| | 1 | ◢ <[wiki:THU120907/Lab7 實作七]> | <[wiki:THU120907 回課程大綱]> ▲ | ◣ |
| | 2 | |
| | 3 | = 實作八 Lab 8 = |
| | 4 | [[PageOutline]] |
| | 5 | {{{ |
| | 6 | #!html |
| | 7 | <div style="text-align: center;"><big style="font-weight: bold;"><big>練習豬的拉丁語<br/>Pig Latin in Practice</big></big></div> |
| | 8 | }}} |
| | 9 | |
| | 10 | {{{ |
| | 11 | #!text |
| | 12 | 以下練習,請連線至 hadoop.nchc.org.tw 操作。底下的 hXXXX 等於您的用戶名稱。 |
| | 13 | }}} |
| | 14 | |
| | 15 | == Aggregation (Local Mode) == |
| | 16 | |
| | 17 | {{{ |
| | 18 | ~$ wget http://hadoop.nchc.org.tw/excite-small.log |
| | 19 | ~$ pig -x local |
| | 20 | grunt> log = LOAD 'excite-small.log' AS (user, timestamp, query); |
| | 21 | grunt> grpd = GROUP log BY user; |
| | 22 | grunt> cntd = FOREACH grpd GENERATE group, COUNT(log); |
| | 23 | grunt> STORE cntd INTO 'lab8_out1'; |
| | 24 | grunt> quit |
| | 25 | ~$ head lab8_out1 |
| | 26 | }}} |
| | 27 | |
| | 28 | == Filter (Local Mode) == |
| | 29 | |
| | 30 | {{{ |
| | 31 | ~$ pig -x local |
| | 32 | grunt> log = LOAD 'excite-small.log' AS (user, timestamp, query); |
| | 33 | grunt> grpd = GROUP log BY user; |
| | 34 | grunt> cntd = FOREACH grpd GENERATE group, COUNT(log) AS cnt; |
| | 35 | grunt> fltrd = FILTER cntd BY cnt > 50; |
| | 36 | grunt> STORE fltrd INTO 'lab8_out2'; |
| | 37 | grunt> quit |
| | 38 | ~$ head lab8_out2 |
| | 39 | }}} |
| | 40 | |
| | 41 | == Sorting (Local Mode) == |
| | 42 | |
| | 43 | {{{ |
| | 44 | ~$ pig -x local |
| | 45 | grunt> log = LOAD 'excite-small.log' AS (user, timestamp, query); |
| | 46 | grunt> grpd = GROUP log BY user; |
| | 47 | grunt> cntd = FOREACH grpd GENERATE group, COUNT(log) AS cnt; |
| | 48 | grunt> fltrd = FILTER cntd BY cnt > 50; |
| | 49 | grunt> srtd = ORDER fltrd BY cnt; |
| | 50 | grunt> STORE srtd INTO 'lab8_out3'; |
| | 51 | grunt> quit |
| | 52 | ~$ head lab8_out3 |
| | 53 | }}} |
| | 54 | |
| | 55 | == Connect Pig to Hadoop (Full Distributed Mode) == |
| | 56 | |
| | 57 | {{{ |
| | 58 | ~$ hadoop fs -put excite-small.log . |
| | 59 | ~$ pig |
| | 60 | grunt> log = LOAD 'excite-small.log' AS (user, timestamp, query); |
| | 61 | grunt> grpd = GROUP log BY user; |
| | 62 | grunt> cntd = FOREACH grpd GENERATE group, COUNT(log); |
| | 63 | grunt> STORE cntd INTO 'lab8_out1'; |
| | 64 | grunt> quit |
| | 65 | ~$ hadoop fs -cat lab8_out1/part-00000 |
| | 66 | }}} |