File: infofun.f

package info (click to toggle)
sparskit 2.0.0-4
  • links: PTS, VCS
  • area: main
  • in suites: bullseye
  • size: 4,348 kB
  • sloc: fortran: 15,253; makefile: 260; sh: 136; ansic: 76
file content (800 lines) | stat: -rw-r--r-- 30,960 bytes parent folder | download | duplicates (5)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
c----------------------------------------------------------------------c
c                          S P A R S K I T                             c
c----------------------------------------------------------------------c
c                  INFORMATION ROUTINES. INFO MODULE                   c
c----------------------------------------------------------------------c
c bandwidth :  Computes  ml     = lower_bandwidth(A)                   c
c                        mu     = upper_bandwidth(A)                   c
c                        iband  = max_bandwidth(A)                     c
c                        bndav  = average_bandwidth(A)                 c
c nonz      :  Computes  nzmaxc = max_column_length(A)                 c
c                        nzminc = min_column_length(A)                 c
c                        nzmaxr = max_row_length(A)                    c
c                        nzminr = min_row_length(A)                    c
c                        nzcol  = zero_column_number(A)                c
c                        nzrow  = zero_row_number(A)                   c
c diag_domi :  Computes  ddomc  = diag_domi_column_percentage(A)       c
c                        ddomr  = diag_domi_row_percentage(A)          c
c frobnorm  :  Computes  Fnorm  = Frobenius_norm(A)                    c
c ansym     :  Computes  fas    = sym_part_Frobenius_norm(A)           c
c                        fan    = nonsym_part_Frobenius_norm(A)        c
c                        imatch = matching_elements_number(A)          c
c                        av     = relative_sym_match(A)                c
c distaij   :  Computes  dist   = average_dist_of_a(i,j)(A)            c
c                        std    = standard_deviation(A)                c
c skyline   :  Computes  nsky   = nonzero_number_in_skyline(A)         c
c distdiag  :  Computes  dist   = element_number_in_eachdiag(A)        c
c bandpart  :  Computes  band   = bandwidth_width(A)                   c
c n_imp_diag:  Computes  ndiag  = important_diag_number(A)             c
c nonz_lud  :  Computes  nlower = nonzero_number_of_lower_part(A)      c
c                        nupper = nonzero_number_of_upper_part(A)      c
c                        ndiag  = nonzero_number_of_maindiag(A)        c
c avnz_col  :  Computes  av     = average_nonzero_number_in_column(A)  c
c                        st     = standard_deviation(A)                c
c vbrinfo   :  Print info on matrix in variable block row format       c
c----------------------------------------------------------------------c
      subroutine bandwidth(n,ja, ia, ml, mu, iband, bndav)
      implicit none 
      integer n, ml, mu, iband
      integer ja(*), ia(n+1)
      real*8 bndav
c-----------------------------------------------------------------------
c this routine computes the lower, upper, maximum, and average 
c bandwidths.     revised -- July 12, 2001  -- bug fix -- YS. 
c-----------------------------------------------------------------------
c On Entry:
c----------
c n     = integer. column dimension of matrix
c a     = real array containing the nonzero elements of the matrix
c         the elements are stored by columns in order
c         (i.e. column i comes before column i+1, but the elements
c         within each column can be disordered).
