Using the following sample code I'm trying to remove rows that have not had columns changed based on the scalar #CompareFields
create table #ResultSet_fields(
claimId int,
adjustmentVersion int,
ServiceDateFrom date,
ServiceDateTo date,
ProcedureCode varchar(10),
PlaceOfService varchar(3)
)
declare #CompareFields varchar(max)
select #CompareFields = 'ProcedureCode,PlaceOfService'
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1010,1,'5/5/2015','5/5/2015',92213,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1010,2,'5/5/2015','5/5/2015',92213,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1010,3,'5/5/2015','5/5/2015',92214,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1011,1,'5/5/2015','5/5/2015',5555,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1012,1,'5/7/2015','5/7/2015',66666,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1012,2,'5/7/2015','5/7/2015',66666,13
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1013,1,'5/7/2015','5/7/2015',99999,11
insert into #ResultSet_fields (ClaimId,adjustmentVersion,ServiceDateFrom,ServiceDateTo,ProcedureCode,PlaceOfService) select 1014,1,'5/9/2015','5/9/2015',99999,11
I only want rows that have had a column change of 'ProcedureCode' and 'PlaceOfService' between adjustment versions.
ex. since the row with ClaimId '1010' had a 'ProcedureCode' change and row with claimId '1012' had a 'PlaceOfService' change I would only like those rows to remain in the Result Set.
I have thought about using the EXCEPT clause or UPDATE() trigger but I'm having trouble forming the syntax.
Can someone point me to the right direction on how to accomplish this?
You can use LAG to get the previous row's value to compare to.
SELECT
ClaimId,
adjustmentVersion,
ServiceDateFrom,
ServiceDateTo,
ProcedureCode,
PlaceOfService
FROM (
SELECT *,
ServiceDateFrom_prev = LAG(rs.ServiceDateFrom) OVER (PARTITION BY rs.ClaimId ORDER BY rs.adjustmentVersion),
ServiceDateTo_prev = LAG(rs.ServiceDateTo ) OVER (PARTITION BY rs.ClaimId ORDER BY rs.adjustmentVersion),
ProcedureCode_prev = LAG(rs.ProcedureCode ) OVER (PARTITION BY rs.ClaimId ORDER BY rs.adjustmentVersion),
PlaceOfService_prev = LAG(rs.PlaceOfService ) OVER (PARTITION BY rs.ClaimId ORDER BY rs.adjustmentVersion)
FROM #ResultSet_fields rs
) rs
WHERE (
#CompareFields LIKE '%ServiceDateFrom%' AND ServiceDateFrom <> ServiceDateFrom_prev
OR #CompareFields LIKE '%ServiceDateTo%' AND ServiceDateTo <> ServiceDateTo_prev
OR #CompareFields LIKE '%ProcedureCode%' AND ProcedureCode <> ProcedureCode_prev
OR #CompareFields LIKE '%PlaceOfService%' AND ProcedureCode <> PlaceOfService_prev
);
If you want to use indexes, or you have a lot of columns ot compare, you can use dynamic SQL
DECLARE #lagCols nvarchar(max), #whereFilters nvarchar(max);
SELECT
#lagCols = STRING_AGG(CAST(
' ' + QUOTENAME(c.name + '_chg') + ' = LAG(rs.' + QUOTENAME(c.name) + ') OVER (PARTITION BY rs.ClaimId ORDER BY rs.adjustmentVersion)
' AS nvarchar(max)), ',')
,#whereFilters = STRING_AGG(CAST(
QUOTENAME(c.name + '_chg') + ' <> ' + QUOTENAME(c.name)
AS nvarchar(max)), ' OR
')
FROM STRING_SPLIT(#CompareFields, ',') s
JOIN tempdb.sys.columns c ON c.name = TRIM(s.value) -- make sure to get the right database
WHERE c.object_id = OBJECT_ID('tempdb..#ResultSet_fields');
DECLARE #sql nvarchar(max) = '
SELECT
ClaimId,
adjustmentVersion,
ServiceDateFrom,
ServiceDateTo,
ProcedureCode,
PlaceOfService
FROM (
SELECT *,
' + #lagCols + '
FROM #ResultSet_fields rs
) rs
WHERE (
' + #whereFilters + '
);
';
PRINT #sql; -- for testing
EXEC sp_executesql #sql;
Related
I'm trying to convert my result table which consist of multiple rows into multiple columns. below are my sample result before and after:
Before:
After:
Got a some idea from here but still no luck.
