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
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
|
<!doctype HTML public "-//W3C//DTD HTML 4.0 Frameset//EN">
<html>
<head>
<title>Chapter 7: HDF5 Dataspaces and Partial I/O</title>
<!--(Meta)==========================================================-->
<!--(Links)=========================================================-->
<!--( Begin styles definition )=====================================-->
<link href="ed_styles/NewUGelect.css" rel="stylesheet" type="text/css">
<!--( End styles definition )=======================================-->
</head>
<body>
<!-- #BeginLibraryItem "/ed_libs/Copyright.lbi" -->
<!--
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Copyright by The HDF Group. *
* Copyright by the Board of Trustees of the University of Illinois. *
* All rights reserved. *
* *
* This file is part of HDF5. The full HDF5 copyright notice, including *
* terms governing use, modification, and redistribution, is contained in *
* the files COPYING and Copyright.html. COPYING can be found at the root *
* of the source code distribution tree; Copyright.html can be found at the *
* root level of an installed copy of the electronic HDF5 document set and *
* is linked from the top-level documents page. It can also be found at *
* http://www.hdfgroup.org/HDF5/doc/Copyright.html. If you do not have *
* access to either file, you may request a copy from help@hdfgroup.org. *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
-->
<!-- #EndLibraryItem --><!-- HEADER LEFT "HDF5 User's Guide" -->
<!-- HEADER RIGHT "HDF5 Dataspaces and Partial I/O" -->
<!--( TOC )=========================================================-->
<SCRIPT language="JavaScript">
<!--
document.writeln ('\
<table x-use-null-cells\
align="right"\
width="240"\
cellspacing="0"\
class="tocTable">\
<tr valign="top"> \
<td class="tocTableHeaderCell" colspan="2"> \
<span class="TableHead">Chapter Contents</span></td>\
</tr>\
-->
<!-- Table Version 3 -->\
<!--
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#Intro">1.</a></td>\
<td class="tocTableContentCell3">\
<a href="#Intro">Introduction</a></td> \
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#DSpaceFunctSums">2.</a></td>\
<td class="tocTableContentCell3">\
<a href="#DSpaceFunctSums">Dataspace (H5S) Function Summaries</a></td>\
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#DefDataObjs">3.1</a></td>\
<td class="tocTableContentCell3">\
<a href="#DefDataObjs">Dataspace Objects</a></td> \
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#ProgModel">3.2</a></td>\
<td class="tocTableContentCell3">\
<a href="#ProgModel">Programming Model</a></td>\
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#DTransfer">4.</a></td>\
<td class="tocTableContentCell3">\
<a href="#DTransfer">Dataspaces and Data Transfer</a></td> \
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#DSelectTransfer">5.</a></td>\
<td class="tocTableContentCell3">\
<a href="#DSelectTransfer">Selection Operations and Data Transfer</a></td> \
</tr>\
<tr valign="top"> \
<td class="tocTableContentCell2"> \
<a href="#DRegions">6.</a></td>\
<td class="tocTableContentCell3">\
<a href="#DRegions">References to Dataset Regions</a>\
</td>\
</tr>\
\
<tr valign="top"> \
<td class="tocTableContentCell"> \
-->
<!-- editingComment -- "tocTableContentCell" and "tocTableContentCell4" \
-->\
<!-- are the table-closing cell class.\
<td class="tocTableContentCell2"> \
-->\
<!--
<a href="#Programs">7.</a></td>\
<td class="tocTableContentCell4">\
<a href="#Programs">Sample Programs</a>\
</td></tr>\
</table>\
')
-->
</SCRIPT>
<!--(End TOC)=======================================================-->
<div align="center">
<a name="TOP">
<h2>Chapter 7<br /><font size="6">HDF5 Dataspaces and Partial I/O</font></h2>
</a>
</div>
<!--
<dir>
<h1 class="editingcomment" align="center">- - - DRAFT - - -</h1>
</dir>
-->
<a name="Intro">
<h3>7.1. Introduction</h3>
</a>
<p>The HDF5 <em>dataspace</em> is a required component of an HDF5 dataset
or attribute definition. The dataspace defines the size and shape of the
dataset or attribute raw data. In other words,
a dataspace defines the number of dimensions and the size of each dimension
of the multidimensional array in which the raw data is represented.
The dataspace must be defined when the dataset or attribute is created.</p>
<p>The <em>dataspace</em> is also used during dataset I/O operations,
defining the elements of the dataset that participate in the I/O operation.</p>
<p>This chapter explains the <em>dataspace</em> object and
its use in dataset and attribute creation and data transfer.
It also describes selection operations on a dataspace used to
implement sub-setting, sub-sampling, and scatter-gather access to datasets.
</p>
<p>The rest of this chapter is structured as follows:</p>
<ul>
<li>Section 2, “Dataspace Function Summaries,”
provides a categorized list of dataspace functions,
also known as the H5S APIs</li>
<li>Section 3, “Definition of Dataspace Objects and
the Dataspace Programming Model,”
describes dataspace objects and the programming model,
including the creation and use of dataspaces</li>
<li>Section 4, “Dataspaces and Data Transfer,”
describes the use of dataspaces in data transfer</li>
<li>Section 5, “Dataspace Selection Operations and Data
Transfer,” describes selection operations on dataspaces
and their usage in data transfer</li>
<li>Section 6, “References to Dataset Regions,”
briefly discusses references to dataset regions</li>
<li>Section 7, “Sample Programs,”
contains the full programs from which several of the code samples
in this chapter were derived</li>
</ul>
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="DSpaceFunctSums">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<br />
<!-- NEW PAGE -->
<a name="DSpaceFunctSums">
<h3 class="pagebefore">7.2. Dataspace (H5S) Function Summaries</h3>
</a>
<p>This section provides a reference list of dataspace functions,
the H5S APIs, with brief descriptions.
The functions are presented in the following catagories:</p>
<ul>
<li>Dataspace management functions</li>
<li>Dataspace query functions</li>
<li>Dataspace selection functions: hyperslabs</li>
<li>Dataspace selection functions: points</li>
</ul>
<p>The rest of the chapter will provide examples and explanations
of how to use these functions.</p>
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="3" align="left" valign="bottom">
<b>Function Listing 1.
Dataspace management functions</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td>
<b>C Function<br />Fortran Function</b>
</td><td> </td>
<td>
<b>Purpose</b>
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Screate<br />h5screate_f</code>
</td><td> </td>
<td>
Creates a new dataspace of a specified type.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Scopy<br />h5scopy_f</code>
</td><td> </td>
<td>
Creates an exact copy of a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sclose<br />h5sclose_f</code>
</td><td> </td>
<td>
Releases and terminates access to a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sdecode<br />h5sdecode_f</code>
</td><td> </td>
<td>
Decode a binary object description of a dataspace and return a new
object identifier.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sencode<br />h5sencode</code>
</td><td> </td>
<td>
Encode a dataspace object description into a binary buffer.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Screate_simple<br />h5screate_simple_f</code>
</td><td> </td>
<td>
Creates a new simple dataspace and opens it for access.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sis_simple<br />h5sis_simple_f</code>
</td><td> </td>
<td>
Determines whether a dataspace is a simple dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sextent_copy<br />h5sextent_copy_f</code>
</td><td> </td>
<td>
Copies the extent of a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sextent_equal<br />h5sextent_equal_f</code>
</td><td> </td>
<td>
Determines whether two dataspace extents are equal.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sset_extent_simple<br />h5sset_extent_simple_f</code>
</td><td> </td>
<td>
Sets or resets the size of an existing dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sset_extent_none<br />h5sset_extent_none_f</code>
</td><td> </td>
<td>
Removes the extent from a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
</table>
<br />
<br />
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="3" align="left" valign="bottom">
<b>Function Listing 2. Dataspace query functions</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td>
<b>C Function<br />Fortran Function</b>
</td><td> </td>
<td>
<b>Purpose</b>
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_simple_extent_dims<br />h5sget_simple_extent_dims_f</code>
</td><td> </td>
<td>
Retrieves dataspace dimension size and maximum size.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_simple_extent_ndims<br />h5sget_simple_extent_ndims_f</code>
</td><td> </td>
<td>
Determines the dimensionality of a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_simple_extent_npoints<br />
h5sget_simple_extent_npoints_f</code>
</td><td> </td>
<td>
Determines the number of elements in a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_simple_extent_type<br />h5sget_simple_extent_type_f</code>
</td><td> </td>
<td>
Determine the current class of a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
</table>
<br />
<br />
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="3" align="left" valign="bottom">
<b>Function Listing 3. Dataspace selection functions: hyperslabs
</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td>
<b>C Function<br />Fortran Function</b>
</td><td> </td>
<td>
<b>Purpose</b>
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Soffset_simple<br />h5soffset_simple_f</code>
</td><td> </td>
<td>
Sets the offset of a simple dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_type<br />h5sget_select_type_f</code>
</td><td> </td>
<td>
Determines the type of the dataspace selection.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_hyper_nblocks<br />
h5sget_select_hyper_nblocks_f</code>
</td><td> </td>
<td>
Get number of hyperslab blocks.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_hyper_blocklist<br />
h5sget_select_hyper_blocklist_f</code>
</td><td> </td>
<td>
Gets the list of hyperslab blocks currently selected.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_bounds<br />h5sget_select_bounds_f</code>
</td><td> </td>
<td>
Gets the bounding box containing the current selection.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sselect_all<br />h5sselect_all_f</code>
</td><td> </td>
<td>
Selects the entire dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sselect_none<br />h5sselect_none_f</code>
</td><td> </td>
<td>
Resets the selection region to include no elements.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sselect_valid<br />h5sselect_valid_f</code>
</td><td> </td>
<td>
Verifies that the selection is within the extent of the dataspace.