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c
c on return
c----------
c ml    = lower bandwidth as defined by
c        ml = max(i-j | all  a(i,j).ne. 0)
c mu    = upper bandwidth as defined by
c        mu = max ( j-i | all  a(i,j).ne. 0 )
c iband =  maximum bandwidth as defined by
c         iband = Max (  Max [ j | a(i,j) .ne. 0 ] - 
c                        Min [ j | a(i,j) .ne. 0 ] )
c bndav = Average Bandwidth          
c-----------------------------------------------------------------------
c     locals
      integer max 
      integer j0, j1, jminc, jmaxc, i
c-----------------------------------------------------------------------
      ml = -n
      mu = -n
      bndav = 0.0d0
      iband = 0 
      do 10 i=1,n
         j0 = ia(i)
         j1 = ia(i+1) - 1
         jminc = ja(j0)
         jmaxc = ja(j1)
         ml = max(ml,i-jminc)
         mu = max(mu,jmaxc-i)
         iband = max(iband,jmaxc-jminc+1)
         bndav = bndav+real( jmaxc-jminc+1)
 10   continue
      bndav = bndav/real(n)
      return
c-----end-of-bandwidth--------------------------------------------------
c-----------------------------------------------------------------------
      end
c-----------------------------------------------------------------------
      subroutine nonz(n,sym, ja, ia, iao, nzmaxc, nzminc, 
     *     nzmaxr, nzminr, nzcol, nzrow)
      implicit none 
      integer n , nzmaxc, nzminc, nzmaxr, nzminr, nzcol, nzrow
      integer ja(*), ia(n+1), iao(n+1)
      logical sym
c----------------------------------------------------------------------
c     this routine computes maximum numbers of nonzero elements 
c     per column/row, minimum numbers of nonzero elements per column/row, 
c     and  numbers of zero columns/rows.
c----------------------------------------------------------------------
c     On Entry:
c----------
c     n     = integer column dimension of matrix
c     ja    = integer array containing the row indices of elements in a
c     ia    = integer array containing of length n+1 containing the
c     pointers to the beginning of the columns in arrays a and ja.
c     It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c     iao   = similar array for the transpose of the matrix. 
c     sym   = logical variable indicating whether or not the matrix is 
c     stored in symmetric mode.
c     on return
c----------
c     nzmaxc = max length of columns
c     nzminc = min length of columns 
c     nzmaxr = max length of rows
c nzminr = min length of rows
c     nzcol  = number of zero columns
c     nzrow = number of zero rows
c-----------------------------------------------------------------------
      integer  i, j0, j0r, j1r, indiag, k, j1, lenc, lenr 
c
      nzmaxc = 0
      nzminc = n
      nzmaxr = 0
      nzminr = n
      nzcol  = 0
      nzrow  = 0
c-----------------------------------------------------------------------
      do 10 i = 1, n
         j0 = ia(i)
         j1 = ia(i+1) 
         j0r = iao(i)
         j1r = iao(i+1)
         indiag = 0
         do 20 k=j0, j1-1 
            if (ja(k) .eq. i) indiag = 1
 20      continue
c         
         lenc = j1-j0
         lenr = j1r-j0r
c         
         if (sym) lenc = lenc + lenr - indiag
         if (lenc .le. 0) nzcol = nzcol +1
         nzmaxc = max0(nzmaxc,lenc)
         nzminc = min0(nzminc,lenc)
         if (lenr .le. 0) nzrow = nzrow+1
         nzmaxr = max0(nzmaxr,lenr)
         nzminr = min0(nzminr,lenr)
 10   continue
      return
      end
c-----------------------------------------------------------------------
      subroutine diag_domi(n,sym,valued,a, ja,ia,ao,jao, iao, 
     *     ddomc, ddomr)
      implicit none 
      real*8 a(*), ao(*), ddomc, ddomr
      integer n, ja(*), ia(n+1), jao(*), iao(n+1)
      logical sym, valued
c-----------------------------------------------------------------
c     this routine computes the percentage of weakly diagonally 
c     dominant rows/columns
c-----------------------------------------------------------------
c     on entry:
c     ---------
c     n     = integer column dimension of matrix
c a     = real array containing the nonzero elements of the matrix
c     the elements are stored by columns in order
c     (i.e. column i comes before column i+1, but the elements
c     within each column can be disordered).
c     ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c     pointers to the beginning of the columns in arrays a and ja.
c     It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c     ao    = real array containing the nonzero elements of the matrix
c     the elements are stored by rows in order
c     (i.e. row i comes before row i+1, but the elements
c     within each row can be disordered).
c ao,jao, iao, 
c     structure for transpose of a 
c     sym   = logical variable indicating whether or not the matrix is
c     symmetric.
c     valued= logical equal to .true. if values are provided and .false.
c         if only the pattern of the matrix is provided. (in that
c     case a(*) and ao(*) are dummy arrays.
