aboutsummaryrefslogtreecommitdiffstats
path: root/lib/stdlib/doc/src/sofs.xml
blob: 8c8ae51262435f3b615a885dd33e7bc54f5f9d9b (plain) (blame)
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
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
<?xml version="1.0" encoding="latin1" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">

<erlref>
  <header>
    <copyright>
      <year>2001</year><year>2010</year>
      <holder>Ericsson AB. All Rights Reserved.</holder>
    </copyright>
    <legalnotice>
      The contents of this file are subject to the Erlang Public License,
      Version 1.1, (the "License"); you may not use this file except in
      compliance with the License. You should have received a copy of the
      Erlang Public License along with this software. If not, it can be
      retrieved online at http://www.erlang.org/.

      Software distributed under the License is distributed on an "AS IS"
      basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
      the License for the specific language governing rights and limitations
      under the License.

    </legalnotice>

    <title>sofs</title>
    <prepared>Hans Bolinder</prepared>
    <responsible>nobody</responsible>
    <docno></docno>
    <approved>nobody</approved>
    <checked>no</checked>
    <date>2001-08-25</date>
    <rev>PA1</rev>
    <file>sofs.sgml</file>
  </header>
  <module>sofs</module>
  <modulesummary>Functions for Manipulating Sets of Sets</modulesummary>
  <description>
    <p>The <c>sofs</c> module implements operations on finite sets and
      relations represented as sets. Intuitively, a set is a
      collection of elements; every element belongs to the set, and
      the set contains every element.</p>
    <p>Given a set A and a sentence S(x), where x is a free variable,
      a new set B whose elements are exactly those elements of A for
      which S(x) holds can be formed, this is denoted B&nbsp;=
      {x&nbsp;in&nbsp;A&nbsp;: S(x)}. Sentences are expressed using
      the logical operators "for some" (or "there exists"), "for all",
      "and", "or", "not". If the existence of a set containing all the
      specified elements is known (as will always be the case in this
      module), we write B&nbsp;= {x&nbsp;: S(x)}. </p>
    <p>The <em>unordered set</em> containing the elements a, b and c
      is denoted {a,&nbsp;b,&nbsp;c}. This notation is not to be
      confused with tuples. The <em>ordered pair</em> of a and b, with
      first <em>coordinate</em> a and second coordinate b, is denoted
      (a,&nbsp;b). An ordered pair is an <em>ordered set</em> of two
      elements. In this module ordered sets can contain one, two or
      more elements, and parentheses are used to enclose the elements.
      Unordered sets and ordered sets are orthogonal, again in this
      module; there is no unordered set equal to any ordered set.</p>
    <p>The set that contains no elements is called the <em>empty set</em>.
      If two sets A and B contain the same elements, then A 
      is <marker id="equal"></marker><em>equal</em> to B, denoted
      A&nbsp;=&nbsp;B. Two ordered sets are equal if they contain the
      same number of elements and have equal elements at each
      coordinate. If a set A contains all elements that B contains,
      then B is a <marker id="subset"></marker><em>subset</em> of A. 
      The <marker id="union"></marker><em>union</em> of two sets A and B is 
      the smallest set that contains all elements of A and all elements of
      B. The <marker id="intersection"></marker><em>intersection</em> of two
      sets A and B is the set that contains all elements of A that
      belong to B. 
      Two sets are <marker id="disjoint"></marker><em>disjoint</em> if their
      intersection is the empty set. 
      The <marker id="difference"></marker><em>difference</em> of
      two sets A and B is the set that contains all elements of A that
      do not belong to B. 
      The <marker id="symmetric_difference"></marker><em>symmetric 
      difference</em> of 
      two sets is the set that contains those element that belong to
      either of the two sets, but not both. 
      The <marker id="union_n"></marker><em>union</em> of a collection 
      of sets is the smallest set that contains all the elements that 
      belong to at least one set of the collection. 
      The <marker id="intersection_n"></marker><em>intersection</em> of 
      a non-empty collection of sets is the set that contains all elements
      that belong to every set of the collection.</p>
    <p>The <marker id="Cartesian_product"></marker><em>Cartesian 
      product</em> of
      two sets X and Y, denoted X&nbsp;&times;&nbsp;Y, is the set
      {a&nbsp;: a&nbsp;= (x,&nbsp;y) for some x&nbsp;in&nbsp;X and for
      some y&nbsp;in&nbsp;Y}. 
      A <marker id="relation"></marker><em>relation</em> is a subset of
      X&nbsp;&times;&nbsp;Y. Let R be a relation. The fact that
      (x,&nbsp;y) belongs to R is written as x&nbsp;R&nbsp;y. Since
      relations are sets, the definitions of the last paragraph
      (subset, union, and so on) apply to relations as well. 
      The <marker id="domain"></marker><em>domain</em> of R is the 
      set {x&nbsp;: x&nbsp;R&nbsp;y for some y&nbsp;in&nbsp;Y}. 
      The <marker id="range"></marker><em>range</em> of R is the 
      set {y&nbsp;: x&nbsp;R&nbsp;y for some x&nbsp;in&nbsp;X}. 
      The <marker id="converse"></marker><em>converse</em> of R is the 
      set {a&nbsp;: a&nbsp;= (y,&nbsp;x) for some 
      (x,&nbsp;y)&nbsp;in&nbsp;R}. If A is a subset of X, then 
      the <marker id="image"></marker><em>image</em> of
      A under R is the set {y&nbsp;: x&nbsp;R&nbsp;y for some
      x&nbsp;in&nbsp;A}, and if B is a subset of Y, then 
      the <marker id="inverse_image"></marker><em>inverse image</em> of B is 
      the set {x&nbsp;: x&nbsp;R&nbsp;y for some y&nbsp;in&nbsp;B}. If R is a
      relation from X to Y and S is a relation from Y to Z, then 
      the <marker id="relative_product"></marker><em>relative product</em> of
      R and S is the relation T from X to Z defined so that x&nbsp;T&nbsp;z
      if and only if there exists an element y in Y such that
      x&nbsp;R&nbsp;y and y&nbsp;S&nbsp;z. 
      The <marker id="restriction"></marker><em>restriction</em> of R to A is
      the set S defined so that x&nbsp;S&nbsp;y if and only if there exists an
      element x in A such that x&nbsp;R&nbsp;y. If S is a restriction
      of R to A, then R is 
      an <marker id="extension"></marker><em>extension</em> of S to X. 
      If X&nbsp;=&nbsp;Y then we call R a relation <em>in</em> X. 
      The <marker id="field"></marker><em>field</em> of a relation R in X
      is the union of the domain of R and the range of R. 
      If R is a relation in X, and
      if S is defined so that x&nbsp;S&nbsp;y if x&nbsp;R&nbsp;y and
      not x&nbsp;=&nbsp;y, then S is 
      the <marker id="strict_relation"></marker><em>strict</em> relation 
      corresponding to
      R, and vice versa, if S is a relation in X, and if R is defined
      so that x&nbsp;R&nbsp;y if x&nbsp;S&nbsp;y or x&nbsp;=&nbsp;y,
      then R is the <marker id="weak_relation"></marker><em>weak</em> relation
      corresponding to S. A relation R in X is <em>reflexive</em> if
      x&nbsp;R&nbsp;x for every element x of X; it is
      <em>symmetric</em> if x&nbsp;R&nbsp;y implies that
      y&nbsp;R&nbsp;x; and it is <em>transitive</em> if
      x&nbsp;R&nbsp;y and y&nbsp;R&nbsp;z imply that x&nbsp;R&nbsp;z.</p>
    <p>A <marker id="function"></marker><em>function</em> F is a relation, a
      subset of X&nbsp;&times;&nbsp;Y, such that the domain of F is
      equal to X and such that for every x in X there is a unique
      element y in Y with (x,&nbsp;y) in F. The latter condition can
      be formulated as follows: if x&nbsp;F&nbsp;y and x&nbsp;F&nbsp;z
      then y&nbsp;=&nbsp;z. In this module, it will not be required
      that the domain of F be equal to X for a relation to be
      considered a function. Instead of writing
      (x,&nbsp;y)&nbsp;in&nbsp;F or x&nbsp;F&nbsp;y, we write
      F(x)&nbsp;=&nbsp;y when F is a function, and say that F maps x
      onto y, or that the value of F at x is y. Since functions are
      relations, the definitions of the last paragraph (domain, range,
      and so on) apply to functions as well. If the converse of a
      function F is a function F', then F' is called 
      the <marker id="inverse"></marker><em>inverse</em> of F. 
      The relative product of two functions F1 and F2 is called 
      the <marker id="composite"></marker><em>composite</em> of F1 and F2
      if the range of F1 is a subset of the domain of F2. </p>
    <p>Sometimes, when the range of a function is more important than
      the function itself, the function is called a <em>family</em>.
      The domain of a family is called the <em>index set</em>, and the
      range is called the <em>indexed set</em>. If x is a family from
      I to X, then x[i] denotes the value of the function at index i.