</td>
</tr>
<!-- NEW PAGE -->
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sselect_hyperslab<br />h5sselect_hyperslab_f</code>
</td><td> </td>
<td>
Selects a hyperslab region to add to the current selected region.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
</table>
<br />
<br />
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="3" align="left" valign="bottom">
<b>Function Listing 4. Dataspace selection functions: points
</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td>
<b>C Function<br />Fortran Function</b>
</td><td> </td>
<td>
<b>Purpose</b>
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_npoints<br />h5sget_select_npoints_f</code>
</td><td> </td>
<td>
Determines the number of elements in a dataspace selection.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_elem_npoints<br />h5sget_select_elem_npoints_f</code>
</td><td> </td>
<td>
Gets the number of element points in the current selection.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sget_select_elem_pointlist<br />
h5sget_select_elem_pointlist_f</code>
</td><td> </td>
<td>
Gets the list of element points currently selected.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>
<code>H5Sselect_elements<br />h5sselect_elements_f</code>
</td><td> </td>
<td>
Selects array elements to be included in the selection for a dataspace.
</td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
</table>
<br />
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="DefDataObjs">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<br />
<!-- NEW PAGE -->
<a name="DefDataObjs">
<h3 class="pagebefore">7.3. Definition of Dataspace Objects and
the Dataspace Programming Model</h3></a>
<p>This section introduces the notion of the HDF5 dataspace object
and a programming model for creating and working with dataspaces.</p>
<h4>7.3.1. Dataspace Objects</h4>
<p>An HDF5 dataspace is a required component of an HDF5 dataset or attribute.
A dataspace defines the size and the shape of a dataset’s or an
attribute’s raw data.
Currently, HDF5 supports the following types of the dataspaces: </p>
<ul>
<li>Scalar dataspaces</li>
<li>Simple dataspaces</li>
<li>Null dataspaces</li>
</ul>
<p>A <em>scalar dataspace</em>, <code>H5S_SCALAR</code>,
represents just one element, a scalar.
Note that the datatype of this one element may be very complex,
e.g., a compound structure with members being of
any allowed HDF5 datatype, including
multidimensional arrays, strings, and nested compound structures.
By convention, the rank of a scalar dataspace is always
<code>0</code> (zero); think of it geometrically as a single,
dimensionless point, though that point may be complex.</p>
<p>A <i>simple dataspace</i>, <code>H5S_SIMPLE</code>,
is a multidimensional array of elements.
The dimensionality of the dataspace (or the rank of the array)
is fixed and is defined at creation time.
The size of each dimension can grow during the life time of the dataspace
from the <i>current size</i> up to the <i>maximum size</i>.
Both the current size and the maximum size are specified at
creation time.
The sizes of dimensions at any particular time in the life of a dataspace
are called the <i>current dimensions</i>, or the <i>dataspace extent</i>.
They can be queried along with the maximum sizes.</p>
<p>A <em>null dataspace</em>, <code>H5S_NULL</code>,
contains no data elements.
Note that no selections can be applied to a null dataset
as there is nothing to select.</p>
<p>As shown in the UML diagram in the figure below, an HDF5 simple
dataspace object has three attributes: the rank or number of dimensions;
the current sizes, expressed as an array of length
<span class="codeVar">rank</span> with each element of the array
denoting the current size of the corresponding dimension; and the
maximum sizes, expressed as an array of length
<span class="codeVar">rank</span> with each element of the array
denoting the maximum size of the corresponding dimension. </p>
<table width="400" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<table border="1">
<tr><td align="center">
<code>Simple dataspace</code>
</td></tr>
<tr><td align="left">
<code>
rank:int<br />
current_size:hsize_t[rank] <br />
maximum_size:hsize_t[rank]</code>
</td></tr>
<tr><td align="left">
</td></tr>
</table>
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 1. A simple dataspace</b> <br />
A simple dataspace is defined by its rank,
the current size of each dimension, and
the maximum size of each dimension.
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The size of a current dimension cannot be greater than
the maximum size, which can be unlimited,
specified as <code>H5S_UNLIMITED</code>.
Note that while the HDF5 file format and library impose no maximum size
on an unlimited dimension, practically speaking its size
will always be limited to the biggest integer available on the
particular system being used. </p>
<!-- editingComment
<span class="editingComment">[ [ [ Prior excessively casual phrasing replaced (...the caveat that the value of infinity is limited to...). ] ] ]</span>)
-->
<p>Dataspace rank is restricted to 32,
the standard limit in C on the rank of an array,
in the current implementation of the HDF5 Library.
The HDF5 file format, on the other hand, allows any rank up to the
maximum integer value on the system, so the library restriction can be
raised in the future if higher dimensionality is required.</p>
<p>Note that most of the time Fortran applications calling HDF5 will
work with dataspaces of rank less than or equal to seven,
since seven is the maximum number of dimensions in a Fortran array.
But dataspace rank is not limited to seven for Fortran applications.</p>
<!-- editingComment
<span class="editingComment">[ [ [ or "But with the use of XXX, Fortran applications can [easily?] work with dataspace rank of up to 32." ] ] ]</span>)
-->
<p>The current dimensions of a dataspace, also referred to as the
dataspace extent, define the bounding box for dataset elements that
can participate in I/O operations.</p>
<a name="ProgModel">
<h4>7.3.2. Programming Model</h4>
</a>
<p>
The programming model for creating and working with HDF5 dataspaces
can be summarized as follows:</p>
<ol><li>Create a dataspace</li>
<li>Use the dataspace to create a dataset in the file or
to describe a data array in memory</li>
<li>Modify the dataspace to define dataset elements that will
participate in I/O operations</li>
<li>Use the modified dataspace while reading/writing dataset raw
data or to create a region reference</li>
<li>Close the dataspace when no longer needed</li>
</ol>
<p>The rest of this section will address
steps 1, 2, and 5 of the programming model;
steps 3 and 4 will be discussed in later sections of this chapter.</p>
<h4>7.3.2.1. Creating a Dataspace</h4>
<p>A dataspace can be created by calling the
<span class="codeText">H5Screate</span> function
(<span class="codeText">h5screate_f</span> in Fortran).
Since the definition of a simple dataspace requires the specification of
dimensionality (or rank) and initial and maximum dimension sizes,
the HDF5 Library provides a <i>convenience</i> API,
<span class="codeText">H5Screate_simple</span>
(<span class="codeText">h5screate_simple_f</span>)
to create a simple dataspace in one step.</p>
<p>The following examples illustrate the usage of these APIs.</p>
<h4>7.3.2.2. Creating a Scalar Dataspace</h4>
<p>A scalar dataspace is created with the <code>H5Screate</code>
or the <code>h5screate_f</code> function.</p>
<p>In C:</p>
<pre>
hid_t space_id;
. . .
space_id = H5Screate(H5S_SCALAR);
</pre>
<p>In Fortran:</p>
<pre>
INTEGER(HID_T) :: space_id
. . .
CALL h5screate_f(H5S_SCALAR_F, space_id, error)
</pre>
<p>As mentioned above, the dataspace will contain only one element.
Scalar dataspaces are used more often for describing attributes
that have just one value, e.g. the attribute
<span class="codeText">temperature</span> with the value
<span class="codeText">celsius</span> is used to indicate that the
dataset with this attribute stores temperature values using the celsius scale.</p>
<h4>7.3.2.3. Creating a Null Dataspace</h4>
<p>A null dataspace is created with the <code>H5Screate</code>
or the <code>h5screate_f</code> function.</p>
<p>In C:</p>
<pre>
hid_t space_id;
. . .
space_id = H5Screate(H5S_NULL);
</pre>
<p>In Fortran: </p>
(<code>H5S_NULL</code> not yet implemented in Fortran.)
<pre>
INTEGER(HID_T) :: space_id
. . .
CALL h5screate_f(H5S_NULL_F, space_id, error)
</pre>
As mentioned above, the dataspace will contain no elements.
<!--
NEED MORE INFO.
SPECIFICALLY, HOW ARE SUCH DATASPACES USED?
AND WHAT ATTRIBUTES ARE RELEVANT?
-->
<h4>7.3.2.4. Creating a Simple Dataspace</h4>
<p>Let’s assume that an application wants to store a
two-dimensional array of data, A(20,100).
During the life of the application, the first dimension of the array
can grow up to 30; there is no restriction on the size of the
second dimension.
The following steps are used to declare a dataspace for the dataset
in which the array data will be stored.</p>
<p>In C:</p>
<pre>
hid_t space_id;
int rank = 2;
hsize_t current_dims[2] = {20, 100};
hsize_t max_dims[2] = {30, H5S_UNLIMITED};
. . .
space_id = H5Screate(H5S_SIMPLE);
H5Sset_extent_simple(space_id,rank,current_dims,max_dims);
</pre>
<p>In Fortran:</p>
<pre>
INTEGER(HID_T) :: space_id
INTEGER :: rank = 2
INTEGER(HSIZE_T) :: current dims = /( 20, 100)/
INTEGER(HSIZE_T) :: max_dims = /(30, H5S_UNLIMITED_F)/
INTEGER error
. . .
