Reputation: 51
I have a large raw data table. It's long and I'm trying to sort of transpose it. I'm joining two select statements from within it. As such I get multiple columns with the same name. It's a full outer join and I would like to have the two separate columns of the same name as one column. Since it is an outer join I don't want to pick just one tables column for it either like select t1.c1
Thanks!
SELECT *
FROM (SELECT * FROM [LabData].[dbo].[FermHourlyDCSData] where Attribute='Urea') P
full outer JOIN
(SELECT * FROM [LabData].[dbo].[FermHourlyDCSData] where Attribute='Water to Mash Total Water') FPD ON
P.[TimeStamp] = FPD.[TimeStamp]
and P.Site = FPD.Site
and P.Element = FPD.Element
Actual:
Site Attribute Timestamp Value Site Attribute Timestamp Value
AD Urea 1/1/2019 127 Null Null Null Null
Null Null Null Null AD Water 1/1/2019 7.5
Expected/Desired:
Site Attribute Timestamp Value Value
AD Urea 1/1/2019 127 Null
AD Water 1/1/2019 Null 7.5
Upvotes: 0
Views: 3072
Reputation: 1269953
Maybe I'm missing something, but you seem to want a much simpler query:
select Site, Attribute, Timestamp,
(case when Attribute = 'Urea' then Value end) as value_u,
(case when Attribute = 'Water to Mash Total Water' then Value end) as value_2
from [LabData].[dbo].[FermHourlyDCSData]
where Attribute in ('Urea', 'Water to Mash Total Water')
Upvotes: 0
Reputation: 1468
Try this, it's not very pretty, but it does work:
SELECT
[Site] = ISNULL(P.[Site], FPD.[Site]),
[Attribute] = ISNULL(P.[Attribute], FPD.[Attribute]),
[Timestamp] = ISNULL(P.[Timestamp], FPD.[Timestamp]),
[Value] = ISNULL(P.[Value], FPD.[Value]),
[Element] =ISNULL(P.[Element], FPD.[Element])
FROM (SELECT * FROM [dbo].[FermHourlyDCSData] where Attribute='Urea') P
full outer JOIN
(SELECT * FROM [dbo].[FermHourlyDCSData] where Attribute='Water to Mash Total Water') FPD ON
P.[TimeStamp] = FPD.[TimeStamp]
and P.Site = FPD.Site
and P.Element = FPD.Element
Upvotes: 1
Reputation: 83
ISNULL is what you should use for this
ISNULL(p.Site,fpd.Site) as [Site]
Upvotes: 1