      The notation "a family in X" is used for such a family. When the
      indexed set is a set of subsets of a set X, then we call x 
      a <marker id="family"></marker><em>family of subsets</em> of X. If x
      is a family of subsets of X, then the union of the range of x is
      called the <em>union of the family</em> x. If x is non-empty
      (the index set is non-empty), 
      the <em>intersection of the family</em> x is the intersection of 
      the range of x. In this
      module, the only families that will be considered are families
      of subsets of some set X; in the following the word "family"
      will be used for such families of subsets.</p>
    <p>A <marker id="partition"></marker><em>partition</em> of a set X is a
      collection S of non-empty subsets of X whose union is X and
      whose elements are pairwise disjoint. A relation in a set is an
      <em>equivalence relation</em> if it is reflexive, symmetric and
      transitive. If R is an equivalence relation in X, and x is an
      element of X, 
      the <marker id="equivalence_class"></marker><em>equivalence 
      class</em> of x with respect to R is the set of all those
      elements y of X for which x&nbsp;R&nbsp;y holds. The equivalence
      classes constitute a partitioning of X. Conversely, if C is a
      partition of X, then the relation that holds for any two
      elements of X if they belong to the same equivalence class, is
      an equivalence relation induced by the partition C. If R is an
      equivalence relation in X, then 
      the <marker id="canonical_map"></marker><em>canonical map</em> is
      the function that maps every element of X onto its equivalence class.
      </p>
    <p>Relations as defined above (as sets of ordered pairs) will from
      now on be referred to as <em>binary relations</em>. We call a
      set of ordered sets (x[1],&nbsp;...,&nbsp;x[n]) 
      an <em>(n-ary) relation</em>, and say that the relation is a subset of
      the <marker id="Cartesian_product_tuple"></marker>Cartesian product
      X[1]&nbsp;&times;&nbsp;...&nbsp;&times;&nbsp;X[n] where x[i] is
      an element of X[i], 1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n. 
      The <marker id="projection"></marker><em>projection</em> of an n-ary
      relation R onto coordinate i is the set {x[i]&nbsp;:
      (x[1],&nbsp;...,&nbsp;x[i],&nbsp;...,&nbsp;x[n]) in R for some
      x[j]&nbsp;in&nbsp;X[j], 1&nbsp;&lt;=&nbsp;j&nbsp;&lt;=&nbsp;n
      and not i&nbsp;=&nbsp;j}. The projections of a binary relation R
      onto the first and second coordinates are the domain and the
      range of R respectively. The relative product of binary
      relations can be generalized to n-ary relations as follows. Let
      TR be an ordered set (R[1],&nbsp;...,&nbsp;R[n]) of binary
      relations from X to Y[i] and S a binary relation from
      (Y[1]&nbsp;&times;&nbsp;...&nbsp;&times;&nbsp;Y[n]) to Z. 
      The <marker id="tuple_relative_product"></marker><em>relative 
      product</em> of
      TR and S is the binary relation T from X to Z defined so that
      x&nbsp;T&nbsp;z if and only if there exists an element y[i] in
      Y[i] for each 1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n such that
      x&nbsp;R[i]&nbsp;y[i] and
      (y[1],&nbsp;...,&nbsp;y[n])&nbsp;S&nbsp;z. Now let TR be a an
      ordered set (R[1],&nbsp;...,&nbsp;R[n]) of binary relations from
      X[i] to Y[i] and S a subset of
      X[1]&nbsp;&times;&nbsp;...&nbsp;&times;&nbsp;X[n]. 
      The <marker id="multiple_relative_product"></marker><em>multiple 
      relative product</em> of TR and and S is defined to be the 
      set {z&nbsp;: z&nbsp;= ((x[1],&nbsp;...,&nbsp;x[n]), (y[1],...,y[n]))
      for some (x[1],&nbsp;...,&nbsp;x[n])&nbsp;in&nbsp;S and for some
      (x[i],&nbsp;y[i]) in R[i],
      1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n}. 
      The <marker id="natural_join"></marker><em>natural join</em> of 
      an n-ary relation R
      and an m-ary relation S on coordinate i and j is defined to be
      the set {z&nbsp;: z&nbsp;= (x[1],&nbsp;...,&nbsp;x[n],&nbsp;
      y[1],&nbsp;...,&nbsp;y[j-1],&nbsp;y[j+1],&nbsp;...,&nbsp;y[m])
      for some (x[1],&nbsp;...,&nbsp;x[n])&nbsp;in&nbsp;R and for some
      (y[1],&nbsp;...,&nbsp;y[m])&nbsp;in&nbsp;S such that
      x[i]&nbsp;=&nbsp;y[j]}.</p>
    <p><marker id="sets_definition"></marker>The sets recognized by this 
      module will be represented by elements of the relation Sets, defined as
      the smallest set such that:</p>
    <list type="bulleted">
      <item>for every atom T except '_' and for every term X,
       (T,&nbsp;X) belongs to Sets (<em>atomic sets</em>);
      </item>
      <item>(['_'],&nbsp;[]) belongs to Sets (the <em>untyped empty set</em>);
      </item>
      <item>for every tuple T&nbsp;= {T[1],&nbsp;...,&nbsp;T[n]} and
       for every tuple X&nbsp;= {X[1],&nbsp;...,&nbsp;X[n]}, if
       (T[i],&nbsp;X[i]) belongs to Sets for every
       1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n then (T,&nbsp;X) belongs
       to Sets (<em>ordered sets</em>);
      </item>
      <item>for every term T, if X is the empty list or a non-empty
       sorted list [X[1],&nbsp;...,&nbsp;X[n]] without duplicates
       such that (T,&nbsp;X[i]) belongs to Sets for every
       1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n, then ([T],&nbsp;X)
       belongs to Sets (<em>typed unordered sets</em>).</item>
    </list>
    <p>An <marker id="external_set"></marker><em>external set</em> is an
      element of the range of Sets. 
      A <marker id="type"></marker><em>type</em>
      is an element of the domain of Sets. If S is an element
      (T,&nbsp;X) of Sets, then T is 
      a <marker id="valid_type"></marker><em>valid type</em> of X, 
      T is the type of S, and X is the external set 
      of S. <seealso marker="#from_term">from_term/2</seealso> creates a 
      set from a type and an Erlang term turned into an external set.</p>
    <p>The actual sets represented by Sets are the elements of the
      range of the function Set from Sets to Erlang terms and sets of
      Erlang terms:</p>
    <list type="bulleted">
      <item>Set(T,Term)&nbsp;= Term, where T is an atom;</item>
      <item>Set({T[1],&nbsp;...,&nbsp;T[n]},&nbsp;{X[1],&nbsp;...,&nbsp;X[n]})
       &nbsp;= (Set(T[1],&nbsp;X[1]),&nbsp;...,&nbsp;Set(T[n],&nbsp;X[n]));</item>
      <item>Set([T],&nbsp;[X[1],&nbsp;...,&nbsp;X[n]])
       &nbsp;= {Set(T,&nbsp;X[1]),&nbsp;...,&nbsp;Set(T,&nbsp;X[n])};</item>
      <item>Set([T],&nbsp;[])&nbsp;= {}.</item>
    </list>
    <p>When there is no risk of confusion, elements of Sets will be
      identified with the sets they represent. For instance, if U is
      the result of calling <c>union/2</c> with S1 and S2 as
      arguments, then U is said to be the union of S1 and S2. A more
      precise formulation would be that Set(U) is the union of Set(S1)
      and Set(S2).</p>
    <p>The types are used to implement the various conditions that
      sets need to fulfill. As an example, consider the relative
      product of two sets R and S, and recall that the relative
      product of R and S is defined if R is a binary relation to Y and
      S is a binary relation from Y. The function that implements the relative
      product, <seealso marker="#relprod_impl">relative_product/2</seealso>, checks 
      that the arguments represent binary relations by matching [{A,B}]
      against the type of the first argument (Arg1 say), and [{C,D}]
      against the type of the second argument (Arg2 say). The fact
      that [{A,B}] matches the type of Arg1 is to be interpreted as
      Arg1 representing a binary relation from X to Y, where X is
      defined as all sets Set(x) for some element x in Sets the type
      of which is A, and similarly for Y. In the same way Arg2 is
      interpreted as representing a binary relation from W to Z.
