sql 数据处理时join字段慎重选择--避免出现double数据！

前几天做BI Tabular表格模型的时候，数据中心核对数据发现模型展现数据比实际数据要多，经过核查之后，发现原来我是一个不经意，在做数据选取的时候，inner join的字段CITY_NAME_CN选取错误。

首先对比事实表和我数据源选取的数据量差异：正常（8080945条记录）

SELECT
      count(YYYYMMDD)
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK]  --

我的数据源sql:选取（8110909条记录）

SELECT  [GUID]
      ,[YYYYMMDD]
      ,[BRAND]
      ,[BRAND_GROUP]
      ,[MODEL]
      ,[CHANNEL]
      ,[SALES_UNIT]
      ,[SALES_VALUE]
      ,[PC_TYPE]
      ,[DESIGN]
      ,[PROCESSOR_BRAND]
      ,[PROCESSOR]
      ,[PROCESSOR_NUMBE]
      ,[STORAGE_CAPAC]
      ,[RAM_MB]
      ,[DVD_WRITER]
      ,[SIZE_INCH]
      ,[GPU_BRAND]
      ,[GPU_MODEL]
      ,[ONBOARD_GRAPHIC]
      ,[V_RAM]
      ,[OS_PLANTFORM]
      ,[MICROSOFT_OS]
      ,[CATEGORIES]
      ,[PPC_PRICE_BRAND]
      ,[DPC_PRICE_BRAND]
      ,a.[CITY_NAME_CN]
      ,a.[CITY_NAME_EN]
      ,ISNULL(b.CITY_CODE,0)AS CITY_CODE
      ,[BU_GROUP_CODE]
      ,A.CITY_FULL_NAME_CN
      ,B.CITY_FULL_NAME
      ,Form_Factor
,CITY_GROUP
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] A
   LEFT JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_NAME_CN=B.CITY_NAME_CN

SELECT
      count([YYYYMMDD])
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] A
   LEFT JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_NAME_CN=B.CITY_NAME_CN

由此可见，问题出在我的事实表A join 维度表B 的时候出现问题，出现了一对多的问题。

首先，我要查一下维表CB_GEOGRAPHY对于字段CITY_NAME_CN（乡镇名称）是不是有重复，

SELECT a.CITY_NAME_CN,count(1)as num
   FROM [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] a
   --where a.CITY_NAME_CN=N'九江'
   group by a.CITY_NAME_CN having count(1)>1 

aaarticlea/png;base64,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" alt="" />

想一下很容易理解，地理维度表中，不同地区的乡镇名可能出现重名。所以特设了一个CITY_FULL_NAME的字段。那么join之后究竟是哪些字段出现了重复呢，

   SELECT b.CITY_FULL_NAME_CN,B.CITY_NAME_CN FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] B inner join
   (SELECT a.CITY_NAME_CN,count(1)as num
   FROM [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] a
   --where a.CITY_NAME_CN=N'九江'
   group by a.CITY_NAME_CN having count(1)>1 )aa  ON aa.CITY_NAME_CN=B.CITY_NAME_CN

aaarticlea/png;base64,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" alt="" />

发现这些记录数为：29964，而且全是出在'九江'这个地方，8080945-8110909=-29964，BINGO！

故：出现了一对多的情况，join之后出现了DOUBLE的数据，

SELECT
      A.CITY_FULL_NAME_CN
      ,B.CITY_FULL_NAME
      ,A.CITY_NAME_CN
      ,B.CITY_NAME_CN
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] A
   inner JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_NAME_CN=B.CITY_NAME_CN
   where b.CITY_NAME_CN=N'九江' --

SELECT
      A.CITY_FULL_NAME_CN
      ,B.CITY_FULL_NAME
      ,A.CITY_NAME_CN
      ,B.CITY_NAME_CN
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] A
   LEFT JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_FULL_NAME_CN=B.CITY_FULL_NAME
   where b.CITY_NAME_CN=N'九江'   ----29964

修改之后，LEFT JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_FULL_NAME_CN=B.CITY_FULL_NAME，okay！

SELECT  [GUID]
      ,[YYYYMMDD]
      ,[BRAND]
      ,[BRAND_GROUP]
      ,[MODEL]
   ,[CHANNEL]
      ,CASE [CHANNEL]WHEN 'NON-ITS'THEN'3C'WHEN 'INTERNET B2C'THEN'NC'WHEN 'ASSB'THEN'TC'WHEN 'CS'THEN'TC'ELSE'TC' END AS CHANNEL_TYPE_CODE
      ,[SALES_UNIT]
      ,[SALES_VALUE]
      ,[PC_TYPE]
      ,[DESIGN]
      ,[PROCESSOR_BRAND]
      ,[PROCESSOR]
      ,[PROCESSOR_NUMBE]
      ,[STORAGE_CAPAC]
      ,[RAM_MB]
      ,[DVD_WRITER]
      ,[SIZE_INCH]
      ,[GPU_BRAND]
      ,[GPU_MODEL]
      ,[ONBOARD_GRAPHIC]
      ,[V_RAM]
      ,[OS_PLANTFORM]
      ,[MICROSOFT_OS]
      ,[CATEGORIES]
      ,[PPC_PRICE_BRAND]
      ,[DPC_PRICE_BRAND]
      ,a.[CITY_NAME_CN]
      ,a.[CITY_NAME_EN]
      ,ISNULL(b.CITY_CODE,0)AS CITY_CODE
      ,[BU_GROUP_CODE]
      ,A.[CITY_FULL_NAME_CN]
      ,Form_Factor
      ,isnull(CITY_GROUP,'')as CITY_GROUP
  FROM [DM_ACCN_T1].[dbo].[FS_MS_GFK] A
   LEFT JOIN  [DM_ACCN_T1].[dbo].[CB_GEOGRAPHY] B ON A.CITY_FULL_NAME_CN=B.CITY_FULL_NAME