how to remove the duplicate rows when we used group by method? - sql-server

Question

i used the group by the method for distinct the two column values. But still i got the duplicate values in the column rows.
SELECT MRD_NO,RESOURCE_NAME,
Diagnosis =
STUFF((SELECT DISTINCT ', ' +
(case when Diagnosis is null and OTHER_DIAGONSIS is not null then OTHER_DIAGONSIS else Diagnosis end)
FROM EMR_master b
WHERE b.MRD_NO = a.MRD_NO
FOR XML PATH('')), 1, 2, '')
FROM EMR_master a
where
a.TREATMENT_CODE in ('CC','PO','SRE','REG')
group by MRD_NO,RESOURCE_NAME
order by RESOURCE_NAME,MRD_NO
my output contains the duplicate values in the MRD_NO column, tell me how to remove the duplicate values.
my output
MRD_NO | RESOURCE_NAME | Diagnosis
123 | james | retina
126 | peter | throat pain
129 | Murugan | fever
129 | william | fever
130 | william | retina
i need like this output
MRD_NO | RESOURCE_NAME | Diagnosis
123 | james | retina
126 | peter | throat pain
129 | Murugan | fever
130 | william | retina
note: i got duplicates MRD_NO 129 with two resource name(Murugan,william),
so i need to eliminate the william and get unique MRD_NO

Answer 1

If you don't want to much work around and getting desired output you can do with temptable
SELECT MRD_NO,RESOURCE_NAME,
STUFF((SELECT DISTINCT ', ' +
(case when Diagnosis is null and OTHER_DIAGONSIS is not null then OTHER_DIAGONSIS else Diagnosis end) as 'Diagnosis'
into #tmpEmrDetail
FROM EMR_master b
WHERE b.MRD_NO = a.MRD_NO
FOR XML PATH('')), 1, 2, '')
FROM EMR_master a
where
a.TREATMENT_CODE in ('CC','PO','SRE','REG')
group by MRD_NO,RESOURCE_NAME
order by RESOURCE_NAME,MRD_NO
select distinct * from #tmpEmrDetail
drop table #tmpEmrDetail
or
; WITH ctetbl AS
(
SELECT MRD_NO,RESOURCE_NAME,
Diagnosis =
STUFF((SELECT DISTINCT ', ' +
(case when Diagnosis is null and OTHER_DIAGONSIS is not null then OTHER_DIAGONSIS else Diagnosis end)
FROM EMR_master b
WHERE b.MRD_NO = a.MRD_NO
FOR XML PATH('')), 1, 2, '')
FROM EMR_master a
where
a.TREATMENT_CODE in ('CC','PO','SRE','REG')
group by MRD_NO,RESOURCE_NAME
)
SELECT *
FROM ctetbl order by RESOURCE_NAME,MRD_NO

Answer 2

Try to choose a column in DISTINCT operator:
SELECT MRD_NO,RESOURCE_NAME,
Diagnosis =
STUFF((SELECT DISTINCT b.MRD_NO + ', ' +
(case
when Diagnosis is null and OTHER_DIAGONSIS is not null
then OTHER_DIAGONSIS else Diagnosis end)
FROM EMR_master b
WHERE b.MRD_NO = a.MRD_NO
GROUP BY b.MRD_NO, b.RESOURCE_NAME
FOR XML PATH('')), 1, 2, '')
FROM EMR_master a
where
a.TREATMENT_CODE in ('CC','PO','SRE','REG')
group by MRD_NO,RESOURCE_NAME
order by RESOURCE_NAME,MRD_NO
E.g.:
DECLARE #table TABLE
(
EmpID int,
EmpName varchar(50),
DateOfJoin datetime,
DateOfLeaving DATETIME,
StatusName VARCHAR(50)
)
INSERT INTO #table
(
EmpID,
EmpName,
DateOfJoin,
DateOfLeaving,
StatusName
)
VALUES
(1, 'XYZ', '2015-10-01', '2017-09-26', 'De-ACTIVE')
,(2, 'ABC', '2018-01-01', NULL, 'ACTIVE')
,(3, 'XYZ', '2018-10-15', NULL, 'ACTIVE')
and query:
SELECT
hey = STUFF((
SELECT
DISTINCT t.EmpName + ', '
FROM #table t
FOR XML PATH('')), 1, 2, '')
FROM #table t
OUTPUT:
hey
C, XYZ,
C, XYZ,
C, XYZ,
UPDATE:
If you would like just to remove duplicates, then you should just use GROUP BY to field where you want to get rid of duplicates.
SELECT
MRD_NO
, RESOURCE_NAME
, Diagnosis
FROM YourTable
GROUP BY MRD_NO
, RESOURCE_NAME
, Diagnosis