/usr/lib/python2.7/dist-packages/biom/table.py is in python-biom-format 2.1.5+dfsg-7.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 | #!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
BIOM Table (:mod:`biom.table`)
==============================
The biom-format project provides rich ``Table`` objects to support use of the
BIOM file format. The objects encapsulate matrix data (such as OTU counts) and
abstract the interaction away from the programmer.
.. currentmodule:: biom.table
Classes
-------
.. autosummary::
:toctree: generated/
Table
Examples
--------
First, lets create a toy table to play around with. For this example, we're
going to construct a 10x4 `Table`, or one that has 10 observations and 4
samples. Each observation and sample will be given an arbitrary but unique
name. We'll also add on some metadata.
>>> import numpy as np
>>> from biom.table import Table
>>> data = np.arange(40).reshape(10, 4)
>>> sample_ids = ['S%d' % i for i in range(4)]
>>> observ_ids = ['O%d' % i for i in range(10)]
>>> sample_metadata = [{'environment': 'A'}, {'environment': 'B'},
... {'environment': 'A'}, {'environment': 'B'}]
>>> observ_metadata = [{'taxonomy': ['Bacteria', 'Firmicutes']},
... {'taxonomy': ['Bacteria', 'Firmicutes']},
... {'taxonomy': ['Bacteria', 'Proteobacteria']},
... {'taxonomy': ['Bacteria', 'Proteobacteria']},
... {'taxonomy': ['Bacteria', 'Proteobacteria']},
... {'taxonomy': ['Bacteria', 'Bacteroidetes']},
... {'taxonomy': ['Bacteria', 'Bacteroidetes']},
... {'taxonomy': ['Bacteria', 'Firmicutes']},
... {'taxonomy': ['Bacteria', 'Firmicutes']},
... {'taxonomy': ['Bacteria', 'Firmicutes']}]
>>> table = Table(data, observ_ids, sample_ids, observ_metadata,
... sample_metadata, table_id='Example Table')
Now that we have a table, let's explore it at a high level first.
>>> table
10 x 4 <class 'biom.table.Table'> with 39 nonzero entries (97% dense)
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S0 S1 S2 S3
O0 0.0 1.0 2.0 3.0
O1 4.0 5.0 6.0 7.0
O2 8.0 9.0 10.0 11.0
O3 12.0 13.0 14.0 15.0
O4 16.0 17.0 18.0 19.0
O5 20.0 21.0 22.0 23.0
O6 24.0 25.0 26.0 27.0
O7 28.0 29.0 30.0 31.0
O8 32.0 33.0 34.0 35.0
O9 36.0 37.0 38.0 39.0
>>> print table.ids() # doctest: +NORMALIZE_WHITESPACE
['S0' 'S1' 'S2' 'S3']
>>> print table.ids(axis='observation') # doctest: +NORMALIZE_WHITESPACE
['O0' 'O1' 'O2' 'O3' 'O4' 'O5' 'O6' 'O7' 'O8' 'O9']
>>> print table.nnz # number of nonzero entries
39
While it's fun to just poke at the table, let's dig deeper. First, we're going
to convert `table` into relative abundances (within each sample), and then
filter `table` to just the samples associated with environment 'A'. The
filtering gets fancy: we can pass in an arbitrary function to determine what
samples we want to keep. This function must accept a sparse vector of values,
the corresponding ID and the corresponding metadata, and should return ``True``
or ``False``, where ``True`` indicates that the vector should be retained.
>>> normed = table.norm(axis='sample', inplace=False)
>>> filter_f = lambda values, id_, md: md['environment'] == 'A'
>>> env_a = normed.filter(filter_f, axis='sample', inplace=False)
>>> print env_a # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S0 S2
O0 0.0 0.01
O1 0.0222222222222 0.03
O2 0.0444444444444 0.05
O3 0.0666666666667 0.07
O4 0.0888888888889 0.09
O5 0.111111111111 0.11
O6 0.133333333333 0.13
O7 0.155555555556 0.15
O8 0.177777777778 0.17
O9 0.2 0.19
But, what if we wanted individual tables per environment? While we could just
perform some fancy iteration, we can instead just rely on `Table.partition` for
these operations. `partition`, like `filter`, accepts a function. However, the
`partition` method only passes the corresponding ID and metadata to the
function. The function should return what partition the data are a part of.
Within this example, we're also going to sum up our tables over the partitioned
samples. Please note that we're using the original table (ie, not normalized)
here.
>>> part_f = lambda id_, md: md['environment']
>>> env_tables = table.partition(part_f, axis='sample')
>>> for partition, env_table in env_tables:
... print partition, env_table.sum('sample')
A [ 180. 200.]
B [ 190. 210.]
For this last example, and to highlight a bit more functionality, we're going
to first transform the table such that all multiples of three will be retained,
while all non-multiples of three will get set to zero. Following this, we'll
then collpase the table by taxonomy, and then convert the table into
presence/absence data.
First, let's setup the transform. We're going to define a function that takes
the modulus of every value in the vector, and see if it is equal to zero. If it
is equal to zero, we'll keep the value, otherwise we'll set the value to zero.
>>> transform_f = lambda v,i,m: np.where(v % 3 == 0, v, 0)
>>> mult_of_three = tform = table.transform(transform_f, inplace=False)
>>> print mult_of_three # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S0 S1 S2 S3
O0 0.0 0.0 0.0 3.0
O1 0.0 0.0 6.0 0.0
O2 0.0 9.0 0.0 0.0
O3 12.0 0.0 0.0 15.0
O4 0.0 0.0 18.0 0.0
O5 0.0 21.0 0.0 0.0
O6 24.0 0.0 0.0 27.0
O7 0.0 0.0 30.0 0.0
O8 0.0 33.0 0.0 0.0
O9 36.0 0.0 0.0 39.0
Next, we're going to collapse the table over the phylum level taxon. To do
this, we're going to define a helper variable for the index position of the
phylum (see the construction of the table above). Next, we're going to pass
this to `Table.collapse`, and since we want to collapse over the observations,
we'll need to specify 'observation' as the axis.
>>> phylum_idx = 1
>>> collapse_f = lambda id_, md: '; '.join(md['taxonomy'][:phylum_idx + 1])
>>> collapsed = mult_of_three.collapse(collapse_f, axis='observation')
>>> print collapsed # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S0 S1 S2 S3
Bacteria; Firmicutes 7.2 6.6 7.2 8.4
Bacteria; Bacteroidetes 12.0 10.5 0.0 13.5
Bacteria; Proteobacteria 4.0 3.0 6.0 5.0
Finally, let's convert the table to presence/absence data.
>>> pa = collapsed.pa()
>>> print pa # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S0 S1 S2 S3
Bacteria; Firmicutes 1.0 1.0 1.0 1.0
Bacteria; Bacteroidetes 1.0 1.0 0.0 1.0
Bacteria; Proteobacteria 1.0 1.0 1.0 1.0
"""
# -----------------------------------------------------------------------------
# Copyright (c) 2011-2013, The BIOM Format Development Team.
#
# Distributed under the terms of the Modified BSD License.
#
# The full license is in the file COPYING.txt, distributed with this software.
# -----------------------------------------------------------------------------
from __future__ import division
import numpy as np
from copy import deepcopy
from datetime import datetime
from json import dumps
from functools import reduce
from operator import itemgetter, add
from future.builtins import zip
from future.utils import viewitems
from collections import defaultdict, Hashable, Iterable
from numpy import ndarray, asarray, zeros, newaxis
from scipy.sparse import coo_matrix, csc_matrix, csr_matrix, isspmatrix, vstack
from future.utils import string_types
from biom.exception import TableException, UnknownAxisError, UnknownIDError
from biom.util import (get_biom_format_version_string,
get_biom_format_url_string, flatten, natsort,
prefer_self, index_list, H5PY_VLEN_STR, HAVE_H5PY,
__format_version__)
from biom.err import errcheck
from ._filter import _filter
from ._transform import _transform
from ._subsample import _subsample
__author__ = "Daniel McDonald"
__copyright__ = "Copyright 2011-2013, The BIOM Format Development Team"
__credits__ = ["Daniel McDonald", "Jai Ram Rideout", "Greg Caporaso",
"Jose Clemente", "Justin Kuczynski", "Adam Robbins-Pianka",
"Joshua Shorenstein", "Jose Antonio Navas Molina",
"Jorge CaƱardo Alastuey"]
__license__ = "BSD"
__url__ = "http://biom-format.org"
__maintainer__ = "Daniel McDonald"
__email__ = "daniel.mcdonald@colorado.edu"
MATRIX_ELEMENT_TYPE = {'int': int, 'float': float, 'unicode': str,
u'int': int, u'float': float, u'unicode': str}
def general_parser(x):
return x
def vlen_list_of_str_parser(value):
"""Parses the taxonomy value"""
new_value = []
for v in value:
if v:
if isinstance(v, bytes):
v = v.decode('utf8')
new_value.append(v)
return new_value if new_value else None
def general_formatter(grp, header, md, compression):
"""Creates a dataset for a general atomic type category"""
test_val = md[0][header]
shape = (len(md),)
name = 'metadata/%s' % header
if isinstance(test_val, string_types):
grp.create_dataset(name, shape=shape,
dtype=H5PY_VLEN_STR,
data=[m[header].encode('utf8') for m in md],
compression=compression)
else:
grp.create_dataset(
'metadata/%s' % header, shape=(len(md),),
data=[m[header] for m in md],
compression=compression)
def vlen_list_of_str_formatter(grp, header, md, compression):
"""Creates a (N, ?) vlen str dataset"""
# It is possible that the value for some sample/observation
# is None. In that case, we still need to see them as
# iterables, but their length will be 0
iterable_checks = []
lengths = []
for m in md:
if m[header] is None:
iterable_checks.append(True)
else:
iterable_checks.append(
isinstance(m.get(header, []), Iterable))
lengths.append(len(m[header]))
if not np.all(iterable_checks):
raise TypeError(
"Category %s not formatted correctly. Did you pass"
" --process-obs-metadata taxonomy when converting "
" from tsv?")
max_list_len = max(lengths)
shape = (len(md), max_list_len)
data = np.empty(shape, dtype=object)
for i, m in enumerate(md):
if m[header] is None:
continue
value = np.asarray(m[header])
data[i, :len(value)] = [v.encode('utf8') for v in value]
# Change the None entries on data to empty strings ""
data = np.where(data == np.array(None), "", data)
grp.create_dataset(
'metadata/%s' % header, shape=shape,
dtype=H5PY_VLEN_STR, data=data,
compression=compression)
class Table(object):
"""The (canonically pronounced 'teh') Table.
Give in to the power of the Table!
"""
def __init__(self, data, observation_ids, sample_ids,
observation_metadata=None, sample_metadata=None,
table_id=None, type=None, create_date=None, generated_by=None,
observation_group_metadata=None, sample_group_metadata=None,
**kwargs):
self.type = type
self.table_id = table_id
self.create_date = create_date
self.generated_by = generated_by
self.format_version = __format_version__
if not isspmatrix(data):
shape = (len(observation_ids), len(sample_ids))
input_is_dense = kwargs.get('input_is_dense', False)
self._data = Table._to_sparse(data, input_is_dense=input_is_dense,
shape=shape)
else:
self._data = data
# using object to allow for variable length strings
self._sample_ids = np.asarray(sample_ids, dtype=object)
self._observation_ids = np.asarray(observation_ids, dtype=object)
if sample_metadata is not None:
self._sample_metadata = tuple(sample_metadata)
else:
self._sample_metadata = None
if observation_metadata is not None:
self._observation_metadata = tuple(observation_metadata)
else:
self._observation_metadata = None
self._sample_group_metadata = sample_group_metadata
self._observation_group_metadata = observation_group_metadata
errcheck(self)
# These will be set by _index_ids()
self._sample_index = None
self._obs_index = None
self._cast_metadata()
self._index_ids()
def _index_ids(self):
"""Sets lookups {id:index in _data}.
Should only be called in constructor as this modifies state.
"""
self._sample_index = index_list(self._sample_ids)
self._obs_index = index_list(self._observation_ids)
def _index(self, axis='sample'):
"""Return the index lookups of the given axis
Parameters
----------
axis : {'sample', 'observation'}, optional
Axis to get the index dict. Defaults to 'sample'
Returns
-------
dict
lookups {id:index}
Raises
------
UnknownAxisError
If provided an unrecognized axis.
"""
if axis == 'sample':
return self._sample_index
elif axis == 'observation':
return self._obs_index
else:
raise UnknownAxisError(axis)
def _conv_to_self_type(self, vals, transpose=False, dtype=None):
"""For converting vectors to a compatible self type"""
if dtype is None:
dtype = self.dtype
if isspmatrix(vals):
return vals
else:
return Table._to_sparse(vals, transpose, dtype)
@staticmethod
def _to_dense(vec):
"""Converts a row/col vector to a dense numpy array.
Always returns a 1-D row vector for consistency with numpy iteration
over arrays.
"""
dense_vec = np.asarray(vec.todense())
if vec.shape == (1, 1):
# Handle the special case where we only have a single element, but
# we don't want to return a numpy scalar / 0-d array. We still want
# to return a vector of length 1.
return dense_vec.reshape(1)
else:
return np.squeeze(dense_vec)
@staticmethod
def _to_sparse(values, transpose=False, dtype=float, input_is_dense=False,
shape=None):
"""Try to return a populated scipy.sparse matrix.
NOTE: assumes the max value observed in row and col defines the size of
the matrix.
