hive join基礎

left join 是left outer join的簡寫，left join默認是outer屬性的。

Inner Join

Inner Join 邏輯運算符返回知足第一個（頂端）輸入與第二個（底端）輸入聯接的每一行。這個和用select查詢多表是同樣的效果，因此不多用到；

outer join則會返回每一個知足第一個（頂端）輸入與第二個（底端）輸入的聯接的行。它還返回任何在第二個輸入中沒有匹配行的第一個輸入中的行。

關鍵就是後面那句，返回的多一些。因此一般意義上的left join就是left outer join

CREATE TABLE t1(

name string,

age   int

)

PARTITIONED BY( hour string)

ROW FORMAT DELIMITED

FIELDS TERMINATED BY '\t'

COLLECTION ITEMS TERMINATED BY ':'

LINES TERMINATED BY '\n'

STORED AS textFILE;

CREATE TABLE t2(

name string,

sex   int

)

PARTITIONED BY( hour string)

ROW FORMAT DELIMITED

FIELDS TERMINATED BY '\t'

COLLECTION ITEMS TERMINATED BY ':'

LINES TERMINATED BY '\n'

STORED AS textFILE; 

因爲第二列是int，可是咱們插入的一個字符型，因此會顯示NULL

b       c

hive：

b       NULL    2011111111

LOAD DATA LOCAL INPATH '/opt/smc/xuanli/data/t1.txt' OVERWRITE INTO TABLE test.t1 partition (hour='2011111112');

LOAD DATA LOCAL INPATH '/opt/smc/xuanli/data/t2.txt' OVERWRITE INTO TABLE test.t2 partition (hour='2011111111');

select t1.*,t2.* from t1 left join t2  on (t1.hour=t2.hour and t1.name=t2.name);

select t1.*,t2.* from t1 right join t2  on (t1.hour=t2.hour and t1.name=t2.name);

select t1.*,t2.* from t1 join t2  on (t1.hour=t2.hour and t1.name=t2.name);

select t1.*,t2.* from t1 INNER  join t2  on (t1.hour=t2.hour and t1.name=t2.name);

select t1.*,t2.* from t1 full join t2  on (t1.hour=t2.hour and t1.name=t2.name);

SELECT t1.name  FROM t1 LEFT SEMI JOIN t2  on (t1.hour=t2.hour and t1.name=t2.name);

在t1存在。t2不存在

select t1.*,t2.* from t1 left join t2  on (t1.hour=t2.hour and t1.name=t2.name) where t1.name is not null and t2.name is null;

select t1.*,t3.* from t1 left join (select name,hour from test.t2) t3  on (t1.hour=t3.hour and t1.name=t3.name) where t1.name is not null and t3.name is null;