CALL h5screate_f(H5S_SIMPLE_F, space_id, error)
CALL h5sset_extent_simple_f(space_id, rank, current_dims, max_dims, error)
</pre>
<p>Alternatively, the convenience APIs
<span class="codeText">H5Screate_simple</span>/<span class="codeText">h5screate_simple_f</span>
can replace the <span class="codeText">H5Screate</span>/<span class="codeText">h5screate_f</span>
and <span class="codeText">H5Sset_extent_simple</span>/<span class="codeText">h5sset_extent_simple_f</span>
calls.</p>
<p>In C:</p>
<pre>
space_id = H5Screate_simple(rank, current_dims, max_dims);
</pre>
<p>In Fortran:</p>
<pre>
CALL h5screate_simple_f(rank, current_dims, space_id, error, max_dims)
</pre>
<p>In this example, a dataspace with current dimensions of 20 by 100
is created. The first dimension can be extended only up to 30. The
second dimension, however, is declared unlimited; it can be extended
up to the largest available integer value on the system.</p>
<p>Note that when there is a difference between the current dimensions
and the maximum dimensions of an array, then chunking storage must be
used. In other words, if the number of dimensions may change over the
life of the dataset, then chunking must be used. If the array dimensions
are fixed (if the number of current dimensions is equal to the maximum
number of dimensions when the dataset is created), then contiguous
storage can be used. See the “
<a href="10_Datasets.html#DTransfer">Data Transfer</a>”
section in the “<a href="UG_frame10Datasets.html" target=_top>
Datasets</a>” chapter.</p>
<p>Maximum dimensions can be the same as current dimensions.
In such a case, the sizes of dimensions cannot be changed during the life
of the dataspace object.
In C, <span class="codeText">NULL</span> can be used to indicate to
the <span class="codeText">H5Screate_simple</span> and
<span class="codeText">H5Sset_extent_simple</span> functions that the
maximum sizes of all dimensions are the same as the current sizes.
In Fortran, the maximum size parameter is optional for
<span class="codeText">h5screate_simple_f</span> and can be omitted
when the sizes are the same.</p>
<p>In C:</p>
<pre>
space_id = H5Screate_simple(rank, current_dims, NULL);
</pre>
<p>In Fortran:</p>
<pre>
CALL h5screate_f(rank, current_dims, space_id, error)
</pre>
<p>The created dataspace will have current and maximum dimensions
of 20 and 100 correspondingly,
and the sizes of those dimensions cannot be changed.</p>
<h4>7.3.2.5. C versus Fortran Dataspaces</h4>
<p>Dataspace dimensions are numbered from 1 to
<span class="codeVar">rank</span>. HDF5 uses C storage conventions,
assuming that the last listed dimension is the fastest-changing dimension
and the first-listed dimension is the slowest changing.
<!-- editingComment
<span class="editingComment">[ [ [ Fortran, on the other hand, .... ? ] ] ]</span>
-->
The HDF5 file format storage layout specification adheres to the C convention
and the HDF5 Library adheres to the same convention
when storing dataspace dimensions in the file.
This affects how C programs and tools interpret data written
from Fortran programs and vice versa.
The example below illustrates the issue.</p>
<p>When a Fortran application describes a dataspace to store an array
as A(20,100), it specifies the value of the first dimension to be 20
and the second to be 100.
Since Fortran stores data by columns,
the first-listed dimension with the value 20 is the fastest-changing
dimension and the last-listed dimension with the value 100 is the
slowest-changing.
In order to adhere to the HDF5 storage convention,
the HDF5 Fortran wrapper transposes dimensions,
so the first dimension becomes the last.
The dataspace dimensions stored in the file will be 100,20 instead
of 20,100 in order to correctly describe the Fortran data that is
stored in 100 columns, each containing 20 elements.</p>
<p>When a Fortran application reads the data back,
the HDF5 Fortran wrapper transposes the dimensions once more,
returning the first dimension to be 20 and the second to be 100,
describing correctly the sizes of the array that should be used
to read data in the Fortran array A(20,100).</p>
<p>When a C application reads data back,
the dimensions will come out as 100 and 20,
correctly describing the size of the array to read data into,
since the data was written as 100 records of 20 elements each.
Therefore C tools such as <span class="codeText">h5dump</span>
and <span class="codeText">h5ls</span> always display
transposed dimensions and values for the data written
by a Fortran application. </p>
<p>Consider the following simple example of equivalent
C 3 x 5 and Fortran 5 x 3 arrays.
As illustrated in the figure <!-- formerly Figure 3 -->below,
a C applications will store
a 3 x 5 2-dimensional array as three 5-element rows.
In order to store the same data in the same order,
a Fortran application must view the array as as a 5 x 3 array with
three 5-element columns.
The dataspace of this dataset, as written from Fortran,
will therefore be described as 5 x 3 in the application
but stored and described in the file according to the C convention
as a 3 x 5 array.
This ensures that C and Fortran applications will always read
the data in the order in which it was written.
The HDF5 Fortran interface handles this transposition automatically.</p>
<p>In C (from <a href="#h5_write_c"><code>h5_write.c</code></a>):</p>
<pre>
#define NX 3 /* dataset dimensions */
#define NY 5
. . .
int data[NX][NY]; /* data to write */
. . .
/*
* Data and output buffer initialization.
*/
for (j = 0; j < NX; j++) {
for (i = 0; i < NY; i++)
data[j][i] = i + 1 + j*NY;
}
/*
* 1 2 3 4 5
* 6 7 8 9 10
* 11 12 13 14 15
*/
. . .
dims[0] = NX;
dims[1] = NY;
dataspace = H5Screate_simple(RANK, dims, NULL);
</pre>
<!-- NEW PAGE -->
<p>In Fortran (from <a href="#h5_write_f90"><code>h5_write.f90</code>)</a>:</p>
<pre>
INTEGER, PARAMETER :: NX = 3
INTEGER, PARAMETER :: NY = 5
. . .
INTEGER(HSIZE_T), DIMENSION(2) :: dims = (/3,5/) ! Dataset dimensions
---
INTEGER :: data(NX,NY)
. . .
!
! Initialize data
!
do i = 1, NX
do j = 1, NY
data(i,j) = j + (i-1)*NY
enddo
enddo
!
! Data
!
! 1 2 3 4 5
! 6 7 8 9 10
! 11 12 13 14 15
. . .
CALL h5screate_simple_f(rank, dims, dspace_id, error)
</pre>
<p>In Fortran (from <a href=#h5_write_tr_f90>
<code>h5_write_tr.f90</code>):</a></p>
<pre>
INTEGER, PARAMETER :: NX = 3
INTEGER, PARAMETER :: NY = 5
. . .
INTEGER(HSIZE_T), DIMENSION(2) :: dims = (/NY, NX/) ! Dataset dimensions
. . .
!
! Initialize data
!
do i = 1, NY
do j = 1, NX
data(i,j) = i + (j-1)*NY
enddo
enddo
!
! Data
!
! 1 6 11
! 2 7 12
! 3 8 13
! 4 9 14
! 5 10 15
. . .
CALL h5screate_simple_f(rank, dims, dspace_id, error)
</pre>
<br />
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<br />
<table border="0" width="95%">
<tr align="left" valign="top">
<td width="35%">
<p>A dataset stored by a<br />C program in a 3 x 5 array:</p>
<table border="1" width="100%">
<tr align="center">
<td width="20%">1</td>
<td width="20%">2</td>
<td width="20%">3</td>
<td width="20%">4</td>
<td width="20%">5</td>
</tr>
<tr align="center">
<td width="20%">6</td>
<td width="20%">7</td>
<td width="20%">8</td>
<td width="20%">9</td>
<td width="20%">10</td>
</tr>
<tr align="center">
<td width="20%">11</td>
<td width="20%">12</td>
<td width="20%">13</td>
<td width="20%">14</td>
<td width="20%">15</td>
</tr>
</table>
</td>
<td width="15%"> </td>
<td width="45%">
<p>The same dataset stored by a<br />Fortran program in a 5 x 3 array:</p>
<table border="1" width="48%">
<tr align="center">
<td width="33%">1</td>
<td width="34%">6</td>
<td width="33%">11</td>
</tr>
<tr align="center">
<td width="33%">2</td>
<td width="34%">7</td>
<td width="33%">12</td>
</tr>
<tr align="center">
<td width="33%">3</td>
<td width="34%">8</td>
<td width="33%">13</td>
</tr>
<tr align="center">
<td width="33%">4</td>
<td width="34%">9</td>
<td width="33%">14</td>
</tr>
<tr align="center">
<td width="33%">5</td>
<td width="34%">10</td>
<td width="33%">15</td>
</tr>
</table>
</td>
<td width="5%"> </td>
</tr>
</table>
<table width="95%" align="center">
<tr>
<td align="left"><br />
The left-hand dataset above as written to an HDF5
file from C or the right-hand dataset as written from Fortran:
</td>
</tr>
</table>
<br />
<table border="1" width="95%">
<tr>
<td align="center" width="6.5%">1</td>
<td align="center" width="6.5%">2</td>
<td align="center" width="6.5%">3</td>
<td align="center" width="6.5%">4</td>
<td align="center" width="6.5%">5</td>
<td align="center" width="6.5%">6</td>
<td align="center" width="6.5%">7</td>
<td align="center" width="6.5%">8</td>
<td align="center" width="6.5%">9</td>
<td align="center" width="6.5%">10</td>
<td align="center" width="6.5%">11</td>
<td align="center" width="6.5%">12</td>
<td align="center" width="6.5%">13</td>
<td align="center" width="6.5%">14</td>
<td align="center" width="6.5%">15</td>
</tr>
</table>
<table width="95%" align="center">
<tr>
<td align="left"><br />
The left-hand dataset above as written to an HDF5
file from Fortran:
</td>
</tr>
</table>
<br />
<table border="1" width="95%">
<tr>
<td align="center" width="6.5%">1</td>
<td align="center" width="6.5%">6</td>
<td align="center" width="6.5%">11</td>
<td align="center" width="6.5%">2</td>
<td align="center" width="6.5%">7</td>
<td align="center" width="6.5%">12</td>
<td align="center" width="6.5%">3</td>
<td align="center" width="6.5%">8</td>
<td align="center" width="6.5%">13</td>
<td align="center" width="6.5%">4</td>
<td align="center" width="6.5%">9</td>
<td align="center" width="6.5%">14</td>
<td align="center" width="6.5%">5</td>
<td align="center" width="6.5%">10</td>
<td align="center" width="6.5%">15</td>
</tr>
</table>
<br />
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 2. Comparing C and Fortran dataspaces
<!-- formerly Figure 3 --></b><br />
The HDF5 Library stores arrays along the fastest-changing
dimension. This approach is often referred to as being “in
C order.” C, C++, and Java work with arrays in row-major
order. In other words, the row, or the last dimension, is the
fastest-changing dimension. Fortran, on the other hand, handles
arrays in column-major order making the column, or the first
dimension, the fastest-changing dimension. Therefore, the C and
Fortran arrays illustrated in the top portion of this figure are
stored identically in an HDF5 file. This ensures that data
written by any language can be meaningfully read, interpreted,
and manipulated by any other.
