我将总结我的架构和我需要的结果：
Table A(released and rejected):
day, specimenid, status,bunch of other unneeded fields....
1,AA123, released
1,AA132, released
2,BB234, rejected
3,BB444, released
3,AA444, rejected
Table B(pending):
day, specimenid, bunch of other unneeded columns
1, BB333
1, H234
2, C333
3, F333
我需要的结果：
day, count, status
1, 2, released
1, 2, pending
2, 1, rejected
2, 1, pending
3, 1, released
3, 1, rejected
3, 1, pending
换句话说，对于每一天，我需要获得每个状态的总计数。
我这样做：
select day, count(specimenid) from tableA where status='released', 'released'
union all
select day, count(specimenid) from tableA where status='rejected', 'rejected'
union all
select day, count(specimenid) from tableB, 'pending'
它现在正在运作，但我认为它可能有点草率。
这是我对此的完整怪物查询：
WITH cte(rejected) AS (SELECT DISTINCT [Specimen ID]
FROM dbo.QuickLabDump
WHERE (Outcome = 'REJECTED') AND ([Specimen ID] IS NOT NULL))
SELECT CONVERT(VARCHAR(8), a.[Date Entered], 1) AS [Full Date], DATEPART(yy, a.[Date Entered]) AS [Year Entered], LEFT(DATENAME(MONTH, a.[Date Entered]), 3)
AS [Month Entered], DATEPART(dd, a.[Date Entered]) AS [Day Entered], CASE WHEN DATEPART(WEEKDAY, [DATE entered])
= 1 THEN 'Sun' WHEN DATEPART(WEEKDAY, [DATE entered]) = 2 THEN 'Mon' WHEN DATEPART(WEEKDAY, [DATE entered])
= 3 THEN 'Tus' WHEN DATEPART(WEEKDAY, [DATE entered]) = 4 THEN 'Wed' WHEN DATEPART(WEEKDAY, [DATE entered])
= 5 THEN 'Thu' WHEN DATEPART(WEEKDAY, [DATE entered]) = 6 THEN 'Fri' WHEN DATEPART(WEEKDAY, [DATE entered]) = 7 THEN 'Sat' END AS DOW,
CONVERT(VARCHAR(8), DATEADD(D, - (1 * DATEPART(dw, a.[Date Entered])) + 6, a.[Date Entered]), 1) AS [Week Ending], COUNT(DISTINCT a.[Specimen ID])
AS CountAccns, c.SALESREP AS [Sales Rep], c.NPI AS MLNPI, e.NPIb AS IMSNPI, e.SpecialtyPrimaryCodeb AS [IMS Specialty Primary Code],
sm1.specialtydesc as [Specialty Primary Description],
e.SpecialtySecondaryCodeb AS [IMS Specialty Secondary Code],
sm2.specialtydesc as [Specialty Secondary Description],
e.SpecialtyTertiaryCodeb AS [IMS Specialty Tertiary Code],
sm3.specialtydesc as [Specialty Tertiary Description],
e.ProfessionalID1b AS [IMS Professional ID 1], a.[Requesting Physician] AS Physician, a.[Practice Code], b.[MLIS Practice ID] AS [MLIS Code], a.[Practice Name],
c.DATEESTABLISHED AS [Date Established], c.PRACTICEADDRESS1 AS Address, c.PRACTICEADDRESS2 AS Address2, c.PRACTICECITY AS City,
c.PRACTICESTATE AS State, d.[MLIS Status] AS Status, MAX(a.[Order Count]) AS [order count],
'Released' as [Release Status]
FROM dbo.QuickLabDump AS a LEFT OUTER JOIN
dbo.qlmlismapping AS b ON b.[Quicklab ID] = a.[Practice Code] LEFT OUTER JOIN
dbo.PracticeandPhysician AS c ON a.[Requesting Physician] = c.DOCTORFIRSTNAME + ' ' + c.DOCTORLASTNAME AND
a.[Practice Code] = c.PRACTICECODE LEFT OUTER JOIN
dbo.IMSData AS e ON c.NPI = e.NPIb LEFT OUTER JOIN
dbo.QLMLISInfo AS d ON b.[MLIS Practice ID] = d.[MLIS Practice ID]
left outer JOIN
dbo.SpecialtyMapping sm1 ON e.SpecialtyPrimaryCodeb = sm1.specialtyabbrev
left outer JOIN
dbo.SpecialtyMapping sm2 ON e.SpecialtySecondaryCodeb = sm2.specialtyabbrev
left outer JOIN
dbo.SpecialtyMapping sm3 ON e.SpecialtyTertiaryCodeb = sm3.specialtyabbrev
WHERE (a.[Date Entered] > '20110101') AND (NOT EXISTS