"""
# if it is a vector
if isinstance(values, ndarray) and len(values.shape) == 1:
if transpose:
mat = nparray_to_sparse(values[:, newaxis], dtype)
else:
mat = nparray_to_sparse(values, dtype)
return mat
if isinstance(values, ndarray):
if transpose:
mat = nparray_to_sparse(values.T, dtype)
else:
mat = nparray_to_sparse(values, dtype)
return mat
# the empty list
elif isinstance(values, list) and len(values) == 0:
return coo_matrix((0, 0))
# list of np vectors
elif isinstance(values, list) and isinstance(values[0], ndarray):
mat = list_nparray_to_sparse(values, dtype)
if transpose:
mat = mat.T
return mat
# list of dicts, each representing a row in row order
elif isinstance(values, list) and isinstance(values[0], dict):
mat = list_dict_to_sparse(values, dtype)
if transpose:
mat = mat.T
return mat
# list of scipy.sparse matrices, each representing a row in row order
elif isinstance(values, list) and isspmatrix(values[0]):
mat = list_sparse_to_sparse(values, dtype)
if transpose:
mat = mat.T
return mat
elif isinstance(values, dict):
mat = dict_to_sparse(values, dtype, shape)
if transpose:
mat = mat.T
return mat
elif isinstance(values, list) and isinstance(values[0], list):
if input_is_dense:
d = coo_matrix(values)
mat = coo_arrays_to_sparse((d.data, (d.row, d.col)),
dtype=dtype, shape=shape)
else:
mat = list_list_to_sparse(values, dtype, shape=shape)
return mat
elif isspmatrix(values):
mat = values
if transpose:
mat = mat.transpose()
return mat
else:
raise TableException("Unknown input type")
def _cast_metadata(self):
"""Casts all metadata to defaultdict to support default values.
Should be called after any modifications to sample/observation
metadata.
"""
def cast_metadata(md):
"""Do the actual casting"""
default_md = []
# if we have a list of [None], set to None
if md is not None:
if md.count(None) == len(md):
return None
if md is not None:
for item in md:
d = defaultdict(lambda: None)
if isinstance(item, dict):
d.update(item)
elif item is None:
pass
else:
raise TableException("Unable to cast metadata: %s" %
repr(item))
default_md.append(d)
return tuple(default_md)
return md
self._sample_metadata = cast_metadata(self._sample_metadata)
self._observation_metadata = cast_metadata(self._observation_metadata)
self._sample_group_metadata = (
self._sample_group_metadata
if self._sample_group_metadata else None)
self._observation_group_metadata = (
self._observation_group_metadata
if self._observation_group_metadata else None)
@property
def shape(self):
"""The shape of the underlying contingency matrix"""
return self._data.shape
@property
def dtype(self):
"""The type of the objects in the underlying contingency matrix"""
return self._data.dtype
@property
def nnz(self):
"""Number of non-zero elements of the underlying contingency matrix"""
return self._data.nnz
@property
def matrix_data(self):
"""The sparse matrix object"""
return self._data
def length(self, axis='sample'):
"""Return the length of an axis
Parameters
----------
axis : {'sample', 'observation'}, optional
The axis to operate on
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> from biom import example_table
>>> print example_table.length(axis='sample')
3
>>> print example_table.length(axis='observation')
2
"""
if axis not in ('sample', 'observation'):
raise UnknownAxisError(axis)
return self.shape[1] if axis == 'sample' else self.shape[0]
def add_group_metadata(self, group_md, axis='sample'):
"""Take a dict of group metadata and add it to an axis
Parameters
----------
group_md : dict of tuples
`group_md` should be of the form ``{category: (data type, value)``
axis : {'sample', 'observation'}, optional
The axis to operate on
Raises
------
UnknownAxisError
If provided an unrecognized axis.
"""
if axis == 'sample':
if self._sample_group_metadata is not None:
self._sample_group_metadata.update(group_md)
else:
self._sample_group_metadata = group_md
elif axis == 'observation':
if self._observation_group_metadata is not None:
self._observation_group_metadata.update(group_md)
else:
self._observation_group_metadata = group_md
else:
raise UnknownAxisError(axis)
def add_metadata(self, md, axis='sample'):
"""Take a dict of metadata and add it to an axis.
Parameters
----------
md : dict of dict
`md` should be of the form ``{id: {dict_of_metadata}}``
axis : {'sample', 'observation'}, optional
The axis to operate on
"""
metadata = self.metadata(axis=axis)
if metadata is not None:
for id_, md_entry in viewitems(md):
if self.exists(id_, axis=axis):
idx = self.index(id_, axis=axis)
metadata[idx].update(md_entry)
else:
ids = self.ids(axis=axis)
if axis == 'sample':
self._sample_metadata = tuple(
[md[id_] if id_ in md else None for id_ in ids])
elif axis == 'observation':
self._observation_metadata = tuple(
[md[id_] if id_ in md else None for id_ in ids])
else:
raise UnknownAxisError(axis)
self._cast_metadata()
def __getitem__(self, args):
"""Handles row or column slices
Slicing over an individual axis is supported, but slicing over both
axes at the same time is not supported. Partial slices, such as
`foo[0, 5:10]` are not supported, however full slices are supported,
such as `foo[0, :]`.
Parameters
----------
args : tuple or slice
The specific element (by index position) to return or an entire
row or column of the data.
Returns
-------
float or spmatrix
A float is return if a specific element is specified, otherwise a
spmatrix object representing a vector of sparse data is returned.
Raises
------
IndexError
- If the matrix is empty
- If the arguments do not appear to be a tuple
- If a slice on row and column is specified
- If a partial slice is specified
Notes
-----
Switching between slicing rows and columns is inefficient. Slicing of
rows requires a CSR representation, while slicing of columns requires a
CSC representation, and transforms are performed on the data if the
data are not in the required representation. These transforms can be
expensive if done frequently.
.. shownumpydoc
"""
if self.is_empty():
raise IndexError("Cannot retrieve an element from an empty/null "
"table.")
try:
row, col = args
except:
raise IndexError("Must specify (row, col).")
if isinstance(row, slice) and isinstance(col, slice):
raise IndexError("Can only slice a single axis.")
if isinstance(row, slice):
if row.start is None and row.stop is None:
return self._get_col(col)
else:
raise IndexError("Can only handle full : slices per axis.")
elif isinstance(col, slice):
if col.start is None and col.stop is None:
return self._get_row(row)
else:
raise IndexError("Can only handle full : slices per axis.")
else:
if self._data.getformat() == 'coo':
self._data = self._data.tocsr()
return self._data[row, col]
def _get_row(self, row_idx):
"""Return the row at ``row_idx``.
A row vector will be returned as a scipy.sparse matrix in csr format.
Notes
-----
Switching between slicing rows and columns is inefficient. Slicing of
rows requires a CSR representation, while slicing of columns requires a
CSC representation, and transforms are performed on the data if the
data are not in the required representation. These transforms can be
expensive if done frequently.
"""
self._data = self._data.tocsr()
return self._data.getrow(row_idx)
def _get_col(self, col_idx):
"""Return the column at ``col_idx``.
A column vector will be returned as a scipy.sparse matrix in csc
format.
Notes
-----
Switching between slicing rows and columns is inefficient. Slicing of
rows requires a CSR representation, while slicing of columns requires a
CSC representation, and transforms are performed on the data if the
data are not in the required representation. These transforms can be
expensive if done frequently.
"""
self._data = self._data.tocsc()
return self._data.getcol(col_idx)
def reduce(self, f, axis):
"""Reduce over axis using function `f`
Parameters
----------
f : function
The function to use for the reduce operation
axis : {'sample', 'observation'}
The axis on which to operate
Returns
-------
numpy.array
A one-dimensional array representing the reduced rows
(observations) or columns (samples) of the data matrix
Raises
------
UnknownAxisError
If `axis` is neither "sample" nor "observation"
TableException
If the table's data matrix is empty
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 table
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'foo': 'bar'}, {'x': 'y'}], None)
Create a reduce function
>>> func = lambda x, y: x + y
Reduce table on samples
>>> table.reduce(func, 'sample') # doctest: +NORMALIZE_WHITESPACE
array([ 1., 3., 43.])
Reduce table on observations
>>> table.reduce(func, 'observation') # doctest: +NORMALIZE_WHITESPACE
array([ 1., 46.])
"""
if self.is_empty():
raise TableException("Cannot reduce an empty table")
# np.apply_along_axis might reduce type conversions here and improve
# speed. am opting for reduce right now as I think its more readable
return asarray([reduce(f, v) for v in self.iter_data(axis=axis)])
def sum(self, axis='whole'):
"""Returns the sum by axis
Parameters
----------
axis : {'whole', 'sample', 'observation'}, optional
The axis on which to operate.
Returns
-------
numpy.array or float
If `axis` is "whole", returns an float representing the whole
table sum. If `axis` is either "sample" or "observation", returns a
numpy.array that holds a sum for each sample or observation,
respectively.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Add all values in the table:
>>> table.sum()
array(47.0)
Add all values per sample:
>>> table.sum(axis='sample') # doctest: +NORMALIZE_WHITESPACE
array([ 1., 3., 43.])
Add all values per observation:
>>> table.sum(axis='observation') # doctest: +NORMALIZE_WHITESPACE
array([ 1., 46.])
"""
if axis == 'whole':
axis = None
elif axis == 'sample':
axis = 0
elif axis == 'observation':
axis = 1
else:
raise UnknownAxisError(axis)
matrix_sum = np.squeeze(np.asarray(self._data.sum(axis=axis)))
# We only want to return a scalar if the whole matrix was summed.
if axis is not None and matrix_sum.shape == ():
matrix_sum = matrix_sum.reshape(1)
return matrix_sum
def transpose(self):
"""Transpose the contingency table
The returned table will be an entirely new table, including copies of
the (transposed) data, sample/observation IDs and metadata.
Returns
-------
Table
Return a new table that is the transpose of caller table.
"""
sample_md_copy = deepcopy(self.metadata())
obs_md_copy = deepcopy(self.metadata(axis='observation'))
if self._data.getformat() == 'lil':
# lil's transpose method doesn't have the copy kwarg, but all of
# the others do.
self._data = self._data.tocsr()
# sample ids and observations are reversed becuase we trasposed
return self.__class__(self._data.transpose(copy=True),
self.ids()[:], self.ids(axis='observation')[:],
sample_md_copy, obs_md_copy, self.table_id)
def head(self, n=5, m=5):
"""Get the first n rows and m columns from self
Parameters
----------
n : int, optional
The number of rows (observations) to get. This number must be
greater than 0. If not specified, 5 rows will be retrieved.
m : int, optional
The number of columns (samples) to get. This number must be
greater than 0. If not specified, 5 columns will be
retrieved.
Notes
-----
Like `head` for Linux like systems, requesting more rows (or columns)
than exists will silently work.
Raises
------
IndexError
If `n` or `m` are <= 0.
Returns
-------
Table
The subset table.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
>>> data = np.arange(100).reshape(5, 20)
>>> obs_ids = ['O%d' % i for i in range(1, 6)]
>>> samp_ids = ['S%d' % i for i in range(1, 21)]
>>> table = Table(data, obs_ids, samp_ids)
>>> print table.head() # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3 S4 S5
O1 0.0 1.0 2.0 3.0 4.0
O2 20.0 21.0 22.0 23.0 24.0
O3 40.0 41.0 42.0 43.0 44.0
O4 60.0 61.0 62.0 63.0 64.0
O5 80.0 81.0 82.0 83.0 84.0
"""
if n <= 0:
raise IndexError("n cannot be <= 0.")
if m <= 0:
raise IndexError("m cannot be <= 0.")
row_ids = self.ids(axis='observation')[:n]
col_ids = self.ids(axis='sample')[:m]
table = self.filter(row_ids, axis='observation', inplace=False)
return table.filter(col_ids, axis='sample')
def group_metadata(self, axis='sample'):
"""Return the group metadata of the given axis
Parameters
----------
axis : {'sample', 'observation'}, optional
Axis to search for the group metadata. Defaults to 'sample'
Returns
-------
dict
The corresponding group metadata for the given axis
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table, with group observation metadata and no group
sample metadata:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> group_observation_md = {'tree': ('newick', '(O1:0.3,O2:0.4);')}
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... observation_group_metadata=group_observation_md)
Get the observation group metadata:
>>> table.group_metadata(axis='observation')
{'tree': ('newick', '(O1:0.3,O2:0.4);')}
Get the sample group metadata:
>> table.group_metadata()
None
"""
if axis == 'sample':
return self._sample_group_metadata
elif axis == 'observation':
return self._observation_group_metadata
else:
raise UnknownAxisError(axis)
def ids(self, axis='sample'):
"""Return the ids along the given axis
Parameters
----------
axis : {'sample', 'observation'}, optional
Axis to return ids from. Defaults to 'sample'
Returns
-------
1-D numpy array
The ids along the given axis
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Get the ids along the observation axis:
>>> print table.ids(axis='observation')
['O1' 'O2']
Get the ids along the sample axis:
>>> print table.ids()
['S1' 'S2' 'S3']
"""
if axis == 'sample':
return self._sample_ids
elif axis == 'observation':
return self._observation_ids
else:
raise UnknownAxisError(axis)
def update_ids(self, id_map, axis='sample', strict=True, inplace=True):
"""Update the ids along the given axis
Parameters
----------
id_map : dict
Mapping of old to new ids
axis : {'sample', 'observation'}, optional
Axis to search for `id`. Defaults to 'sample'
strict : bool, optional
If ``True``, raise an error if an id is present in the given axis
but is not a key in ``id_map``. If False, retain old identifier
for ids that are present in the given axis but are not keys in
``id_map``.
inplace : bool, optional
If ``True`` the ids are updated in ``self``; if ``False`` the ids
are updated in a new table is returned.
Returns
-------
Table
Table object where ids have been updated.
Raises
------
UnknownAxisError
If provided an unrecognized axis.
TableException
If an id from ``self`` is not in ``id_map`` and ``strict`` is
``True``.
Examples
--------
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Define a mapping of old to new sample ids:
>>> id_map = {'S1':'s1.1', 'S2':'s2.2', 'S3':'s3.3'}
Get the ids along the sample axis in the table:
>>> print table.ids(axis='sample')
['S1' 'S2' 'S3']
Update the sample ids and get the ids along the sample axis in the
updated table:
>>> updated_table = table.update_ids(id_map, axis='sample')
>>> print updated_table.ids(axis='sample')
['s1.1' 's2.2' 's3.3']
"""
updated_ids = zeros(self.ids(axis=axis).size, dtype=object)
for idx, old_id in enumerate(self.ids(axis=axis)):
if strict and old_id not in id_map:
raise TableException(
"Mapping not provided for %s identifier: %s. If this "
"identifier should not be updated, pass strict=False."
