由于 Group By，每日平均值计算不正确

Question

我有以下 SQL，我试图在其中获取一个月内总计数的日平均值。

SELECT A.BUSINESS_UNIT, A.INV_ITEM_ID, C.MFG_ID, C.MFG_ITM_ID, 
 A.ADJUST_TYPE, A.QTY_BASE, A.UNIT_MEASURE_STD,
 A.DT_TIMESTAMP, B.DESCR254_MIXED, A.TRANSACTION_GROUP, 
 (CONVERT(CHAR(10),A.TRANSACTION_DATE,121)), A.DEPTID, B.PRICE_LIST, 
 A.DESTIN_GL_BU,  A.QTY_BASE* B.PRICE_LIST, (COUNT(*) * 1.0 / 31 ) AS 'Daily 
 Average'
FROM PS_TRANSACTION_INV A, (PS_PURCH_ITEM_ATTR B LEFT OUTER JOIN  
 PS_ITEM_MFG C ON  B.SETID = C.SETID AND B.INV_ITEM_ID = C.INV_ITEM_ID ), 
 PS_CM_ACCTG_DIST_D D 
WHERE ( A.BUSINESS_UNIT = '11MMS' 
 AND A.INV_ITEM_ID = B.INV_ITEM_ID 
 AND C.PREFERRED_MFG = 'Y' 
 AND A.TRANSACTION_DATE >= '05-01-2018' 
 AND A.TRANSACTION_DATE <= '05-31-2018' 
 AND A.BUSINESS_UNIT = D.BUSINESS_UNIT 
 AND A.INV_ITEM_ID = D.INV_ITEM_ID 
 AND D.TRANSACTION_GROUP = A.TRANSACTION_GROUP 
 AND A.INV_ITEM_ID = '1')
GROUP BY A.BUSINESS_UNIT, A.INV_ITEM_ID, C.MFG_ID, C.MFG_ITM_ID, 
 A.ADJUST_TYPE, A.QTY_BASE, A.UNIT_MEASURE_STD,
 A.DT_TIMESTAMP, B.DESCR254_MIXED, A.TRANSACTION_GROUP, 
 (CONVERT(CHAR(10),A.TRANSACTION_DATE,121)), A.DEPTID, B.PRICE_LIST, 
 A.DESTIN_GL_BU,  A.QTY_BASE* B.PRICE_LIST

根据上面的数据运行，我得到 36 行（我的计数）的输出，我将其除以 31（天），并期望日平均值为 1.161，但是我的输出是每行每日平均值的行是 .032258（我确定为 1.16 / 36）。非聚合列在分组依据中，但以某种方式扭曲了每日平均值。

如果我运行仅对计算出的平均值进行查询 (COUNT(*) * 1.0 / 31 ) AS 'Daily Average'，那么我会得到正确的日平均值 1.161，因此我知道它与分组依据有关。提前致谢！

编辑：

运行这个查询 SELECT (计数(*) * 1.0 / 31 ) 作为 'Daily Average' FROM PS_TRANSACTION_INV A, (PS_PURCH_ITEM_ATTR B 左外连接<br> PS_ITEM_MFG C ON B.SETID = C.SETID AND B.INV_ITEM_ID = C.INV_ITEM_ID ), PS_CM_ACCTG_DIST_DD WHERE ( A.BUSINESS_UNIT = '11MMS' 和 A.INV_ITEM_ID = B.INV_ITEM_ID 和 C.PREFERRED_MFG = 'Y' AND A.TRANSACTION_DATE >= '05-01-2018' AND A.TRANSACTION_DATE <= '05-31-2018' 和 A.BUSINESS_UNIT = D.BUSINESS_UNIT 和 A.INV_ITEM_ID = D.INV_ITEM_ID 和 D.TRANSACTION_GROUP = A.TRANSACTION_GROUP AND A.INV_ITEM_ID = '1') 产生 1.161 的正确结果，就在我将它添加到上面的查询中时，我的结果不正确。这就是我试图在主查询中模仿的内容。

Answer 1

当您按所有列添加分组后，count 的结果将始终为 1，这就是您得到的结果。

完成您所要求的最简单方法是将第二个查询添加为主查询的子查询，到那时您可能会丢失该组，如下所示：

SELECT A.BUSINESS_UNIT, A.INV_ITEM_ID, C.MFG_ID, C.MFG_ITM_ID, 
 A.ADJUST_TYPE, A.QTY_BASE, A.UNIT_MEASURE_STD,
 A.DT_TIMESTAMP, B.DESCR254_MIXED, A.TRANSACTION_GROUP, 
 (CONVERT(CHAR(10),A.TRANSACTION_DATE,121)), A.DEPTID, B.PRICE_LIST, 
 A.DESTIN_GL_BU,  A.QTY_BASE* B.PRICE_LIST,  
 (SELECT COUNT(*) * 1.0 / 31
     FROM PS_TRANSACTION_INV A, (PS_PURCH_ITEM_ATTR B LEFT OUTER JOIN
      PS_ITEM_MFG C ON  B.SETID = C.SETID AND B.INV_ITEM_ID = C.INV_ITEM_ID ), 
      PS_CM_ACCTG_DIST_D D 
     WHERE ( A.BUSINESS_UNIT = '11MMS' 
      AND A.INV_ITEM_ID = B.INV_ITEM_ID 
      AND C.PREFERRED_MFG = 'Y' 
      AND A.TRANSACTION_DATE >= '05-01-2018' 
      AND A.TRANSACTION_DATE <= '05-31-2018' 
      AND A.BUSINESS_UNIT = D.BUSINESS_UNIT 
      AND A.INV_ITEM_ID = D.INV_ITEM_ID 
      AND D.TRANSACTION_GROUP = A.TRANSACTION_GROUP 
      AND A.INV_ITEM_ID = '1') 
 )  AS 'Daily Average'
FROM PS_TRANSACTION_INV A, (PS_PURCH_ITEM_ATTR B LEFT OUTER JOIN  
 PS_ITEM_MFG C ON  B.SETID = C.SETID AND B.INV_ITEM_ID = C.INV_ITEM_ID ), 
 PS_CM_ACCTG_DIST_D D 
WHERE ( A.BUSINESS_UNIT = '11MMS' 
 AND A.INV_ITEM_ID = B.INV_ITEM_ID 
 AND C.PREFERRED_MFG = 'Y' 
 AND A.TRANSACTION_DATE >= '05-01-2018' 
 AND A.TRANSACTION_DATE <= '05-31-2018' 
 AND A.BUSINESS_UNIT = D.BUSINESS_UNIT 
 AND A.INV_ITEM_ID = D.INV_ITEM_ID 
 AND D.TRANSACTION_GROUP = A.TRANSACTION_GROUP 
 AND A.INV_ITEM_ID = '1')

由于 Group By，每日平均值计算不正确

Daily average not calculating correctly as a result of Group By

sql

sql-server-2014