Update 1:
select #cols = STUFF((select ',' + QUOTENAME(institution) + ',' + QUOTENAME(intstatus)
From #temp
group by refno,frmstatus
FOR XML PATH(''), TYPE).value('.', 'NVARCHAR(MAX)'),1,1,'')
i have no idea how to insert each column into the pivot query
set #query = 'select refno, frmstatus,' + #cols + '
from (
select refno, frmstatus, institution, intstatus from #temp
) x
pivot
(
???????
)
Please try the below query:
CREATE TABLE #temp(refno nvarchar(20), firmstatus nvarchar(20), institution nvarchar(20), intstatus nvarchar(20), ranking int)
DECLARE #qu NVARCHAR(MAX), #pcol NVARCHAR(MAX)
INSERT INTO #temp
SELECT
*,
ROW_NUMBER() OVER (PARTITION BY refno ORDER BY institution ASC) AS ranking
FROM temp
SELECT #pcol=
STUFF((
SELECT
DISTINCT N', Institution'+ CAST (ranking AS NVARCHAR(25)) +', '+ N'Intstatus'+ CAST (ranking AS NVARCHAR(25))
FROM #temp
FOR XML PATH('')),1,1,'')
SET #qu=N'SELECT refno, firmstatus,'+ #pcol +
N' FROM
(
select refno,firmstatus,ColData,colheader+ CAST(ranking as varchar) as colnames from
(select * from #temp)s
UNPIVOT
(ColData for colheader in ([institution], [intstatus])) up
)S
PIVOT
(MAX(ColData) FOR colnames IN ('+#pcol +N')) AS piv'
EXEC sp_executesql #qu -- execute the dynamic sql
DROP TABLE #temp -- remove the temp table
The temp table in above script was created like below
--create table temp( refno nvarchar(20), firmstatus nvarchar(20), institution nvarchar(20), intstatus nvarchar(20))
--insert into temp values
--('AAA/1','Active','InstA','Ongoing'),
--('AAA/1','Active','InstB','Ongoing'),
--('AAA/1','Active','InstC','Ongoing'),
--('AAA/2','Active','InstA','Ongoing'),
--('AAA/2','Active','InstB','Ongoing')
Result received:
If repeated columns number is small and restricted, you can use simple solution:
WITH A AS (
SELECT *,
ROW_NUMBER() OVER(PARTITION BY refno, fmstatus ORDER BY institution) n
FROM myTable
),
B AS (
SELECT refno, fmstatus,
CASE WHEN n=1 THEN institution END institution,
CASE WHEN n=1 THEN intstatus END intstatus,
CASE WHEN n=2 THEN institution END institution1,
CASE WHEN n=2 THEN intstatus END intstatus1,
CASE WHEN n=3 THEN institution END institution2,
CASE WHEN n=3 THEN intstatus END intstatus2
FROM A
)
SELECT refno, fmstatus,
MAX(institution) institution,
MAX(intstatus) intstatus,
MAX(institution1) institution1,
MAX(intstatus1) intstatus1,
MAX(institution2) institution2,
MAX(intstatus2) intstatus2
FROM B
GROUP BY refno, fmstatus
Otherwise use PIVOT
I have the following situation (heavily abstracted, please ignore bad design):
CREATE TABLE dbo.PersonTest (Id INT, name VARCHAR(255))
INSERT INTO dbo.PersonTest
(Id, name )
VALUES (1, 'Pete')
, (1, 'Marie')
, (2, 'Sam')
, (2, 'Daisy')
I am looking for the following result:
Id Name1 Name2
1 Marie Pete
2 Daisy Sam
So, for each Id, the rows should be merged.
Getting this result I used the following query:
WITH PersonRN AS
(
SELECT *
, ROW_NUMBER() OVER(PARTITION BY Id ORDER BY name) RN
FROM dbo.PersonTest
)
SELECT PT1.Id
, PT1.name Name1
, PT2.name Name2
FROM PersonRN AS PT1
LEFT JOIN PersonRN AS PT2 -- Left join in case there's only 1 name
ON PT2.Id = PT1.Id
AND PT2.RN = 2
WHERE PT1.RN = 1
Which works perfectly fine.
My question is: Is this the best way (best in terms of performance and resilience)? If, for example, one of these Id's has a third name, this third name is ignored by my query. I'm thinking the best way to deal with that would be dynamic SQL, which would be fine, but if it can be done without dynamic, I would prefer that.