      Finally it is checked that B matches C, which is sufficient to
      ensure that W is equal to Y. The untyped empty set is handled
      separately: its type, ['_'], matches the type of any unordered
      set.</p>
    <p>A few functions of this module (<c>drestriction/3</c>,
      <c>family_projection/2</c>, <c>partition/2</c>,
      <c>partition_family/2</c>, <c>projection/2</c>,
      <c>restriction/3</c>, <c>substitution/2</c>) accept an Erlang
      function as a means to modify each element of a given unordered
      set. Such a function, called SetFun in the following, can be
      specified as a functional object (fun), a tuple
      <c>{external,&nbsp;Fun}</c>, or an integer. If SetFun is
      specified as a fun, the fun is applied to each element of the
      given set and the return value is assumed to be a set. If SetFun
      is specified as a tuple <c>{external, Fun}</c>, Fun is applied
      to the external set of each element of the given set and the
      return value is assumed to be an external set. Selecting the
      elements of an unordered set as external sets and assembling a
      new unordered set from a list of external sets is in the present
      implementation more efficient than modifying each element as a
      set. However, this optimization can only be utilized when the
      elements of the unordered set are atomic or ordered sets. It
      must also be the case that the type of the elements matches some
      clause of Fun (the type of the created set is the result of
      applying Fun to the type of the given set), and that Fun does
      nothing but selecting, duplicating or rearranging parts of the
      elements. Specifying a SetFun as an integer I is equivalent to
      specifying <c>{external, fun(X)&nbsp;-> element(I,&nbsp;X)&nbsp;end}</c>,
      but is to be preferred since it makes it possible to handle this
      case even more efficiently. Examples of SetFuns:</p>
    <pre>
{sofs, union}
fun(S) -> sofs:partition(1, S) end
{external, fun(A) -> A end}
{external, fun({A,_,C}) -> {C,A} end}
{external, fun({_,{_,C}}) -> C end}
{external, fun({_,{_,{_,E}=C}}) -> {E,{E,C}} end}
2</pre>
    <p>The order in which a SetFun is applied to the elements of an
      unordered set is not specified, and may change in future
      versions of sofs.</p>
    <p>The execution time of the functions of this module is dominated
      by the time it takes to sort lists. When no sorting is needed,
      the execution time is in the worst case proportional to the sum
      of the sizes of the input arguments and the returned value. A
      few functions execute in constant time: <c>from_external</c>,
      <c>is_empty_set</c>, <c>is_set</c>, <c>is_sofs_set</c>,
      <c>to_external</c>, <c>type</c>.</p>
    <p>The functions of this module exit the process with a
      <c>badarg</c>, <c>bad_function</c>, or <c>type_mismatch</c>
      message when given badly formed arguments or sets the types of
      which are not compatible.</p>
    <p>When comparing external sets the operator <c>==/2</c> is used.</p>
    <p><em>Types</em></p>
    <pre>
anyset() = -&nbsp;an unordered, ordered or atomic set&nbsp;-
binary_relation() = -&nbsp;a binary relation&nbsp;-
bool() = true | false
external_set() = -&nbsp;an external set&nbsp;-
family() = -&nbsp;a family (of subsets)&nbsp;-
function() = -&nbsp;a function&nbsp;-
ordset() = -&nbsp;an ordered set&nbsp;-
relation() = -&nbsp;an n-ary relation&nbsp;-
set() = -&nbsp;an unordered set&nbsp;-
set_of_sets() = -&nbsp;an unordered set of set()&nbsp;-
set_fun() = integer() >= 1
          | {external, fun(external_set()) -> external_set()}
          | fun(anyset()) -> anyset()
spec_fun() = {external, fun(external_set()) -> bool()}
           | fun(anyset()) -> bool()
type() = -&nbsp;a type&nbsp;- </pre>
  </description>
  <funcs>
    <func>
      <name>a_function(Tuples [, Type]) -> Function</name>
      <fsummary>Create a function.</fsummary>
      <type>
        <v>Function = function()</v>
        <v>Tuples = [tuple()]</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Creates a <seealso marker="#function">function</seealso>.
          <c>a_function(F,&nbsp;T)</c> is equivalent to
          <c>from_term(F,&nbsp;T)</c>, if the result is a function. If
          no <seealso marker="#type">type</seealso> is explicitly
          given, <c>[{atom,&nbsp;atom}]</c> is used as type of the
          function.</p>
      </desc>
    </func>
    <func>
      <name>canonical_relation(SetOfSets) -> BinRel</name>
      <fsummary>Return the canonical map.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>SetOfSets = set_of_sets()</v>
      </type>
      <desc>
        <p>Returns the binary relation containing the elements
          (E,&nbsp;Set) such that Set belongs to SetOfSets and E
          belongs to Set. If SetOfSets is 
	  a <seealso marker="#partition">partition</seealso> of a set X and
	  R is the equivalence relation in X induced by SetOfSets, then the
          returned relation is 
	  the <seealso marker="#canonical_map">canonical map</seealso> from 
	  X onto the equivalence classes with respect to R.</p>
        <pre>
1> <input>Ss = sofs:from_term([[a,b],[b,c]]),</input>
<input>CR = sofs:canonical_relation(Ss),</input>
<input>sofs:to_external(CR).</input>
[{a,[a,b]},{b,[a,b]},{b,[b,c]},{c,[b,c]}]</pre>
      </desc>
    </func>
    <func>
      <name>composite(Function1, Function2) -> Function3</name>
      <fsummary>Return the composite of two functions.</fsummary>
      <type>
        <v>Function1 = Function2 = Function3 = function()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#composite">composite</seealso> of 
	  the functions Function1 and Function2.</p>
        <pre>
1> <input>F1 = sofs:a_function([{a,1},{b,2},{c,2}]),</input>
<input>F2 = sofs:a_function([{1,x},{2,y},{3,z}]),</input>
<input>F = sofs:composite(F1, F2),</input>
<input>sofs:to_external(F).</input>
[{a,x},{b,y},{c,y}]</pre>
      </desc>
    </func>
    <func>
      <name>constant_function(Set, AnySet) -> Function</name>
      <fsummary>Create the function that maps each element of a 
        set onto another set.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>Function = function()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Creates the <seealso marker="#function">function</seealso>
          that maps each element of the set Set onto AnySet.</p>
        <pre>
1> <input>S = sofs:set([a,b]),</input>
<input>E = sofs:from_term(1),</input>
<input>R = sofs:constant_function(S, E),</input>
<input>sofs:to_external(R).</input>
[{a,1},{b,1}]</pre>
      </desc>
    </func>
    <func>
      <name>converse(BinRel1) -> BinRel2</name>
      <fsummary>Return the converse of a binary relation.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#converse">converse</seealso>
          of the binary relation BinRel1.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{2,b},{3,a}]),</input>
<input>R2 = sofs:converse(R1),</input>
<input>sofs:to_external(R2).</input>
[{a,1},{a,3},{b,2}]</pre>
      </desc>
    </func>
    <func>
      <name>difference(Set1, Set2) -> Set3</name>
      <fsummary>Return the difference of two sets.</fsummary>
      <type>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#difference">difference</seealso> of 
	  the sets Set1 and Set2.</p>
      </desc>
    </func>
    <func>
      <name>digraph_to_family(Graph [, Type]) -> Family</name>
      <fsummary>Create a family from a directed graph.</fsummary>
      <type>
        <v>Graph = digraph() -&nbsp;see digraph(3)&nbsp;-</v>
        <v>Family = family()</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Creates a <seealso marker="#family">family</seealso> from
          the directed graph Graph. Each vertex a of Graph is
          represented by a pair (a,&nbsp;{b[1],&nbsp;...,&nbsp;b[n]})
          where the b[i]'s are the out-neighbours of a. If no type is
          explicitly given, [{atom,&nbsp;[atom]}] is used as type of
          the family. It is assumed that Type is 
	  a <seealso marker="#valid_type">valid type</seealso> of the 
	  external set of the family.</p>
        <p>If G is a directed graph, it holds that the vertices and
          edges of G are the same as the vertices and edges of
          <c>family_to_digraph(digraph_to_family(G))</c>.</p>
      </desc>
    </func>
    <func>
      <name>domain(BinRel) -> Set</name>
      <fsummary>Return the domain of a binary relation.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#domain">domain</seealso> of 
	  the binary relation BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{1,a},{1,b},{2,b},{2,c}]),</input>
<input>S = sofs:domain(R),</input>
<input>sofs:to_external(S).</input>
[1,2]</pre>
      </desc>
    </func>
    <func>
      <name>drestriction(BinRel1, Set) -> BinRel2</name>
      <fsummary>Return a restriction of a binary relation.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the difference between the binary relation BinRel1
          and the <seealso marker="#restriction">restriction</seealso>
          of BinRel1 to Set.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{2,b},{3,c}]),</input>
<input>S = sofs:set([2,4,6]),</input>
<input>R2 = sofs:drestriction(R1, S),</input>
<input>sofs:to_external(R2).</input>
[{1,a},{3,c}]</pre>
        <p><c>drestriction(R,&nbsp;S)</c> is equivalent to
          <c>difference(R,&nbsp;restriction(R,&nbsp;S))</c>.</p>
      </desc>
    </func>
    <func>
      <name>drestriction(SetFun, Set1, Set2) -> Set3</name>
      <fsummary>Return a restriction of a relation.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns a subset of Set1 containing those elements that do
          not yield an element in Set2 as the result of applying
          SetFun.</p>
        <pre>
1> <input>SetFun = {external, fun({_A,B,C}) -> {B,C} end},</input>
<input>R1 = sofs:relation([{a,aa,1},{b,bb,2},{c,cc,3}]),</input>
<input>R2 = sofs:relation([{bb,2},{cc,3},{dd,4}]),</input>
<input>R3 = sofs:drestriction(SetFun, R1, R2),</input>
<input>sofs:to_external(R3).</input>
[{a,aa,1}]</pre>
        <p><c>drestriction(F,&nbsp;S1,&nbsp;S2)</c> is equivalent to
          <c>difference(S1,&nbsp;restriction(F,&nbsp;S1,&nbsp;S2))</c>.</p>
      </desc>
    </func>
    <func>
      <name>empty_set() -> Set</name>
      <fsummary>Return the untyped empty set.</fsummary>
      <type>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#sets_definition">untyped empty 
          set</seealso>. <c>empty_set()</c> is equivalent to
          <c>from_term([],&nbsp;['_'])</c>.</p>
      </desc>
    </func>
    <func>
      <name>extension(BinRel1, Set, AnySet) -> BinRel2</name>
      <fsummary>Extend the domain of a binary relation.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#extension">extension</seealso> of 
	  BinRel1 such that
          for each element E in Set that does not belong to the
          <seealso marker="#domain">domain</seealso> of BinRel1,
          BinRel2 contains the pair (E,&nbsp;AnySet).</p>
        <pre>
1> <input>S = sofs:set([b,c]),</input>
<input>A = sofs:empty_set(),</input>
<input>R = sofs:family([{a,[1,2]},{b,[3]}]),</input>
<input>X = sofs:extension(R, S, A),</input>
<input>sofs:to_external(X).</input>
[{a,[1,2]},{b,[3]},{c,[]}]</pre>
      </desc>
    </func>
    <func>
      <name>family(Tuples [, Type]) -> Family</name>
      <fsummary>Create a family of subsets.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>Tuples = [tuple()]</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Creates a <seealso marker="#family">family of subsets</seealso>. 