% (axis, old_id))
updated_ids[idx] = id_map.get(old_id, old_id)
# prepare the result object and update the ids along the specified
# axis
result = self if inplace else self.copy()
if axis == 'sample':
result._sample_ids = updated_ids
else:
result._observation_ids = updated_ids
result._index_ids()
# check for errors (specifically, we want to esnsure that duplicate
# ids haven't been introduced)
errcheck(result)
return result
def _get_sparse_data(self, axis='sample'):
"""Returns the internal data in the correct sparse representation
Parameters
----------
axis : {'sample', 'observation'}, optional
Axis to search for `id`. Defaults to 'sample'
Returns
-------
sparse matrix
The data in csc (axis='sample') or csr (axis='observation')
representation
"""
if axis == 'sample':
return self._data.tocsc()
elif axis == 'observation':
return self._data.tocsr()
else:
raise UnknownAxisError(axis)
def metadata(self, id=None, axis='sample'):
"""Return the metadata of the identified sample/observation.
Parameters
----------
id : str
ID of the sample or observation whose index will be returned.
axis : {'sample', 'observation'}
Axis to search for `id`.
Returns
-------
defaultdict or None
The corresponding metadata ``defaultdict`` or ``None`` of that axis
does not have metadata.
Raises
------
UnknownAxisError
If provided an unrecognized axis.
UnknownIDError
If provided an unrecognized sample/observation ID.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table, with observation metadata and no sample
metadata:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'foo': 'bar'}, {'x': 'y'}], None)
Get the metadata of the observation with ID "O2":
>>> # casting to `dict` as the return is `defaultdict`
>>> dict(table.metadata('O2', 'observation'))
{'x': 'y'}
Get the metadata of the sample with ID "S1":
>>> table.metadata('S1', 'sample') is None
True
"""
if axis == 'sample':
md = self._sample_metadata
elif axis == 'observation':
md = self._observation_metadata
else:
raise UnknownAxisError(axis)
if id is None:
return md
idx = self.index(id, axis=axis)
return md[idx] if md is not None else None
def index(self, id, axis):
"""Return the index of the identified sample/observation.
Parameters
----------
id : str
ID of the sample or observation whose index will be returned.
axis : {'sample', 'observation'}
Axis to search for `id`.
Returns
-------
int
Index of the sample/observation identified by `id`.
Raises
------
UnknownAxisError
If provided an unrecognized axis.
UnknownIDError
If provided an unrecognized sample/observation ID.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Get the index of the observation with ID "O2":
>>> table.index('O2', 'observation')
1
Get the index of the sample with ID "S1":
>>> table.index('S1', 'sample')
0
"""
idx_lookup = self._index(axis=axis)
if id not in idx_lookup:
raise UnknownIDError(id, axis)
return idx_lookup[id]
def get_value_by_ids(self, obs_id, samp_id):
"""Return value in the matrix corresponding to ``(obs_id, samp_id)``
Parameters
----------
obs_id : str
The ID of the observation
samp_id : str
The ID of the sample
Returns
-------
float
The data value corresponding to the specified matrix position
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'Z3'])
Retrieve the number of counts for observation `O1` in sample `Z3`.
>>> print table.get_value_by_ids('O2', 'Z3')
42.0
See Also
--------
Table.data
"""
return self[self.index(obs_id, 'observation'),
self.index(samp_id, 'sample')]
def __str__(self):
"""Stringify self
Default str output for a Table is just row/col ids and data values
"""
return self.delimited_self()
def __repr__(self):
"""Returns a high-level summary of the table's properties
Returns
-------
str
A string detailing the shape, class, number of nonzero entries, and
table density
"""
rows, cols = self.shape
return '%d x %d %s with %d nonzero entries (%d%% dense)' % (
rows, cols, repr(self.__class__), self.nnz,
self.get_table_density() * 100
)
def exists(self, id, axis="sample"):
"""Returns whether id exists in axis
Parameters
----------
id: str
id to check if exists
axis : {'sample', 'observation'}, optional
The axis to check
Returns
-------
bool
``True`` if `id` exists, ``False`` otherwise
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Check whether sample ID is in the table:
>>> table.exists('S1')
True
>>> table.exists('S4')
False
Check whether an observation ID is in the table:
>>> table.exists('O1', 'observation')
True
>>> table.exists('O3', 'observation')
False
"""
return id in self._index(axis=axis)
def delimited_self(self, delim=u'\t', header_key=None, header_value=None,
metadata_formatter=str,
observation_column_name=u'#OTU ID'):
"""Return self as a string in a delimited form
Default str output for the Table is just row/col ids and table data
without any metadata
Including observation metadata in output: If ``header_key`` is not
``None``, the observation metadata with that name will be included
in the delimited output. If ``header_value`` is also not ``None``, the
observation metadata will use the provided ``header_value`` as the
observation metadata name (i.e., the column header) in the delimited
output.
``metadata_formatter``: a function which takes a metadata entry and
returns a formatted version that should be written to file
``observation_column_name``: the name of the first column in the output
table, corresponding to the observation IDs. For example, the default
will look something like:
#OTU ID\tSample1\tSample2
OTU1\t10\t2
OTU2\t4\t8
"""
def to_utf8(i):
if isinstance(i, bytes):
return i.decode('utf8')
else:
return str(i)
if self.is_empty():
raise TableException("Cannot delimit self if I don't have data...")
samp_ids = delim.join([to_utf8(i) for i in self.ids()])
# 17 hrs of straight programming later...
if header_key is not None:
if header_value is None:
raise TableException(
"You need to specify both header_key and header_value")
if header_value is not None:
if header_key is None:
raise TableException(
"You need to specify both header_key and header_value")
if header_value:
output = [u'# Constructed from biom file',
u'%s%s%s\t%s' % (observation_column_name, delim,
samp_ids, header_value)]
else:
output = ['# Constructed from biom file',
'%s%s%s' % (observation_column_name, delim, samp_ids)]
obs_metadata = self.metadata(axis='observation')
for obs_id, obs_values in zip(self.ids(axis='observation'),
self._iter_obs()):
str_obs_vals = delim.join(map(str, self._to_dense(obs_values)))
obs_id = to_utf8(obs_id)
if header_key and obs_metadata is not None:
md = obs_metadata[self._obs_index[obs_id]]
md_out = metadata_formatter(md.get(header_key, None))
output.append(
u'%s%s%s\t%s' %
(obs_id, delim, str_obs_vals, md_out))
else:
output.append(u'%s%s%s' % (obs_id, delim, str_obs_vals))
return '\n'.join(output)
def is_empty(self):
"""Check whether the table is empty
Returns
-------
bool
``True`` if the table is empty, ``False`` otherwise
"""
if not self.ids().size or not self.ids(axis='observation').size:
return True
else:
return False
def __iter__(self):
"""See ``biom.table.Table.iter``"""
return self.iter()
def _iter_samp(self):
"""Return sample vectors of data matrix vectors"""
for c in range(self.shape[1]):
# this pulls out col vectors but need to convert to the expected
# row vector
colvec = self._get_col(c)
yield colvec.transpose(copy=True)
def _iter_obs(self):
"""Return observation vectors of data matrix"""
for r in range(self.shape[0]):
yield self._get_row(r)
def get_table_density(self):
"""Returns the fraction of nonzero elements in the table.
Returns
-------
float
The fraction of nonzero elements in the table
"""
density = 0.0
if not self.is_empty():
density = (self.nnz /
(len(self.ids()) * len(self.ids(axis='observation'))))
return density
def descriptive_equality(self, other):
"""For use in testing, describe how the tables are not equal"""
if not isinstance(other, self.__class__):
return "Tables are not of comparable classes"
if not self.type == other.type:
return "Tables are not the same type"
if not np.array_equal(self.ids(axis='observation'),
other.ids(axis='observation')):
return "Observation IDs are not the same"
if not np.array_equal(self.ids(), other.ids()):
return "Sample IDs are not the same"
if not np.array_equal(self.metadata(axis='observation'),
other.metadata(axis='observation')):
return "Observation metadata are not the same"
if not np.array_equal(self.metadata(), other.metadata()):
return "Sample metadata are not the same"
if not self._data_equality(other._data):
return "Data elements are not the same"
return "Tables appear equal"
def __eq__(self, other):
"""Equality is determined by the data matrix, metadata, and IDs"""
if not isinstance(other, self.__class__):
return False
if self.type != other.type:
return False
if not np.array_equal(self.ids(axis='observation'),
other.ids(axis='observation')):
return False
if not np.array_equal(self.ids(), other.ids()):
return False
if not np.array_equal(self.metadata(axis='observation'),
other.metadata(axis='observation')):
return False
if not np.array_equal(self.metadata(), other.metadata()):
return False
if not self._data_equality(other._data):
return False
return True
def _data_equality(self, other):
"""Return ``True`` if both matrices are equal.
Matrices are equal iff the following items are equal:
- shape
- dtype
- size (nnz)
- matrix data (more expensive, so checked last)
The sparse format does not need to be the same between the two
matrices. ``self`` and ``other`` will be converted to csr format if
necessary before performing the final comparison.
"""
if self._data.shape != other.shape:
return False
if self._data.dtype != other.dtype:
return False
if self._data.nnz != other.nnz:
return False
self._data = self._data.tocsr()
other = other.tocsr()
if (self._data != other).nnz > 0:
return False
return True
def __ne__(self, other):
return not (self == other)
def data(self, id, axis='sample', dense=True):
"""Returns data associated with an `id`
Parameters
----------
id : str
ID of the samples or observations whose data will be returned.
axis : {'sample', 'observation'}
Axis to search for `id`.
dense : bool, optional
If ``True``, return data as dense
Returns
-------
np.ndarray or scipy.sparse.spmatrix
np.ndarray if ``dense``, otherwise scipy.sparse.spmatrix
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> from biom import example_table
>>> example_table.data('S1', axis='sample')
array([ 0., 3.])
See Also
--------
Table.get_value_by_ids
"""
if axis == 'sample':
data = self[:, self.index(id, 'sample')]
elif axis == 'observation':
data = self[self.index(id, 'observation'), :]
else:
raise UnknownAxisError(axis)
if dense:
return self._to_dense(data)
else:
return data
def copy(self):
"""Returns a copy of the table"""
return self.__class__(self._data.copy(),
self.ids(axis='observation').copy(),
self.ids().copy(),
deepcopy(self.metadata(axis='observation')),
deepcopy(self.metadata()),
self.table_id,
type=self.type)
def iter_data(self, dense=True, axis='sample'):
"""Yields axis values
Parameters
----------
dense : bool, optional
Defaults to ``True``. If ``False``, yield compressed sparse row or
compressed sparse columns if `axis` is 'observation' or 'sample',
respectively.
axis : {'sample', 'observation'}, optional
Axis to iterate over.
Returns
-------
generator
Yields list of values for each value in `axis`
Raises
------
UnknownAxisError
If axis other than 'sample' or 'observation' passed
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
>>> data = np.arange(30).reshape(3,10) # 3 X 10 OTU X Sample table
>>> obs_ids = ['o1', 'o2', 'o3']
>>> sam_ids = ['s%i' %i for i in range(1,11)]
>>> bt = Table(data, observation_ids=obs_ids, sample_ids=sam_ids)
Lets find the sample with the largest sum
>>> sample_gen = bt.iter_data(axis='sample')
>>> max_sample_count = max([sample.sum() for sample in sample_gen])
>>> print max_sample_count
57.0
"""
if axis == "sample":
for samp_v in self._iter_samp():
if dense:
yield self._to_dense(samp_v)
else:
yield samp_v
elif axis == "observation":
for obs_v in self._iter_obs():
if dense:
yield self._to_dense(obs_v)
else:
yield obs_v
else:
raise UnknownAxisError(axis)
def iter(self, dense=True, axis='sample'):
"""Yields ``(value, id, metadata)``
Parameters
----------
dense : bool, optional
Defaults to ``True``. If ``False``, yield compressed sparse row or
compressed sparse columns if `axis` is 'observation' or 'sample',
respectively.
axis : {'sample', 'observation'}, optional
The axis to iterate over.
Returns
-------
GeneratorType
A generator that yields (values, id, metadata)
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'Z3'])
Iter over samples and keep those that start with an Z:
>>> [(values, id, metadata)
... for values, id, metadata in table.iter() if id[0]=='Z']
[(array([ 1., 42.]), 'Z3', None)]
Iter over observations and add the 2nd column of the values
>>> col = [values[1] for values, id, metadata in table.iter()]
>>> sum(col)
46.0
"""
ids = self.ids(axis=axis)
metadata = self.metadata(axis=axis)
if axis == 'sample':
iter_ = self._iter_samp()
elif axis == 'observation':
iter_ = self._iter_obs()
else:
raise UnknownAxisError(axis)
if metadata is None:
metadata = (None,) * len(ids)
iter_ = self.iter_data(axis=axis, dense=dense)
return zip(iter_, ids, metadata)
def iter_pairwise(self, dense=True, axis='sample', tri=True, diag=False):
"""Pairwise iteration over self
Parameters
----------
dense : bool, optional
Defaults to ``True``. If ``False``, yield compressed sparse row or
compressed sparse columns if `axis` is 'observation' or 'sample',
respectively.
axis : {'sample', 'observation'}, optional
The axis to iterate over.
tri : bool, optional
If ``True``, just yield [i, j] and not [j, i]
diag : bool, optional
If ``True``, yield [i, i]
Returns
-------
GeneratorType
Yields [(val_i, id_i, metadata_i), (val_j, id_j, metadata_j)]
Raises
------
UnknownAxisError
Examples
--------
>>> from biom import example_table
By default, only the upper triangle without the diagonal of the
resulting pairwise combinations is yielded.
>>> iter_ = example_table.iter_pairwise()
>>> for (val_i, id_i, md_i), (val_j, id_j, md_j) in iter_:
... print id_i, id_j
S1 S2
S1 S3
S2 S3
The full pairwise combinations can also be yielded though.