Aside from dynamic PIVOT, you can do this using Dynamic Crosstab, which I prefer for readability.
SQL Fiddle
DECLARE #sql1 VARCHAR(1000) = '',
#sql2 VARCHAR(1000) = '',
#sql3 VARCHAR(1000) = ''
DECLARE #max INT
SELECT TOP 1 #max = COUNT(*) FROM PersonTest GROUP BY ID ORDER BY COUNT(*) DESC
SELECT #sql1 =
'SELECT
ID' + CHAR(10)
SELECT #sql2 = #sql2 +
' , MAX(CASE WHEN RN =' + CONVERT(VARCHAR(5), RN)
+ ' THEN name END) AS ' + QUOTENAME('Name' + CONVERT(VARCHAR(5), RN)) + CHAR(10)
FROM(
SELECT TOP(#max)
ROW_NUMBER() OVER(ORDER BY (SELECT NULL)) AS RN
FROM sys.columns
)t
ORDER BY RN
SELECT #sql3 =
'FROM(
SELECT *,
RN = ROW_NUMBER() OVER(PARTITION BY ID ORDER BY name)
FROM PersonTest
)t
GROUP BY ID
ORDER BY ID'
PRINT (#sql1 + #sql2 + #sql3)
EXEC (#sql1 + #sql2 + #sql3)
Preamble: I've read through the three questions/answers here,here, and here, with big ups to #cade-roux. This all stemmed from trying to use the following data in a 2005 SSRS matrix that, I believe, doesn't work because I want to show a member having to take a test multiple times, and SSRS seems to require the aggregate where I want to show all dates.
I get the following results in my table, which seems to be showing all the data correctly:
How do I change the code below to show a) the "tests" at the top of each column with b) if it's called for, the multiple dates that test was taken?
Here's the code I have to produce the table, above. Much of it is commented out as I was just trying to get the pivot to work, but you may notice I am also trying to specify which test column comes first.
CREATE TABLE #tmp ( ---THIS WORKS BUT TESTS ARE VERTICAL
[TEST] [varchar](30) NOT NULL,
[ED] [datetime] NOT NULL
)
--WHERE THE TEST AND ED COME FROM
INSERT #TMP
SELECT DISTINCT
-- N.FULL_NAME
-- , CONVERT(VARCHAR(30), AM.CREATEDATE, 101) AS ACCOUNT_CLAIMED
-- , N.EMAIL
-- , NULL AS 'BAD EMAIL'
-- , CONVERT(VARCHAR(30), AC.EFFECTIVE_DATE, 101) AS EFFECTIVE_DATE
AC.PRODUCT_CODE AS TEST
, CONVERT(VARCHAR(30), AC.EFFECTIVE_DATE, 101) AS ED
-- , CASE
-- WHEN AC.PRODUCT_CODE = 'NewMem_Test' THEN '9'
-- WHEN AC.PRODUCT_CODE = 'NM_Course1' THEN '1'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course1' THEN '2'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course2' THEN '3'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course3' THEN '4'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course4' THEN '5'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course5' THEN '6'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course6' THEN '7'
-- WHEN AC.PRODUCT_CODE = 'NMEP_Course7' THEN '8'
-- END AS 'COLUMN_ORDER'
FROM NAME N
JOIN USERMAIN UM
ON N.ID = UM.CONTACTMASTER
JOIN formTransLog TL
ON UM.USERID = TL.USERNAME
JOIN anet_Users AU
ON UM.USERID = AU.USERNAME
JOIN anet_Membership AM
ON AU.USERID = AM.USERID
JOIN ACTIVITY AC
ON N.ID = AC.ID
AND AC.ACTIVITY_TYPE = 'COURSE'
AND AC.PRODUCT_CODE LIKE 'N%'
--ORDER BY 1, 7
DECLARE #sql AS varchar(max)
DECLARE #pivot_list AS varchar(max) -- Leave NULL for COALESCE technique
DECLARE #select_list AS varchar(max) -- Leave NULL for COALESCE technique
SELECT #pivot_list = COALESCE(#pivot_list + ', ', '') + '[' + CONVERT(varchar, PIVOT_CODE) + ']'
,#select_list = COALESCE(#select_list + ', ', '') + '[' + CONVERT(varchar, PIVOT_CODE) + '] AS [col_' + CONVERT(varchar, PIVOT_CODE) + ']'
FROM (
SELECT DISTINCT PIVOT_CODE
FROM (
SELECT TEST, ED, ROW_NUMBER() OVER (PARTITION BY TEST ORDER BY ED) AS PIVOT_CODE
FROM #tmp
) AS rows
) AS PIVOT_CODES
SET #sql = '
;WITH p AS (
SELECT TEST, ED, ROW_NUMBER() OVER (PARTITION BY TEST ORDER BY ED) AS PIVOT_CODE
FROM #tmp
)
SELECT TEST, ' + #select_list + '
FROM p
PIVOT (
MIN(ED)
FOR PIVOT_CODE IN (
' + #pivot_list + '
)
) AS pvt
'
PRINT #sql
EXEC (#sql)
EDIT:
The goal is to have the report in SSRS look like this:
I was able to produce the results you were looking for by adding in a number (RowNum) to the query underneath the PIVOT operator. It doesn't have to be in the final query (though you might want it for client-side sorting), but by having it in the underlying layer the PIVOT operation treats that number like a member of a GROUP BY clause.