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<!-- editingComment this entire section needs to be -->
<!-- written properly then reincluded. -->
<!--
<h4>Extending a dataspace</h4>
<p>
<span class="editingComment">[ [ [ Text here describing the extension of the dataset, per the following diagram. ] ]</span>
<p>
<table x-use-null-cells
width="600"
cellspacing="0"
class="fullImgTable"
align="center">
<tr valign="top">
<td class="fullImgTableImgCell" align="center">
<img src="Images/Dspace_fig03.jpg">
</td></tr>
<tr>
<td align="center" class="fullImgTableCapCell">
<span class="figurenumber">Figure 4</span>
Extend the simple dataspace array to 12x130
-->
<!-- The next 10 lines were already commented out -->
<!-- before this entire section was removed. -->
<!--
<p align="left">
To extend (conceptual):
<br />
<code>err = H5Sset_extent_simple(id,rank,new_dims,max_dims)</code>
<br />
To extend (concrete):
<br />
<code>err = H5Sset_extent_simple(id,2,[12,130],[H5S_UNLIMITED,200])</code>
-->
<!--
</td>
</tr>
</table>
-->
<h4>7.3.2.6. Finding Dataspace Charateristics</h4>
<p>The HDF5 Library provides several APIs designed to query
the characteristics of a dataspace.</p>
<p>The function <span class="codeText">H5Sis_simple</span>
(<span class="codeText">h5sis_simple_f</span>)
returns information about the type of a dataspace.
This function is rarely used and
currently supports only simple and scalar dataspaces.</p>
<!-- editingComment
<span class="editingComment">[ [ [ Isn't that all of them? What other types are there? ] ] ]</span>
-->
<p>To find out the dimensionality, or rank, of a dataspace,
use <span class="codeText">H5Sget_simple_extent_ndims</span>
(<span class="codeText">h5sget_simple_extent_ndims_f</span>).
<span class="codeText">H5Sget_simple_extent_dims</span>
can also be used to find out the rank.
See the example below.
If both functions return <span class="codeText">0</span> for the
value of <span class="codeVar">rank</span>, then the dataspace is scalar.</p>
<p>To query the sizes of the current and maximum dimensions,
use <span class="codeText">H5Sget_simple_extent_dims</span>
(<span class="codeText">h5sget_simple_extent_dims_f)</span>. </p>
<p>The following example illustrates querying the rank and dimensions
of a dataspace using these functions. </p>
<!-- NEW PAGE -->
<p>In C:</p>
<pre>
hid_t space_id;
int rank;
hsize_t *current_dims;
hsize_t *max_dims;
---------
rank=H5Sget_simple_extent_ndims(space_id);
(or rank=H5Sget_simple_extent_dims(space_id, NULL, NULL);)
current_dims= (hsize_t)malloc(rank*sizeof(hsize_t));
max_dims=(hsize_t)malloc(rank*sizeof(hsize_t));
H5Sget_simple_extent_dims(space_id, current_dims, max_dims);
Print values here for the previous example
</pre>
<!-- editingComment
<p>In Fortran:
<pre>
Example ??????????
</pre>
-->
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="DTransfer">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<br />
<!-- NEW PAGE -->
<a name="DTransfer">
<h3 class="pagebefore">7.4. Dataspaces and Data Transfer</h3>
</a>
<p>Read and write operations transfer data between an HDF5 file on
disk and in memory. The shape that the array data takes in the file
and in memory may be the same, but HDF5 also allows users the ability
to represent data in memory in a different shape than in the file.
If the shape of an array in the file and in memory will be the same,
then the same dataspace definition can be used for both. If the shape
of an array in memory needs to be different than the shape in the file,
then the dataspace definition for the shape of the array in memory can
be changed. During a read operation, the array will be read into the
different shape in memory, and during a write operation, the array
will be written to the file in the shape specified by the dataspace
in the file. The only qualification is that the number of elements
read or written must be the same in both the source and the destination
dataspaces.</p>
<p>Item a in the figure below shows a simple example of a read operation
in which the data is stored as a 3 by 4 array in the file (item b) on
disk, but the program wants it to be a 4 by 3 array in memory. This
is accomplished by setting the memory dataspace to describe the
desired memory layout, as in item c. The read operation reads the
data in the file array into the memory array.</p>
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig4.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 3. Data layout before and after a read operation
<!-- formerly Figure 4 --></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<!-- NEW PAGE -->
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig5.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 4. Moving data from disk to memory
<!-- formerly Figure 5 --></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>Both the source and destination are stored as contiguous blocks of storage
with the elements in the order specified by the <em>dataspace</em>.
The figure above <!-- formerly Figure 5 -->shows one way the elements might
be organized. In item a<!-- formerly Figure 5a-->,
the elements are stored as 3 blocks of 4 elements. The destination is
an array of 12 elements in memory (see item c<!-- formerly Figure 5c -->).
As the figure suggests, the transfer reads the disk blocks into a
memory buffer (see item b<!-- formerly Figure 5b-->), and then
writes the elements to the correct locations in memory. A similar
process occurs in reverse when data is written to disk.</p>
<h4>7.4.1. Data Selection</h4>
<p>In addition to rearranging data, the transfer may select the data
elements from the source and destination.</p>
<p>Data selection is implemented by creating a <em>dataspace</em>
object that describes the selected elements (within the hyper rectangle)
rather than the whole array. Two <em>dataspace</em> objects with
selections can be used in data transfers to read selected elements
from the source and write selected elements to the destination. When
data is transferred using the dataspace object, only the selected
elements will be transferred.</p>
<p>This can be used to implement partial I/O, including:</p>
<ul>
<li>Sub-setting - reading part of a large dataset</li>
<li>Sampling - reading selected elements (e.g., every second element)
of a dataset</li>
<li>Scatter-gather - read non-contiguous elements into contiguous locations
(gather) or read contiguous elements into non-contiguous locations
(scatter) or both</li>
</ul>
<p>To use selections, the following steps are followed:</p>
<ol>
<li>Get or define the dataspace for the source and destination</li>
<li>Specify one or more selections for source and destination dataspaces</li>
<li>Transfer data using the dataspaces with selections</li>
</ol>
<!-- NEW PAGE -->
<p>A selection is created by applying one or more selections to a
<em>dataspace</em>. A selection may override any other selections
(<code>H5T_SELECT_SET</code>)
or may be “Ored” with previous selections on the same dataspace
(<code>H5T_SELECT_OR</code>).
In the latter case, the resulting selection is the union of the selection and
all previously selected selections. Arbitrary sets of points from a dataspace
can be selected by specifying an appropriate set of selections.</p>
<p>Two selections are used in data transfer, so the source and destination
must be compatible, as described below.</p>
<p>There are two forms of selection, hyperslab and point. A selection must
be either a point selection or a set of hyperslab selections.
Selections cannot be mixed.</p>
<p>The definition of a selection within a dataspace, not the data
in the selection, cannot be saved to the file unless the selection
definition is saved as a region reference. See the <a href="#DRegions">
References to Dataset Regions</a> section for more information.</p>
<h4>7.4.1.1. Hyperslab selection</h4>
<p>A hyperslab is a selection of elements from a hyper rectangle.
An HDF5 hyperslab is a rectangular pattern defined by four arrays. The
four arrays are summarized in the table below <!-- formerly Table 1-->. </p>
<p>The <em>offset</em> defines the origin of the hyperslab in the original
dataspace.</p>
<p>The <em>stride</em> is the number of elements to increment between
selected elements. A stride of ‘1’ is every element, a
stride of ‘2’ is every second element, etc. Note that there
may be a different stride for each dimension of the dataspace. The
default stride is 1.</p>
<p>The <em>count</em> is the number of elements in the hyperslab selection.