>>> iter_ = example_table.iter_pairwise(tri=False, diag=True)
>>> for (val_i, id_i, md_i), (val_j, id_j, md_j) in iter_:
... print id_i, id_j
S1 S1
S1 S2
S1 S3
S2 S1
S2 S2
S2 S3
S3 S1
S3 S2
S3 S3
"""
metadata = self.metadata(axis=axis)
ids = self.ids(axis=axis)
if metadata is None:
metadata = (None,) * len(ids)
ind = np.arange(len(ids))
diag_v = 1 - diag # for offseting tri_f, where a 0 includes the diag
if tri:
def tri_f(idx):
return ind[idx+diag_v:]
else:
def tri_f(idx):
return np.hstack([ind[:idx], ind[idx+diag_v:]])
for idx, i in enumerate(ind):
id_i = ids[i]
md_i = metadata[i]
data_i = self.data(id_i, axis=axis, dense=dense)
for j in tri_f(idx):
id_j = ids[j]
md_j = metadata[j]
data_j = self.data(id_j, axis=axis, dense=dense)
yield ((data_i, id_i, md_i), (data_j, id_j, md_j))
def sort_order(self, order, axis='sample'):
"""Return a new table with `axis` in `order`
Parameters
----------
order : iterable
The desired order for axis
axis : {'sample', 'observation'}, optional
The axis to operate on
Returns
-------
Table
A table where the observations or samples are sorted according to
`order`
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[1, 0, 4], [1, 3, 0]])
>>> table = Table(data, ['O2', 'O1'], ['S2', 'S1', 'S3'])
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S2 S1 S3
O2 1.0 0.0 4.0
O1 1.0 3.0 0.0
Sort the table using a list of samples:
>>> sorted_table = table.sort_order(['S2', 'S3', 'S1'])
>>> print sorted_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S2 S3 S1
O2 1.0 4.0 0.0
O1 1.0 0.0 3.0
Additionally you could sort the table's observations:
>>> sorted_table = table.sort_order(['O1', 'O2'], axis="observation")
>>> print sorted_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S2 S1 S3
O1 1.0 3.0 0.0
O2 1.0 0.0 4.0
"""
md = []
vals = []
metadata = self.metadata(axis=axis)
if axis == 'sample':
for id_ in order:
cur_idx = self.index(id_, 'sample')
vals.append(self._to_dense(self[:, cur_idx]))
if metadata is not None:
md.append(metadata[cur_idx])
if not md:
md = None
return self.__class__(self._conv_to_self_type(vals,
transpose=True),
self.ids(axis='observation')[:], order[:],
self.metadata(axis='observation'), md,
self.table_id, self.type)
elif axis == 'observation':
for id_ in order:
cur_idx = self.index(id_, 'observation')
vals.append(self[cur_idx, :])
if metadata is not None:
md.append(metadata[cur_idx])
if not md:
md = None
return self.__class__(self._conv_to_self_type(vals),
order[:], self.ids()[:],
md, self.metadata(), self.table_id,
self.type)
else:
raise UnknownAxisError(axis)
def sort(self, sort_f=natsort, axis='sample'):
"""Return a table sorted along axis
Parameters
----------
sort_f : function, optional
Defaults to ``biom.util.natsort``. A function that takes a list of
values and sorts it
axis : {'sample', 'observation'}, optional
The axis to operate on
Returns
-------
biom.Table
A table whose samples or observations are sorted according to the
`sort_f` function
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table:
>>> data = np.asarray([[1, 0, 4], [1, 3, 0]])
>>> table = Table(data, ['O2', 'O1'], ['S2', 'S1', 'S3'])
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S2 S1 S3
O2 1.0 0.0 4.0
O1 1.0 3.0 0.0
Sort the order of samples in the table using the default natural
sorting:
>>> new_table = table.sort()
>>> print new_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O2 0.0 1.0 4.0
O1 3.0 1.0 0.0
Sort the order of observations in the table using the default natural
sorting:
>>> new_table = table.sort(axis='observation')
>>> print new_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S2 S1 S3
O1 1.0 3.0 0.0
O2 1.0 0.0 4.0
Sort the samples in reverse order using a custom sort function:
>>> sort_f = lambda x: list(sorted(x, reverse=True))
>>> new_table = table.sort(sort_f=sort_f)
>>> print new_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S3 S2 S1
O2 4.0 1.0 0.0
O1 0.0 1.0 3.0
"""
return self.sort_order(sort_f(self.ids(axis=axis)), axis=axis)
def filter(self, ids_to_keep, axis='sample', invert=False, inplace=True):
"""Filter a table based on a function or iterable.
Parameters
----------
ids_to_keep : iterable, or function(values, id, metadata) -> bool
If a function, it will be called with the values of the
sample/observation, its id (a string) and the dictionary
of metadata of each sample/observation, and must return a
boolean. If it's an iterable, it must be a list of ids to
keep.
axis : {'sample', 'observation'}, optional
It controls whether to filter samples or observations and
defaults to "sample".
invert : bool, optional
Defaults to ``False``. If set to ``True``, discard samples or
observations where `ids_to_keep` returns True
inplace : bool, optional
Defaults to ``True``. Whether to return a new table or modify
itself.
Returns
-------
biom.Table
Returns itself if `inplace`, else returns a new filtered table.
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table, with observation metadata and sample
metadata:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'full_genome_available': True},
... {'full_genome_available': False}],
... [{'sample_type': 'a'}, {'sample_type': 'a'},
... {'sample_type': 'b'}])
Define a function to keep only samples with sample_type == 'a'. This
will drop sample S3, which has sample_type 'b':
>>> filter_fn = lambda val, id_, md: md['sample_type'] == 'a'
Get a filtered version of the table, leaving the original table
untouched:
>>> new_table = table.filter(filter_fn, inplace=False)
>>> print table.ids()
['S1' 'S2' 'S3']
>>> print new_table.ids()
['S1' 'S2']
Using the same filtering function, discard all samples with sample_type
'a'. This will keep only sample S3, which has sample_type 'b':
>>> new_table = table.filter(filter_fn, inplace=False, invert=True)
>>> print table.ids()
['S1' 'S2' 'S3']
>>> print new_table.ids()
['S3']
Filter the table in-place using the same function (drop all samples
where sample_type is not 'a'):
>>> table.filter(filter_fn)
2 x 2 <class 'biom.table.Table'> with 2 nonzero entries (50% dense)
>>> print table.ids()
['S1' 'S2']
Filter out all observations in the table that do not have
full_genome_available == True. This will filter out observation O2:
>>> filter_fn = lambda val, id_, md: md['full_genome_available']
>>> table.filter(filter_fn, axis='observation')
1 x 2 <class 'biom.table.Table'> with 0 nonzero entries (0% dense)
>>> print table.ids(axis='observation')
['O1']
"""
table = self if inplace else self.copy()
metadata = table.metadata(axis=axis)
ids = table.ids(axis=axis)
index = self._index(axis=axis)
axis = table._axis_to_num(axis=axis)
arr = table._data
arr, ids, metadata = _filter(arr,
ids,
metadata,
index,
ids_to_keep,
axis,
invert=invert)
table._data = arr
if axis == 1:
table._sample_ids = ids
table._sample_metadata = metadata
elif axis == 0:
table._observation_ids = ids
table._observation_metadata = metadata
table._index_ids()
errcheck(table)
return table
def partition(self, f, axis='sample'):
"""Yields partitions
Parameters
----------
f : function
`f` is given the ID and metadata of the vector and must return
what partition the vector is part of.
axis : {'sample', 'observation'}, optional
The axis to iterate over
Returns
-------
GeneratorType
A generator that yields (partition, `Table`)
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
>>> from biom.util import unzip
Create a 2x3 BIOM table, with observation metadata and sample
metadata:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'full_genome_available': True},
... {'full_genome_available': False}],
... [{'sample_type': 'a'}, {'sample_type': 'a'},
... {'sample_type': 'b'}])
Define a function to bin by sample_type
>>> f = lambda id_, md: md['sample_type']
Partition the table and view results
>>> bins, tables = table.partition(f)
>>> print bins[1] # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 0.0 0.0
O2 1.0 3.0
>>> print tables[1] # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S3
O1 1.0
O2 42.0
"""
partitions = {}
# conversion of vector types is not necessary, vectors are not
# being passed to an arbitrary function
for vals, id_, md in self.iter(dense=False, axis=axis):
part = f(id_, md)
# try to make it hashable...
if not isinstance(part, Hashable):
part = tuple(part)
if part not in partitions:
partitions[part] = [[], [], []]
partitions[part][0].append(id_)
partitions[part][1].append(vals)
partitions[part][2].append(md)
md = self.metadata(axis=self._invert_axis(axis))
for part, (ids, values, metadata) in viewitems(partitions):
if axis == 'sample':
data = self._conv_to_self_type(values, transpose=True)
samp_ids = ids
samp_md = metadata
obs_ids = self.ids(axis='observation')[:]
obs_md = md[:] if md is not None else None
elif axis == 'observation':
data = self._conv_to_self_type(values, transpose=False)
obs_ids = ids
obs_md = metadata
samp_ids = self.ids()[:]
samp_md = md[:] if md is not None else None
yield part, Table(data, obs_ids, samp_ids, obs_md, samp_md,
self.table_id, type=self.type)
def collapse(self, f, collapse_f=None, norm=True, min_group_size=1,
include_collapsed_metadata=True, one_to_many=False,
one_to_many_mode='add', one_to_many_md_key='Path',
strict=False, axis='sample'):
"""Collapse partitions in a table by metadata or by IDs
Partition data by metadata or IDs and then collapse each partition into
a single vector.
If `include_collapsed_metadata` is ``True``, the metadata for the
collapsed partition will be a category named 'collapsed_ids', in which
a list of the original ids that made up the partition is retained
The remainder is only relevant to setting `one_to_many` to ``True``.
If `one_to_many` is ``True``, allow vectors to collapse into multiple
bins if the metadata describe a one-many relationship. Supplied
functions must allow for iteration support over the metadata key and
must return a tuple of (path, bin) as to describe both the path in the
hierarchy represented and the specific bin being collapsed into. The
uniqueness of the bin is _not_ based on the path but by the name of the
bin.
The metadata value for the corresponding collapsed column may include
more (or less) information about the collapsed data. For example, if
collapsing "FOO", and there are vectors that span three associations A,
B, and C, such that vector 1 spans A and B, vector 2 spans B and C and
vector 3 spans A and C, the resulting table will contain three
collapsed vectors:
- A, containing original vectors 1 and 3
- B, containing original vectors 1 and 2
- C, containing original vectors 2 and 3
If a vector maps to the same partition multiple times, it will be
counted multiple times.
There are two supported modes for handling one-to-many relationships
via `one_to_many_mode`: ``add`` and `divide`. ``add`` will add the
vector counts to each partition that the vector maps to, which may
increase the total number of counts in the output table. ``divide``
will divide a vectors's counts by the number of metadata that the
vector has before adding the counts to each partition. This will not
increase the total number of counts in the output table.
If `one_to_many_md_key` is specified, that becomes the metadata
key that describes the collapsed path. If a value is not specified,
then it defaults to 'Path'.
If `strict` is specified, then all metadata pathways operated on
must be indexable by `metadata_f`.
`one_to_many` and `norm` are not supported together.
`one_to_many` and `collapse_f` are not supported together.
`one_to_many` and `min_group_size` are not supported together.
A final note on space consumption. At present, the `one_to_many`
functionality requires a temporary dense matrix representation.
Parameters
----------
f : function
Function that is used to determine what partition a vector belongs
to
collapse_f : function, optional
Function that collapses a partition in a one-to-one collapse. The
expected function signature is:
dense or sparse_vector <- collapse_f(Table, axis)
Defaults to a pairwise add.
norm : bool, optional
Defaults to ``True``. If ``True``, normalize the resulting table
min_group_size : int, optional
Defaults to ``1``. The minimum size of a partition when performing
a one-to-one collapse
include_collapsed_metadata : bool, optional
Defaults to ``True``. If ``True``, retain the collapsed metadata
keyed by the original IDs of the associated vectors
one_to_many : bool, optional
Defaults to ``False``. Perform a one-to-many collapse
one_to_many_mode : {'add', 'divide'}, optional
The way to reduce two vectors in a one-to-many collapse
one_to_many_md_key : str, optional
Defaults to "Path". If `include_collapsed_metadata` is ``True``,
store the original vector metadata under this key
strict : bool, optional
Defaults to ``False``. Requires full pathway data within a
one-to-many structure
axis : {'sample', 'observation'}, optional
The axis to collapse
Returns
-------
Table
The collapsed table
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a ``Table``
>>> dt_rich = Table(
... np.array([[5, 6, 7], [8, 9, 10], [11, 12, 13]]),
... ['1', '2', '3'], ['a', 'b', 'c'],
... [{'taxonomy': ['k__a', 'p__b']},
... {'taxonomy': ['k__a', 'p__c']},
... {'taxonomy': ['k__a', 'p__c']}],
... [{'barcode': 'aatt'},
... {'barcode': 'ttgg'},
... {'barcode': 'aatt'}])
>>> print dt_rich # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID a b c
1 5.0 6.0 7.0
2 8.0 9.0 10.0
3 11.0 12.0 13.0
Create Function to determine what partition a vector belongs to
>>> bin_f = lambda id_, x: x['taxonomy'][1]
>>> obs_phy = dt_rich.collapse(
... bin_f, norm=False, min_group_size=1,
... axis='observation').sort(axis='observation')
>>> print obs_phy # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID a b c
p__b 5.0 6.0 7.0
p__c 19.0 21.0 23.0
"""
collapsed_data = []
collapsed_ids = []
if include_collapsed_metadata:
collapsed_md = []
else:
collapsed_md = None
if one_to_many_mode not in ['add', 'divide']:
raise ValueError("Unrecognized one-to-many mode '%s'. Must be "
"either 'add' or 'divide'." % one_to_many_mode)
# transpose is only necessary in the one-to-one case
# new_data_shape is only necessary in the one-to-many case
# axis_slice is only necessary in the one-to-many case
def axis_ids_md(t):
return (t.ids(axis=axis), t.metadata(axis=axis))
if axis == 'sample':
transpose = True
def new_data_shape(ids, collapsed):
return (len(ids), len(collapsed))
def axis_slice(lookup, key):
return (slice(None), lookup[key])
elif axis == 'observation':
transpose = False
def new_data_shape(ids, collapsed):
return (len(collapsed), len(ids))
def axis_slice(lookup, key):
return (lookup[key], slice(None))
else:
raise UnknownAxisError(axis)
if one_to_many:
if norm:
raise AttributeError(
"norm and one_to_many are not supported together")
# determine the collapsed pathway
# we drop all other associated metadata
new_md = {}
md_count = {}
for id_, md in zip(*axis_ids_md(self)):
md_iter = f(id_, md)
num_md = 0
while True:
try:
pathway, partition = next(md_iter)
except IndexError:
# if a pathway is incomplete
if strict:
# bail if strict
err = "Incomplete pathway, ID: %s, metadata: %s" %\
(id_, md)
raise IndexError(err)
else:
# otherwise ignore
continue
except StopIteration:
