What I dislike about PIVOT is that you have to repeat UNIONs several times depending on the metric. This makes it difficult to change and error-prone as you may...

Frank's answer above removes the reference to the CODE_CHARGECREDIT table and removes c.PEOPLE_ORG_ID (which is non-existent) from the GROUP BY.  Idk if that makes sense.  Fixing the (many) errors and...

Horses for courses and everybody thinks their way is the best.  Why not have a sense of humor about it?  I agree Pivot and Unpivot are kludge compared to conditional...

Idk why don't you give it a try and let us know

... since I write SQL now not a...

Ha I think's the other way around.  CASE WHEN is an abomination of over wordiness that never should've been created.  First, back in the day you used to have to...

Imo IIF gets hated on because it was available first in MS Access.  IIF gets interpreted back to "CASE WHEN" so it's really the same (which is a shame because...

It also works without the tally table and just using CROSS APPLY

select
  po.ProductCategory, po.ProductSubCategory, v.Region, v.OrdersPlaced,
  (v.OrdersPlaced*100.0/SUM(v.OrdersPlaced) over (partition by po.ProductCategory)) PercentPlaced
from #po po
 ...

Both work.  The tally based is probably more efficient depending on scale.  It didn't occur to use a tally table...

Ok ok I changed the decode to 'South Africa' and now it does work.  Misdiagnosed due to misleading code?  Yea, if it scans once and there's no dependency then it...

Wait a sec.  North Africa is in the code twice!

The bottom grouping is the code with tally ordered by 1,2,3.  The top is UNION ALL with same ordering

When CROSS JOIN with the tally table the resulting ordering may not align with Region.  It's not known which region will be 1 or 2 or ... unless joined back...

When the code posted above runs it produces the following output:

CountryStatenamesurnameagegender
USAlabamaJohnDon45NULL
CanadaTorontoMarkBagwell35Male

This is exactly what was requested no?  Is it a requirement the query be dynamic because the...

drop table if exists dbo.TestAZURE_JSON;
go
CREATE TABLE dbo.TestAZURE_JSON (
RowID int IDENTITY(1,1) NOT NULL PRIMARY KEY CLUSTERED,
Country varchar(10),
[State] varchar(10),
[JSON_Value] nvarchar(max));
go

INSERT dbo.TestAZURE_JSON(Country, [State], [JSON_Value]) VALUES
('US', 'Alabama', '{"name":"John","surname":"Don","age":45}'),
('Canada', 'Toronto', '{"name":"Mark","surname":"Bagwell","age":35,"gender":"Male"}');

select 
...

• This reply was modified 5 years, 1 month ago by Steve Collins. Reason: made JSON nvarchar(max)