Please look through my sample SQL below and let me know if this matches your criteria.
CREATE TABLE #TMP
(
Name VARCHAR(10),
Test VARCHAR(20),
EffectiveDate DATETIME
)
INSERT INTO #TMP (Name, Test, EffectiveDate)
SELECT 'Jane', 'NM_Course1', '01/17/2014' UNION
SELECT 'Jane', 'NMEP_Course1', '12/19/2013' UNION
SELECT 'Jane', 'NMEP_Course1', '12/20/2013' UNION
SELECT 'Jane', 'NMEP_Course2', '12/19/2013' UNION
SELECT 'Jane', 'NMEP_Course2', '12/22/2013' UNION
SELECT 'Jane', 'NMEP_Course2', '01/05/2014' UNION
SELECT 'John', 'NM_Course1', '01/17/2014' UNION
SELECT 'John', 'NMEP_Course1', '01/11/2014'
DECLARE #sql AS varchar(max)
DECLARE #pivot_list AS varchar(max) -- Leave NULL for COALESCE technique
DECLARE #select_list AS varchar(max) -- Leave NULL for COALESCE technique
SELECT #pivot_list = COALESCE(#pivot_list + ', ', '') + '[' + CONVERT(varchar, PIVOT_CODE) + ']'
,#select_list = COALESCE(#select_list + ', ', '') + '[' + CONVERT(varchar, PIVOT_CODE) + '] AS [col_' + CONVERT(varchar, PIVOT_CODE) + ']'
FROM (
SELECT DISTINCT PIVOT_CODE
FROM (
SELECT TEST AS PIVOT_CODE
FROM #tmp
) AS rows
) AS PIVOT_CODES
SET #sql = '
SELECT Name, ' + #select_list + '
FROM
(
SELECT b.Name, RowNum, b.EffectiveDate, b.TEST AS PIVOT_CODE
FROM
(
SELECT Name, Test, EffectiveDate, ROW_NUMBER() OVER (PARTITION BY NAME, TEST ORDER BY EffectiveDate) RowNum
FROM #Tmp
) b
) p
PIVOT (
MIN(EffectiveDate)
FOR PIVOT_CODE IN (
' + #pivot_list + '
)
) AS pvt
ORDER BY Name, RowNum
'
PRINT #sql
EXEC (#sql)
DROP TABLE #TMP
I'm storing some queries in a table column so I can execute them later passing some parameters.
But it has been really annoying to format the query into an Update sentence, because of the special characters.
For Example:
SELECT * FROM MOUNTAINS WHERE MON_NAME='PALMA' AND MON_DESC LIKE '%TRANVULCANIA%'
Then I need the string just for the udpate query:
UPDATE QUERIES
SET QUE_SEL='SELECT * FROM MOUNTAINS WHERE MON_NAME='''+'PALMA'+''' AND MON_DESC LIKE '''+'%TRANVULCANIA%'+''' '
WHERE QUE_ID=1
as you can see the first ' must be replaced for '''+' but the next door ' must be replaced by '+'''
This is the query I'm working on:
DECLARE #QUERY VARCHAR(MAX)
SELECT #QUERY='SELECT * FROM QUERIES WHERE QUE_NOMBRE='''+'PRUEBA 1'+''' '
SELECT
t.r.value('.', 'varchar(255)') AS token
, ROW_NUMBER() OVER (ORDER BY (SELECT 1)) AS id
FROM (
SELECT myxml = CAST('<t>' + REPLACE(#QUERY, '''', '</t><t>''</t><t>') + '</t>' AS XML)
) p
CROSS APPLY myxml.nodes('/t') t(r)
this is the result:
token id
-------------------------------------------------- --------------------
SELECT * FROM QUERIES WHERE QUE_NOMBRE= 1
' 2
PRUEBA 1 3
' 4
5
Now I want a column that tell me when to open and when to close and then I can set the final replace.