	  <c>family(F,&nbsp;T)</c> is equivalent to
          <c>from_term(F,&nbsp;T)</c>, if the result is a family. If
          no <seealso marker="#type">type</seealso> is explicitly
          given, <c>[{atom,&nbsp;[atom]}]</c> is used as type of the
          family.</p>
      </desc>
    </func>
    <func>
      <name>family_difference(Family1, Family2) -> Family3</name>
      <fsummary>Return the difference of two families.</fsummary>
      <type>
        <v>Family1 = Family2 = Family3 = family()</v>
      </type>
      <desc>
        <p>If Family1 and Family2 
	  are <seealso marker="#family">families</seealso>, then 
	  Family3 is the family
          such that the index set is equal to the index set of
          Family1, and Family3[i] is the difference between Family1[i]
          and Family2[i] if Family2 maps i, Family1[i] otherwise.</p>
        <pre>
1> <input>F1 = sofs:family([{a,[1,2]},{b,[3,4]}]),</input>
<input>F2 = sofs:family([{b,[4,5]},{c,[6,7]}]),</input>
<input>F3 = sofs:family_difference(F1, F2),</input>
<input>sofs:to_external(F3).</input>
[{a,[1,2]},{b,[3]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_domain(Family1) -> Family2</name>
      <fsummary>Return a family of domains.</fsummary>
      <type>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          and Family1[i] is a binary relation for every i in the index
          set of Family1, then Family2 is the family with the same
          index set as Family1 such that Family2[i] is 
	  the <seealso marker="#domain">domain</seealso> of Family1[i].</p>
        <pre>
1> <input>FR = sofs:from_term([{a,[{1,a},{2,b},{3,c}]},{b,[]},{c,[{4,d},{5,e}]}]),</input>
<input>F = sofs:family_domain(FR),</input>
<input>sofs:to_external(F).</input>
[{a,[1,2,3]},{b,[]},{c,[4,5]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_field(Family1) -> Family2</name>
      <fsummary>Return a family of fields.</fsummary>
      <type>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          and Family1[i] is a binary relation for every i in the index
          set of Family1, then Family2 is the family with the same
          index set as Family1 such that Family2[i] is 
	  the <seealso marker="#field">field</seealso> of Family1[i].</p>
        <pre>
1> <input>FR = sofs:from_term([{a,[{1,a},{2,b},{3,c}]},{b,[]},{c,[{4,d},{5,e}]}]),</input>
<input>F = sofs:family_field(FR),</input>
<input>sofs:to_external(F).</input>
[{a,[1,2,3,a,b,c]},{b,[]},{c,[4,5,d,e]}]</pre>
        <p><c>family_field(Family1)</c> is equivalent to
          <c>family_union(family_domain(Family1), family_range(Family1))</c>.</p>
      </desc>
    </func>
    <func>
      <name>family_intersection(Family1) -> Family2</name>
      <fsummary>Return the intersection of a family
         of sets of sets.</fsummary>
      <type>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          and Family1[i] is a set of sets for every i in the index set
          of Family1, then Family2 is the family with the same index
          set as Family1 such that Family2[i] is 
	  the <seealso marker="#intersection_n">intersection</seealso> of
          Family1[i].</p>
        <p>If Family1[i] is an empty set for some i, then the process
          exits with a <c>badarg</c> message.</p>
        <pre>
1> <input>F1 = sofs:from_term([{a,[[1,2,3],[2,3,4]]},{b,[[x,y,z],[x,y]]}]),</input>
<input>F2 = sofs:family_intersection(F1),</input>
<input>sofs:to_external(F2).</input>
[{a,[2,3]},{b,[x,y]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_intersection(Family1, Family2) -> Family3</name>
      <fsummary>Return the intersection of two families.</fsummary>
      <type>
        <v>Family1 = Family2 = Family3 = family()</v>
      </type>
      <desc>
        <p>If Family1 and Family2 
	  are <seealso marker="#family">families</seealso>, then Family3 
	  is the family such that the index set is the intersection of
          Family1's and Family2's index sets, and Family3[i] is the
          intersection of Family1[i] and Family2[i].</p>
        <pre>
1> <input>F1 = sofs:family([{a,[1,2]},{b,[3,4]},{c,[5,6]}]),</input>
<input>F2 = sofs:family([{b,[4,5]},{c,[7,8]},{d,[9,10]}]),</input>
<input>F3 = sofs:family_intersection(F1, F2),</input>
<input>sofs:to_external(F3).</input>
[{b,[4]},{c,[]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_projection(SetFun, Family1) -> Family2</name>
      <fsummary>Return a family of modified subsets.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Family1 = Family2 = family()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          then Family2 is the family with the same index set as
          Family1 such that Family2[i] is the result of calling SetFun
          with Family1[i] as argument.</p>
        <pre>
1> <input>F1 = sofs:from_term([{a,[[1,2],[2,3]]},{b,[[]]}]),</input>
<input>F2 = sofs:family_projection({sofs, union}, F1),</input>
<input>sofs:to_external(F2).</input>
[{a,[1,2,3]},{b,[]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_range(Family1) -> Family2</name>
      <fsummary>Return a family of ranges.</fsummary>
      <type>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          and Family1[i] is a binary relation for every i in the index
          set of Family1, then Family2 is the family with the same
          index set as Family1 such that Family2[i] is 
	  the <seealso marker="#range">range</seealso> of Family1[i].</p>
        <pre>
1> <input>FR = sofs:from_term([{a,[{1,a},{2,b},{3,c}]},{b,[]},{c,[{4,d},{5,e}]}]),</input>
<input>F = sofs:family_range(FR),</input>
<input>sofs:to_external(F).</input>
[{a,[a,b,c]},{b,[]},{c,[d,e]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_specification(Fun, Family1) -> Family2</name>
      <fsummary>Select a subset of a family using a predicate.</fsummary>
      <type>
        <v>Fun = spec_fun()</v>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>,
          then Family2 is 
	  the <seealso marker="#restriction">restriction</seealso> of 
	  Family1 to those elements i of the
          index set for which Fun applied to Family1[i] returns
          <c>true</c>. If Fun is a tuple <c>{external,&nbsp;Fun2}</c>,
          Fun2 is applied to 
	  the <seealso marker="#external_set">external set</seealso> of 
	  Family1[i], otherwise Fun is applied to Family1[i].</p>
        <pre>
1> <input>F1 = sofs:family([{a,[1,2,3]},{b,[1,2]},{c,[1]}]),</input>
<input>SpecFun = fun(S) -> sofs:no_elements(S) =:= 2 end,</input>
<input>F2 = sofs:family_specification(SpecFun, F1),</input>
<input>sofs:to_external(F2).</input>
[{b,[1,2]}]</pre>
      </desc>
    </func>
    <func>
      <name>family_to_digraph(Family [, GraphType]) -> Graph</name>
      <fsummary>Create a directed graph from a family.</fsummary>
      <type>
        <v>Graph = digraph()</v>
        <v>Family = family()</v>
        <v>GraphType = -&nbsp;see digraph(3)&nbsp;-</v>
      </type>
      <desc>
        <p>Creates a directed graph from 
	  the <seealso marker="#family">family</seealso> Family. For each 
	  pair (a,&nbsp;{b[1],&nbsp;...,&nbsp;b[n]}) of Family, the vertex
          a as well the edges (a,&nbsp;b[i]) for
          1&nbsp;&lt;=&nbsp;i&nbsp;&lt;=&nbsp;n are added to a newly
          created directed graph.</p>
        <p>If no graph type is given, <c>digraph:new/1</c> is used for
          creating the directed graph, otherwise the GraphType
          argument is passed on as second argument to
          <c>digraph:new/2</c>.</p>
        <p>It F is a family, it holds that F is a subset of
          <c>digraph_to_family(family_to_digraph(F),&nbsp;type(F))</c>.
