union和join是需要聯合多張表時常見的關聯詞,具體概念我就不說了,想知道上網查就行,因為我也記不準確。
先說差別:union對兩張表的操作是合并數據條數,等于是縱向的,要求是兩張表字段必須是相同的(Schema of both sides of union should match.)。也就是說如果A表中有三條數據,B表中有兩條數據,那么A union B就會有五條數據。說明一下union 和union all的差別,對于union如果存在相同的數據記錄會被合并,而union all不會合并相同的數據記錄,該有多少條記錄就會有多少條記錄。例如在mysql下執行以下語句:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
select * from tmp_libingxue_a; name number libingxue 1001 yuwen 1002 select * from tmp_libingxue_b; name number libingxue 1001 feiyao 1003 select * from tmp_libingxue_a union select * from tmp_libingxue_b; libingxue 1001 yuwen 1002 feiyao 1003 select * from tmp_libingxue_a union all select * from tmp_libingxue_b; libingxue 1001 yuwen 1002 libingxue 1001 feiyao 1003 |
但是這樣在hive里面是不能執行的,執行select * from tmp_libingxue_a union all select * from tmp_libingxue_b;會failed,hive中union必須在子查詢中進行。如
1
|
select * from ( select * from tmp_yuwen_a union all select * from tmp_yuwen_b) t1; |
注意,必須是union all,單獨用union它會提示你缺少ALL,而且后面的t1必須寫,你可以寫成a或者b,但是一定要寫,不寫會出錯。
而join則是偏于橫向的聯合,僅僅是偏向于,等下詳細說明。join跟union比起來顯得更寬松,對兩個表的字段不做要求,沒有限制條件的join等于兩個表的笛卡爾乘積,所有join需要有限制條件來約束,經過限制的join就是橫向的擴張了。對于滿足限制條件的join會被提取出來,不滿足的直接過濾掉。用法可以很靈活,下面有兩個簡單的例子:
1
2
|
select * from ( select * from tmp_yuwen_a)t1 join ( select * from tmp_yuwen_b) t2; select * from tmp_yuwen_a t1 join ( select * from tmp_yuwen_b) t2; |
left outer join和right outer join用法類似,區別就是left outer join會把左邊表的字段全部選擇出來,右邊表的字段把符合條件的也選擇出來,不滿足的全部置空,也就是說以左邊表為參照。right outer join同理以右邊表為參照。這三個join之間的差別說過很多次,網上也有更詳細的解釋,不再贅述。
相同點:在某些特定的情況下,可以用join實現union all的功能,這種情況是有條件的,當出現這種情況的時候選擇union all還是group by就可以看情況或者看兩者的消耗而決定。sql雖然就在那么幾個關鍵詞,但變化多端、功能強大,只要能實現想要的功能,怎么用隨便你。需求情況sql簡單重現如下
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
drop table tmp_libingxue_resource; create external table if not exists tmp_libingxue_resource( user_id string, shop_id string, auction_id string, search_time string )partitioned by (pt string) row format delimited fields terminated by '\t' lines terminated by '\n' stored as sequencefile; drop table tmp_libingxue_result; create external table if not exists tmp_libingxue_result( user_id string, shop_id string, auction_id string, search_time string )partitioned by (pt string) row format delimited fields terminated by '\t' lines terminated by '\n' stored as sequencefile; insert overwrite table tmp_libingxue_result where (pt=20041104) select * from tmp_libingxue_resource; |
1
2
|
sudo -u taobao hadoop dfs -rmr /group/tbads/warehouse/tmp_libingxue_result/pt =20041104 sudo -u taobao hadoop jar /home/taobao/dataqa/framework/DailyReport .jar com.alimama.loganalyzer.tool.SeqFileLoader tmp_libingxue_resource.txt hdfs: //v039182 .sqa.cm4:54310 /group/tbads/warehouse/tmp_libingxue_result/pt =20041104 /part-00000 |
1
|
hive> select * from tmp_libingxue_resource; |
1
2
3
|
OK 2001 0 11 101 20041104 2002 0 11 102 20041104 |
1
|
hive> select * from tmp_libingxue_result; |
1
2
3
|
OK 2001 0 12 103 20041104 2002 0 12 104 20041104 |
1
2
3
4
5
6
|
select user_id,shop_id, max (auction_id), max (search_time) from ( select * from tmp_libingxue_resource union all select * from tmp_libingxue_result )t1 group by user_id,shop_id; |
1
2
|
2001 0 12 103 2002 0 12 104 |
1
2
3
4
5
6
|
select t1.user_id,t1.shop_id,t2.auction_id,t2.search_time from ( select * from tmp_libingxue_resource) t1 join ( select * from tmp_libingxue_result) t2 on t1.user_id=t2.user_id and t1.shop_id=t2.shop_id; |
1
2
|
2001 0 12 103 2002 0 12 104 |
通過前面的介紹,使用UNION對表的結果集進行并運算與使用JOIN對多表進行連接,二者有本質的不同。
下面給出一個使用UNION運算符連接二表記錄的運算實例。
典型的二表記錄的UNION運算
假定有兩個表Table3和Table4,其包含的列和數據分別如下所示。
Table1數據庫表
Table2數據庫表
Table1表和Table2表具有相同的列結構,因此可以使用UNION運算符連接兩個表的記錄集,得到的連接結果如下表所示。
使用UNION連接Table3表和Table4表的記錄
上述連接過程的實現代碼可表示如下:
1
2
3
4
5
|
SELECT * FROM Table1 UNION SELECT * FROM Table2 |