } 1991 } 1992 1993 /* read in nonzero values */ 1994 ierr = PetscBinaryRead(fd,a->a,nz,PETSC_SCALAR); CHKERRQ(ierr); 1995 1996 /* set matrix "i" values */ 1997 a->i[0] = -shift; 1998 for ( i=1; i<= M; i++ ) { 1999 a->i[i] = a->i[i-1] + rowlengths[i-1]; 2000 a->ilen[i-1] = rowlengths[i-1]; 2001 } 2002 PetscFree(rowlengths); 2003 2004 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 2005 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); 2006 PetscFunctionReturn(0); 2007 } 2008 2009 #undef __FUNC__ 2010 #define __FUNC__ "MatEqual_SeqAIJ" 2011 int MatEqual_SeqAIJ(Mat A,Mat B, PetscTruth* flg) 2012 { 2013 Mat_SeqAIJ *a = (Mat_SeqAIJ *)A->data, *b = (Mat_SeqAIJ *)B->data; 2014 2015 PetscFunctionBegin; 2016 if (B->type !=MATSEQAIJ)SETERRQ(PETSC_ERR_ARG_INCOMP,0,"Matrices must be same type"); 2017 2018 /* If the matrix dimensions are not equal, or no of nonzeros or shift */ 2019 if ((a->m != b->m ) || (a->n !=b->n) ||( a->nz != b->nz)|| 2020 (a->indexshift != b->indexshift)) { 2021 *flg = PETSC_FALSE; PetscFunctionReturn(0); 2022 } 2023 2024 /* if the a->i are the same */ 2025 if (PetscMemcmp(a->i,b->i,(a->m+1)*sizeof(int))) { 2026 *flg = PETSC_FALSE; PetscFunctionReturn(0); 2027 } 2028 2029 /* if a->j are the same */ 2030 if (PetscMemcmp(a->j, b->j, (a->nz)*sizeof(int))) { 2031 *flg = PETSC_FALSE; PetscFunctionReturn(0); 2032 } 2033 2034 /* if a->a are the same */ 2035 if (PetscMemcmp(a->a, b->a, (a->nz)*sizeof(Scalar))) { 2036 *flg = PETSC_FALSE; PetscFunctionReturn(0); 2037 } 2038 *flg = PETSC_TRUE; 2039 PetscFunctionReturn(0); 2040 2041 } 2042