          Equality holds if <c>union_of_family(F)</c> is a subset of
          <c>domain(F)</c>.</p>
        <p>Creating a cycle in an acyclic graph exits the process with
          a <c>cyclic</c> message.</p>
      </desc>
    </func>
    <func>
      <name>family_to_relation(Family) -> BinRel</name>
      <fsummary>Create a binary relation from a family.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>BinRel = binary_relation()</v>
      </type>
      <desc>
        <p>If Family is a <seealso marker="#family">family</seealso>,
          then BinRel is the binary relation containing all pairs
          (i,&nbsp;x) such that i belongs to the index set of Family
          and x belongs to Family[i].</p>
        <pre>
1> <input>F = sofs:family([{a,[]}, {b,[1]}, {c,[2,3]}]),</input>
<input>R = sofs:family_to_relation(F),</input>
<input>sofs:to_external(R).</input>
[{b,1},{c,2},{c,3}]</pre>
      </desc>
    </func>
    <func>
      <name>family_union(Family1) -> Family2</name>
      <fsummary>Return the union of a family of sets of sets.</fsummary>
      <type>
        <v>Family1 = Family2 = family()</v>
      </type>
      <desc>
        <p>If Family1 is a <seealso marker="#family">family</seealso>
          and Family1[i] is a set of sets for each i in the index set
          of Family1, then Family2 is the family with the same index
          set as Family1 such that Family2[i] is 
	  the <seealso marker="#union_n">union</seealso> of Family1[i].</p>
        <pre>
1> <input>F1 = sofs:from_term([{a,[[1,2],[2,3]]},{b,[[]]}]),</input>
<input>F2 = sofs:family_union(F1),</input>
<input>sofs:to_external(F2).</input>
[{a,[1,2,3]},{b,[]}]</pre>
        <p><c>family_union(F)</c> is equivalent to
          <c>family_projection({sofs,union},&nbsp;F)</c>.</p>
      </desc>
    </func>
    <func>
      <name>family_union(Family1, Family2) -> Family3</name>
      <fsummary>Return the union of two families.</fsummary>
      <type>
        <v>Family1 = Family2 = Family3 = family()</v>
      </type>
      <desc>
        <p>If Family1 and Family2 
	  are <seealso marker="#family">families</seealso>, then Family3 
	  is the family such that the index set is the union of Family1's 
	  and Family2's index sets, and Family3[i] is the union of
          Family1[i] and Family2[i] if both maps i, Family1[i] or
          Family2[i] otherwise.</p>
        <pre>
1> <input>F1 = sofs:family([{a,[1,2]},{b,[3,4]},{c,[5,6]}]),</input>
<input>F2 = sofs:family([{b,[4,5]},{c,[7,8]},{d,[9,10]}]),</input>
<input>F3 = sofs:family_union(F1, F2),</input>
<input>sofs:to_external(F3).</input>
[{a,[1,2]},{b,[3,4,5]},{c,[5,6,7,8]},{d,[9,10]}]</pre>
      </desc>
    </func>
    <func>
      <name>field(BinRel) -> Set</name>
      <fsummary>Return the field of a binary relation.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#field">field</seealso> of the
          binary relation BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{1,a},{1,b},{2,b},{2,c}]),</input>
<input>S = sofs:field(R),</input>
<input>sofs:to_external(S).</input>
[1,2,a,b,c]</pre>
        <p><c>field(R)</c> is equivalent 
	  to <c>union(domain(R), range(R))</c>.</p>
      </desc>
    </func>
    <func>
      <name>from_external(ExternalSet, Type) -> AnySet</name>
      <fsummary>Create a set.</fsummary>
      <type>
        <v>ExternalSet = external_set()</v>
        <v>AnySet = anyset()</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Creates a set from the <seealso marker="#external_set">external 
	  set</seealso> ExternalSet
          and the <seealso marker="#type">type</seealso> Type. It is
          assumed that Type is a <seealso marker="#valid_type">valid 
	  type</seealso> of ExternalSet.</p>
      </desc>
    </func>
    <func>
      <name>from_sets(ListOfSets) -> Set</name>
      <fsummary>Create a set out of a list of sets.</fsummary>
      <type>
        <v>Set = set()</v>
        <v>ListOfSets = [anyset()]</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#sets_definition">unordered 
	  set</seealso> containing the sets of the list ListOfSets.</p>
        <pre>
1> <input>S1 = sofs:relation([{a,1},{b,2}]),</input>
<input>S2 = sofs:relation([{x,3},{y,4}]),</input>
<input>S = sofs:from_sets([S1,S2]),</input>
<input>sofs:to_external(S).</input>
[[{a,1},{b,2}],[{x,3},{y,4}]]</pre>
      </desc>
    </func>
    <func>
      <name>from_sets(TupleOfSets) -> Ordset</name>
      <fsummary>Create an ordered set out of a tuple of sets.</fsummary>
      <type>
        <v>Ordset = ordset()</v>
        <v>TupleOfSets = tuple-of(anyset())</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#sets_definition">ordered 
	  set</seealso> containing the sets of the non-empty tuple 
	  TupleOfSets.</p>
      </desc>
    </func>
    <func>
      <name>from_term(Term [, Type]) -> AnySet</name>
      <fsummary>Create a set.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>Term = term()</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p><marker id="from_term"></marker>Creates an element 
	  of <seealso marker="#sets_definition">Sets</seealso> by 
	  traversing the term Term, sorting lists, removing duplicates and
	  deriving or verifying a <seealso marker="#valid_type">valid 
	  type</seealso> for the so obtained external set. An
          explicitly given <seealso marker="#type">type</seealso> Type
          can be used to limit the depth of the traversal; an atomic
          type stops the traversal, as demonstrated by this example
          where "foo" and {"foo"} are left unmodified:</p>
        <pre>
1> <input>S = sofs:from_term([{{"foo"},[1,1]},{"foo",[2,2]}], [{atom,[atom]}]),</input>
<input>sofs:to_external(S).</input>
[{{"foo"},[1]},{"foo",[2]}]</pre>
        <p><c>from_term</c> can be used for creating atomic or ordered
          sets. The only purpose of such a set is that of later
          building unordered sets since all functions in this module
          that <em>do</em> anything operate on unordered sets.
          Creating unordered sets from a collection of ordered sets
          may be the way to go if the ordered sets are big and one
          does not want to waste heap by rebuilding the elements of
          the unordered set. An example showing that a set can be
          built "layer by layer":</p>
        <pre>
1> <input>A = sofs:from_term(a),</input>
<input>S = sofs:set([1,2,3]),</input>
<input>P1 = sofs:from_sets({A,S}),</input>
<input>P2 = sofs:from_term({b,[6,5,4]}),</input>
<input>Ss = sofs:from_sets([P1,P2]),</input>
<input>sofs:to_external(Ss).</input>
[{a,[1,2,3]},{b,[4,5,6]}]</pre>
        <p>Other functions that create sets are <c>from_external/2</c>
          and <c>from_sets/1</c>. Special cases of <c>from_term/2</c>
          are <c>a_function/1,2</c>, <c>empty_set/0</c>,
          <c>family/1,2</c>, <c>relation/1,2</c>, and <c>set/1,2</c>.</p>
      </desc>
    </func>
    <func>
      <name>image(BinRel, Set1) -> Set2</name>
      <fsummary>Return the image of a set under a binary relation.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#image">image</seealso> of the
          set Set1 under the binary relation BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{1,a},{2,b},{2,c},{3,d}]),</input>
<input>S1 = sofs:set([1,2]),</input>
<input>S2 = sofs:image(R, S1),</input>
<input>sofs:to_external(S2).</input>
[a,b,c]</pre>
      </desc>
    </func>
    <func>
      <name>intersection(SetOfSets) -> Set</name>
      <fsummary>Return the intersection of a set of sets.</fsummary>
      <type>
        <v>Set = set()</v>
        <v>SetOfSets = set_of_sets()</v>
      </type>
      <desc>
        <p>Returns 
	  the <seealso marker="#intersection_n">intersection</seealso> of 
	  the set of sets SetOfSets.</p>
        <p>Intersecting an empty set of sets exits the process with a
          <c>badarg</c> message.</p>
      </desc>
    </func>
    <func>
      <name>intersection(Set1, Set2) -> Set3</name>
      <fsummary>Return the intersection of two sets.</fsummary>
      <type>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns 
	  the <seealso marker="#intersection">intersection</seealso> of 
	  Set1 and Set2.</p>
      </desc>
    </func>
    <func>
      <name>intersection_of_family(Family) -> Set</name>
      <fsummary>Return the intersection of a family.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the intersection of 
	  the <seealso marker="#family">family</seealso> Family.</p>
        <p>Intersecting an empty family exits the process with a
          <c>badarg</c> message.</p>
        <pre>
1> <input>F = sofs:family([{a,[0,2,4]},{b,[0,1,2]},{c,[2,3]}]),</input>
<input>S = sofs:intersection_of_family(F),</input>
<input>sofs:to_external(S).</input>
[2]</pre>
      </desc>
    </func>
    <func>
      <name>inverse(Function1) -> Function2</name>
      <fsummary>Return the inverse of a function.</fsummary>
