sql server data warehouse: Measure with "Distinct Count" Aggregate Function and Null value -- sql server data warehouse

You're in the

sql server data warehouse

group:

Measure with "Distinct Count" Aggregate Function and Null value

Paulo Andre Ortega Ribeiro
12/15/2005 5:02:39 AM

sql server data warehouse:

I Have a cube with many measures. One of them use as "Aggregate
Function" the function "Distinct Count". But some records of fact table
has null value in the field of this measure. In SQL Server, when I
execute a "SELECT COUNT (DISTINCT field)" command, it eliminates the
Null value, as showed in the next exemple:

Table X

field A
1122
2345
4567
1122
1122
2345
null
null

When a execute the command:
SELECT COUNT (DISTINCT A) FROM X
I recieve "3" as result.

But when I create a measure over the field A in a cube and I choose the
aggregate funtion "Distinct Count", I recieve "4" as result. I think
that Analysis Services is considering "null" as one of distinct values
of the field A, but I don´t want that, because it makes no sense for me.

Is there a way to eliminate the null value in Analysis Services, to show
"3" as the result in the same way I receive in SQL Server?

Thanks,

Paulo

Re: Measure with "Distinct Count" Aggregate Function and Null value

Dino Hsu
12/15/2005 6:56:17 AM

I am new to SQL server, but experieced in Oracle.
My top of mind test would be:
select count(distinct *) from x => 4
select count(distinct a) from x => 3
It's standard SQL behaviour, I think.
pls verify, tks

Re: Measure with "Distinct Count" Aggregate Function and Null value

Paulo Andre Ortega Ribeiro
12/15/2005 8:03:49 AM

When a execute the command "select count(distinct *) from x" , I recieve
"4" as result and when a execute the command "select count(distinct a)
from x", I recieve "3" as result.

But my problem is not with SQL Server. My problem is in Analysis
Services. I created a measure over the field "a", with the aggregate
function "Distinct Count". But the result in Analysis Services is "4"
and not "3", as I expected.

Do you have a clue?

Thanks,

Paulo

Re: Measure with "Distinct Count" Aggregate Function and Null value

Jéjé
12/15/2005 11:59:55 PM

try to create a new cube (copy/paste the current one)
then create a view on the database like:
select * from mytable where mycolumntodcount IS NOT NULL

use this view in the new cube.
this cube will contain only the dcount measure and the null values are
ignored.

[quoted text, click to view]

Re: Measure with "Distinct Count" Aggregate Function and Null value

Paulo Andre Ortega Ribeiro
12/16/2005 6:28:47 AM

I tried this solution and It works. But I don´t want to create a new
cube. That will be my last option. Is not possible to ignore the null
vules with "distinct count" as the aggregate function? Is not there a
propriety that I can configure?

Thanks,

Paulo

Re: Measure with "Distinct Count" Aggregate Function and Null valu

Erik Veerman
12/16/2005 9:29:13 AM

Even if you didn't have the NULL challenge with Distinct Count, creating a
separate cube for the Distinct Count is a best practice. There are
aggregation and query performance benefits.

The dimensional granularity to the distinct count cube (with the NOT NULL
filter) would be identical to the other cube, so any queries against the
virtual cube would not require any special calculations.

Erik
----
Erik Veerman
erik (at) solidqualitylearning.com

[quoted text, click to view]

Re: Measure with "Distinct Count" Aggregate Function and Null valu

Erik Veerman
12/16/2005 11:28:02 AM

Here another idea:
Add a hidden measure "NULL_Exists" that's a 0 or 1 based on whether the
distinct count column is NULL or not (Null=1). Use the MAX aggregate type
(the result telling you whether there was a null or not across the
dimensionality you are looking at)
Use a calculated measure that = [DistinctCountMeasure] - [NULL_Exists]

So, if there wasn't a NULL, the NULL_Exists returns a 0 and the distinct
count isn't changed. If there was a NULL, then the result is the distinct
count - 1.

----
Erik Veerman
erik (at) solidqualitylearning.com

[quoted text, click to view]

Re: Measure with "Distinct Count" Aggregate Function and Null value

Jéjé
12/16/2005 1:10:34 PM

the only other option I can propose is:
add a dimension in the cube, add a column in the fact table called "ToCount"
which contains a Y/N or 0/1 value linked to this new dimension.
case when MyColumn is null then 'N' else 'Y' end as ToCount

hide the dimension.
Rename the dcount measure to HiddenDcount, hide this measure
create a calculated measure which is:
(measures.HiddenDCount, MyDummyDimension.&[Y])
this ignore the null values.

[quoted text, click to view]

Got something to say? Click here to post a reply to this thread.

Don't see what you're looking for? Try a search.

View other messages in the sql server data warehouse group.

View Other Months
January 2003
March 2003
July 2003
August 2003
September 2003
October 2003
November 2003
December 2003
January 2004
February 2004
March 2004
April 2004
May 2004
June 2004
July 2004
August 2004
September 2004
October 2004
November 2004
December 2004
January 2005
February 2005
March 2005
April 2005
May 2005
June 2005
July 2005
August 2005
September 2005
October 2005
November 2005
December 2005
January 2006
February 2006
March 2006
April 2006
May 2006
June 2006
July 2006
August 2006
September 2006
October 2006
November 2006
December 2006
January 2007
February 2007
March 2007
April 2007
May 2007
June 2007
July 2007
August 2007
September 2007
October 2007
November 2007

home · blog · groups

Questions? Comments? Contact the dn