Adapting the solution given by #rivarolle
DECLARE #QUERY VARCHAR(MAX)
DECLARE #FORMATTED varchar(max)
SELECT #QUERY='SELECT * FROM QUERIES WHERE QUE_NOMBRE='''+'PRUEBA 1'+''''
;WITH TOKENS AS(
SELECT
t.r.value('.', 'varchar(MAX)') AS token
, ROW_NUMBER() OVER (ORDER BY (SELECT 1)) AS Id
FROM (
SELECT myxml = CAST('<t>' + REPLACE(#QUERY, '''', '</t><t>''</t><t>') + '</t>' AS XML)
) p
CROSS APPLY myxml.nodes('/t') t(r)
)
,
Tokens2 as (
SELECT
TOKENS.token as token
,quotes.row%2 as tipoapostrofe
from Tokens
left join (select row_number() over( order by Id asc) as row, a.* FROM (SELECT * from Tokens) a where Token = '''') quotes
on quotes.Id = Tokens.Id
)
SELECT #FORMATTED = STUFF((
SELECT ' ' + REPLACE(token,'''',CASE tipoapostrofe WHEN 1 THEN '''''''+''' WHEN 0 THEN '''+''''''' ELSE '' END) AS [text()]
FROM Tokens2
FOR XML PATH('')
), 1, 1, '')
print #FORMATTED
This Works, just need a function for cleaning XML special characters and another for putting back, and the Dynamic queries are printed ready for an update.
I think its not necessary to replace an apostrophe with '''+' to open and '+''' to close, I made some probes and you can exec a query that you replace opening and closing apostrophes with the same.. for example '''+' for open and '''+' for close.
So the query would be:
DECLARE #QUERY VARCHAR(MAX)
DECLARE #FORMATTED varchar(max)
SELECT #QUERY='SELECT * FROM QUERIES WHERE QUE_NOMBRE='''+'PRUEBA 1'+''''
SELECT #FORMATTED= STUFF((
SELECT ' ' +
(SELECT
CASE
WHEN t.r.value('.', 'varchar(250)')='''' THEN REPLACE(t.r.value('.', 'varchar(250)'), '''','''''''+''')
ELSE t.r.value('.', 'varchar(250)')
END
) AS [text()]
-- , ROW_NUMBER() OVER (ORDER BY (SELECT 1)) AS id
FROM (
SELECT myxml = CAST('<t>' + REPLACE(#QUERY, '''', '</t><t>''</t><t>') + '</t>' AS XML)
) p
CROSS APPLY myxml.nodes('/t') t(r)
FOR XML PATH('')
), 1, 1, '')
SET #FORMATTED=REPLACE(#FORMATTED,' ','')
PRINT #FORMATTED
then I get:
SELECT * FROM QUERIES WHERE QUE_NOMBRE= '''+' PRUEBA 1 '''+'
then I copy into a variable and execute
DECLARE #VAR VARCHAR(500)
SET #VAR='SELECT * FROM QUERIES WHERE QUE_NOMBRE='''+'PRUEBA 1'''+' '
EXEC(#VAR)
It Works for very simple queries, but with longer and complicated queries it doesn't works..
Assuming your token table is Tokens(Token, Id, Position):
update Tokens
set position = quotes.row%2
from Tokens
left join (select row_number() over( order by Id asc) as row, a.* FROM (SELECT * from Tokens) a where Token = '''') quotes
on quotes.Id = Tokens.Id
The position column will have a value of 1 for starting quote and 0 for closing quote. NULL for the rest.
I need a query that will return a table where each column is the count of distinct values in the columns of another table.
I know how to count the distinct values in one column:
select count(distinct columnA) from table1;
I suppose that I could just make this a really long select clause:
select count(distinct columnA), count(distinct columnB), ... from table1;
but that isn't very elegant and it's hardcoded. I'd prefer something more flexible.
This code should give you all the columns in 'table1' with the respective distinct value count for each one as data.