break
new_md[partition] = pathway
num_md += 1
md_count[id_] = num_md
idx_lookup = {part: i for i, part in enumerate(sorted(new_md))}
# We need to store floats, not ints, as things won't always divide
# evenly.
dtype = float if one_to_many_mode == 'divide' else self.dtype
new_data = zeros(new_data_shape(self.ids(self._invert_axis(axis)),
new_md),
dtype=dtype)
# for each vector
# for each bin in the metadata
# for each partition associated with the vector
for vals, id_, md in self.iter(axis=axis):
md_iter = f(id_, md)
while True:
try:
pathway, part = next(md_iter)
except IndexError:
# if a pathway is incomplete
if strict:
# bail if strict, should never get here...
err = "Incomplete pathway, ID: %s, metadata: %s" %\
(id_, md)
raise IndexError(err)
else:
# otherwise ignore
continue
except StopIteration:
break
if one_to_many_mode == 'add':
new_data[axis_slice(idx_lookup, part)] += vals
else:
new_data[axis_slice(idx_lookup, part)] += \
vals / md_count[id_]
if include_collapsed_metadata:
# reassociate pathway information
for k, i in sorted(viewitems(idx_lookup), key=itemgetter(1)):
collapsed_md.append({one_to_many_md_key: new_md[k]})
# get the new sample IDs
collapsed_ids = [k for k, i in sorted(viewitems(idx_lookup),
key=itemgetter(1))]
# convert back to self type
data = self._conv_to_self_type(new_data)
else:
if collapse_f is None:
def collapse_f(t, axis):
return t.reduce(add, axis)
for part, table in self.partition(f, axis=axis):
axis_ids, axis_md = axis_ids_md(table)
if len(axis_ids) < min_group_size:
continue
redux_data = collapse_f(table, self._invert_axis(axis))
if norm:
redux_data /= len(axis_ids)
collapsed_data.append(self._conv_to_self_type(redux_data))
collapsed_ids.append(part)
if include_collapsed_metadata:
# retain metadata but store by original id
collapsed_md.append({'collapsed_ids': axis_ids.tolist()})
# tmp_md = {}
# for id_, md in izip(axis_ids, axis_md):
# tmp_md[id_] = md
# collapsed_md.append(tmp_md)
data = self._conv_to_self_type(collapsed_data, transpose=transpose)
# if the table is empty
errcheck(self, 'empty')
md = self.metadata(axis=self._invert_axis(axis))
if axis == 'sample':
sample_ids = collapsed_ids
sample_md = collapsed_md
obs_ids = self.ids(axis='observation')[:]
obs_md = md if md is not None else None
else:
sample_ids = self.ids()[:]
obs_ids = collapsed_ids
obs_md = collapsed_md
sample_md = md if md is not None else None
return Table(data, obs_ids, sample_ids, obs_md, sample_md,
self.table_id, type=self.type)
def _invert_axis(self, axis):
"""Invert an axis"""
if axis == 'sample':
return 'observation'
elif axis == 'observation':
return 'sample'
else:
return UnknownAxisError(axis)
def _axis_to_num(self, axis):
"""Convert str axis to numerical axis"""
if axis == 'sample':
return 1
elif axis == 'observation':
return 0
else:
raise UnknownAxisError(axis)
def min(self, axis='sample'):
"""Get the minimum nonzero value over an axis
Parameters
----------
axis : {'sample', 'observation', 'whole'}, optional
Defaults to "sample". The axis over which to calculate minima.
Returns
-------
scalar of self.dtype or np.array of self.dtype
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> from biom import example_table
>>> print example_table.min(axis='sample')
[ 3. 1. 2.]
"""
if axis not in ['sample', 'observation', 'whole']:
raise UnknownAxisError(axis)
if axis == 'whole':
min_val = np.inf
for data in self.iter_data(dense=False):
# only min over the actual nonzero values
min_val = min(min_val, data.data.min())
else:
min_val = zeros(len(self.ids(axis=axis)), dtype=self.dtype)
for idx, data in enumerate(self.iter_data(dense=False, axis=axis)):
min_val[idx] = data.data.min()
return min_val
def max(self, axis='sample'):
"""Get the maximum nonzero value over an axis
Parameters
----------
axis : {'sample', 'observation', 'whole'}, optional
Defaults to "sample". The axis over which to calculate maxima.
Returns
-------
scalar of self.dtype or np.array of self.dtype
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> from biom import example_table
>>> print example_table.max(axis='observation')
[ 2. 5.]
"""
if axis not in ['sample', 'observation', 'whole']:
raise UnknownAxisError(axis)
if axis == 'whole':
max_val = -np.inf
for data in self.iter_data(dense=False):
# only min over the actual nonzero values
max_val = max(max_val, data.data.max())
else:
max_val = np.empty(len(self.ids(axis=axis)), dtype=self.dtype)
for idx, data in enumerate(self.iter_data(dense=False, axis=axis)):
max_val[idx] = data.data.max()
return max_val
def subsample(self, n, axis='sample', by_id=False):
"""Randomly subsample without replacement.
Parameters
----------
n : int
Number of items to subsample from `counts`.
axis : {'sample', 'observation'}, optional
The axis to sample over
by_id : boolean, optional
If `False`, the subsampling is based on the counts contained in the
matrix (e.g., rarefaction). If `True`, the subsampling is based on
the IDs (e.g., fetch a random subset of samples). Default is
`False`.
Returns
-------
biom.Table
A subsampled version of self
Raises
------
ValueError
If `n` is less than zero.
Notes
-----
Subsampling is performed without replacement. If `n` is greater than
the sum of a given vector, that vector is omitted from the result.
Adapted from `skbio.math.subsample`, see biom-format/licenses for more
information about scikit-bio.
This code assumes absolute abundance if `by_id` is False.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
>>> table = Table(np.array([[0, 2, 3], [1, 0, 2]]), ['O1', 'O2'],
... ['S1', 'S2', 'S3'])
Subsample 1 item over the sample axis by value (e.g., rarefaction):
>>> print table.subsample(1).sum(axis='sample')
[ 1. 1. 1.]
Subsample 2 items over the sample axis, note that 'S1' is filtered out:
>>> ss = table.subsample(2)
>>> print ss.sum(axis='sample')
[ 2. 2.]
>>> print ss.ids()
['S2' 'S3']
Subsample by IDs over the sample axis. For this example, we're going to
randomly select 2 samples and do this 100 times, and then print out the
set of IDs observed.
>>> ids = set([tuple(table.subsample(2, by_id=True).ids())
... for i in range(100)])
>>> print sorted(ids)
[('S1', 'S2'), ('S1', 'S3'), ('S2', 'S3')]
"""
if n < 0:
raise ValueError("n cannot be negative.")
table = self.copy()
if by_id:
ids = table.ids(axis=axis).copy()
np.random.shuffle(ids)
subset = set(ids[:n])
table.filter(lambda v, i, md: i in subset)
else:
data = table._get_sparse_data()
_subsample(data, n)
table._data = data
table.filter(lambda v, i, md: v.sum() > 0, axis=axis)
inv_axis = self._invert_axis(axis)
table.filter(lambda v, i, md: v.sum() > 0, axis=inv_axis)
return table
def pa(self, inplace=True):
"""Convert the table to presence/absence data
Parameters
----------
inplace : bool, optional
Defaults to ``False``
Returns
-------
Table
Returns itself if `inplace`, else returns a new presence/absence
table.
Examples
--------
>>> from biom.table import Table
>>> import numpy as np
Create a 2x3 BIOM table
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'])
Convert to presence/absence data
>>> _ = table.pa()
>>> print table.data('O1', 'observation')
[ 0. 0. 1.]
>>> print table.data('O2', 'observation')
[ 1. 1. 1.]
"""
def transform_f(data, id_, metadata):
return np.where(data != 0, 1., 0.)
return self.transform(transform_f, inplace=inplace)
def transform(self, f, axis='sample', inplace=True):
"""Iterate over `axis`, applying a function `f` to each vector.
Only non null values can be modified and the density of the
table can't increase. However, zeroing values is fine.
Parameters
----------
f : function(data, id, metadata) -> new data
A function that takes three values: an array of nonzero
values corresponding to each observation or sample, an
observation or sample id, and an observation or sample
metadata entry. It must return an array of transformed
values that replace the original values.
axis : {'sample', 'observation'}, optional
The axis to operate on. Can be "sample" or "observation".
inplace : bool, optional
Defaults to ``True``. Whether to return a new table or modify
itself.
Returns
-------
biom.Table
Returns itself if `inplace`, else returns a new transformed table.
Raises
------
UnknownAxisError
If provided an unrecognized axis.
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 table
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'foo': 'bar'}, {'x': 'y'}], None)
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 1.0
O2 1.0 3.0 42.0
Create a transform function
>>> f = lambda data, id_, md: data / 2
Transform to a new table on samples
>>> table2 = table.transform(f, 'sample', False)
>>> print table2 # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 0.5
O2 0.5 1.5 21.0
`table` hasn't changed
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 1.0
O2 1.0 3.0 42.0
Tranform in place on observations
>>> table3 = table.transform(f, 'observation', True)
`table` is different now
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 0.5
O2 0.5 1.5 21.0
but the table returned (`table3`) is the same as `table`
>>> print table3 # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 0.5
O2 0.5 1.5 21.0
"""
table = self if inplace else self.copy()
metadata = table.metadata(axis=axis)
ids = table.ids(axis=axis)
arr = table._get_sparse_data(axis=axis)
axis = table._axis_to_num(axis)
_transform(arr, ids, metadata, f, axis)
arr.eliminate_zeros()
table._data = arr
return table
def norm(self, axis='sample', inplace=True):
"""Normalize in place sample values by an observation, or vice versa.
Parameters
----------
axis : {'sample', 'observation'}, optional
The axis to use for normalization
inplace : bool, optional
Defaults to ``True``. If ``True``, performs the normalization in
place. Otherwise, returns a new table with the noramlization
applied.
Returns
-------
biom.Table
The normalized table
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x2 table:
>>> data = np.asarray([[2, 0], [6, 1]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2'])
Get a version of the table normalized on the 'sample' axis, leaving the
original table untouched:
>>> new_table = table.norm(inplace=False)
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 2.0 0.0
O2 6.0 1.0
>>> print new_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 0.25 0.0
O2 0.75 1.0
Get a version of the table normalized on the 'observation' axis,
again leaving the original table untouched:
>>> new_table = table.norm(axis='observation', inplace=False)
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 2.0 0.0
O2 6.0 1.0
>>> print new_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 1.0 0.0
O2 0.857142857143 0.142857142857
Do the same normalization on 'observation', this time in-place:
>>> table.norm(axis='observation')
2 x 2 <class 'biom.table.Table'> with 3 nonzero entries (75% dense)
>>> print table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 1.0 0.0
O2 0.857142857143 0.142857142857
"""
def f(val, id_, _):
return val / float(val.sum())
return self.transform(f, axis=axis, inplace=inplace)
def nonzero(self):
"""Yields locations of nonzero elements within the data matrix
Returns
-------
generator
Yields ``(observation_id, sample_id)`` for each nonzero element
"""
csr = self._data.tocsr()
samp_ids = self.ids()
obs_ids = self.ids(axis='observation')
indptr = csr.indptr
indices = csr.indices
for row_idx in range(indptr.size - 1):
start = indptr[row_idx]
end = indptr[row_idx+1]
obs_id = obs_ids[row_idx]
for col_idx in indices[start:end]:
yield (obs_id, samp_ids[col_idx])
def nonzero_counts(self, axis, binary=False):
"""Get nonzero summaries about an axis
Parameters
----------
axis : {'sample', 'observation', 'whole'}
The axis on which to count nonzero entries
binary : bool, optional
Defaults to ``False``. If ``False``, return number of nonzero
entries. If ``True``, sum the values of the entries.
Returns
-------
numpy.array
Counts in index order to the axis
"""
if binary:
dtype = 'int'
def op(x):
return x.nonzero()[0].size
else:
dtype = self.dtype
def op(x):
return x.sum()
if axis in ('sample', 'observation'):
# can use np.bincount for CSMat or ScipySparse
result = zeros(len(self.ids(axis=axis)), dtype=dtype)
for idx, vals in enumerate(self.iter_data(axis=axis)):
result[idx] = op(vals)
else:
result = zeros(1, dtype=dtype)
for vals in self.iter_data():
result[0] += op(vals)
return result
def _union_id_order(self, a, b):
"""Determines merge order for id lists A and B"""
all_ids = list(a[:])
all_ids.extend(b[:])
new_order = {}
idx = 0
for id_ in all_ids:
if id_ not in new_order:
new_order[id_] = idx
idx += 1
return new_order
def _intersect_id_order(self, a, b):
"""Determines the merge order for id lists A and B"""
all_b = set(b[:])
new_order = {}
idx = 0
for id_ in a:
if id_ in all_b:
new_order[id_] = idx
idx += 1
return new_order
def merge(self, other, sample='union', observation='union',
sample_metadata_f=prefer_self,
observation_metadata_f=prefer_self):
"""Merge two tables together
The axes, samples and observations, can be controlled independently.
Both can work on either "union" or "intersection".
`sample_metadata_f` and `observation_metadata_f` define how to
merge metadata between tables. The default is to just keep the metadata
associated to self if self has metadata otherwise take metadata from
other. These functions are given both metadata dicts and must return
a single metadata dict
Parameters
----------
other : biom.Table
The other table to merge with this one
sample : {'union', 'intersection'}, optional
observation : {'union', 'intersection'}, optional
sample_metadata_f : function, optional
Defaults to ``biom.util.prefer_self``. Defines how to handle sample
metadata during merge.
obesrvation_metadata_f : function, optional
Defaults to ``biom.util.prefer_self``. Defines how to handle
observation metdata during merge.
