分析函數--FIRST_VALUE，LAST_VALUE，LAG，LEAD，ROW_NUMBER

一 ，FIRST_VALUE,LAST_VALUER的使用
1，FIRST_VALUE 是指返回組中數據窗口的第一個值
2，LAST_VALUE 是指返回組中數據窗口的最後一個值（注意點：不受排序的影響,後面發現是因爲沒有開窗導致的）

例子：
SELECT DEPARTMENT_ID,
FIRST_NAME || ' ' || LAST_NAME EMPLOYEE_NAME,
HIRE_DATE,
SALARY,
FIRST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY) AS LOWEST_SAL,
FIRST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY DESC) AS HIGHEST_SAL,
LAST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY) AS LAST_SAL,
LAST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY DESC) AS LAST_SAL_DESC
FROM HR.EMPLOYEES


上面的例子，並不能得到期待結果。
原因是沒有指定開窗子句，
order by會加上一個默認窗口RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW.
正確的寫法：
增加上指定窗口爲所有行--ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING

SELECT DEPARTMENT_ID,
FIRST_NAME || ' ' || LAST_NAME EMPLOYEE_NAME,
HIRE_DATE,
SALARY,
FIRST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY) AS LOWEST_SAL,
FIRST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY DESC) AS HIGHEST_SAL,
LAST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY
ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING） AS LAST_SAL,
LAST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY DESC
LAST_VALUE(FIRST_NAME || ' ' || LAST_NAME || ':'|| SALARY) OVER(PARTITION BY DEPARTMENT_ID ORDER BY SALARY DESC
ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING） AS LAST_SAL_DESC
FROM HR.EMPLOYEES


二，LAG,LEAD函數的使用
1，LAG可以訪問結果集中的其它行而不用進行自連接。它允許去處理遊標，就好像遊標
是一個數組一樣。在給定組中可參考當前行之前的行，這樣就可以從組中與當前行一起選擇
以前的行。Offset 是一個正整數，其默認值爲 1，若索引超出窗口的範圍，就返回默認值（默
認返回的是組中第一行），其相反的函數是 LEAD
2,LEAD 與 LAG 相反，LEAD 可以訪問組中當前行之後的行。Offset 是一個正整數，
其默認值爲 1，若索引超出窗口的範圍，就返回默認值（默認返回的是組中第一行）

例子：
SELECT FIRST_NAME || ' ' || LAST_NAME EMPLOYEE_NAME,
HIRE_DATE,
SALARY,
LAG(SALARY, 1, 0) OVER(ORDER BY HIRE_DATE ) AS 比當前前一個的薪酬,
LEAD(SALARY, 1, 0) OVER(ORDER BY HIRE_DATE ) AS 比當前後一個的薪酬
FROM HR.EMPLOYEES



三，ROW_NUMBE的使用
1，ROW_NUMBE是指 返回有序組中一行的偏移量，從而可用於按特定標準排序的行號。
也可以利用他排除重複的數據

例子：

SELECT DEPARTMENT_ID,
FIRST_NAME || ' ' || LAST_NAME EMPLOYEE_NAME,
EMPLOYEE_ID,
ROW_NUMBER() OVER(PARTITION BY DEPARTMENT_ID ORDER BY EMPLOYEE_ID) AS EMP_ID
FROM HR.EMPLOYEES