When the stride is 1, the selection is a hyper rectangle with a corner at
the offset and size count[0] by count[1] by.... When stride is greater
than one, the hyperslab bounded by the offset and the corners defined
by stride[n] * count[n].</p>
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="2" align="left" valign="bottom">
<b>Table 1. Hyperslab elements</b></td>
</tr>
<tr><td colspan="2"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td width="15%"><b>Parameter</b></td>
<td width="85%"><b>Description</b></td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Offset</td>
<td>The starting location for the hyperslab.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Stride</td>
<td>The number of elements to separate each element or block
to be selected.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Count</td>
<td>The number of elements or blocks to select along each
dimension.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Block</td>
<td>The size of the block selected from the dataspace.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="3" /></td></tr>
</table>
<br />
<p>The <em>block</em> is a count on the number of repetitions of the
hyperslab. The default block size is ‘1’, which is one
hyperslab. A block of 2 would be two hyperslabs in that dimension,
with the second starting at offset[n]+ (count[n] * stride[n]) + 1.</p>
<p>A hyperslab can be used to access a sub-set of a large dataset.
The figure below <!-- formerly Figure 6 -->shows an example of a hyperslab
that reads a rectangle from the middle of a larger two dimensional
array. The destination is the same shape as the source.</p>
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig6.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 5. Access a sub-set of data
with a hyperslab<!-- formerly Figure 6--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>Hyperslabs can be combined to select complex regions of the source and
destination. The figure below <!-- formerly Figure 7 -->shows an example
of a transfer from one non-rectangular
region into another non-rectangular region. The source is defined as the
union of two hyperslabs, and the destination is the union of three
hyperslabs.</p>
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig7.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 6. Build complex regions with
hyperslab unions<!-- formerly Figure 7--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<!-- NEW PAGE -->
<p>Hyperslabs may also be used to collect or scatter data from regular
patterns. The figure below <!-- formerly Figure 8 -->shows an example
where the source is a repeating pattern of blocks, and the destination
is a single, one dimensional array.</p>
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig8.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 7. Use hyperslabs to combine or disperse data
<!-- formerly Figure 8--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.4.1.2. Select Points</h4>
<p>The second type of selection is an array of points, i.e., coordinates.
Essentially, this selection is a list of all the points to include.
The figure below <!-- formerly Figure 9 -->shows an example of a transfer
of seven elements from a two dimensional dataspace to a three
dimensional dataspace using a point selection to specify the points.</p>
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig9.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 8. Point selection
<!-- formerly Figure 9--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.4.1.3. Rules for Defining Selections</h4>
<p>A selection must have the same number of dimensions (rank) as the
dataspace it is applied to, although it may select from only a small
region, e.g., a plane from a 3D dataspace. Selections do not affect the
extent of the <em>dataspace</em>, the selection may be larger than
the <em>dataspace</em>. The boundaries of selections are reconciled
with the extent at the time of the data transfer.</p>
<h4>7.4.1.4. Data Transfer with Selections</h4>
<p>A data transfer (read or write) with selections is the same as any
read or write, except the source and destination <em>dataspace</em>
have compatible selections. </p>
<p>During the data transfer, the following steps are executed by the
library:</p>
<ul>
<li>The source and destination <em>dataspaces</em> are checked to assure
that the selections are compatible.</li>
<ul>
<li>Each selection must be within the current extent of the <em>dataspace</em>.
A selection may be defined to extend outside the current extent of the
<em>dataspace</em>, but the <em>dataspace</em> cannot be accessed if the
selection is not valid at the time of the access.</li>
<li>The total number of points selected in the source and destination must
be the same. Note that the dimensionality of the source and destination can
be different (e.g., the source could be 2D, the destination 1D or 3D), and
the shape can be different, but the number of elements selected must be
the same.</li>
</ul>
<li>The data is transferred, element by element.</li>
</ul>
<p>Selections have an iteration order for the points selected, which can be
any permutation of the dimensions involved (defaulting to ‘C’
array order) or a specific order for the selected points, for selections
composed of single array elements with <code>H5Sselect_elements</code>.</p>
<p>The elements of the selections are transferred in row-major, or C order.
That is, it is assumed that the first dimension varies slowest, the second
next slowest, and so forth. For hyperslab selections, the order can be any
permutation of the dimensions involved (defaulting to ‘C’
array order). When multiple hyperslabs are combined, the hyperslabs
are coalesced into contiguous reads and writes.</p>
<p>In the case of point selections, the points are read and written in the
order specified.</p>
<h4>7.4.2. Programming Model</h4>
<h4>7.4.2.1. Selecting Hyperslabs</h4>
<p>Suppose we want to read a 3x4 hyperslab from a dataset in a file beginning
at the element <1,2> in the dataset, and read it into a 7 x 7 x 3 array
in memory. See the figure below. <!-- formerly (Figure 10).--> In order
to do this, we must create a dataspace that describes the overall rank
and dimensions of the dataset in the file as well as the position and
size of the hyperslab that we are extracting from that dataset.</p>
<!-- NEW PAGE -->
<table width="400" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig10.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 9. Selecting a hyperslab
<!-- formerly Figure 10--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The code in the first example below <!-- formerly Figure 11 -->
illustrates the selection of the hyperslab in the
file dataspace. The second example below <!-- formerly Figure 12 -->shows
the definition of the destination dataspace in
memory. Since the in-memory dataspace has three dimensions, the hyperslab is
an array with three dimensions with the last dimension being
1: <3,4,1>. The third example below <!-- formerly Figure 13 -->
shows the read using the source and destination <em>dataspaces</em>
with selections.</p>
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
/*
* get the file dataspace.
*/
dataspace = H5Dget_space(dataset); /* dataspace identifier */
/*
* Define hyperslab in the dataset.
*/
offset[0] = 1;
offset[1] = 2;
count[0] = 3;
count[1] = 4;
status = H5Sselect_hyperslab(dataspace, H5S_SELECT_SET, offset, NULL,
count, NULL);</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 1. Selecting a hyperslab
<!-- formerly Figure 11--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
/*
* Define memory dataspace.
*/
dimsm[0] = 7;
dimsm[1] = 7;
dimsm[2] = 3;
memspace = H5Screate_simple(3,dimsm,NULL);
/*
* Define memory hyperslab.
*/
offset_out[0] = 3;
offset_out[1] = 0;
offset_out[2] = 0;
count_out[0] = 3;
count_out[1] = 4;
count_out[2] = 1;
status = H5Sselect_hyperslab(memspace, H5S_SELECT_SET, offset_out, NULL,
count_out, NULL);</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 2. Defining the destination memory
<!-- formerly Figure 12--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
ret = H5Dread(dataset, H5T_NATIVE_INT, memspace, dataspace, H5P_DEFAULT,
data);</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 3. A sample read specifying source
and destination dataspaces
<!-- formerly Figure 13--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.4.2.2. Example with Strides and Blocks</h4>
<p>Consider an 8 x 12 dataspace into which we want to write eight 3 x 2
blocks in a two dimensional array
from a source dataspace in memory that is a 50-element one dimensional
array.
See the figure below.<!-- formerly (Figure 14).--></p>
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<table>
<tr>
<td width="80"> </td>
<td width="440">a) The source is a 1D array with 50 elements</td>
<td width="80"> </td>
</tr>
<tr valign="top"><td> </td>
<td align="center">
<img src="Images/Dspace_fig14.JPG">
</td><td> </td></tr>
<tr>
<td width="80"> </td>
<td width="440">b) The destination on disk is a 2D array
with 48 selected elements</td>
<td width="80"> </td>
</tr>
</table>
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 10. Write from a one dimensional
array to a two dimensional array<!-- Figure 14--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The example below <!-- formerly Figure 15 -->shows code to write 48
elements from the one dimensional array to the
file dataset starting with the second element in vector. The destination
hyperslab has the following parameters: offset=(0,1), stride=(4,3),
count=(2,4), block=(3,2). The source has the parameters: offset=(1),
stride=(1), count=(48), block=(1). After these operations, the file
dataspace will have the values shown in item b in the figure above
<!-- formerly Figure 14-->. Notice that the values are inserted in
the file dataset in row-major order.</p>
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
/* Select hyperslab for the dataset in the file, using 3 x 2 blocks, (4,3) stride
* (2,4) count starting at the position (0,1).
*/
offset[0] = 0; offset[1] = 1;
stride[0] = 4; stride[1] = 3;
count[0] = 2; count[1] = 4;
block[0] = 3; block[1] = 2;
ret = H5Sselect_hyperslab(fid, H5S_SELECT_SET, offset, stride, count, block);
/*
* Create dataspace for the first dataset.
*/
mid1 = H5Screate_simple(MSPACE1_RANK, dim1, NULL);
/*
* Select hyperslab.
* We will use 48 elements of the vector buffer starting at the second element.
* Selected elements are 1 2 3 . . . 48
*/
offset[0] = 1;
stride[0] = 1;
count[0] = 48;
block[0] = 1;
ret = H5Sselect_hyperslab(mid1, H5S_SELECT_SET, offset, stride, count, block);
/*
* Write selection from the vector buffer to the dataset in the file.