c     
c     ON RETURN
c----------
c     ddomc = percentage of weakly diagonally dominant columns
c     ddomr = percentage of weakly diagonally dominant rows
c-------------------------------------------------------------------
c     locals
      integer i, j0, j1, k, j 
      real*8 aii, dsumr, dsumc 
c     number of diagonally dominant columns
c     real arithmetic used to avoid problems.. YS. 03/27/01 
      ddomc = 0.0  
c     number of diagonally dominant rows
      ddomr = 0.0 
      do 10 i = 1, n
         j0 = ia(i)
         j1 = ia(i+1) - 1
         if (valued) then
            aii = 0.0d0
            dsumc = 0.0d0
            do 20 k=j0,j1
               j = ja(k) 
               if (j .eq. i) then
                  aii = abs(a(k))
               else
                  dsumc = dsumc + abs(a(k))
               endif
 20         continue
            dsumr = 0.0d0
            if (.not. sym) then
               do 30 k=iao(i), iao(i+1)-1
                  if (jao(k) .ne. i) dsumr = dsumr+abs(ao(k))
 30            continue 
            else
               dsumr = dsumc
            endif
            if (dsumc .le. aii) ddomc = ddomc + 1.0
            if (dsumr .le. aii) ddomr = ddomr + 1.0
         endif
 10   continue
      ddomr = ddomr / real(n)
      ddomc = ddomc / real(n)
      return
c-----------------------------------------------------------------------
c--------end-of-diag_moni-----------------------------------------------
      end
c-----------------------------------------------------------------------
      subroutine frobnorm(n,sym,a,ja,ia,Fnorm)
      implicit none 
      integer n 
      real*8 a(*),Fnorm
      integer ja(*),ia(n+1)
      logical sym
c--------------------------------------------------------------------------
c     this routine computes the Frobenius norm of A.
c--------------------------------------------------------------------------
c     on entry:
c-----------
c n      = integer colum dimension of matrix
c a      = real array containing the nonzero elements of the matrix
c          the elements are stored by columns in order
c          (i.e. column i comes before column i+1, but the elements
c          within each column can be disordered).
c ja     = integer array containing the row indices of elements in a.
c ia     = integer array containing of length n+1 containing the
c          pointers to the beginning of the columns in arrays a and 
c          ja. It is assumed that ia(*)= 1 and ia(n+1)  = nnz +1.
c sym    = logical variable indicating whether or not the matrix is
c          symmetric.
c
c on return
c-----------
c Fnorm  = Frobenius norm of A.
c--------------------------------------------------------------------------
      real*8 Fdiag
      integer i, k 
      Fdiag = 0.0
      Fnorm = 0.0
      do i =1,n
         do k = ia(i), ia(i+1)-1
            if (ja(k) .eq. i) then
               Fdiag = Fdiag + a(k)**2
            else
               Fnorm = Fnorm + a(k)**2
            endif
         enddo 
      enddo 
      if (sym) then
         Fnorm = 2*Fnorm +Fdiag
      else
        Fnorm = Fnorm + Fdiag
      endif
      Fnorm = sqrt(Fnorm)
      return
      end
c----------------------------------------------------------------------
      subroutine ansym(n,sym,a,ja,ia,ao,jao,iao,imatch,
     *     av,fas,fan)
c---------------------------------------------------------------------
c     this routine computes the Frobenius norm of the symmetric and
c     non-symmetric parts of A, computes number of matching elements
c     in symmetry and relative symmetry match. 
c---------------------------------------------------------------------
c on entry:
c----------
c n   = integer column dimension of matrix
c a   = real array containing the nonzero elements of the matrix
c       the elements are stored by columns in order
c       (i.e. column i comes before column i+1, but the elements
c       within each column can be disordered).
c ja  = integer array containing the row indices of elements in a
c ia  = integer array containing of length n+1 containing the
c       pointers to the beginning of the columns in arrays a and ja.
c       It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c sym = logical variable indicating whether or not the matrix is
c       symmetric.
c on return
c----------
c fas   = Frobenius norm of symmetric part
c fan   = Frobenius norm of non-symmetric part
c imatch = number of matching elements in symmetry
c av     = relative symmetry match (symmetry = 1)
c ao,jao,iao = transpose of A just as a, ja, ia contains 
c              information of A.
