MySQL - database query - use of aggregate function, aggregate query, grouping query

Aggregate functions ( Statistical function )

SQL Allow calculation of data in the table , Take each column of data in the table as a whole , Do the longitudinal calculation .

column Indicates the field name , Perform some operations on a certain field

MAX、MIN、COUNT Null values are not calculated

COUNT(*) Returns the number of rows in all columns , Contains a null value

among COUNT Function can be used for any data type ( Because it only records the number of lines ), and SUM 、AVG Functions can only evaluate numeric types ,MAX and MIN Can be used for numerical 、 String or date time data type .

1. Count the number of employees in the enterprise  

SELECT COUNT(*) FROM emp;

Be careful COUNT() Null values are not included

SELECT COUNT(mgr) FROM emp;

2. Count the average salary of employees in the enterprise

SELECT AVG(sal) FROM emp;

3. Query the maximum salary of employees in the enterprise

SELECT MAX(sal) FROM emp;

4. Query the minimum wage of employees in the enterprise

SELECT MIN(sal) FROM emp;

5. Calculate the sum of wages for all sales

SELECT sal FROM emp WHERE job='salesman';

SELECT SUM(sal) FROM emp WHERE job='salesman';

Group query

Aggregate functions are actually statistical functions .

Next, let's learn about grouping statistical query .

One thing to declare is that , Although statistical functions are generally combined with grouping queries , But there must be some cases of using it alone . for example ： Basic paging operations when doing report display , Be sure to find out all the data .

So what is grouping ？

The concept of grouping is actually very common in life , For example, there are the following requirements ：

• 1. In a class , Ask men and women to compete in a debate

• 2. In the company , Each department is required to compete in a tug of war

For the first demand , Suppose there is a student table , Then there must be a gender field in the student table , The gender can only be male or female .

And in the company , If you want to group departments , There must be duplication in the contents of a department column . Departments and employees have a one to many relationship , A Department corresponds to multiple employees .

Grouping is meaningful only for fields with duplicate data , We can check emp surface , Find the field positions that can be grouped job And department number deptno.

First, employee number empno It can't be repeated , If I repeat , It's the same person , full name ename Generally, there will be no repetition

grammar

GROUP BY Then write which field to group

HAVING Filter the grouped data after grouping

WHERE Filter the data before grouping

HAVING Filter the data after grouping

SELECT  Grouping field / Aggregate functions  FROM  Table name  [WHERE  Conditions ] GROUP BY  Grouping field  [HAVING  Conditions after grouping ];

Query requirement

1. Count the number of people in each job , Output all jobs , Then group it , Multiple repetitions are in a group , Finally, count the number of people in each group

SELECT job,COUNT(*) FROM emp GROUP BY job;

2. Count the minimum and maximum wages for each position

--  Count out each position 
SELECT job FROM emp GROUP BY job;

SELECT job,MIN(sal),MAX(sal) FROM emp GROUP BY job;

The above two queries realize the basic operation of grouping , And these codes are written in standard format . But in the Group , Personally, I think the most troublesome thing is the restrictions of grouping operation .

matters needing attention

1. If a query ( non-existent GROUP BY Clause ), So in SELECT Only statistical functions are allowed in Clauses , No other fields are allowed .

When there is a statistical function , No other fields are allowed .

# Wrong statement 
SELECT ename,COUNT(job) FROM emp;
# The right sentence 
SELECT COUNT(job) FROM emp;

2. In statistical query ( There is GROUP BY Clause ),SELECT Only statistical functions and grouping fields are allowed in Clauses (GROUP BY Fields defined later ), No other fields are allowed .

# Wrong statement 
SELECT ename,COUNT(job) FROM emp GROUP BY job;
# The right sentence 
SELECT deptno,COUNT(job) FROM emp GROUP BY job;

When grouping later , On one principle ：GROUP BY The allowed fields in the clause are SELECT Allowed fields in Clause .

Query requirement

1. Display the salary of different positions less than 1500 The number of people

--  Show different positions 
SELECT job FROM emp GROUP BY job;  

--  Show the number of people in different positions  
SELECT job,COUNT(*) FROM emp GROUP BY job;  

--  Display the salary of different positions less than 1500 The number of people  
SELECT job,COUNT(*) FROM emp WHERE sal<1500 GROUP BY job;

2. Add one more restriction , Only show the number of people greater than 2 Of , In fact, No 1 The result of the question will be screened again

SELECT job,COUNT(*) FROM emp WHERE sal<1500 GROUP BY job HAVING COUNT(*)>2;

3. Show non salespeople (salesman) The sum of the job name and the monthly salary of employees engaged in the same job , And the total monthly salary of employees engaged in the same work shall be greater than or equal to 5000

SELECT job FROM emp WHERE job!='salesman';

SELECT job,SUM(sal) FROM emp WHERE job!='salesman' GROUP BY job;

SELECT job,SUM(sal) AS total FROM emp WHERE job!='salesman' GROUP BY job HAVING total>=5000;

WHERE and HAVING difference

• WHERE： Is to use... Before grouping ( There can be no GROUP BY), Statistical functions are not allowed .

• HAVING： Is used after grouping ( Must be combined with GROUP BY), Statistical functions are allowed .