      <type>
        <v>Function1 = Function2 = function()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#inverse">inverse</seealso>
          of the function Function1.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{2,b},{3,c}]),</input>
<input>R2 = sofs:inverse(R1),</input>
<input>sofs:to_external(R2).</input>
[{a,1},{b,2},{c,3}]</pre>
      </desc>
    </func>
    <func>
      <name>inverse_image(BinRel, Set1) -> Set2</name>
      <fsummary>Return the inverse image of a set under 
        a binary relation.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#inverse_image">inverse 
	  image</seealso> of Set1 under the binary relation BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{1,a},{2,b},{2,c},{3,d}]),</input>
<input>S1 = sofs:set([c,d,e]),</input>
<input>S2 = sofs:inverse_image(R, S1),</input>
<input>sofs:to_external(S2).</input>
[2,3]</pre>
      </desc>
    </func>
    <func>
      <name>is_a_function(BinRel) -> Bool</name>
      <fsummary>Test for a function.</fsummary>
      <type>
        <v>Bool = bool()</v>
        <v>BinRel = binary_relation()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if the binary relation BinRel is a
          <seealso marker="#function">function</seealso> or the
          untyped empty set, <c>false</c> otherwise.</p>
      </desc>
    </func>
    <func>
      <name>is_disjoint(Set1, Set2) -> Bool</name>
      <fsummary>Test for disjoint sets.</fsummary>
      <type>
        <v>Bool = bool()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if Set1 and Set2 
	  are <seealso marker="#disjoint">disjoint</seealso>, <c>false</c>
          otherwise.</p>
      </desc>
    </func>
    <func>
      <name>is_empty_set(AnySet) -> Bool</name>
      <fsummary>Test for an empty set.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>Bool = bool()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if Set is an empty unordered set,
          <c>false</c> otherwise.</p>
      </desc>
    </func>
    <func>
      <name>is_equal(AnySet1, AnySet2) -> Bool</name>
      <fsummary>Test two sets for equality.</fsummary>
      <type>
        <v>AnySet1 = AnySet2 = anyset()</v>
        <v>Bool = bool()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if the AnySet1 and AnySet2 
	  are <seealso marker="#equal">equal</seealso>, <c>false</c> 
	  otherwise. This example shows that <c>==/2</c> is used when
          comparing sets for equality:</p>
        <pre>
1> <input>S1 = sofs:set([1.0]),</input>
<input>S2 = sofs:set([1]),</input>
<input>sofs:is_equal(S1, S2).</input>
true</pre>
      </desc>
    </func>
    <func>
      <name>is_set(AnySet) -> Bool</name>
      <fsummary>Test for an unordered set.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>Bool = bool()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if AnySet is 
	  an <seealso marker="#sets_definition">unordered set</seealso>, and
          <c>false</c> if AnySet is an ordered set or an atomic set.</p>
      </desc>
    </func>
    <func>
      <name>is_sofs_set(Term) -> Bool</name>
      <fsummary>Test for an unordered set.</fsummary>
      <type>
        <v>Bool = bool()</v>
        <v>Term = term()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if Term is 
          an <seealso marker="#sets_definition">unordered set</seealso>, an
          ordered set or an atomic set, <c>false</c> otherwise.</p>
      </desc>
    </func>
    <func>
      <name>is_subset(Set1, Set2) -> Bool</name>
      <fsummary>Test two sets for subset.</fsummary>
      <type>
        <v>Bool = bool()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if Set1 is 
	  a <seealso marker="#subset">subset</seealso> of Set2, <c>false</c>
          otherwise.</p>
      </desc>
    </func>
    <func>
      <name>is_type(Term) -> Bool</name>
      <fsummary>Test for a type.</fsummary>
      <type>
        <v>Bool = bool()</v>
        <v>Term = term()</v>
      </type>
      <desc>
        <p>Returns <c>true</c> if the term Term is 
	  a <seealso marker="#type">type</seealso>.</p>
      </desc>
    </func>
    <func>
      <name>join(Relation1, I, Relation2, J) -> Relation3</name>
      <fsummary>Return the join of two relations.</fsummary>
      <type>
        <v>Relation1 = Relation2 = Relation3 = relation()</v>
        <v>I = J = integer() > 0</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#natural_join">natural 
	  join</seealso> of the relations Relation1 and Relation2 on
          coordinates I and J.</p>
        <pre>
1> <input>R1 = sofs:relation([{a,x,1},{b,y,2}]),</input>
<input>R2 = sofs:relation([{1,f,g},{1,h,i},{2,3,4}]),</input>
<input>J = sofs:join(R1, 3, R2, 1),</input>
<input>sofs:to_external(J).</input>
[{a,x,1,f,g},{a,x,1,h,i},{b,y,2,3,4}]</pre>
      </desc>
    </func>
    <func>
      <name>multiple_relative_product(TupleOfBinRels, BinRel1) -> BinRel2</name>
      <fsummary>Return the multiple relative product of a tuple of binary 
        relations and a relation.</fsummary>
      <type>
        <v>TupleOfBinRels = tuple-of(BinRel)</v>
        <v>BinRel = BinRel1 = BinRel2 = binary_relation()</v>
      </type>
      <desc>
        <p>If TupleOfBinRels is a non-empty tuple
          {R[1],&nbsp;...,&nbsp;R[n]} of binary relations and BinRel1
          is a binary relation, then BinRel2 is 
	  the <seealso marker="#multiple_relative_product">multiple relative 
	  product</seealso> of the ordered set
          (R[i],&nbsp;...,&nbsp;R[n]) and BinRel1.</p>
        <pre>
1> <input>Ri = sofs:relation([{a,1},{b,2},{c,3}]),</input>
<input>R = sofs:relation([{a,b},{b,c},{c,a}]),</input>
<input>MP = sofs:multiple_relative_product({Ri, Ri}, R),</input>
<input>sofs:to_external(sofs:range(MP)).</input>
[{1,2},{2,3},{3,1}]</pre>
      </desc>
    </func>
    <func>
      <name>no_elements(ASet) -> NoElements</name>
      <fsummary>Return the number of elements of a set.</fsummary>
      <type>
        <v>ASet = set() | ordset()</v>
        <v>NoElements = integer() >= 0 </v>
      </type>
      <desc>
        <p>Returns the number of elements of the ordered or unordered
          set ASet.</p>
      </desc>
    </func>
    <func>
      <name>partition(SetOfSets) -> Partition</name>
      <fsummary>Return the coarsest partition given a set of sets.</fsummary>
      <type>
        <v>SetOfSets = set_of_sets()</v>
        <v>Partition = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#partition">partition</seealso> of 
	  the union of the set of sets SetOfSets such that two
	  elements are considered equal if they belong to the same
	  elements of SetOfSets.</p>
        <pre>
1> <input>Sets1 = sofs:from_term([[a,b,c],[d,e,f],[g,h,i]]),</input>
<input>Sets2 = sofs:from_term([[b,c,d],[e,f,g],[h,i,j]]),</input>
<input>P = sofs:partition(sofs:union(Sets1, Sets2)),</input>
<input>sofs:to_external(P).</input>
[[a],[b,c],[d],[e,f],[g],[h,i],[j]]</pre>
      </desc>
    </func>
    <func>
      <name>partition(SetFun, Set) -> Partition</name>
      <fsummary>Return a partition of a set.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Partition = set()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#partition">partition</seealso> of 
          Set such that two elements are considered equal if the
          results of applying SetFun are equal.</p>
        <pre>
1> <input>Ss = sofs:from_term([[a],[b],[c,d],[e,f]]),</input>
<input>SetFun = fun(S) -> sofs:from_term(sofs:no_elements(S)) end,</input>
<input>P = sofs:partition(SetFun, Ss),</input>
<input>sofs:to_external(P).</input>
[[[a],[b]],[[c,d],[e,f]]]</pre>
      </desc>
    </func>
    <func>
      <name>partition(SetFun, Set1, Set2) -> {Set3, Set4}</name>
      <fsummary>Return a partition of a set.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Set1 = Set2 = Set3 = Set4 = set()</v>
      </type>
      <desc>
        <p>Returns a pair of sets that, regarded as constituting a
          set, forms a <seealso marker="#partition">partition</seealso> of 
	  Set1. If the
          result of applying SetFun to an element of Set1 yields an
          element in Set2, the element belongs to Set3, otherwise the
          element belongs to Set4.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{2,b},{3,c}]),</input>
<input>S = sofs:set([2,4,6]),</input>
<input>{R2,R3} = sofs:partition(1, R1, S),</input>
<input>{sofs:to_external(R2),sofs:to_external(R3)}.</input>
{[{2,b}],[{1,a},{3,c}]}</pre>
        <p><c>partition(F,&nbsp;S1,&nbsp;S2)</c> is equivalent to
          <c>{restriction(F,&nbsp;S1,&nbsp;S2), 
          drestriction(F,&nbsp;S1,&nbsp;S2)}</c>.</p>
      </desc>
    </func>
    <func>
      <name>partition_family(SetFun, Set) -> Family</name>
      <fsummary>Return a family indexing a partition.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>SetFun = set_fun()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#family">family</seealso>
