分組函數(shù)

• 什么是分組函數(shù)

  • 分組函數(shù)作用于一組數(shù)據(jù)，并對一組數(shù)據(jù)返回一個值

• 組函數(shù)類型：主要有6種

  • AVG - 平均
  • COUNT - 計數(shù)
  • MAX - 最大
  • MIN - 最小
  • SUM - 求和
  • STDDEV - 方差

• 組函數(shù)語法

SELECT  [column,] group_function(column), ...
FROM    table
[WHERE  condition]
[GROUP BY column]
[ORDER BY column]

• AVG（平均值）和 SUM （合計）函數(shù)
  • 可以對數(shù)值型數(shù)據(jù)使用AVG 和 SUM 函數(shù)。

SELECT AVG(salary), MAX(salary),
              MIN(salary), SUM(salary)
FROM   employees
WHERE  job_id LIKE '%REP%'

• MIN（最小值）和 MAX（最大值）函數(shù)
  • 可以對任意數(shù)據(jù)類型的數(shù)據(jù)使用 MIN 和 MAX 函數(shù)。

SELECT MIN(hire_date), MAX(hire_date)
FROM    employees;

• COUNT（計數(shù)）函數(shù)
  • COUNT(*) 返回表中記錄總數(shù),適用于任意數(shù)據(jù)類型。

SELECT COUNT(*)
FROM    employees
WHERE  department_id = 50;

• COUNT(expr) 返回expr不為空的記錄總數(shù)。

SELECT COUNT(commission_pct)
FROM   employees
WHERE  department_id = 50;

• 組函數(shù)與空值：組函數(shù)忽略空值。

SELECT AVG(commission_pct)
FROM   employees;

-- 
Select avg(commission_pct),sum(commission_pct)/107,
sum(commission_pct)/count(commission_pct)
From employees;
-- 查看結(jié)果的不同

• 在組函數(shù)中使用NVL函數(shù)
  • NVL函數(shù)使分組函數(shù)無法忽略空值

SELECT AVG(NVL(commission_pct, 0))
FROM   employees

• DISTINCT 關(guān)鍵字
  • COUNT(DISTINCT expr)返回expr非空且不重復(fù)的記錄總數(shù)

SELECT COUNT(DISTINCT department_id)
FROM   employees

• 分組數(shù)據(jù) ：GROUP BY 子句語法
  • 可以使用GROUP BY子句將表中的數(shù)據(jù)分成若干組

SELECT  column, group_function(column)
FROM  table
[WHERE  condition]
[GROUP BY group_by_expression]
[ORDER BY column];

• 明確：WHERE一定放在FROM后面
• 在SELECT 列表中所有未包含在組函數(shù)中的列都應(yīng)該包含
  在 GROUP BY 子句中

SELECT   department_id, AVG(salary)
FROM     employees
GROUP BY department_id 

• 包含在 GROUP BY 子句中的列不必包含在SELECT 列表中

SELECT   AVG(salary)
FROM     employees
GROUP BY department_id

• 使用多個列分組
  • 在GROUP BY子句中包含多個列

SELECT   department_id dept_id, job_id, SUM(salary)
FROM     employees
GROUP BY department_id, job_id 

• 非法使用組函數(shù)
  • 所有包含于SELECT 列表中，而未包含于組函數(shù)中的列都
    必須包含于 GROUP BY 子句中

-- 錯誤的示例代碼
SELECT department_id, COUNT(last_name)
FROM   employees
-- SELECT department_id, COUNT(last_name)
--       *
-- ERROR at line 1:
-- ORA-00937: not a single-group group function

• 不能在 WHERE 子句中使用組函數(shù)。
• 可以在 HAVING 子句中使用組函數(shù)。

-- 錯誤的示例代碼
SELECT   department_id, AVG(salary)
FROM     employees
WHERE    AVG(salary) > 8000
GROUP BY department_id;

-- WHERE  AVG(salary) > 8000
--       *
-- ERROR at line 3:
-- ORA-00934: group function is not allowed here
-- WHERE 子句中不能使用組函數(shù)

• 過濾分組：Having 字句

使用 HAVING 過濾分組:

1. 行已經(jīng)被分組。
2. 使用了組函數(shù)。
3. 滿足HAVING 子句中條件的分組將被顯示。

SELECT  column, group_function
FROM    table
[WHERE  condition]
[GROUP BY group_by_expression]
[HAVING group_condition]
[ORDER BY column];

-- 示例
SELECT   department_id, MAX(salary)
FROM     employees
GROUP BY department_id
HAVING   MAX(salary)>10000 

• 嵌套組函數(shù)
  • 顯示各部門平均工資的最大值

SELECT   MAX(AVG(salary))
FROM     employees
GROUP BY department_id;