Returns
-------
biom.Table
The merged table
Notes
-----
- There is an implicit type conversion to ``float``.
- The return type is always that of ``self``
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x2 table and a 3x2 table:
>>> d_a = np.asarray([[2, 0], [6, 1]])
>>> t_a = Table(d_a, ['O1', 'O2'], ['S1', 'S2'])
>>> d_b = np.asarray([[4, 5], [0, 3], [10, 10]])
>>> t_b = Table(d_b, ['O1', 'O2', 'O3'], ['S1', 'S2'])
Merging the table results in the overlapping samples/observations (see
`O1` and `S2`) to be summed and the non-overlapping ones to be added to
the resulting table (see `S3`).
>>> merged_table = t_a.merge(t_b)
>>> print merged_table # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2
O1 6.0 5.0
O2 6.0 4.0
O3 10.0 10.0
"""
# determine the sample order in the resulting table
if sample is 'union':
new_samp_order = self._union_id_order(self.ids(), other.ids())
elif sample is 'intersection':
new_samp_order = self._intersect_id_order(self.ids(), other.ids())
else:
raise TableException("Unknown sample merge type: %s" % sample)
# determine the observation order in the resulting table
if observation is 'union':
new_obs_order = self._union_id_order(
self.ids(axis='observation'), other.ids(axis='observation'))
elif observation is 'intersection':
new_obs_order = self._intersect_id_order(
self.ids(axis='observation'), other.ids(axis='observation'))
else:
raise TableException(
"Unknown observation merge type: %s" %
observation)
# convert these to lists, no need to be dictionaries and reduces
# calls to items() and allows for pre-caluculating insert order
new_samp_order = sorted(new_samp_order.items(), key=itemgetter(1))
new_obs_order = sorted(new_obs_order.items(), key=itemgetter(1))
# if we don't have any samples, complain loudly. This is likely from
# performing an intersection without overlapping ids
if not new_samp_order:
raise TableException("No samples in resulting table!")
if not new_obs_order:
raise TableException("No observations in resulting table!")
# helper index lookups
other_obs_idx = other._obs_index
self_obs_idx = self._obs_index
other_samp_idx = other._sample_index
self_samp_idx = self._sample_index
# pre-calculate sample order from each table. We only need to do this
# once which dramatically reduces the number of dict lookups necessary
# within the inner loop
other_samp_order = []
self_samp_order = []
for samp_id, nsi in new_samp_order: # nsi -> new_sample_index
other_samp_order.append((nsi, other_samp_idx.get(samp_id, None)))
self_samp_order.append((nsi, self_samp_idx.get(samp_id, None)))
# pre-allocate the a list for placing the resulting vectors as the
# placement id is not ordered
vals = [None for i in range(len(new_obs_order))]
# POSSIBLE DECOMPOSITION
# resulting sample ids and sample metadata
sample_ids = []
sample_md = []
self_sample_md = self.metadata()
other_sample_md = other.metadata()
for id_, idx in new_samp_order:
sample_ids.append(id_)
# if we have sample metadata, grab it
if self_sample_md is None or not self.exists(id_):
self_md = None
else:
self_md = self_sample_md[self_samp_idx[id_]]
# if we have sample metadata, grab it
if other_sample_md is None or not other.exists(id_):
other_md = None
else:
other_md = other_sample_md[other_samp_idx[id_]]
sample_md.append(sample_metadata_f(self_md, other_md))
# POSSIBLE DECOMPOSITION
# resulting observation ids and sample metadata
obs_ids = []
obs_md = []
self_obs_md = self.metadata(axis='observation')
other_obs_md = other.metadata(axis='observation')
for id_, idx in new_obs_order:
obs_ids.append(id_)
# if we have observation metadata, grab it
if self_obs_md is None or not self.exists(id_, axis="observation"):
self_md = None
else:
self_md = self_obs_md[self_obs_idx[id_]]
# if we have observation metadata, grab it
if other_obs_md is None or \
not other.exists(id_, axis="observation"):
other_md = None
else:
other_md = other_obs_md[other_obs_idx[id_]]
obs_md.append(observation_metadata_f(self_md, other_md))
# length used for construction of new vectors
vec_length = len(new_samp_order)
# walk over observations in our new order
for obs_id, new_obs_idx in new_obs_order:
# create new vector for matrix values
new_vec = zeros(vec_length, dtype='float')
# This method allows for the creation of a matrix of self type.
# See note above
# new_vec = data_f()
# see if the observation exists in other, if so, pull it out.
# if not, set to the placeholder missing
if other.exists(obs_id, axis="observation"):
other_vec = other.data(obs_id, 'observation')
else:
other_vec = None
# see if the observation exists in self, if so, pull it out.
# if not, set to the placeholder missing
if self.exists(obs_id, axis="observation"):
self_vec = self.data(obs_id, 'observation')
else:
self_vec = None
# short circuit. If other doesn't have any values, then we can just
# take all values from self
if other_vec is None:
for (n_idx, s_idx) in self_samp_order:
if s_idx is not None:
new_vec[n_idx] = self_vec[s_idx]
# short circuit. If self doesn't have any values, then we can just
# take all values from other
elif self_vec is None:
for (n_idx, o_idx) in other_samp_order:
if o_idx is not None:
new_vec[n_idx] = other_vec[o_idx]
else:
# NOTE: DM 7.5.12, no observed improvement at the profile level
# was made on this inner loop by using self_samp_order and
# other_samp_order lists.
# walk over samples in our new order
for samp_id, new_samp_idx in new_samp_order:
# pull out each individual sample value. This is expensive,
# but the vectors are in a different alignment. It is
# possible that this could be improved with numpy take but
# needs to handle missing values appropriately
if samp_id not in self_samp_idx:
self_vec_value = 0
else:
self_vec_value = self_vec[self_samp_idx[samp_id]]
if samp_id not in other_samp_idx:
other_vec_value = 0
else:
other_vec_value = other_vec[other_samp_idx[samp_id]]
new_vec[new_samp_idx] = self_vec_value + other_vec_value
# convert our new vector to self type as to make sure we don't
# accidently force a dense representation in memory
vals[new_obs_idx] = self._conv_to_self_type(new_vec)
return self.__class__(self._conv_to_self_type(vals), obs_ids[:],
sample_ids[:], obs_md, sample_md)
@classmethod
def from_hdf5(cls, h5grp, ids=None, axis='sample', parse_fs=None):
"""Parse an HDF5 formatted BIOM table
If ids is provided, only the samples/observations listed in ids
(depending on the value of axis) will be loaded
The expected structure of this group is below. A few basic definitions,
N is the number of observations and M is the number of samples. Data
are stored in both compressed sparse row (for observation oriented
operations) and compressed sparse column (for sample oriented
operations).
Notes
-----
The expected HDF5 group structure is below. An example of an HDF5 file
in DDL can be found here [3]_.
- ./id : str, an \
arbitrary ID
- ./type : str, the \
table type (e.g, OTU table)
- ./format-url : str, a URL \
that describes the format
- ./format-version : two element \
tuple of int32, major and minor
- ./generated-by : str, what \
generated this file
- ./creation-date : str, ISO \
format
- ./shape : two element \
tuple of int32, N by M
- ./nnz : int32 or \
int64, number of non zero elems
- ./observation : Group
- ./observation/ids : (N,) dataset\
of str or vlen str
- ./observation/matrix : Group
- ./observation/matrix/data : (nnz,) \
dataset of float64
- ./observation/matrix/indices : (nnz,) \
dataset of int32
- ./observation/matrix/indptr : (M+1,) \
dataset of int32
- ./observation/metadata : Group
- [./observation/metadata/foo] : Optional, \
(N,) dataset of any valid HDF5 type in index order with IDs.
- ./observation/group-metadata : Group
- [./observation/group-metadata/foo] : Optional, \
(?,) dataset of group metadata that relates IDs
- [./observation/group-metadata/foo.attrs['data_type']] : attribute of\
the foo dataset that describes contained type (e.g., newick)
- ./sample : Group
- ./sample/ids : (M,) dataset\
of str or vlen str
- ./sample/matrix : Group
- ./sample/matrix/data : (nnz,) \
dataset of float64
- ./sample/matrix/indices : (nnz,) \
dataset of int32
- ./sample/matrix/indptr : (N+1,) \
dataset of int32
- ./sample/metadata : Group
- [./sample/metadata/foo] : Optional, \
(M,) dataset of any valid HDF5 type in index order with IDs.
- ./sample/group-metadata : Group
- [./sample/group-metadata/foo] : Optional, \
(?,) dataset of group metadata that relates IDs
- [./sample/group-metadata/foo.attrs['data_type']] : attribute of\
the foo dataset that describes contained type (e.g., newick)
The '?' character on the dataset size means that it can be of arbitrary
length.
The expected structure for each of the metadata datasets is a list of
atomic type objects (int, float, str, ...), where the index order of
the list corresponds to the index order of the relevant axis IDs.
Special metadata fields have been defined, and they are stored in a
specific way. Currently, the available special metadata fields are:
- taxonomy: (N, ?) dataset of str or vlen str
- KEGG_Pathways: (N, ?) dataset of str or vlen str
- collapsed_ids: (N, ?) dataset of str or vlen str
Parameters
----------
h5grp : a h5py ``Group`` or an open h5py ``File``
ids : iterable
The sample/observation ids of the samples/observations that we need
to retrieve from the hdf5 biom table
axis : {'sample', 'observation'}, optional
The axis to subset on
parse_fs : dict, optional
Specify custom parsing functions for metadata fields. This dict is
expected to be {'metadata_field': function}, where the function
signature is (object) corresponding to a single row in the
associated metadata dataset. The return from this function an
object as well, and is the parsed representation of the metadata.
Returns
-------
biom.Table
A BIOM ``Table`` object
Raises
------
ValueError
If `ids` are not a subset of the samples or observations ids
present in the hdf5 biom table
References
----------
.. [1] http://docs.scipy.org/doc/scipy-0.13.0/reference/generated/sci\
py.sparse.csr_matrix.html
.. [2] http://docs.scipy.org/doc/scipy-0.13.0/reference/generated/sci\
py.sparse.csc_matrix.html
.. [3] http://biom-format.org/documentation/format_versions/biom-2.0.\
html
See Also
--------
Table.to_hdf5
Examples
--------
>>> from biom.table import Table
>>> from biom.util import biom_open
>>> with biom_open('rich_sparse_otu_table_hdf5.biom') as f \
# doctest: +SKIP
>>> t = Table.from_hdf5(f) # doctest: +SKIP
Parse a hdf5 biom table subsetting observations
>>> from biom.util import biom_open # doctest: +SKIP
>>> from biom.parse import parse_biom_table
>>> with biom_open('rich_sparse_otu_table_hdf5.biom') as f \
# doctest: +SKIP
>>> t = Table.from_hdf5(f, ids=["GG_OTU_1"],
... axis='observation') # doctest: +SKIP
"""
if not HAVE_H5PY:
raise RuntimeError("h5py is not in the environment, HDF5 support "
"is not available")
if axis not in ['sample', 'observation']:
raise UnknownAxisError(axis)
if parse_fs is None:
parse_fs = {}
id_ = h5grp.attrs['id']
create_date = h5grp.attrs['creation-date']
generated_by = h5grp.attrs['generated-by']
shape = h5grp.attrs['shape']
type_ = None if h5grp.attrs['type'] == '' else h5grp.attrs['type']
def axis_load(grp):
"""Loads all the data of the given group"""
# fetch all of the IDs
ids = grp['ids'][:]
if ids.size > 0 and isinstance(ids[0], bytes):
ids = np.array([i.decode('utf8') for i in ids])
parser = defaultdict(lambda: general_parser)
parser['taxonomy'] = vlen_list_of_str_parser
parser['KEGG_Pathways'] = vlen_list_of_str_parser
parser['collapsed_ids'] = vlen_list_of_str_parser
parser.update(parse_fs)
# fetch ID specific metadata
md = [{} for i in range(len(ids))]
for category, dset in viewitems(grp['metadata']):
parse_f = parser[category]
data = dset[:]
for md_dict, data_row in zip(md, data):
md_dict[category] = parse_f(data_row)
# If there was no metadata on the axis, set it up as none
md = md if any(md) else None
# Fetch the group metadata
grp_md = {cat: val
for cat, val in grp['group-metadata'].items()}
return ids, md, grp_md
obs_ids, obs_md, obs_grp_md = axis_load(h5grp['observation'])
samp_ids, samp_md, samp_grp_md = axis_load(h5grp['sample'])
# load the data
data_grp = h5grp[axis]['matrix']
h5_data = data_grp["data"]
h5_indices = data_grp["indices"]
h5_indptr = data_grp["indptr"]
# Check if we need to subset the biom table
if ids is not None:
def _get_ids(source_ids, desired_ids):
"""If desired_ids is not None, makes sure that it is a subset
of source_ids and returns the desired_ids array-like and a
boolean array indicating where the desired_ids can be found in
source_ids"""
if desired_ids is None:
ids = source_ids[:]
idx = np.ones(source_ids.shape, dtype=bool)
else:
desired_ids = np.asarray(desired_ids)
# Get the index of the source ids to include
idx = np.in1d(source_ids, desired_ids)
# Retrieve only the ids that we are interested on
ids = source_ids[idx]
# Check that all desired ids have been found on source ids
if ids.shape != desired_ids.shape:
raise ValueError("The following ids could not be "
"found in the biom table: %s" %
(set(desired_ids) - set(ids)))
return ids, idx
# Get the observation and sample ids that we are interested in
samp, obs = (ids, None) if axis == 'sample' else (None, ids)
obs_ids, obs_idx = _get_ids(obs_ids, obs)
samp_ids, samp_idx = _get_ids(samp_ids, samp)
# Get the new matrix shape
shape = (len(obs_ids), len(samp_ids))
# Fetch the metadata that we are interested in
def _subset_metadata(md, idx):
"""If md has data, returns the subset indicated by idx, a
boolean array"""
if md:
md = list(np.asarray(md)[np.where(idx)])
return md
obs_md = _subset_metadata(obs_md, obs_idx)
samp_md = _subset_metadata(samp_md, samp_idx)
# load the subset of the data
idx = samp_idx if axis == 'sample' else obs_idx
keep = np.where(idx)[0]
indptr_indices = sorted([(h5_indptr[i], h5_indptr[i+1])
for i in keep])
# Create the new indptr
indptr_subset = np.array([end - start
for start, end in indptr_indices])
indptr = np.empty(len(keep) + 1, dtype=np.int32)
indptr[0] = 0
indptr[1:] = indptr_subset.cumsum()
data = np.hstack(h5_data[start:end]
for start, end in indptr_indices)
indices = np.hstack(h5_indices[start:end]
for start, end in indptr_indices)
else:
# no subset need, just pass all data to scipy
data = h5_data
indices = h5_indices
indptr = h5_indptr
cs = (data, indices, indptr)
if axis == 'sample':
matrix = csc_matrix(cs, shape=shape)
else:
matrix = csr_matrix(cs, shape=shape)
t = Table(matrix, obs_ids, samp_ids, obs_md or None,
samp_md or None, type=type_, create_date=create_date,
generated_by=generated_by, table_id=id_,
observation_group_metadata=obs_grp_md,
sample_group_metadata=samp_grp_md)
if ids is not None:
# filter out any empty samples or observations which may exist due
# to subsetting
def any_value(vals, id_, md):
return np.any(vals)
axis = 'observation' if axis == 'sample' else 'sample'
t.filter(any_value, axis=axis)
return t
def to_hdf5(self, h5grp, generated_by, compress=True, format_fs=None):
"""Store CSC and CSR in place
The resulting structure of this group is below. A few basic
definitions, N is the number of observations and M is the number of
samples. Data are stored in both compressed sparse row [1]_ (CSR, for
observation oriented operations) and compressed sparse column [2]_
(CSC, for sample oriented operations).