(SELECT 1 AS Expr1
FROM cte AS cte_1
WHERE (rejected = a.[Specimen ID]))) AND (NOT EXISTS
(SELECT 1 AS Expr1
FROM dbo.PendingSpecimens
WHERE ([Specimen ID] = a.[Specimen ID])))
GROUP BY a.[Date Entered], c.SALESREP, c.NPI, e.NPIb, e.SpecialtyPrimaryCodeb, e.SpecialtySecondaryCodeb, e.SpecialtyTertiaryCodeb, e.ProfessionalID1b,
a.[Requesting Physician], a.[Practice Code], b.[MLIS Practice ID], a.[Practice Name], c.DATEESTABLISHED, c.PRACTICEADDRESS1, c.PRACTICEADDRESS2,
c.PRACTICECITY, c.PRACTICESTATE, d.[MLIS Status],sm1.specialtydesc,sm2.specialtydesc,sm3.specialtydesc
union all
select
CONVERT(VARCHAR(8), a.[Date Entered], 1) AS [Full Date],
DATEPART(yy, a.[Date Entered]) AS [Year Entered],
LEFT(DATENAME(MONTH, a.[Date Entered]), 3) AS [Month Entered],
DATEPART(dd, a.[Date Entered]) AS [Day Entered],
CASE WHEN DATEPART(WEEKDAY, [DATE entered])
= 1 THEN 'Sun' WHEN DATEPART(WEEKDAY, [DATE entered]) = 2 THEN 'Mon' WHEN DATEPART(WEEKDAY, [DATE entered])
= 3 THEN 'Tus' WHEN DATEPART(WEEKDAY, [DATE entered]) = 4 THEN 'Wed' WHEN DATEPART(WEEKDAY, [DATE entered])
= 5 THEN 'Thu' WHEN DATEPART(WEEKDAY, [DATE entered]) = 6 THEN 'Fri' WHEN DATEPART(WEEKDAY, [DATE entered]) = 7 THEN 'Sat' END AS DOW,
CONVERT(VARCHAR(8), DATEADD(D, - (1 * DATEPART(dw, a.[Date Entered])) + 6, a.[Date Entered]), 1) AS [Week Ending],
COUNT(DISTINCT a.[Specimen ID]) AS CountAccns,
null AS [Sales Rep],
null AS MLNPI,
null AS IMSNPI,
null AS [IMS Specialty Primary Code],
null as [Specialty Primary Description],
null AS [IMS Specialty Secondary Code],
null as [Specialty Secondary Description],
null AS [IMS Specialty Tertiary Code],
null as [Specialty Tertiary Description],
null AS [IMS Professional ID 1],
null as Physician,
SUBSTRING(agency,1,charindex(' - ',agency,1)-1) as [Practice Code],
b.[MLIS Practice ID] AS [MLIS Code],
SUBSTRING(agency,charindex(' - ',agency,1)+3,len(agency)) as [Practice Name],
c.DATEESTABLISHED AS [Date Established],
c.PRACTICEADDRESS1 AS Address,
c.PRACTICEADDRESS2 AS Address2,
c.PRACTICECITY AS City,
c.PRACTICESTATE AS State,
d.[MLIS Status] AS Status,
null AS [order count],
'Pending' as [Release Status]
from dbo.pendingspecimens AS a LEFT OUTER JOIN
dbo.qlmlismapping AS b ON b.[Quicklab ID] = SUBSTRING(agency,1,charindex(' - ',agency,1))
LEFT OUTER JOIN
dbo.PracticeandPhysician AS c ON SUBSTRING(agency,1,charindex(' - ',agency,1)) = c.PRACTICECODE LEFT OUTER JOIN
dbo.IMSData AS e ON c.NPI = e.NPIb LEFT OUTER JOIN
dbo.QLMLISInfo AS d ON b.[MLIS Practice ID] = d.[MLIS Practice ID]
WHERE (a.[Date Entered] > '20110101')
group by agency,a.[Date Entered], c.SALESREP, c.NPI, e.NPIb, e.SpecialtyPrimaryCodeb, e.SpecialtySecondaryCodeb, e.SpecialtyTertiaryCodeb, e.ProfessionalID1b,
b.[MLIS Practice ID], c.DATEESTABLISHED, c.PRACTICEADDRESS1, c.PRACTICEADDRESS2,
c.PRACTICECITY, c.PRACTICESTATE, d.[MLIS Status]
union all
SELECT CONVERT(VARCHAR(8), a.[Date Entered], 1) AS [Full Date], DATEPART(yy, a.[Date Entered]) AS [Year Entered], LEFT(DATENAME(MONTH, a.[Date Entered]), 3)
AS [Month Entered], DATEPART(dd, a.[Date Entered]) AS [Day Entered], CASE WHEN DATEPART(WEEKDAY, [DATE entered])