DECLARE #TableName VarChar (Max) = 'table1'
DECLARE #SqlString VarChar (Max)
set #SqlString = (
SELECT DISTINCT
'SELECT ' +
RIGHT (ColumnList, LEN (ColumnList)-1) +
' FROM ' + Table_Name
FROM INFORMATION_SCHEMA.COLUMNS COL1
CROSS AppLy (
SELECT ', COUNT (DISTINCT [' + COLUMN_NAME + ']) AS ' + '''' + COLUMN_NAME + ''''
FROM INFORMATION_SCHEMA.COLUMNS COL2
WHERE COL1.TABLE_NAME = COL2.TABLE_NAME
FOR XML PATH ('')
) TableColumns (ColumnList)
WHERE
1=1 AND
COL1.TABLE_NAME = #TableName
)
EXECUTE (#SqlString)
try this (sql server 2005 syntax):
DECLARE #YourTable table (col1 varchar(5)
,col2 int
,col3 datetime
,col4 char(3)
)
insert into #YourTable values ('abcdf',123,'1/1/2009','aaa')
insert into #YourTable values ('aaaaa',456,'1/2/2009','bbb')
insert into #YourTable values ('bbbbb',789,'1/3/2009','aaa')
insert into #YourTable values ('ccccc',789,'1/4/2009','bbb')
insert into #YourTable values ('aaaaa',789,'1/5/2009','aaa')
insert into #YourTable values ('abcdf',789,'1/6/2009','aaa')
;with RankedYourTable AS
(
SELECT
ROW_NUMBER() OVER(PARTITION by col1 order by col1) AS col1Rank
,ROW_NUMBER() OVER(PARTITION by col2 order by col2) AS col2Rank
,ROW_NUMBER() OVER(PARTITION by col3 order by col3) AS col3Rank
,ROW_NUMBER() OVER(PARTITION by col4 order by col4) AS col4Rank
FROM #YourTable
)
SELECT
SUM(CASE WHEN col1Rank=1 THEN 1 ELSE 0 END) AS col1DistinctCount
,SUM(CASE WHEN col2Rank=1 THEN 1 ELSE 0 END) AS col2DistinctCount
,SUM(CASE WHEN col3Rank=1 THEN 1 ELSE 0 END) AS col3DistinctCount
,SUM(CASE WHEN col4Rank=1 THEN 1 ELSE 0 END) AS col4DistinctCount
FROM RankedYourTable
OUTPUT:
col1DistinctCount col2DistinctCount col3DistinctCount col4DistinctCount
----------------- ----------------- ----------------- -----------------
4 3 6 2
(1 row(s) affected)
and it's hardcoded.
It is not hardcoding to provide a field list for a sql statement. It's common and acceptable practice.
This won't necessarily be possible for every field in a table. For example, you can't do a DISTINCT against a SQL Server ntext or image field unless you cast them to other data types and lose some precision.
I appreciate all of the responses. I think the solution that will work best for me in this situation (counting the number of distinct values in each column of a table from an external program that has no knowledge of the table except its name) is as follows:
Run "describe table1" and pull out the column names from the result.
Loop through the column names and create the query to count the distinct values in each column. The query will look something like "select count(distinct columnA), count(distinct columnB), ... from table1".
Raj More's answer works well if you don't need to consider null as a value as count(distinct...) does not count null.
Here is a modification to count values including null by converting values to a string and replacing null with "NULL AS SOME IMPOSSIBLE STRING":
DECLARE #TableName VarChar (1024) = 'tableName'
DECLARE #SqlString VarChar (Max)
set #SqlString = (
SELECT DISTINCT
'SELECT ' +
RIGHT (ColumnList, LEN (ColumnList)-1) +
' FROM ' + Table_Name
FROM INFORMATION_SCHEMA.COLUMNS COL1
CROSS AppLy (
SELECT ', COUNT (DISTINCT coalesce(cast([' + COLUMN_NAME + '] as varchar),
''NULL AS SOME IMPOSSIBLE STRING'')) AS ' + '''' + COLUMN_NAME + ''''
FROM INFORMATION_SCHEMA.COLUMNS COL2
WHERE COL1.TABLE_NAME = COL2.TABLE_NAME
FOR XML PATH ('')
) TableColumns (ColumnList)
WHERE
COL1.TABLE_NAME = #TableName
)
EXECUTE (#SqlString)
DISTINCT is evil. Do COUNT/GROUP BY