【HIVE】(3)联合查询join、时间戳函数、字符串函数

时间:2023-03-08 23:18:43
【HIVE】(3)联合查询join、时间戳函数、字符串函数

数据
t_join1.txt
1,a,1
2,b,2
3,c,4 t_join2.txt
1,2a
2,2b
3,2c 建表、导入:
create table t_join1(id int, name string, cid int) row format delimited fields terminated by ",";
create table t_join2(id int, name string) row format delimited fields terminated by ",";
load data local inpath "/root/example/hive/data/t_join1.txt" into table t_join1;
load data local inpath "/root/example/hive/data/t_join2.txt" into table t_join2; join,多表联合查询
join:只输出on条件相等的行
select * from t_join1 t1 join t_join2 t2 on t1.cid=t2.id; 等同于:
select * from t_join1 t1 inner join t_join2 t2 on t1.cid=t2.id;
select * from t_join1 t1, t_join2 t2 where t1.cid=t2.id; left join,输出on条件相等的行 + 左表剩余的行(这些行对应右表的字段使用NULL)
select * from t_join1 t1 left join t_join2 t2 on t1.cid=t2.id; right join,输出on条件相等的行 + 右表剩余的行(这些行对应左表的字段使用NULL)
select * from t_join1 t1 right join t_join2 t2 on t1.cid=t2.id; full join,输出左右两表所有行,on条件不满足的行,使用NULL
select * from t_join1 t1 full join t_join2 t2 on t1.cid=t2.id; built-in function,内置函数 1.查看系统自带的函数
hive> show functions;
2.显示自带的函数的用法
hive> desc function year;
3.详细显示自带的函数的用法
hive> desc function extended year; * 数值函数:
round(DOUBLE a):四舍五入取整;
round(DOUBLE a, INT d):指定小数点位数;
rand():0~1之间的随机小数;
select round(rand()*a+b):取[b, a+b]之间的整数; * 日期函数:
- 当前时间:
current_date
current_timestamp - 时间戳函数:
from_unixtime(bigint unixtime[, string format]):时间戳转换为时间;
unix_timestamp():获取当前时间戳;
unix_timestamp(string date):时间转换为时间戳; - 分别获取年、月、日:
year(string date)
quarter(date/timestamp/string)
month(string date)
day(string date) dayofmonth(date)
hour(string date)
minute(string date)
second(string date)
weekofyear(string date) - 时间间隔:
datediff(string enddate, string startdate):两段时间相隔的天数;
date_add(date/timestamp/string startdate, tinyint/smallint/int days):增加天数;
date_sub(date/timestamp/string startdate, tinyint/smallint/int days):减少天数; * 条件函数:
if(boolean testCondition, T valueTrue, T valueFalseOrNull)
isnull( a )
isnotnull ( a )
nvl(T value, T default_value):替换空值;
COALESCE(T v1, T v2, ...):返回第一个非空值; 重难点: CASE a WHEN b THEN c [WHEN d THEN e]* [ELSE f] END:该函数可以将多行转换为多行多列;
- 增加yuwen和shuxue列:
select * , case course when "yuwen" then score else 0 end as yuwen, case course when "shuxue" then score else 0 end as shuxue from t_course;
- 使用max分组:
select sid, max(yuwen) as yuwen, max(shuxue) as shuxue from (select * , case course when "yuwen" then score else 0 end as yuwen, case course when "shuxue" then score else 0 end as shuxue from t_course) t group by sid;
- 获取语文成绩比数学成绩好的学生:
select * from (select sid, max(yuwen) as yuwen, max(shuxue) as shuxue from (select * , case course when "yuwen" then score else 0 end as yuwen, case course when "shuxue" then score else 0 end as shuxue from t_course) t group by sid) t where yuwen>shuxue; 还有另一种方法,可以完成上面任务,提示:使用collect_set(),大家可以先想想思路。 CASE WHEN a THEN b [WHEN c THEN d]* [ELSE e] END * 字符串函数:
concat(string|binary A, string|binary B...):字符串连接
concat_ws(string SEP, string A, string B...):使用指定字符连接;
substr(string|binary A, int start, int len):获取子字符串;
substring(string|binary A, int start, int len):同上;
trim(string A):去除两边空格;
lower(string A) lcase(string A) upper(string A) ucase(string A):大小写转换;
split(string str, string pat):字符串分割为数组;
str_to_map(text[, delimiter1, delimiter2]):字符串转换为map;