c-----------------------------------------------------------------------
      implicit real*8 (a-h, o-z)
      real*8 a(*),ao(*),fas,fan,av, Fnorm, st
      integer n, ja(*), ia(n+1), jao(*), iao(n+1),imatch
      logical sym
c-----------------------------------------------------------------------
      nnz    = ia(n+1)-ia(1)
      call csrcsc(n,1,1,a,ja,ia,ao,jao,iao)
      if (sym) goto 7
      st     = 0.0d0
      fas    = 0.0d0
      fan    = 0.0d0
      imatch = 0
      do 6 i=1,n
         k1 = ia(i)
         k2 = iao(i)
         k1max = ia(i+1) - 1
         k2max = iao(i+1) - 1
c     
 5       if (k1 .gt. k1max .or. k2 .gt. k2max) goto 6
c     
         j1 = ja(k1)
         j2 = jao(k2)
         if (j1 .ne. j2 ) goto 51
         fas = fas + (a(k1)+ao(k2))**2
         fan = fan + (a(k1)-ao(k2))**2
         st  = st + a(k1)**2
         imatch = imatch + 1
 51      k1 = k1+1
         k2 = k2+1
         if (j1 .lt. j2)  k2 = k2 - 1
         if (j1 .gt. j2)  k1 = k1 - 1
         goto 5
 6    continue
      fas = 0.25D0 * fas
      fan = 0.25D0 * fan
 7    call frobnorm(n,sym,ao,jao,iao,Fnorm)
      if (sym) then
         imatch = nnz
         fas = Fnorm
         fan = 0.0d0
      else
         if (imatch.eq.nnz) then
            st = 0.0D0
         else
            st = 0.5D0 * (Fnorm**2 - st)
            if (st.lt.0.0D0) st = 0.0D0
         endif
         fas = sqrt(fas + st)
         fan = sqrt(fan + st)
      endif
      av = real(imatch)/real(nnz)
      return
      end
c------end-of-ansym-----------------------------------------------------
c-----------------------------------------------------------------------
      subroutine distaij(n,nnz,sym,ja,ia,dist, std)
      implicit real*8 (a-h, o-z)
      real*8  dist, std
      integer ja(*), ia(n+1)
c-----------------------------------------------------------------------
c     this routine computes the average distance of a(i,j) from diag and
c     standard deviation  for this average.
c-----------------------------------------------------------------------
c On entry :
c-----------
c n     = integer. column dimension of matrix
c nnz   = number of nonzero elements of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c sym   = logical variable indicating whether or not the matrix is
c         symmetric.
c on return
c----------
c dist  = average distance of a(i,j) from diag.
c std   = standard deviation for above average.
c-----------------------------------------------------------------------
c
c distance of an element from diagonal.
c
      dist   = 0.0
      std = 0.0
      do 3 i=1,n
         j0 = ia(i)
         j1 = ia(i+1) - 1
         do 31 k=j0, j1
            j=ja(k)
            dist = dist + real(iabs(j-i) )
 31      continue
 3    continue
      dist = dist/real(nnz)
      do 6 i = 1, n 
         do 61 k=ia(i), ia(i+1) - 1
            std=std+(dist-real(iabs(ja(k)-i)))**2
 61      continue
 6    continue
      std = sqrt(std/ real(nnz))
      return
      end
c-----------------------------------------------------------------------
      subroutine skyline(n,sym,ja,ia,jao,iao,nsky)
      implicit real*8 (a-h, o-z)
      integer n, ja(*), ia(n+1), jao(*), iao(n+1)
      integer nskyl, nskyu, nsky
      logical sym
c-------------------------------------------------------------------
c this routine computes the number of nonzeros in the skyline storage.
c-------------------------------------------------------------------
c
c On entry :
c-----------
c n     = integer. column dimension of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c iao   = integer array containing of length n+1 containing the
c         pointers to the beginning of the rows in arrays ao and jao.
c         It is assumed that iao(*) = 1 and iao(n+1) = nzz+1.
c jao   = integer array containing the column indices of elements in ao.
c sym   = logical variable indicating whether or not the matrix is
c         symmetric.
