Analysis function in SQL

One , Analyze the difference between functions and aggregate functions

Aggregate functions ：
1 Aggregate functions can be used directly . max() ,min(),avg(),count()…
2 The calculation result of the aggregate function returns a value .
3 GROUP BY After grouping , The aggregation function will group fields according to , Each group returns a calculation result .
4 If the aggregate function is to be queried together with other fields , Then the remaining fields must be grouped fields .
5 GROUP BY after , stay SELECT Back Can only query Grouped fields , And aggregate functions , The remaining ungrouped fields are not allowed to be queried directly .
6 If the data does not exist in the table , Not subject to the above conditions .

Analysis function
1 Analysis functions have almost no restrictions .
2 The way of grouping is different , Analysis function Use PARTITION BY Grouping .
3 The number of return values is different ： The analysis function will be based on the number of rows in the table , Each row returns a result of the calculation .
4 ORDER BY In aggregate function , Just a simple sort , In addition to being able to sort in the analysis function , It can also be evaluated cumulatively .

Two , What is the analytic function ？

The analytic function is Oracle Powerful functions for complex report statistics , It can group in the data and then calculate some statistics based on the group , And each row of each group can return a statistical value .

          

3、 ... and , What is the difference between analysis function and aggregate function ？

Common aggregate functions use group by grouping , Each group returns a statistic , And the analysis function uses partition by grouping , And each row of each group can return a statistical value .

              

Four , Analyze the form of the function

The analysis function has a windowing function ：over()

Contains three analysis clauses : grouping (partition by), Sort (order by), window (rows) 

They are used in the following forms ：

over(partition by xxx order by yyy)

over(partition by xxx order by yyy rows between unbounded preceding and current row)                    notes ： Marked in red is the window scope clause

5、 ... and , Common function analysis

1, Sort analysis function

Seek the overall situation TopN Can be used when order by limit Realization , Just ask each group TopN You have to use the sorting analysis function to quickly realize

• row_number  In the same group , The serial number is not repeated , The serial number is absolutely continuous
• rand  In the same group , The serial number is repeated , The next serial number of the repeated serial number will skip
• dense_rank In the same group , The serial number is repeated , The serial number is absolutely continuous

1） Comprehensive case

 Table data 
create table TEST_ROW_NUMBER_OVER(
       id varchar(10) not null,
       name varchar(10) null,
       age varchar(10) null,
       salary int null
);
select * from TEST_ROW_NUMBER_OVER t;
 
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(1,'a',10,8000);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(1,'a2',11,6500);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(2,'b',12,13000);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(2,'b2',13,4500);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(3,'c',14,3000);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(3,'c2',15,20000);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(4,'d',16,30000);
insert into TEST_ROW_NUMBER_OVER(id,name,age,salary) values(5,'d2',17,1800);

 2） Sort query results （ No grouping ）

select id,name,age,salary,row_number()over(order by salary desc) rank
from TEST_ROW_NUMBER_OVER t

 

3)   according to id Grouping sorting

select id,name,age,salary,row_number()over(partition by id order by salary desc) rank
from TEST_ROW_NUMBER_OVER t

 4） Find out the number one data in each group

select * from(select id,name,age,salary,row_number()over(partition by id order by salary desc) rank
from TEST_ROW_NUMBER_OVER t)
where rank <2

2, Aggregate analysis functions

3, Shift analysis function

1)  WITH AS The phrase , Also called subquery part （subquery factoring）, You can define a SQL fragment , The SQL The fragment will be replaced by the whole SQL Statement use .

More with  as See the blog below ：

https://blog.csdn.net/jia718/article/details/88253918

with 
cr as 
( 
    select CountryRegionCode from person.CountryRegion where Name like 'C%' 
)

select * from person.StateProvince where CountryRegionCode in (select * from cr)





 The query results are the same as the following ：


select * from person.StateProvince where CountryRegionCode in 
         (select CountryRegionCode from person.CountryRegion where Name like 'C%')

2）  CTE  

3） lead ()

https://blog.csdn.net/a5276957/article/details/103808997

4) ·lag（）

 https://blog.csdn.net/a5276957/article/details/103808997

4,presto array_agg Analysis function

array_agg(expression): Turn an expression into an array

 

5, Window scope clause

https://www.cnblogs.com/Brambling/p/6721078.html