          Family where the indexed set is 
	  a <seealso marker="#partition">partition</seealso> of Set 
	  such that two elements are considered equal if the results
          of applying SetFun are the same value i. This i is the index
          that Family maps onto 
	  the <seealso marker="#equivalence_class">equivalence 
	  class</seealso>.</p>
        <pre>
1> <input>S = sofs:relation([{a,a,a,a},{a,a,b,b},{a,b,b,b}]),</input>
<input>SetFun = {external, fun({A,_,C,_}) -> {A,C} end},</input>
<input>F = sofs:partition_family(SetFun, S),</input>
<input>sofs:to_external(F).</input>
[{{a,a},[{a,a,a,a}]},{{a,b},[{a,a,b,b},{a,b,b,b}]}]</pre>
      </desc>
    </func>
    <func>
      <name>product(TupleOfSets) -> Relation</name>
      <fsummary>Return the Cartesian product of a tuple of sets.</fsummary>
      <type>
        <v>Relation = relation()</v>
        <v>TupleOfSets = tuple-of(set())</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#Cartesian_product_tuple">Cartesian 
	  product</seealso> of the non-empty tuple of sets
          TupleOfSets. If (x[1],&nbsp;...,&nbsp;x[n]) is an element of
          the n-ary relation Relation, then x[i] is drawn from element
          i of TupleOfSets.</p>
        <pre>
1> <input>S1 = sofs:set([a,b]),</input>
<input>S2 = sofs:set([1,2]),</input>
<input>S3 = sofs:set([x,y]),</input>
<input>P3 = sofs:product({S1,S2,S3}),</input>
<input>sofs:to_external(P3).</input>
[{a,1,x},{a,1,y},{a,2,x},{a,2,y},{b,1,x},{b,1,y},{b,2,x},{b,2,y}]</pre>
      </desc>
    </func>
    <func>
      <name>product(Set1, Set2) -> BinRel</name>
      <fsummary>Return the Cartesian product of two sets.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#Cartesian_product">Cartesian 
          product</seealso> of Set1 and Set2.</p>
        <pre>
1> <input>S1 = sofs:set([1,2]),</input>
<input>S2 = sofs:set([a,b]),</input>
<input>R = sofs:product(S1, S2),</input>
<input>sofs:to_external(R).</input>
[{1,a},{1,b},{2,a},{2,b}]</pre>
        <p><c>product(S1,&nbsp;S2)</c> is equivalent to
          <c>product({S1,&nbsp;S2})</c>.</p>
      </desc>
    </func>
    <func>
      <name>projection(SetFun, Set1) -> Set2</name>
      <fsummary>Return a set of substituted elements.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns the set created by substituting each element of
          Set1 by the result of applying SetFun to the element.</p>
        <p>If SetFun is a number i&nbsp;&gt;=&nbsp;1 and Set1 is a
          relation, then the returned set is 
	  the <seealso marker="#projection">projection</seealso> of Set1 
	  onto coordinate i.</p>
        <pre>
1> <input>S1 = sofs:from_term([{1,a},{2,b},{3,a}]),</input>
<input>S2 = sofs:projection(2, S1),</input>
<input>sofs:to_external(S2).</input>
[a,b]</pre>
      </desc>
    </func>
    <func>
      <name>range(BinRel) -> Set</name>
      <fsummary>Return the range of a binary relation.</fsummary>
      <type>
        <v>BinRel = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#range">range</seealso> of the
          binary relation BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{1,a},{1,b},{2,b},{2,c}]),</input>
<input>S = sofs:range(R),</input>
<input>sofs:to_external(S).</input>
[a,b,c]</pre>
      </desc>
    </func>
    <func>
      <name>relation(Tuples [, Type]) -> Relation</name>
      <fsummary>Create a relation.</fsummary>
      <type>
        <v>N = integer()</v>
        <v>Type = N | type()</v>
        <v>Relation = relation()</v>
        <v>Tuples = [tuple()]</v>
      </type>
      <desc>
        <p>Creates a <seealso marker="#relation">relation</seealso>.
          <c>relation(R,&nbsp;T)</c> is equivalent to
          <c>from_term(R,&nbsp;T)</c>, if T is 
	  a <seealso marker="#type">type</seealso> and the result is a
	  relation. If Type is an integer N, then
          <c>[{atom,&nbsp;...,&nbsp;atom}])</c>, where the size of the
          tuple is N, is used as type of the relation. If no type is
          explicitly given, the size of the first tuple of Tuples is
          used if there is such a tuple. <c>relation([])</c> is
          equivalent to <c>relation([],&nbsp;2)</c>.</p>
      </desc>
    </func>
    <func>
      <name>relation_to_family(BinRel) -> Family</name>
      <fsummary>Create a family from a binary relation.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>BinRel = binary_relation()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#family">family</seealso>
          Family such that the index set is equal to 
	  the <seealso marker="#domain">domain</seealso> of the binary
	  relation BinRel, and Family[i] is 
	  the <seealso marker="#image">image</seealso> of the set of i 
	  under BinRel.</p>
        <pre>
1> <input>R = sofs:relation([{b,1},{c,2},{c,3}]),</input>
<input>F = sofs:relation_to_family(R),</input>
<input>sofs:to_external(F).</input>
[{b,[1]},{c,[2,3]}]</pre>
      </desc>
    </func>
    <func>
      <name>relative_product(TupleOfBinRels [, BinRel1]) -> BinRel2</name>
      <fsummary>Return the relative product of a tuple of binary relations
         and a binary relation.</fsummary>
      <type>
        <v>TupleOfBinRels = tuple-of(BinRel)</v>
        <v>BinRel = BinRel1 = BinRel2 = binary_relation()</v>
      </type>
      <desc>
        <p>If TupleOfBinRels is a non-empty tuple
          {R[1],&nbsp;...,&nbsp;R[n]} of binary relations and BinRel1
          is a binary relation, then BinRel2 is 
	  the <seealso marker="#tuple_relative_product">relative 
	  product</seealso> of the ordered set (R[i],&nbsp;...,&nbsp;R[n]) 
	  and BinRel1.</p>
        <p>If BinRel1 is omitted, the relation of equality between the
          elements of 
	  the <seealso marker="#Cartesian_product_tuple">Cartesian 
	  product</seealso> of the ranges of R[i],
          range&nbsp;R[1]&nbsp;&times;&nbsp;...&nbsp;&times;&nbsp;range&nbsp;R[n],
          is used instead (intuitively, nothing is "lost").</p>
        <pre>
1> <input>TR = sofs:relation([{1,a},{1,aa},{2,b}]),</input>
<input>R1 = sofs:relation([{1,u},{2,v},{3,c}]),</input>
<input>R2 = sofs:relative_product({TR, R1}),</input>
<input>sofs:to_external(R2).</input>
[{1,{a,u}},{1,{aa,u}},{2,{b,v}}]</pre>
        <p>Note that <c>relative_product({R1},&nbsp;R2)</c> is
          different from <c>relative_product(R1,&nbsp;R2)</c>; the
          tuple of one element is not identified with the element
          itself.</p>
      </desc>
    </func>
    <func>
      <name>relative_product(BinRel1, BinRel2) -> BinRel3</name>
      <fsummary>Return the relative product of 
        two binary relations.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = BinRel3 = binary_relation()</v>
      </type>
      <desc>
        <p><marker id="relprod_impl"></marker>Returns
          the <seealso marker="#relative_product">relative 
	  product</seealso> of the binary relations BinRel1 and BinRel2.</p>
      </desc>
    </func>
    <func>
      <name>relative_product1(BinRel1, BinRel2) -> BinRel3</name>
      <fsummary>Return the relative_product of 
        two binary relations.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = BinRel3 = binary_relation()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#relative_product">relative 
	  product</seealso> of 
	  the <seealso marker="#converse">converse</seealso> of the
	  binary relation BinRel1 and the binary relation BinRel2.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{1,aa},{2,b}]),</input>
<input>R2 = sofs:relation([{1,u},{2,v},{3,c}]),</input>
<input>R3 = sofs:relative_product1(R1, R2),</input>
<input>sofs:to_external(R3).</input>
[{a,u},{aa,u},{b,v}]</pre>
        <p><c>relative_product1(R1,&nbsp;R2)</c> is equivalent to
          <c>relative_product(converse(R1),&nbsp;R2)</c>.</p>
      </desc>
    </func>
    <func>
      <name>restriction(BinRel1, Set) -> BinRel2</name>
      <fsummary>Return a restriction of a binary relation.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#restriction">restriction</seealso> of
          the binary relation BinRel1 to Set.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,a},{2,b},{3,c}]),</input>
<input>S = sofs:set([1,2,4]),</input>
<input>R2 = sofs:restriction(R1, S),</input>
<input>sofs:to_external(R2).</input>
[{1,a},{2,b}]</pre>
      </desc>
    </func>
    <func>
      <name>restriction(SetFun, Set1, Set2) -> Set3</name>
      <fsummary>Return a restriction of a set.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns a subset of Set1 containing those elements that
          yield an element in Set2 as the result of applying SetFun.</p>
        <pre>
1> <input>S1 = sofs:relation([{1,a},{2,b},{3,c}]),</input>
<input>S2 = sofs:set([b,c,d]),</input>
<input>S3 = sofs:restriction(2, S1, S2),</input>
<input>sofs:to_external(S3).</input>
[{2,b},{3,c}]</pre>