c on return
c----------
c nsky  = number of nonzeros in skyline storage
c-------------------------------------------------------------------
c
c nskyu = skyline storage for upper part
      nskyu = 0
c nskyl = skyline storage for lower part
      nskyl = 0
      do 10 i=1,n
         j0 = ia(i)
         j0r = iao(i)
         
         jminc = ja(j0)
         jminr = jao(j0r)
         if (sym) jminc = jminr
         
         nskyl = nskyl + i-jminr + 1
         nskyu = nskyu + i-jminc + 1
         
 10   continue
      nsky = nskyl+nskyu-n
      if (sym)  nsky = nskyl
      return
      end
c-----------------------------------------------------------------------
      subroutine distdiag(nrow,ncol,ja,ia,dist)
      implicit real*8 (a-h, o-z)
      integer nrow,ncol,ja(*), ia(nrow+1),dist(*)
c----------------------------------------------------------------------
c this routine computes the numbers of elements in each diagonal. 
c----------------------------------------------------------------------
c On entry :
c-----------
c nrow  = integer. row dimension of matrix
c ncol  = integer. column dimension of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c on return
c----------
c dist  = integer array containing the numbers of elements in each of 
c         the nrow+ncol-1 diagonals of A. dist(k) contains the 
c         number of elements in diagonal '-nrow+k'.  k ranges from 
c         1 to (nrow+ncol-1).
c----------------------------------------------------------------------
      nnz  = ia(nrow+1)-ia(1)
      n2   = nrow+ncol-1
      do 8 i=1, n2
         dist(i) = 0
 8    continue
      do 9 i=1, nrow
         k1 = ia(i)
         k2 = ia(i+1) -1
         do 91 k=k1, k2
            j = ja(k)
            dist(nrow+j-i) = dist(nrow+j-i) +1
 91      continue
 9    continue
      return
      end
c-----------------------------------------------------------------------
      subroutine bandpart(n,ja,ia,dist,nper,band)
      implicit real*8 (a-h, o-z)
      integer n,ja(*), ia(n+1),dist(*)
      integer nper,band
c-------------------------------------------------------------------------
c this routine computes the bandwidth of the banded matrix, which contains
c 'nper' percent of the original matrix.
c-------------------------------------------------------------------------
c On entry :
c-----------
c n     = integer. column dimension of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c dist  = integer array containing the numbers of elements in the 
c         matrix with different distance of row indices and column 
c         indices.
c nper  = percentage of matrix  within the bandwidth
c on return
c----------
c band  = the width of the bandwidth
c----------------------------------------------------------------------
      nnz  = ia(n+1)-ia(1)
      iacc  = dist(n)
      band  = 0
      j     = 0
 10   j     = j+1
      iacc  = iacc + dist(n+j) +dist(n-j) 
      if (iacc*100 .le. nnz*nper) then 
         band = band +1 
         goto 10         
      endif 
      return 
      end 
c-----------------------------------------------------------------------      
      subroutine  n_imp_diag(n,nnz,dist, ipar1,ndiag,ioff,dcount)
      implicit real*8 (a-h, o-z)
      real*8  dcount(*)
      integer n,nnz, dist(*), ndiag, ioff(*), ipar1
c-----------------------------------------------------------------------
c     this routine computes the most important diagonals.
c-----------------------------------------------------------------------
c
c On entry :
c-----------
c n     = integer. column dimension of matrix
c nnz   = number of nonzero elements of matrix
c dist  = integer array containing the numbers of elements in the 
c         matrix with different distance of row indices and column 
c         indices. ipar1 = percentage of nonzero  elements of A that 
c         a diagonal should have in order to be an important diagonal 
c on return
c----------
c ndiag = number of the most important diagonals
c ioff  = the offsets with respect to the main diagonal
c dcount= the accumulated percentages
c-----------------------------------------------------------------------
      n2 = n+n-1
      ndiag = 10
      ndiag = min0(n2,ndiag)
      itot  = 0
      ii = 0
      idiag = 0
c     sort diagonals by decreasing order of weights.