Notes
-----
The expected HDF5 group structure is below. An example of an HDF5 file
in DDL can be found here [3]_.
- ./id : str, an \
arbitrary ID
- ./type : str, the \
table type (e.g, OTU table)
- ./format-url : str, a URL \
that describes the format
- ./format-version : two element \
tuple of int32, major and minor
- ./generated-by : str, what \
generated this file
- ./creation-date : str, ISO \
format
- ./shape : two element \
tuple of int32, N by M
- ./nnz : int32 or \
int64, number of non zero elems
- ./observation : Group
- ./observation/ids : (N,) dataset\
of str or vlen str
- ./observation/matrix : Group
- ./observation/matrix/data : (nnz,) \
dataset of float64
- ./observation/matrix/indices : (nnz,) \
dataset of int32
- ./observation/matrix/indptr : (M+1,) \
dataset of int32
- ./observation/metadata : Group
- [./observation/metadata/foo] : Optional, \
(N,) dataset of any valid HDF5 type in index order with IDs.
- ./observation/group-metadata : Group
- [./observation/group-metadata/foo] : Optional, \
(?,) dataset of group metadata that relates IDs
- [./observation/group-metadata/foo.attrs['data_type']] : attribute of\
the foo dataset that describes contained type (e.g., newick)
- ./sample : Group
- ./sample/ids : (M,) dataset\
of str or vlen str
- ./sample/matrix : Group
- ./sample/matrix/data : (nnz,) \
dataset of float64
- ./sample/matrix/indices : (nnz,) \
dataset of int32
- ./sample/matrix/indptr : (N+1,) \
dataset of int32
- ./sample/metadata : Group
- [./sample/metadata/foo] : Optional, \
(M,) dataset of any valid HDF5 type in index order with IDs.
- ./sample/group-metadata : Group
- [./sample/group-metadata/foo] : Optional, \
(?,) dataset of group metadata that relates IDs
- [./sample/group-metadata/foo.attrs['data_type']] : attribute of\
the foo dataset that describes contained type (e.g., newick)
The '?' character on the dataset size means that it can be of arbitrary
length.
The expected structure for each of the metadata datasets is a list of
atomic type objects (int, float, str, ...), where the index order of
the list corresponds to the index order of the relevant axis IDs.
Special metadata fields have been defined, and they are stored in a
specific way. Currently, the available special metadata fields are:
- taxonomy: (N, ?) dataset of str or vlen str
- KEGG_Pathways: (N, ?) dataset of str or vlen str
- collapsed_ids: (N, ?) dataset of str or vlen str
Parameters
----------
h5grp : `h5py.Group` or `h5py.File`
The HDF5 entity in which to write the BIOM formatted data.
generated_by : str
A description of what generated the table
compress : bool, optional
Defaults to ``True`` means fields will be compressed with gzip,
``False`` means no compression
format_fs : dict, optional
Specify custom formatting functions for metadata fields. This dict
is expected to be {'metadata_field': function}, where the function
signature is (h5py.Group, str, dict, bool) corresponding to the
specific HDF5 group the metadata dataset will be associated with,
the category being operated on, the metadata for the entire axis
being operated on, and whether to enable compression on the
dataset. Anything returned by this function is ignored.
Notes
-----
This method does not return anything and operates in place on h5grp.
See Also
--------
Table.from_hdf5
References
----------
.. [1] http://docs.scipy.org/doc/scipy-0.13.0/reference/generated/sci\
py.sparse.csr_matrix.html
.. [2] http://docs.scipy.org/doc/scipy-0.13.0/reference/generated/sci\
py.sparse.csc_matrix.html
.. [3] http://biom-format.org/documentation/format_versions/biom-2.1.\
html
Examples
--------
>>> from biom.util import biom_open # doctest: +SKIP
>>> from biom.table import Table
>>> from numpy import array
>>> t = Table(array([[1, 2], [3, 4]]), ['a', 'b'], ['x', 'y'])
>>> with biom_open('foo.biom', 'w') as f: # doctest: +SKIP
... t.to_hdf5(f, "example")
"""
if not HAVE_H5PY:
raise RuntimeError("h5py is not in the environment, HDF5 support "
"is not available")
if format_fs is None:
format_fs = {}
def axis_dump(grp, ids, md, group_md, order, compression=None):
"""Store for an axis"""
self._data = self._data.asformat(order)
len_ids = len(ids)
len_indptr = len(self._data.indptr)
len_data = self.nnz
grp.create_group('matrix')
grp.create_dataset('matrix/data', shape=(len_data,),
dtype=np.float64,
data=self._data.data,
compression=compression)
grp.create_dataset('matrix/indices', shape=(len_data,),
dtype=np.int32,
data=self._data.indices,
compression=compression)
grp.create_dataset('matrix/indptr', shape=(len_indptr,),
dtype=np.int32,
data=self._data.indptr,
compression=compression)
if len_ids > 0:
# if we store IDs in the table as numpy arrays then this store
# is cleaner, as is the parse
grp.create_dataset('ids', shape=(len_ids,),
dtype=H5PY_VLEN_STR,
data=[i.encode('utf8') for i in ids],
compression=compression)
else:
# Empty H5PY_VLEN_STR datasets are not supported.
grp.create_dataset('ids', shape=(0, ), data=[],
compression=compression)
# Create the group for the metadata
grp.create_group('metadata')
if md is not None:
formatter = defaultdict(lambda: general_formatter)
formatter['taxonomy'] = vlen_list_of_str_formatter
formatter['KEGG_Pathways'] = vlen_list_of_str_formatter
formatter['collapsed_ids'] = vlen_list_of_str_formatter
formatter.update(format_fs)
# Loop through all the categories
for category in md[0]:
# Create the dataset for the current category,
# putting values in id order
formatter[category](grp, category, md, compression)
# Create the group for the group metadata
grp.create_group('group-metadata')
if group_md:
for key, value in group_md.items():
datatype, val = value
grp_dataset = grp.create_dataset(
'group-metadata/%s' % key,
shape=(1,), dtype=H5PY_VLEN_STR,
data=val, compression=compression)
grp_dataset.attrs['data_type'] = datatype
h5grp.attrs['id'] = self.table_id if self.table_id else "No Table ID"
h5grp.attrs['type'] = self.type if self.type else ""
h5grp.attrs['format-url'] = "http://biom-format.org"
h5grp.attrs['format-version'] = self.format_version
h5grp.attrs['generated-by'] = generated_by
h5grp.attrs['creation-date'] = datetime.now().isoformat()
h5grp.attrs['shape'] = self.shape
h5grp.attrs['nnz'] = self.nnz
compression = None
if compress is True:
compression = 'gzip'
axis_dump(h5grp.create_group('observation'),
self.ids(axis='observation'),
self.metadata(axis='observation'),
self.group_metadata(axis='observation'), 'csr', compression)
axis_dump(h5grp.create_group('sample'), self.ids(),
self.metadata(), self.group_metadata(), 'csc', compression)
@classmethod
def from_json(self, json_table, data_pump=None,
input_is_dense=False):
"""Parse a biom otu table type
Parameters
----------
json_table : dict
A JSON object or dict that represents the BIOM table
data_pump : tuple or None
A secondary source of data
input_is_dense : bool
If `True`, the data contained will be interpretted as dense
Returns
-------
Table
Examples
--------
>>> from biom import Table
>>> json_obj = {"id": "None",
... "format": "Biological Observation Matrix 1.0.0",
... "format_url": "http://biom-format.org",
... "generated_by": "foo",
... "type": "OTU table",
... "date": "2014-06-03T14:24:40.884420",
... "matrix_element_type": "float",
... "shape": [5, 6],
... "data": [[0,2,1.0],
... [1,0,5.0],
... [1,1,1.0],
... [1,3,2.0],
... [1,4,3.0],
... [1,5,1.0],
... [2,2,1.0],
... [2,3,4.0],
... [2,5,2.0],
... [3,0,2.0],
... [3,1,1.0],
... [3,2,1.0],
... [3,5,1.0],
... [4,1,1.0],
... [4,2,1.0]],
... "rows": [{"id": "GG_OTU_1", "metadata": None},
... {"id": "GG_OTU_2", "metadata": None},
... {"id": "GG_OTU_3", "metadata": None},
... {"id": "GG_OTU_4", "metadata": None},
... {"id": "GG_OTU_5", "metadata": None}],
... "columns": [{"id": "Sample1", "metadata": None},
... {"id": "Sample2", "metadata": None},
... {"id": "Sample3", "metadata": None},
... {"id": "Sample4", "metadata": None},
... {"id": "Sample5", "metadata": None},
... {"id": "Sample6", "metadata": None}]
... }
>>> t = Table.from_json(json_obj)
"""
sample_ids = [col['id'] for col in json_table['columns']]
sample_metadata = [col['metadata'] for col in json_table['columns']]
obs_ids = [row['id'] for row in json_table['rows']]
obs_metadata = [row['metadata'] for row in json_table['rows']]
dtype = MATRIX_ELEMENT_TYPE[json_table['matrix_element_type']]
if 'matrix_type' in json_table:
if json_table['matrix_type'] == 'dense':
input_is_dense = True
else:
input_is_dense = False
type_ = json_table['type']
if data_pump is None:
table_obj = Table(json_table['data'], obs_ids, sample_ids,
obs_metadata, sample_metadata,
shape=json_table['shape'],
dtype=dtype,
type=type_,
input_is_dense=input_is_dense)
else:
table_obj = Table(data_pump, obs_ids, sample_ids,
obs_metadata, sample_metadata,
shape=json_table['shape'],
dtype=dtype,
type=type_,
input_is_dense=input_is_dense)
return table_obj
def to_json(self, generated_by, direct_io=None):
"""Returns a JSON string representing the table in BIOM format.
Parameters
----------
generated_by : str
a string describing the software used to build the table
direct_io : file or file-like object, optional
Defaults to ``None``. Must implementing a ``write`` function. If
`direct_io` is not ``None``, the final output is written directly
to `direct_io` during processing.
Returns
-------
str
A JSON-formatted string representing the biom table
"""
if not isinstance(generated_by, string_types):
raise TableException("Must specify a generated_by string")
# Fill in top-level metadata.
if direct_io:
direct_io.write(u'{')
direct_io.write(u'"id": "%s",' % str(self.table_id))
direct_io.write(
u'"format": "%s",' %
get_biom_format_version_string((1, 0))) # JSON table -> 1.0.0
direct_io.write(
u'"format_url": "%s",' %
get_biom_format_url_string())
direct_io.write(u'"generated_by": "%s",' % generated_by)
direct_io.write(u'"date": "%s",' % datetime.now().isoformat())
else:
id_ = u'"id": "%s",' % str(self.table_id)
format_ = u'"format": "%s",' % get_biom_format_version_string(
(1, 0)) # JSON table -> 1.0.0
format_url = u'"format_url": "%s",' % get_biom_format_url_string()
generated_by = u'"generated_by": "%s",' % generated_by
date = u'"date": "%s",' % datetime.now().isoformat()
# Determine if we have any data in the matrix, and what the shape of
# the matrix is.
try:
num_rows, num_cols = self.shape
except:
num_rows = num_cols = 0
has_data = True if num_rows > 0 and num_cols > 0 else False
# Default the matrix element type to test to be an integer in case we
# don't have any data in the matrix to test.
test_element = 0
if has_data:
test_element = self[0, 0]
# Determine the type of elements the matrix is storing.
if isinstance(test_element, int):
matrix_element_type = u"int"
elif isinstance(test_element, float):
matrix_element_type = u"float"
elif isinstance(test_element, string_types):
matrix_element_type = u"str"
else:
raise TableException("Unsupported matrix data type.")