      </desc>
    </func>
    <func>
      <name>set(Terms [, Type]) -> Set</name>
      <fsummary>Create a set of atoms or any type of sets.</fsummary>
      <type>
        <v>Set = set()</v>
        <v>Terms = [term()]</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Creates an <seealso marker="#sets_definition">unordered 
	  set</seealso>. <c>set(L,&nbsp;T)</c> is equivalent to
          <c>from_term(L,&nbsp;T)</c>, if the result is an unordered
          set. If no <seealso marker="#type">type</seealso> is
          explicitly given, <c>[atom]</c> is used as type of the set.</p>
      </desc>
    </func>
    <func>
      <name>specification(Fun, Set1) -> Set2</name>
      <fsummary>Select a subset using a predicate.</fsummary>
      <type>
        <v>Fun = spec_fun()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns the set containing every element of Set1 for which
          Fun returns <c>true</c>. If Fun is a tuple
          <c>{external,&nbsp;Fun2}</c>, Fun2 is applied to the
          <seealso marker="#external_set">external set</seealso> of
          each element, otherwise Fun is applied to each element.</p>
        <pre>
1> <input>R1 = sofs:relation([{a,1},{b,2}]),</input>
<input>R2 = sofs:relation([{x,1},{x,2},{y,3}]),</input>
<input>S1 = sofs:from_sets([R1,R2]),</input>
<input>S2 = sofs:specification({sofs,is_a_function}, S1),</input>
<input>sofs:to_external(S2).</input>
[[{a,1},{b,2}]]</pre>
      </desc>
    </func>
    <func>
      <name>strict_relation(BinRel1) -> BinRel2</name>
      <fsummary>Return the strict relation corresponding to 
        a given relation.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#strict_relation">strict 
          relation</seealso> corresponding to the binary relation BinRel1.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,1},{1,2},{2,1},{2,2}]),</input>
<input>R2 = sofs:strict_relation(R1),</input>
<input>sofs:to_external(R2).</input>
[{1,2},{2,1}]</pre>
      </desc>
    </func>
    <func>
      <name>substitution(SetFun, Set1) -> Set2</name>
      <fsummary>Return a function with a given set as domain.</fsummary>
      <type>
        <v>SetFun = set_fun()</v>
        <v>Set1 = Set2 = set()</v>
      </type>
      <desc>
        <p>Returns a function, the domain of which is Set1. The value
          of an element of the domain is the result of applying SetFun
          to the element.</p>
        <pre>
1> <input>L = [{a,1},{b,2}].</input>
[{a,1},{b,2}]
2> <input>sofs:to_external(sofs:projection(1,sofs:relation(L))).</input>
[a,b]
3> <input>sofs:to_external(sofs:substitution(1,sofs:relation(L))).</input>
[{{a,1},a},{{b,2},b}]
4> <input>SetFun = {external, fun({A,_}=E) -> {E,A} end},</input>
<input>sofs:to_external(sofs:projection(SetFun,sofs:relation(L))).</input>
[{{a,1},a},{{b,2},b}]</pre>
        <p>The relation of equality between the elements of {a,b,c}:</p>
        <pre>
1> <input>I = sofs:substitution(fun(A) -> A end, sofs:set([a,b,c])),</input>
<input>sofs:to_external(I).</input>
[{a,a},{b,b},{c,c}]</pre>
        <p>Let SetOfSets be a set of sets and BinRel a binary
          relation. The function that maps each element Set of
          SetOfSets onto the <seealso marker="#image">image</seealso>
          of Set under BinRel is returned by this function:</p>
        <pre>
images(SetOfSets, BinRel) ->
   Fun = fun(Set) -> sofs:image(BinRel, Set) end,
   sofs:substitution(Fun, SetOfSets).</pre>
        <p>Here might be the place to reveal something that was more
          or less stated before, namely that external unordered sets
          are represented as sorted lists. As a consequence, creating
          the image of a set under a relation R may traverse all
          elements of R (to that comes the sorting of results, the
          image). In <c>images/2</c>, BinRel will be traversed once
          for each element of SetOfSets, which may take too long. The
          following efficient function could be used instead under the
          assumption that the image of each element of SetOfSets under
          BinRel is non-empty:</p>
        <pre>
images2(SetOfSets, BinRel) ->
   CR = sofs:canonical_relation(SetOfSets),
   R = sofs:relative_product1(CR, BinRel),
   sofs:relation_to_family(R).</pre>
      </desc>
    </func>
    <func>
      <name>symdiff(Set1, Set2) -> Set3</name>
      <fsummary>Return the symmetric difference of two sets.</fsummary>
      <type>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#symmetric_difference">symmetric 
          difference</seealso> (or the Boolean sum) of Set1 and Set2.</p>
        <pre>
1> <input>S1 = sofs:set([1,2,3]),</input>
<input>S2 = sofs:set([2,3,4]),</input>
<input>P = sofs:symdiff(S1, S2),</input>
<input>sofs:to_external(P).</input>
[1,4]</pre>
      </desc>
    </func>
    <func>
      <name>symmetric_partition(Set1, Set2) -> {Set3, Set4, Set5}</name>
      <fsummary>Return a partition of two sets.</fsummary>
      <type>
        <v>Set1 = Set2 = Set3 = Set4 = Set5 = set()</v>
      </type>
      <desc>
        <p>Returns a triple of sets: Set3 contains the elements
          of Set1 that do not belong to Set2; Set4 contains the
          elements of Set1 that belong to Set2; Set5 contains the
          elements of Set2 that do not belong to Set1.</p>
      </desc>
    </func>
    <func>
      <name>to_external(AnySet) -> ExternalSet</name>
      <fsummary>Return the elements of a set.</fsummary>
      <type>
        <v>ExternalSet = external_set()</v>
        <v>AnySet = anyset()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#external_set">external 
          set</seealso> of an atomic, ordered or unordered set.</p>
      </desc>
    </func>
    <func>
      <name>to_sets(ASet) -> Sets</name>
      <fsummary>Return a list or a tuple of the elements of set.</fsummary>
      <type>
        <v>ASet = set() | ordset()</v>
        <v>Sets = tuple_of(AnySet) | [AnySet]</v>
      </type>
      <desc>
        <p>Returns the elements of the ordered set ASet as a tuple of
          sets, and the elements of the unordered set ASet as a sorted
          list of sets without duplicates.</p>
      </desc>
    </func>
    <func>
      <name>type(AnySet) -> Type</name>
      <fsummary>Return the type of a set.</fsummary>
      <type>
        <v>AnySet = anyset()</v>
        <v>Type = type()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#type">type</seealso> of an
          atomic, ordered or unordered set.</p>
      </desc>
    </func>
    <func>
      <name>union(SetOfSets) -> Set</name>
      <fsummary>Return the union of a set of sets.</fsummary>
      <type>
        <v>Set = set()</v>
        <v>SetOfSets = set_of_sets()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#union_n">union</seealso> of the
          set of sets SetOfSets.</p>
      </desc>
    </func>
    <func>
      <name>union(Set1, Set2) -> Set3</name>
      <fsummary>Return the union of two sets.</fsummary>
      <type>
        <v>Set1 = Set2 = Set3 = set()</v>
      </type>
      <desc>
        <p>Returns the <seealso marker="#union">union</seealso> of
          Set1 and Set2.</p>
      </desc>
    </func>
    <func>
      <name>union_of_family(Family) -> Set</name>
      <fsummary>Return the union of a family.</fsummary>
      <type>
        <v>Family = family()</v>
        <v>Set = set()</v>
      </type>
      <desc>
        <p>Returns the union of 
          the <seealso marker="#family">family</seealso> Family.</p>
        <pre>
1> <input>F = sofs:family([{a,[0,2,4]},{b,[0,1,2]},{c,[2,3]}]),</input>
<input>S = sofs:union_of_family(F),</input>
<input>sofs:to_external(S).</input>
[0,1,2,3,4]</pre>
      </desc>
    </func>
    <func>
      <name>weak_relation(BinRel1) -> BinRel2</name>
      <fsummary>Return the weak relation corresponding to 
        a given relation.</fsummary>
      <type>
        <v>BinRel1 = BinRel2 = binary_relation()</v>
      </type>
      <desc>
        <p>Returns a subset S of the <seealso marker="#weak_relation">weak
          relation</seealso> W
          corresponding to the binary relation BinRel1. Let F be the
          <seealso marker="#field">field</seealso> of BinRel1. The
          subset S is defined so that x S y if x W y for some x in F
          and for some y in F.</p>
        <pre>
1> <input>R1 = sofs:relation([{1,1},{1,2},{3,1}]),</input>
<input>R2 = sofs:weak_relation(R1),</input>
<input>sofs:to_external(R2).</input>
[{1,1},{1,2},{2,2},{3,1},{3,3}]</pre>
      </desc>
    </func>
  </funcs>

  <section>
    <title>See Also</title>
    <p><seealso marker="dict">dict(3)</seealso>, 
      <seealso marker="digraph">digraph(3)</seealso>,
      <seealso marker="orddict">orddict(3)</seealso>, 
      <seealso marker="ordsets">ordsets(3)</seealso>, 
      <seealso marker="sets">sets(3)</seealso></p>
  </section>
</erlref>