= 1 THEN 'Sun' WHEN DATEPART(WEEKDAY, [DATE entered]) = 2 THEN 'Mon' WHEN DATEPART(WEEKDAY, [DATE entered])
= 3 THEN 'Tus' WHEN DATEPART(WEEKDAY, [DATE entered]) = 4 THEN 'Wed' WHEN DATEPART(WEEKDAY, [DATE entered])
= 5 THEN 'Thu' WHEN DATEPART(WEEKDAY, [DATE entered]) = 6 THEN 'Fri' WHEN DATEPART(WEEKDAY, [DATE entered]) = 7 THEN 'Sat' END AS DOW,
CONVERT(VARCHAR(8), DATEADD(D, - (1 * DATEPART(dw, a.[Date Entered])) + 6, a.[Date Entered]), 1) AS [Week Ending], COUNT(DISTINCT a.[Specimen ID])
AS CountAccns, c.SALESREP AS [Sales Rep], c.NPI AS MLNPI, e.NPIb AS IMSNPI, e.SpecialtyPrimaryCodeb AS [IMS Specialty Primary Code],
sm1.specialtydesc as [Specialty Primary Description],
e.SpecialtySecondaryCodeb AS [IMS Specialty Secondary Code],
sm2.specialtydesc as [Specialty Secondary Description],
e.SpecialtyTertiaryCodeb AS [IMS Specialty Tertiary Code],
sm3.specialtydesc as [Specialty Tertiary Description],
e.ProfessionalID1b AS [IMS Professional ID 1], a.[Requesting Physician] AS Physician, a.[Practice Code], b.[MLIS Practice ID] AS [MLIS Code], a.[Practice Name],
c.DATEESTABLISHED AS [Date Established], c.PRACTICEADDRESS1 AS Address, c.PRACTICEADDRESS2 AS Address2, c.PRACTICECITY AS City,
c.PRACTICESTATE AS State, d.[MLIS Status] AS Status, MAX(a.[Order Count]) AS [order count],
'Rejected' as [Release Status]
FROM dbo.QuickLabDump AS a LEFT OUTER JOIN
dbo.qlmlismapping AS b ON b.[Quicklab ID] = a.[Practice Code] LEFT OUTER JOIN
dbo.PracticeandPhysician AS c ON a.[Requesting Physician] = c.DOCTORFIRSTNAME + ' ' + c.DOCTORLASTNAME AND
a.[Practice Code] = c.PRACTICECODE LEFT OUTER JOIN
dbo.IMSData AS e ON c.NPI = e.NPIb LEFT OUTER JOIN
dbo.QLMLISInfo AS d ON b.[MLIS Practice ID] = d.[MLIS Practice ID]
left outer JOIN
dbo.SpecialtyMapping sm1 ON e.SpecialtyPrimaryCodeb = sm1.specialtyabbrev
left outer JOIN
dbo.SpecialtyMapping sm2 ON e.SpecialtySecondaryCodeb = sm2.specialtyabbrev
left outer JOIN
dbo.SpecialtyMapping sm3 ON e.SpecialtyTertiaryCodeb = sm3.specialtyabbrev
join cte on cte.rejected=a.[Specimen ID]
WHERE (a.[Date Entered] > '20110101')
GROUP BY a.[Date Entered], c.SALESREP, c.NPI, e.NPIb, e.SpecialtyPrimaryCodeb, e.SpecialtySecondaryCodeb, e.SpecialtyTertiaryCodeb, e.ProfessionalID1b,
a.[Requesting Physician], a.[Practice Code], b.[MLIS Practice ID], a.[Practice Name], c.DATEESTABLISHED, c.PRACTICEADDRESS1, c.PRACTICEADDRESS2,
c.PRACTICECITY, c.PRACTICESTATE, d.[MLIS Status],sm1.specialtydesc,sm2.specialtydesc,sm3.specialtydesc
我真的不认为我应该做UNION ALL，而是以更智能的方式构建查询。你能帮我简化查询，这样我就不需要做UNION ALLs了吗？
答案 0 :(得分：3)
怎么样？
select day, count(*) as count, status from TableA
group by day, status
union
select day, count(*) as count, 'pending' from TableB
group by day
如果您在TableA中拥有超过这两个状态，则可以添加where子句来过滤它们。
我认为你不能逃离union。记录在不同的表中，您需要从两个表中获取它们。
它也是union而不是union all。后者将删除重复项，并将需要更多的执行。我们不需要因为我写的查询中没有重复...我不能保证你在问题中粘贴的大量内容：P
答案 1 :(得分：1)
我的想法是在一个简单的CTE中结合你的来源，然后有一个“怪物”查询：
; with cte as ( --union all for source rows without joins and clauses
select * from A
union all
select *,'pending' from B
)
-- "monster" query without the union
select [day]
, [status]
, statusCount=COUNT(*)
from cte
group by [day], [status]
order by [day], [status]