# Fill in details about the matrix.
if direct_io:
direct_io.write(
u'"matrix_element_type": "%s",' %
matrix_element_type)
direct_io.write(u'"shape": [%d, %d],' % (num_rows, num_cols))
else:
matrix_element_type = u'"matrix_element_type": "%s",' % \
matrix_element_type
shape = u'"shape": [%d, %d],' % (num_rows, num_cols)
# Fill in the table type
if self.type is None:
type_ = u'"type": null,'
else:
type_ = u'"type": "%s",' % self.type
if direct_io:
direct_io.write(type_)
# Fill in details about the rows in the table and fill in the matrix's
# data. BIOM 2.0+ is now only sparse
if direct_io:
direct_io.write(u'"matrix_type": "sparse",')
direct_io.write(u'"data": [')
else:
matrix_type = u'"matrix_type": "sparse",'
data = [u'"data": [']
max_row_idx = len(self.ids(axis='observation')) - 1
max_col_idx = len(self.ids()) - 1
rows = [u'"rows": [']
have_written = False
for obs_index, obs in enumerate(self.iter(axis='observation')):
# i'm crying on the inside
if obs_index != max_row_idx:
rows.append(u'{"id": %s, "metadata": %s},' % (dumps(obs[1]),
dumps(obs[2])))
else:
rows.append(u'{"id": %s, "metadata": %s}],' % (dumps(obs[1]),
dumps(obs[2])))
# turns out its a pain to figure out when to place commas. the
# simple work around, at the expense of a little memory
# (bound by the number of samples) is to build of what will be
# written, and then add in the commas where necessary.
built_row = []
for col_index, val in enumerate(obs[0]):
if float(val) != 0.0:
built_row.append(u"[%d,%d,%r]" % (obs_index, col_index,
val))
if built_row:
# if we have written a row already, its safe to add a comma
if have_written:
if direct_io:
direct_io.write(u',')
else:
data.append(u',')
if direct_io:
direct_io.write(u','.join(built_row))
else:
data.append(u','.join(built_row))
have_written = True
# finalize the data block
if direct_io:
direct_io.write(u"],")
else:
data.append(u"],")
# Fill in details about the columns in the table.
columns = [u'"columns": [']
for samp_index, samp in enumerate(self.iter()):
if samp_index != max_col_idx:
columns.append(u'{"id": %s, "metadata": %s},' % (
dumps(samp[1]), dumps(samp[2])))
else:
columns.append(u'{"id": %s, "metadata": %s}]' % (
dumps(samp[1]), dumps(samp[2])))
if rows[0] == u'"rows": [' and len(rows) == 1:
# empty table case
rows = [u'"rows": [],']
columns = [u'"columns": []']
rows = u''.join(rows)
columns = u''.join(columns)
if direct_io:
direct_io.write(rows)
direct_io.write(columns)
direct_io.write(u'}')
else:
return u"{%s}" % ''.join([id_, format_, format_url, matrix_type,
generated_by, date, type_,
matrix_element_type, shape,
u''.join(data), rows, columns])
@staticmethod
def from_tsv(lines, obs_mapping, sample_mapping,
process_func, **kwargs):
"""Parse a tab separated (observation x sample) formatted BIOM table
Parameters
----------
lines : list, or file-like object
The tab delimited data to parse
obs_mapping : dict or None
The corresponding observation metadata
sample_mapping : dict or None
The corresponding sample metadata
process_func : function
A function to transform the observation metadata
Returns
-------
biom.Table
A BIOM ``Table`` object
Examples
--------
Parse tab separated data into a table:
>>> from biom.table import Table
>>> from StringIO import StringIO
>>> tsv = 'a\\tb\\tc\\n1\\t2\\t3\\n4\\t5\\t6'
>>> tsv_fh = StringIO(tsv)
>>> func = lambda x : x
>>> test_table = Table.from_tsv(tsv_fh, None, None, func)
"""
(sample_ids, obs_ids, data, t_md,
t_md_name) = Table._extract_data_from_tsv(lines, **kwargs)
# if we have it, keep it
if t_md is None:
obs_metadata = None
else:
obs_metadata = [{t_md_name: process_func(v)} for v in t_md]
if sample_mapping is None:
sample_metadata = None
else:
sample_metadata = [sample_mapping[sample_id]
for sample_id in sample_ids]
# will override any metadata from parsed table
if obs_mapping is not None:
obs_metadata = [obs_mapping[obs_id] for obs_id in obs_ids]
return Table(data, obs_ids, sample_ids, obs_metadata, sample_metadata)
@staticmethod
def _extract_data_from_tsv(lines, delim='\t', dtype=float, md_parse=None):
"""Parse a classic table into (sample_ids, obs_ids, data, metadata,
name)
Parameters
----------
lines: list or file-like object
delimted data to parse
delim: string
delimeter in file lines
dtype: type
md_parse: function or None
funtion used to parse metdata
Returns
-------
list
sample_ids
list
observation_ids
array
data
list
metadata
string
column name if last column is non-numeric
Notes
------
This is intended to be close to how QIIME classic OTU tables are parsed
with the exception of the additional md_name field
This function is ported from QIIME (http://www.qiime.org), previously
named parse_classic_otu_table. QIIME is a GPL project, but we obtained
permission from the authors of this function to port it to the BIOM
Format project (and keep it under BIOM's BSD license).
.. shownumpydoc
"""
if not isinstance(lines, list):
try:
hasattr(lines, 'seek')
except AttributeError:
raise RuntimeError(
"Input needs to support seek or be indexable")
# find header, the first line that is not empty and does not start
# with a #
header = False
list_index = 0
for line in lines:
if not line.strip():
continue
if not line.startswith('#'):
# Covers the case where the first line is the header
# and there is no indication of it (no comment character)
if not header:
header = line.strip().split(delim)[1:]
data_start = list_index + 1
else:
data_start = list_index
break
list_index += 1
header = line.strip().split(delim)[1:]
# If the first line is the header, then we need to get the next
# line for the "last column" check
if isinstance(lines, list):
line = lines[data_start]
else:
lines.seek(0)
for index in range(0, data_start + 1):
line = lines.readline()
# attempt to determine if the last column is non-numeric, ie, metadata
first_values = line.strip().split(delim)
last_value = first_values[-1]
last_column_is_numeric = True
if '.' in last_value:
try:
float(last_value)
except ValueError:
last_column_is_numeric = False
else:
try:
int(last_value)
except ValueError:
last_column_is_numeric = False
# determine sample ids
if last_column_is_numeric:
md_name = None
metadata = None
samp_ids = header[:]
else:
md_name = header[-1]
metadata = []
samp_ids = header[:-1]
data = []
obs_ids = []
row_number = 0
# Go back to the beginning if it is a file:
if hasattr(lines, 'seek'):
lines.seek(0)
for index in range(0, data_start):
line = lines.readline()
else:
lines = lines[data_start:]
for line in lines:
line = line.strip()
if not line:
continue
if line.startswith('#'):
continue
fields = line.strip().split(delim)
obs_ids.append(fields[0])
if last_column_is_numeric:
values = list(map(dtype, fields[1:]))
else:
values = list(map(dtype, fields[1:-1]))
if md_parse is not None:
metadata.append(md_parse(fields[-1]))
else:
metadata.append(fields[-1])
for column_number in range(0, len(values)):
if values[column_number] != dtype(0):
data.append([row_number, column_number,
values[column_number]])
row_number += 1
return samp_ids, obs_ids, data, metadata, md_name
def to_tsv(self, header_key=None, header_value=None,
metadata_formatter=str, observation_column_name='#OTU ID'):
"""Return self as a string in tab delimited form
Default ``str`` output for the ``Table`` is just row/col ids and table
data without any metadata
Parameters
----------
header_key : str or ``None``, optional
Defaults to ``None``
header_value : str or ``None``, optional
Defaults to ``None``
metadata_formatter : function, optional
Defaults to ``str``. a function which takes a metadata entry and
returns a formatted version that should be written to file
observation_column_name : str, optional
Defaults to "#OTU ID". The name of the first column in the output
table, corresponding to the observation IDs.
Returns
-------
str
tab delimited representation of the Table
Examples
--------
>>> import numpy as np
>>> from biom.table import Table
Create a 2x3 BIOM table, with observation metadata and no sample
metadata:
>>> data = np.asarray([[0, 0, 1], [1, 3, 42]])
>>> table = Table(data, ['O1', 'O2'], ['S1', 'S2', 'S3'],
... [{'foo': 'bar'}, {'x': 'y'}], None)
>>> print table.to_tsv() # doctest: +NORMALIZE_WHITESPACE
# Constructed from biom file
#OTU ID S1 S2 S3
O1 0.0 0.0 1.0
O2 1.0 3.0 42.0
"""
return self.delimited_self(u'\t', header_key, header_value,
metadata_formatter,
observation_column_name)
def coo_arrays_to_sparse(data, dtype=np.float64, shape=None):
"""Map directly on to the coo_matrix constructor
Parameters
----------
data : tuple
data must be (values, (rows, cols))
dtype : type, optional
Defaults to ``np.float64``
shape : tuple or ``None``, optional
Defaults to ``None``. If `shape` is ``None``, shape will be determined
automatically from `data`.
"""
if shape is None:
values, (rows, cols) = data
n_rows = max(rows) + 1
n_cols = max(cols) + 1
else:
n_rows, n_cols = shape
# coo_matrix allows zeros to be added as data, and this affects
# nnz, items, and iteritems. Clean them out here, as this is
# the only time these zeros can creep in.
# Note: coo_matrix allows duplicate entries; the entries will
# be summed when converted. Not really sure how we want to
# handle this generally within BIOM- I'm okay with leaving it
# as undefined behavior for now.
matrix = coo_matrix(data, shape=(n_rows, n_cols), dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def list_list_to_sparse(data, dtype=float, shape=None):
"""Convert a list of lists into a scipy.sparse matrix.
Parameters
----------
data : iterable of iterables
`data` should be in the format [[row, col, value], ...]
dtype : type, optional
defaults to ``float``
shape : tuple or ``None``, optional
Defaults to ``None``. If `shape` is ``None``, shape will be determined
automatically from `data`.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
rows, cols, values = zip(*data)
if shape is None:
n_rows = max(rows) + 1
n_cols = max(cols) + 1
else:
n_rows, n_cols = shape
matrix = coo_matrix((values, (rows, cols)), shape=(n_rows, n_cols),
dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def nparray_to_sparse(data, dtype=float):
"""Convert a numpy array to a scipy.sparse matrix.
Parameters
----------
data : numpy.array
The data to convert into a sparse matrix
dtype : type, optional
Defaults to ``float``. The type of data to be represented.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
if data.shape == (0,):
# an empty vector. Note, this short circuit is necessary as calling
# csr_matrix([], shape=(0, 0), dtype=dtype) will result in a matrix
# has a shape of (1, 0).
return csr_matrix((0, 0), dtype=dtype)
elif data.shape in ((1, 0), (0, 1)) and data.size == 0:
# an empty matrix. This short circuit is necessary for the same reason
# as the empty vector. While a (1, 0) matrix is _empty_, this does
# confound code that assumes that (1, 0) means there might be metadata
# or IDs associated with that singular row
return csr_matrix((0, 0), dtype=dtype)
elif len(data.shape) == 1:
# a vector
shape = (1, data.shape[0])
else:
shape = data.shape
matrix = coo_matrix(data, shape=shape, dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def list_nparray_to_sparse(data, dtype=float):
"""Takes a list of numpy arrays and creates a scipy.sparse matrix.
Parameters
----------
data : iterable of numpy.array
The data to convert into a sparse matrix
dtype : type, optional
Defaults to ``float``. The type of data to be represented.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
matrix = coo_matrix(data, shape=(len(data), len(data[0])), dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def list_sparse_to_sparse(data, dtype=float):
"""Takes a list of scipy.sparse matrices and creates a scipy.sparse mat.
Parameters
----------
data : iterable of scipy.sparse matrices
The data to convert into a sparse matrix
dtype : type, optional
Defaults to ``float``. The type of data to be represented.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
if isspmatrix(data[0]):
if data[0].shape[0] > data[0].shape[1]:
n_cols = len(data)
n_rows = data[0].shape[0]
else:
n_rows = len(data)
n_cols = data[0].shape[1]
else:
all_keys = flatten([d.keys() for d in data])
n_rows = max(all_keys, key=itemgetter(0))[0] + 1
n_cols = max(all_keys, key=itemgetter(1))[1] + 1
if n_rows > n_cols:
n_cols = len(data)
else:
n_rows = len(data)
data = vstack(data)
matrix = coo_matrix(data, shape=(n_rows, n_cols),
dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def list_dict_to_sparse(data, dtype=float):
"""Takes a list of dict {(row,col):val} and creates a scipy.sparse mat.
Parameters
----------
data : iterable of dicts
The data to convert into a sparse matrix
dtype : type, optional
Defaults to ``float``. The type of data to be represented.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
if isspmatrix(data[0]):
if data[0].shape[0] > data[0].shape[1]:
is_col = True
n_cols = len(data)
n_rows = data[0].shape[0]
else:
is_col = False
n_rows = len(data)
n_cols = data[0].shape[1]
else:
all_keys = flatten([d.keys() for d in data])
n_rows = max(all_keys, key=itemgetter(0))[0] + 1
n_cols = max(all_keys, key=itemgetter(1))[1] + 1
if n_rows > n_cols:
is_col = True
n_cols = len(data)
else:
is_col = False
n_rows = len(data)
rows = []
cols = []
vals = []
for row_idx, row in enumerate(data):
for (row_val, col_idx), val in row.items():
if is_col:
# transpose
rows.append(row_val)
cols.append(row_idx)
vals.append(val)
else:
rows.append(row_idx)
cols.append(col_idx)
vals.append(val)
matrix = coo_matrix((vals, (rows, cols)), shape=(n_rows, n_cols),
dtype=dtype)
matrix = matrix.tocsr()
matrix.eliminate_zeros()
return matrix
def dict_to_sparse(data, dtype=float, shape=None):
"""Takes a dict {(row,col):val} and creates a scipy.sparse matrix.
Parameters
----------
data : dict
The data to convert into a sparse matrix
dtype : type, optional
Defaults to ``float``. The type of data to be represented.
Returns
-------
scipy.csr_matrix
The newly generated matrix
"""
if shape is None:
n_rows = max(data.keys(), key=itemgetter(0))[0] + 1
n_cols = max(data.keys(), key=itemgetter(1))[1] + 1
else:
n_rows, n_cols = shape
rows = []
cols = []
vals = []
for (r, c), v in viewitems(data):
rows.append(r)
cols.append(c)
vals.append(v)
return coo_arrays_to_sparse((vals, (rows, cols)),
shape=(n_rows, n_cols), dtype=dtype)
|