 40   jmax = 0
      i    = 1
      do 41 k=1, n2
         j = dist(k)
         if (j .lt. jmax) goto 41
         i = k
         jmax = j
 41   continue
c     permute ----
c     save offsets and accumulated count if diagonal is acceptable
c     (if it has at least ipar1*nnz/100 nonzero elements)
c     quite if no more acceptable diagonals --
c     
      if (jmax*100 .lt. ipar1*nnz) goto 4
      ii = ii+1
      ioff(ii) = i-n
      dist(i)   = - jmax
      itot = itot + jmax
      dcount(ii) = real(100*itot)/real(nnz)
      if (ii .lt. ndiag) goto 40
 4    continue
      ndiag = ii
      return
c-----------------------------------------------------------------------
      end
c-----------------------------------------------------------------------
      subroutine nonz_lud(n,ja,ia,nlower, nupper, ndiag)
      implicit real*8 (a-h, o-z)
      integer n,  ja(*), ia(n+1)
      integer nlower, nupper, ndiag
c-----------------------------------------------------------------------
c this routine computes the number of nonzero elements in strict lower
c part, strict upper part, and main diagonal.
c-----------------------------------------------------------------------
c
c On entry :
c-----------
c n     = integer. column dimension of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c on return
c----------
c nlower= number of nonzero elements in strict lower part
c nupper= number of nonzero elements in strict upper part
c ndiag = number of nonzero elements in main diagonal
c------------------------------------------------------------------- 
c
c number of nonzero elements in upper part
c
      nupper = 0
      ndiag = 0
      
      do 3 i=1,n
c     indiag = nonzero diagonal element indicator
         do 31 k=ia(i), ia(i+1)-1
            j=ja(k)
            if (j .lt. i) nupper = nupper+1
            if (j .eq. i) ndiag = ndiag + 1 
 31      continue
 3    continue
      nlower = ia(n+1)-1-nupper-ndiag
      return
      end
c-----------------------------------------------------------------------
      subroutine avnz_col(n,ja,ia,iao, ndiag, av, st)
      implicit real*8 (a-h, o-z)
      real*8 av, st
      integer n,  ja(*), ia(n+1), iao(n+1)
c---------------------------------------------------------------------
c     this routine computes average number of nonzero elements/column and
c     standard deviation for this average
c---------------------------------------------------------------------
c
c On entry :
c-----------
c n     = integer. column dimension of matrix
c ja    = integer array containing the row indices of elements in a
c ia    = integer array containing of length n+1 containing the
c         pointers to the beginning of the columns in arrays a and ja.
c         It is assumed that ia(*) = 1 and ia(n+1) = nzz+1.
c ndiag = number of the most important diagonals
c On return
c----------
c av    = average number of nonzero elements/column
c st    = standard deviation for this average
c Notes
c---------
c standard deviation will not be correct for symmetric storage. 
c----------------------------------------------------------------------
c     standard deviatioan for the average
      st = 0.0d0
c     average and standard deviation
c     
      av = real(ia(n+1)-1)/real(n)
c     
c     will be corrected later.
c     
      do 3 i=1,n
         j0 = ia(i)
         j1 = ia(i+1) - 1
c     indiag = nonzero diagonal element indicator
         do 31 k=j0, j1
            j=ja(k)
 31      continue
         lenc = j1+1-j0
         st = st + (real(lenc) - av)**2
 3    continue
c     
      st = sqrt( st / real(n) )
      return
      end
c-----------------------------------------------------------------------
c-----------------------------------------------------------------------
      subroutine vbrinfo(nr, nc, kvstr, kvstc, ia, ja, ka, iwk, iout)
c-----------------------------------------------------------------------
      integer nr, nc, kvstr(nr+1), kvstc(nc+1), ia(nr+1), ja(*), ka(*)
      integer iwk(nr+nc+2+nr), iout
c-----------------------------------------------------------------------
c     Print info on matrix in variable block row format.
c-----------------------------------------------------------------------
c     On entry:
c--------------
c     nr,nc   = matrix block row and block column dimension
c     kvstr   = first row number for each block row
c     kvstc   = first column number for each block column
c     ia,ja,ka,a = input matrix in VBR format
c     iout    = unit number for printed output
c
c     On return:
c---------------
c     Printed output to unit number specified in iout.  If a non-square
c     matrix is provided, the analysis will be performed on the block
c     rows, otherwise a row/column conformal partitioning will be used.