*
ret = H5Dwrite(dataset, H5T_NATIVE_INT, midd1, fid, H5P_DEFAULT, vector)</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 4. Write from a one dimensional
array to a two dimensional array
<!-- formerly Figure 15--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.4.2.3. Selecting a Union of Hyperslabs</h4>
<p>The HDF5 Library allows the user to select a union of hyperslabs and
write or read the selection into another selection. The shapes of the
two selections may differ, but the number of elements must be equal. </p>
<!-- NEW PAGE -->
<table width="400" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig16a.jpg">
<br />
<img src="Images/Dspace_fig16b.jpg">
<br />
<img src="Images/Dspace_fig16c.jpg">
</td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 11. Transferring hyperslab unions
<!-- formerly Figure 16--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The figure above <!-- formerly Figure 16 -->shows the transfer of a
selection that is two overlapping hyperslabs
from the dataset into a union of hyperslabs in the memory dataset. Note that
the destination dataset has a different shape from the source dataset.
Similarly, the selection in the memory dataset could have a different shape
than the selected union of hyperslabs in the original file. For simplicity,
the selection is that same shape at the destination.</p>
<p>To implement this transfer, it is necessary to:</p>
<ol>
<li>Get the source dataspace</li>
<li>Define one hyperslab selection for the source</li>
<li>Define a second hyperslab selection, unioned with the first</li>
<li>Get the destination dataspace</li>
<li>Define one hyperslab selection for the destination</li>
<li>Define a second hyperslab seletion, unioned with the first</li>
<li>Execute the data transfer (<code>H5Dread</code> or
<code>H5Dwrite</code>) using the source and
destination dataspaces</li>
</ol><br />
<p>The example below <!-- formerly Figure 17 -->shows example code to
create the selections for the source
dataspace (the file). The first hyperslab is size 3 x 4 and the left upper
corner at the position (1,2). The hyperslab is a simple rectangle, so the
stride and block are 1. The second hyperslab is 6 x 5 at the position (2,4).
The second selection is a union with the first hyperslab
(<code>H5S_SELECT_OR</code>).</p>
<br />
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
fid = H5Dget_space(dataset);
/*
* Select first hyperslab for the dataset in the file.
*
*/
offset[0] = 1; offset[1] = 2;
block[0] = 1; block[1] = 1;
stride[0] = 1; stride[1] = 1;
count[0] = 3; count[1] = 4;
ret = H5Sselect_hyperslab(fid, H5S_SELECT_SET, offset, stride, count, block);
/*
* Add second selected hyperslab to the selection.
*/
offset[0] = 2; offset[1] = 4;
block[0] = 1; block[1] = 1;
stride[0] = 1; stride[1] = 1;
count[0] = 6; count[1] = 5;
ret = H5Sselect_hyperslab(fid, H5S_SELECT_OR, offset, stride, count, block);</pre>
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 5. Select source hyperslabs
<!-- formerly Figure 17 --></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The example below <!-- formerly Figure 18 -->shows example code to
create the selection for the destination
in memory. The steps are similar. In this example, the hyperslabs are the
same shape, but located in different positions in the dataspace. The first
hyperslab is 3 x 4 and starts at (0,0), and the second is 6 x 5 and starts at (1,2).</p>
<p>Finally, the <code>H5Dread</code> call transfers the selected data from
the file dataspace
to the selection in memory.</p>
<p>In this example, the source and destination selections are two
overlapping rectangles. In general, any number of rectangles can be
OR’ed, and they do not have to be contiguous. The order of the
selections does not matter, but the first should use
<code>H5S_SELECT_SET</code>; subsequent selections are unioned
using <code>H5S_SELECT_OR</code>.</p>
<p>It is important to emphasize that the source and destination do not
have to be the same shape (or number of rectangles). As long as the two
selections have the same number of elements, the data can be transferred.</p>
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
/*
* Create memory dataspace.
*/
mid = H5Screate_simple(MSPACE_RANK, mdim, NULL);
/*
* Select two hyperslabs in memory. Hyperslabs has the same
* size and shape as the selected hyperslabs for the file dataspace.
*/
offset[0] = 0; offset[1] = 0;
block[0] = 1; block[1] = 1;
stride[0] = 1; stride[1] = 1;
count[0] = 3; count[1] = 4;
ret = H5Sselect_hyperslab(mid, H5S_SELECT_SET, offset, stride, count, block);
offset[0] = 1; offset[1] = 2;
block[0] = 1; block[1] = 1;
stride[0] = 1; stride[1] = 1;
count[0] = 6; count[1] = 5;
ret = H5Sselect_hyperslab(mid, H5S_SELECT_OR, offset, stride, count, block);
ret = H5Dread(dataset, H5T_NATIVE_INT, mid, fid, H5P_DEFAULT, matrix_out);</pre>
</td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 6. Select destination hyperslabs
<!-- formerly Figure 18--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.4.2.4. Selecting a List of Independent Points</h4>
<p>It is also possible to specify a list of elements to read or write using
the function <code>H5Sselect_elements</code>. The procedure is similar
to hyperslab selections.</p>
<ol>
<li>Get the source dataspace</li>
<li>Set the selected points</li>
<li>Get the destination dataspacev
<li>Set the selected points</li>
<li>Transfer the data using the source and destination dataspaces</li>
</ol><br />
<p>The figure below <!-- formerly Figure 19 -->shows an example where
four values are to be written to four separate points in a two
dimensional dataspace. The source dataspace is a one dimensional
array with the values 53, 59, 61, 67. The destination dataspace
is an 8 x 12 array. The elements are to be written to the points
(0,0), (3,3), (3,5), and (5,6). In this example, the source does not
require a selection. The example below the figure
<!-- formerly Figure 20 -->shows example code to implement this
transfer.</p>
<p>A point selection lists the exact points to be transferred and the order
they will be transferred. The source and destination are required to have
the same number of elements. A point selection can be used with a hyperslab
(e.g., the source could be a point selection and the destination a hyperslab,
or vice versa), so long as the number of elements selected are the same.</p>
<table width="400" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig19a.jpg">
<br />
<img src="Images/Dspace_fig19b.jpg">
<br />
<img src="Images/Dspace_fig19c.jpg">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 12. Write data to separate points
<!-- formerly Figure 19--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
hsize_t dim2[] = {4};
int values[] = {53, 59, 61, 67};
hssize_t coord[4][2]; /* Array to store selected points
from the file dataspace */
/*
* Create dataspace for the second dataset.
*/
mid2 = H5Screate_simple(1, dim2, NULL);
/*
* Select sequence of NPOINTS points in the file dataspace.
*/
coord[0][0] = 0; coord[0][1] = 0;
coord[1][0] = 3; coord[1][1] = 3;
coord[2][0] = 3; coord[2][1] = 5;
coord[3][0] = 5; coord[3][1] = 6;
ret = H5Sselect_elements(fid, H5S_SELECT_SET, NPOINTS,
(const hssize_t **)coord);
ret = H5Dwrite(dataset, H5T_NATIVE_INT, mid2, fid, H5P_DEFAULT, values);</pre>
</td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 7. Write data to separate points
<!-- formerly Figure 20 --></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<!-- NEW PAGE -->
<h4>7.4.2.5. Combinations of Selections</h4>
<p>Selections are a very flexible mechanism for reorganizing data during a
data transfer. With different combinations of <em>dataspaces</em> and
selections, it is possible to implement many kinds of data transfers
including sub-setting, sampling, and reorganizing the data. The table below
<!-- formerly Table 2 -->gives some example combinations of
source and destination, and the operations they implement.</p>
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="3" align="left" valign="bottom">
<b>Table 2. Selection operations</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td width="38%"><b>Source</b></td>
<td width="37%"><b>Destination</b></td>
<td width="25%"><b>Operation</b></td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>All</td>
<td>All</td>
<td>Copy whole array</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>All</td>
<td>All (different shape)</td>
<td>Copy and reorganize array</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Hyperslab</td>
<td>All</td>
<td>Sub-set</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Hyperslab</td>
<td>Hyperslab (same shape)</td>
<td>Selection</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Hyperslab</td>
<td>Hyperslab (different shape)</td>
<td>Select and rearrange</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Hyperslab with stride or block</td>
<td>All or hyperslab with stride 1</td>
<td>Sub-sample, scatter</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Hyperslab</td>
<td>Points</td>
<td>Scatter</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Points</td>
<td>Hyperslab or all</td>
<td>Gather</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Points</td>
<td>Points (same)</td>
<td>Selection</td>
</tr>
<tr><td colspan="3"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td>Points</td>
<td>Points (different)</td>
<td>Reorder points</td>
</tr>
<tr><td colspan="3"><hr color="green" size="3" /></td></tr>
</table>
<br />
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="DSelectTransfer">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<br />
<a name="DSelectTransfer">
<h3 class="pagebefore">7.5. Dataspace Selection Operations and Data Transfer</h3>
</a>
<p><em>This section is under construction.</em></p>
<br />
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="DRegions">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<!-- NEW PAGE -->
<a name="DRegions">
<h3 class="pagebefore">7.6. References to Dataset Regions</h3>
</a>
<p>Another use of selections is to store a reference to a region of a
dataset. An HDF5 object reference object is a pointer to an object
(dataset, group, or committed datatype) in the file. A selection can be
used to create a pointer to a set of selected elements of a
<em>dataset</em>, called a region reference. The selection can be
either a point selection or a hyperslab selection. </p>
<!-- editingComment
<span class="editingComment">
WORKING TOWARD AN IMPROVED PARAGRAPH:
In addition to the object reference, HDF5 also provides a regions reference.
An HDF5 Region Reference is a pointer to a selection within a dataset.
The selection can be either a point or hyperslab selection.
</span>
-->
A more complete description of region references can be found in the
chapter “<a href="11_Datatypes.html">HDF5 Datatypes</a>.”
<p>A region reference is an object maintained by the HDF5 Library.
The region reference can be stored in a dataset or attribute, and then read.
The dataset or attribute is defined to have the special datatype,
<code>H5T_STD_REF_DSETREG</code>. </p>
<p>To discover the elements and/or read the data, the region reference can
be dereferenced. The <code>H5Rdefrerence</code> call returns an
identifier for the <em>dataset</em>, and then the selected dataspace can
be retrieved with <code>H5Rget_select</code> call. The selected
<em>dataspace</em> can be used to read the selected data elements.</p>
<h4>7.6.1. Example Uses for Region References</h4>
<p>Region references are used to implement stored pointers to data within
a dataset. For example, features in a large dataset might be indexed
by a table. See the figure below<!-- formerly Figure 21-->. This table
could be stored as an HDF5 dataset with a compound datatype, for example,
with a field for the name of the feature and a region reference
to point to the feature in the dataset. See the second figure below.
<!-- formerly Figure 22--></p>
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig21.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 13. Features indexed by a table
<!-- formerly Figure 21--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<!-- NEW PAGE -->
<table width="500" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
<img src="Images/Dspace_fig22.JPG">
</td></tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 14. Storing the table with a
compound datatype<!-- formerly Figure 22--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<h4>7.6.2. Creating References to Regions</h4>
<p>To create a region reference:</p>
<ol>
<li>Create or open the dataset that contains the region</li>
<li>Get the dataspace for the dataset</li>
<li>Define a selection that specifies the region</li>
<li>Create a region reference using the dataset and dataspace with
selection</li>
<li>Write the region reference(s) to the desired dataset or attribute</li>
</ol>
<p>The figure below <!-- formerly Figure 23 -->shows a diagram of a file
with three datasets. Dataset D1 and D2 are two dimensional arrays of
integers. Dataset R1 is a one dimensional array of references to
regions in D1 and D2. The regions can be any valid selection
of the dataspace of the target dataset.</p>
<!-- NEW PAGE -->
<table width="400" cellspacing="0" align="center">
<tr valign="top">
<td align="center">
<hr color="green" size="3"/>
a) 1 D array of region pointers, <br />each pointer refers to a
<br />selection in one Dataset.<br />
<img src="Images/Dspace_fig23.JPG">
</td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left" >
<b>Figure 15. A file with three datasets
<!-- formerly Figure 23--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>The example below <!-- formerly Figure 24 -->shows code to
create the array of region references.
The references are created in an array of type <code>hdset_reg_ref_t</code>.
Each region is defined as a selection on the dataspace of the dataset,
and a reference is created using <code>H5Rcreate()</code>. The call
to <code>H5Rcreate()</code> specifies the file, dataset, and the
dataspace with selection.</p>
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
/* create an array of 4 region references */
hdset_reg_ref_t ref[4];
/*
* Create a reference to the first hyperslab in the first Dataset.
*/
offset[0] = 1; offset[1] = 1;
count[0] = 3; count[1] = 2;
status = H5Sselect_hyperslab(space_id, H5S_SELECT_SET, offset, NULL,
count, NULL);
status = H5Rcreate(&ref[0], file_id, "D1", H5R_DATASET_REGION,
space_id);
/*
* The second reference is to a union of hyperslabs in the first
* Dataset
*/
offset[0] = 5; offset[1] = 3;
count[0] = 1; count[1] = 4;
status = H5Sselect_none(space_id);
status = H5Sselect_hyperslab(space_id, H5S_SELECT_SET,offset,
NULL, count, NULL);
offset[0] = 6; offset[1] = 5;
count[0] = 1; count[1] = 2;
status = H5Sselect_hyperslab(space_id, H5S_SELECT_OR, offset, NULL,
count, NULL);
status = H5Rcreate(&ref[1], file_id, "D1", H5R_DATASET_REGION,
space_id);
/*
* the fourth reference is to a selection of points in the first
* Dataset
*/
status = H5Sselect_none(space_id);
coord[0][0] = 4; coord[0][1] = 4;
coord[1][0] = 2; coord[1][1] = 6;
coord[2][0] = 3; coord[2][1] = 7;
coord[3][0] = 1; coord[3][1] = 5;
coord[4][0] = 5; coord[4][1] = 8;
status = H5Sselect_elements(space_id, H5S_SELECT_SET,num_points,
(const hssize_t **)coord);
status = H5Rcreate(&ref[3], file_id, "D1", H5R_DATASET_REGION,
space_id);
/*
* the third reference is to a hyperslab in the second Dataset
*/
offset[0] = 0; offset[1] = 0;
count[0] = 4; count[1] = 6;
status = H5Sselect_hyperslab(space_id2, H5S_SELECT_SET, offset, NULL,
count, NULL);
status = H5Rcreate(&ref[2], file_id, "D2", H5R_DATASET_REGION,
space_id2);</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 8. Create an array of region references
<!-- formerly Figure 24--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<p>When all the references are created, the array of references is written
to the dataset R1. The dataset is declared to have datatype
<code>H5T_STD_REF_DSETREG</code>. See the example below.
<!-- formerly Figure 25--></p>
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
Hsize_t dimsr[1];
dimsr[0] = 4;
/*
* Dataset with references.
*/
spacer_id = H5Screate_simple(1, dimsr, NULL);
dsetr_id = H5Dcreate(file_id, "R1", H5T_STD_REF_DSETREG,
spacer_id, H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT);
/*
* Write dataset with the references.
*/
status = H5Dwrite(dsetr_id, H5T_STD_REF_DSETREG, H5S_ALL, H5S_ALL,
H5P_DEFAULT,ref);</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 9. Write the array of references to a dataset
<!-- formerly Figure 25--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<p>When creating region references, the following rules are enforced.</p>
<ul>
<li>The selection must be a valid selection for the target <em>dataset</em>,
just as when transferring data</li>
<li>The <em>dataset</em> must exist in the file when the reference is
created (<code>H5Rcreate</code>)</li>
<li>The target <em>dataset</em> must be in the same file as the
stored reference</li>
</ul>
<br />
<h4>7.6.3. Reading References to Regions</h4>
<p>To retrieve data from a region reference, the reference must be read from
the file, and then the data can be retrieved. The steps are:</p>
<ol>
<li>Open the dataset or attribute containing the reference objects</li>
<li>Read the reference object(s)</li>
<li>For each region reference, get the dataset (<code>H5R_dereference</code>) and
dataspace (<code>H5Rget_space</code>)</li>
<li>Use the dataspace and datatype to discover what space is needed to
store the data, allocate the correct storage and create a dataspace
and datatype to define the memory data layout</li>
</ol>
<p>The example below <!-- formerly Figure 26 -->shows code to read an
array of region references from a
dataset, and then read the data from the first selected region. Note that the
region reference has information that records the dataset (within the file)
and the selection on the <em>dataspace</em> of the <em>dataset</em>.
After dereferencing the regions reference, the <em>datatype</em>,
number of points, and some aspects of the selection can be discovered.
(For a union of hyperslabs, it may not be possible to determine the exact
set of hyperslabs that has been combined.) The table below the code example
<!-- formerly Table 3 -->shows the inquiry functions.</p>
<!-- NEW PAGE -->
<p>When reading data from a region reference, the following rules are
enforced:</p>
<ul>
<li>The target <em>dataset</em> must be present and accessible in the
file</li>
<li>The selection must be a valid selection for the <em>dataset</em></li>
</ul>
<br />
<table width="600" cellspacing="0" align="center">
<tr valign="top">
<td align="left">
<hr color="green" size="3"/>
<pre>
dsetr_id = H5Dopen (file_id, "R1", H5P_DEFAULT);
status = H5Dread(dsetr_id, H5T_STD_REF_DSETREG, H5S_ALL, H5S_ALL,
H5P_DEFAULT, ref_out);
/*
* Dereference the first reference.
* 1) get the dataset (H5Rdereference)
* 2) get the selected dataspace (H5Rget_region)
*/
dsetv_id = H5Rdereference(dsetr_id, H5R_DATASET_REGION,
&ref_out[0]);
space_id = H5Rget_region(dsetr_id, H5R_DATASET_REGION,&ref_out[0]);
/*
* Discover how many points and shape of the data
*/
ndims = H5Sget_simple_extent_ndims(space_id);
H5Sget_simple_extent_dims(space_id,dimsx,NULL);
/*
* Read and display hyperslab selection from the dataset.