c
c     Work space:
c----------------
c     iwk(1:nb+1) = conformal block partitioning
c        (nb is unknown at start but is no more than nr+nc)
c     iwk(nb+2:nb+2+nr) = frequency of each blocksize
c     The workspace is not assumed to be initialized to zero, nor is it
c     left that way.
c
c-----------------------------------------------------------------------
c-----local variables
      integer n, nb, nnz, nnzb, i, j, neq, max, num
      character*101 tmpst
      integer bsiz(10), freq(10)
c-----------------------------------------------------------------------
      n = kvstr(nr+1)-1
      nnz = ka(ia(nr+1)) - ka(1)
      nnzb = ia(nr+1) - ia(1)
      write (iout, 96)
      write (iout, 100) n, nnz, real(nnz)/real(n)
      write (iout, 101) nr, nnzb, real(nnzb)/real(nr)
c-----if non-square matrix, do analysis on block rows,
c     else do analysis on conformal partitioning
      if (kvstr(nr+1) .ne. kvstc(nc+1)) then
         write (iout, 103)
         do i = 1, nr+1
            iwk(i) = kvstr(i)
         enddo
         nb = nr
      else
         call kvstmerge(nr, kvstr, nc, kvstc, nb, iwk)
         if ((nr .ne. nc) .or. (nc .ne. nb)) write (iout, 104) nb
      endif
c-----accumulate frequencies of each blocksize
      max = 1
      iwk(1+nb+2) = 0
      do i = 1, nb
         neq = iwk(i+1) - iwk(i)
         if (neq .gt. max) then
            do j = max+1, neq
               iwk(j+nb+2) = 0
            enddo
            max = neq
         endif
         iwk(neq+nb+2) = iwk(neq+nb+2) + 1
      enddo
c-----store largest 10 of these blocksizes
      num = 0
      do i = max, 1, -1
         if ((iwk(i+nb+2) .ne. 0) .and. (num .lt. 10)) then
            num = num + 1
            bsiz(num) = i
            freq(num) = iwk(i+nb+2)
         endif
      enddo
c-----print information about blocksizes
      write (iout, 109) num
      write (tmpst,'(10i6)') (bsiz(j),j=1,num)
      write (iout,110) num,tmpst
      write (tmpst,'(10i6)') (freq(j),j=1,num)
      write (iout,111) tmpst
      write (iout, 96)
c-----------------------------------------------------------------------
 99    format (2x,38(2h *))
 96    format (6x,' *',65(1h-),'*')
 100   format(
     * 6x,' *  Number of rows                                   = ',
     * i10,'  *'/
     * 6x,' *  Number of nonzero elements                       = ',
     * i10,'  *'/
     * 6x,' *  Average number of nonzero elements/Row           = ',
     * f10.4,'  *')
 101   format(
     * 6x,' *  Number of block rows                             = ',
     * i10,'  *'/
     * 6x,' *  Number of nonzero blocks                         = ',
     * i10,'  *'/
     * 6x,' *  Average number of nonzero blocks/Block row       = ',
     * f10.4,'  *')
 103   format(
     * 6x,' *  Non-square matrix.                                 ',
     *     '            *'/
     * 6x,' *  Performing analysis on block rows.                 ',
     *     '            *')
 104   format(
     * 6x,' *  Non row-column conformal partitioning supplied.    ',
     *     '            *'/
     * 6x,' *  Using conformal partitioning.  Number of bl rows = ',
     * i10,'  *')
 109   format(
     * 6x,' *  Number of different blocksizes                   = ',
     * i10,'  *')
 110   format(6x,' *  The ', i2, ' largest dimension nodes',
     *     ' have dimension    : ',10x,'  *',/,
     * 6x,' *',a61,3x,' *')
 111   format(6x,' *  The frequency of nodes these ',
     *     'dimensions are      : ',10x,'  *',/,
     * 6x,' *',a61,3x,' *')
c---------------------------------
      return
      end
c-----------------------------------------------------------------------
c-----------------------end-of-vbrinfo----------------------------------