*/
dimsy[0] = H5Sget_select_npoints(space_id);
spacex_id = H5Screate_simple(1, dimsy, NULL);
status = H5Dread(dsetv_id, H5T_NATIVE_INT, H5S_ALL, space_id,
H5P_DEFAULT, data_out);
printf("Selected hyperslab: ");
for (i = 0; i < 8; i++)
{
printf("\n");
for (j = 0; j < 10; j++)
printf("%d ", data_out[i][j]);
}
printf("\n");</pre></td>
</tr>
<tr><td><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td align="left">
<b>Example 10. Read an array of region references, and then
read from the first selection
<!-- formerly Figure 26--></b>
<hr color="green" size="3"/></td>
</tr>
</table>
<br />
<br />
<!-- NEW PAGE -->
<table width="600" cellspacing="0" align="center" cellpadding="0">
<tr valign="bottom">
<td colspan="2" align="left" valign="bottom">
<b>Table 3. The inquiry functions</b></td>
</tr>
<tr><td colspan="2"><hr color="green" size="3" /></td></tr>
<tr valign="top">
<td width="45%"><b>Function</b></td>
<td width="55%"><b>Information</b></td>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_npoints</code></td>
<td>The number of elements in the selection (hyperslab
or point selection).</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_bounds</code></td>
<td>The bounding box that encloses the selected
points (hyperslab or point selection).</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_hyper_nblocks</code></td>
<td>The number of blocks in the selection.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_hyper_blocklist</code></td>
<td>A list of the blocks in the selection.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_elem_npoints</code></td>
<td>The number of points in the selection.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="1" /></td></tr>
<tr valign="top">
<td><code>H5Sget_select_elem_pointlist</code></td>
<td>The points.</td>
</tr>
<tr><td colspan="2"><hr color="green" size="3" /></td></tr>
</table>
<br />
<SCRIPT language="JavaScript">
<!--
document.writeln ("
<a name="Programs">
<div align="right">
<a href="#TOP"><font size=-1>(Top)</font></a>
</div>
</a>
");
-->
</SCRIPT>
<br />
<!-- NEW PAGE -->
<a name="Programs">
<h3 class="pagebefore">7.7. Sample Programs</h3>
</a>
This section contains the full programs from which several of the
code examples in this chapter were derived.
The <code>h5dump</code> output from the program’s output file
immediately follows each program.
<h4>7.7.1. <a name="h5_write_c"><code>h5_write.c</code></a></h4>
<pre>
----------
#include "hdf5.h"
#define H5FILE_NAME "SDS.h5"
#define DATASETNAME "C Matrix"
#define NX 3 /* dataset dimensions */
#define NY 5
#define RANK 2
int
main (void)
{
hid_t file, dataset; /* file and dataset identifiers */
hid_t datatype, dataspace; /* identifiers */
hsize_t dims[2]; /* dataset dimensions */
herr_t status;
int data[NX][NY]; /* data to write */
int i, j;
/*
* Data and output buffer initialization.
*/
for (j = 0; j < NX; j++) {
for (i = 0; i < NY; i++)
data[j][i] = i + 1 + j*NY;
}
/*
* 1 2 3 4 5
* 6 7 8 9 10
* 11 12 13 14 15
*/
/*
* Create a new file using H5F_ACC_TRUNC access,
* default file creation properties, and default file
* access properties.
*/
file = H5Fcreate(H5FILE_NAME, H5F_ACC_TRUNC, H5P_DEFAULT, H5P_DEFAULT);
/*
* Describe the size of the array and create the data space for fixed
* size dataset.
*/
dims[0] = NX;
dims[1] = NY;
dataspace = H5Screate_simple(RANK, dims, NULL);
/*
* Create a new dataset within the file using defined dataspace and
* datatype and default dataset creation properties.
*/
dataset = H5Dcreate(file, DATASETNAME, H5T_NATIVE_INT, dataspace,
H5P_DEFAULT, H5P_DEFAULT, H5P_DEFAULT);</pre>
<!-- NEW PAGE -->
<pre>
/*
* Write the data to the dataset using default transfer properties.
*/
status = H5Dwrite(dataset, H5T_NATIVE_INT, H5S_ALL, H5S_ALL,
H5P_DEFAULT, data);
/*
* Close/release resources.
*/
H5Sclose(dataspace);
H5Dclose(dataset);
H5Fclose(file);
return 0;
}
SDS.out
-------
HDF5 "SDS.h5" {
GROUP "/" {
DATASET "C Matrix" {
DATATYPE H5T_STD_I32BE
DATASPACE SIMPLE { ( 3, 5 ) / ( 3, 5 ) }
DATA {
1, 2, 3, 4, 5,
6, 7, 8, 9, 10,
11, 12, 13, 14, 15
}
}
}
}
</pre>
<br />
<h4>7.7.2. <a name="h5_write_f90"><code>h5_write.f90</code></a></h4>
<pre>
------------
PROGRAM DSETEXAMPLE
USE HDF5 ! This module contains all necessary modules
IMPLICIT NONE
CHARACTER(LEN=7), PARAMETER :: filename = "SDSf.h5" ! File name
CHARACTER(LEN=14), PARAMETER :: dsetname = "Fortran Matrix" ! Dataset name
INTEGER, PARAMETER :: NX = 3
INTEGER, PARAMETER :: NY = 5
INTEGER(HID_T) :: file_id ! File identifier
INTEGER(HID_T) :: dset_id ! Dataset identifier
INTEGER(HID_T) :: dspace_id ! Dataspace identifier
INTEGER(HSIZE_T), DIMENSION(2) :: dims = (/3,5/) ! Dataset dimensions
INTEGER :: rank = 2 ! Dataset rank
INTEGER :: data(NX,NY)
INTEGER :: error ! Error flag
INTEGER :: i, j</pre>
<pre>
!
! Initialize data
!
do i = 1, NX
do j = 1, NY
data(i,j) = j + (i-1)*NY
enddo
enddo
!
! Data
!
! 1 2 3 4 5
! 6 7 8 9 10
! 11 12 13 14 15
!
! Initialize FORTRAN interface.
!
CALL h5open_f(error)
!
! Create a new file using default properties.
!
CALL h5fcreate_f(filename, H5F_ACC_TRUNC_F, file_id, error)
!
! Create the dataspace.
!
CALL h5screate_simple_f(rank, dims, dspace_id, error)
!
! Create and write dataset using default properties.
!
CALL h5dcreate_f(file_id, dsetname, H5T_NATIVE_INTEGER, dspace_id, &
dset_id, error, H5P_DEFAULT_F, H5P_DEFAULT_F, &
H5P_DEFAULT_F)
CALL h5dwrite_f(dset_id, H5T_NATIVE_INTEGER, data, dims, error)
!
! End access to the dataset and release resources used by it.
!
CALL h5dclose_f(dset_id, error)
!
! Terminate access to the data space.
!
CALL h5sclose_f(dspace_id, error)
!
! Close the file.
!
CALL h5fclose_f(file_id, error)
!
! Close FORTRAN interface.
!
CALL h5close_f(error)
END PROGRAM DSETEXAMPLE</pre>
<pre>
SDSf.out
--------
HDF5 "SDSf.h5" {
GROUP "/" {
DATASET "Fortran Matrix" {
DATATYPE H5T_STD_I32BE
DATASPACE SIMPLE { ( 5, 3 ) / ( 5, 3 ) }
DATA {
1, 6, 11,
2, 7, 12,
3, 8, 13,
4, 9, 14,
5, 10, 15
}
}
}
}
</pre>
<br />
<h4>7.7.3. <a name="h5_write_tr_f90"><code>h5_write_tr.f90</code></a></h4>
<pre>
---------------
PROGRAM DSETEXAMPLE
USE HDF5 ! This module contains all necessary modules
IMPLICIT NONE
CHARACTER(LEN=10), PARAMETER :: filename = "SDSf_tr.h5" ! File name
CHARACTER(LEN=24), PARAMETER :: dsetname = "Fortran Transpose Matrix"
! Dataset name
INTEGER, PARAMETER :: NX = 3
INTEGER, PARAMETER :: NY = 5
INTEGER(HID_T) :: file_id ! File identifier
INTEGER(HID_T) :: dset_id ! Dataset identifier
INTEGER(HID_T) :: dspace_id ! Dataspace identifier
INTEGER(HSIZE_T), DIMENSION(2) :: dims = (/NY, NX/) ! Dataset dimensions
INTEGER :: rank = 2 ! Dataset rank
INTEGER :: data(NY,NX)
INTEGER :: error ! Error flag
INTEGER :: i, j
!
! Initialize data
!
do i = 1, NY
do j = 1, NX
data(i,j) = i + (j-1)*NY
enddo
enddo
!
! Data
!
! 1 6 11
! 2 7 12
! 3 8 13
! 4 9 14
! 5 10 15
!
! Initialize FORTRAN interface.
!
CALL h5open_f(error)
!
! Create a new file using default properties.
!
CALL h5fcreate_f(filename, H5F_ACC_TRUNC_F, file_id, error)
!
! Create the dataspace.
!
CALL h5screate_simple_f(rank, dims, dspace_id, error)
!
! Create and write dataset using default properties.
!
CALL h5dcreate_f(file_id, dsetname, H5T_NATIVE_INTEGER, dspace_id, &
dset_id, error, H5P_DEFAULT_F, H5P_DEFAULT_F, &
H5P_DEFAULT_F)
CALL h5dwrite_f(dset_id, H5T_NATIVE_INTEGER, data, dims, error)
!
! End access to the dataset and release resources used by it.
!
CALL h5dclose_f(dset_id, error)
!
! Terminate access to the data space.
!
CALL h5sclose_f(dspace_id, error)
!
! Close the file.
!
CALL h5fclose_f(file_id, error)
!
! Close FORTRAN interface.
!
CALL h5close_f(error)
END PROGRAM DSETEXAMPLE</pre>
<!-- NEW PAGE -->
<pre>
SDSf_tr.out
-----------
HDF5 "SDSf_tr.h5" {
GROUP "/" {
DATASET "Fortran Transpose Matrix" {
DATATYPE H5T_STD_I32LE
DATASPACE SIMPLE { ( 3, 5 ) / ( 3, 5 ) }
DATA {
1, 2, 3, 4, 5,
6, 7, 8, 9, 10,
11, 12, 13, 14, 15
}
}
}
}</pre>
<br /><br />
</body>
</html>
|
