dbTalk Databases Forums  

Problem with Calc Cell and Hidden Measures

microsoft.public.sqlserver.olap microsoft.public.sqlserver.olap


Discuss Problem with Calc Cell and Hidden Measures in the microsoft.public.sqlserver.olap forum.



Reply
 
Thread Tools Display Modes
  #1  
Old   
Dave Fackler
 
Posts: n/a

Default Problem with Calc Cell and Hidden Measures - 08-20-2003 , 08:26 AM






I believe I've uncovered a bug with calculated cells, but
would like comments or suggestions before I open a support
call with Microsoft. Here's the situation.

I have a cube where all of the base measures are hidden
and calculated members in the [Measures] dimension expose
the data to the end user in various ways. I've added a
calc cell to the cube in order to handle some calculations
within a dimension. Since I want the calc cell to resolve
before the calculated members do, I have it's solve order
set lower than all of the calculated members.

The problem is that in this situation, the cells affected
by the calc cell do not show any data. It appears to be
some conflict between the calc cell and the hidden
measures. If I make a measure visible, then data appears
for any calculated member (measure) based on the visible
base measure.

You should be able to recreate this problem by doing the
following:

1. In the [Sales] cube in [Foodmart 2000], set the
Visible property for all of the base measures to false.
Save and process the cube. You should now only have the
two calculated members ([Profit] and [Sales Average])
showing in the [Measures] dimension.

2. Set the solve order for both of the calculated members
to something higher than 0. Save the cube.

3. Add a calc cell to the cube. For the calculation
subcube definition, use the descendants of the [Time].
[1998] member (i.e., DESCENDANTS([Time].&[1998])). For
the calculation value, use the following formula to make
[1998] values 10% higher than their corresponding [1997]
values: ParallelPeriod([Year]) * 1.1. Name the calc cell
soemthing like "Forecast 1998". Save the cube. The
default solve order for the calc cell will be 0 -- which
should now be lower than the calculated members.

Now browse the cube and put the [Time] dimension on rows
and the [Measures] dimension on columns. You should see
data for 1997, but not for 1998. If you make the base
measures visible again, the forecasted numbers for 1998
will appear.

Can anyone out there try this and confirm that it happens
the same for them? Want to make sure I'm not going
crazy. Also, if anyone has any ideas on what the issue
might be or work-arounds (or an explanation that what I'm
trying to do is just bad or wrong), please chime in!!

Thanks (and sorry for the long post)...
Dave Fackler


Reply With Quote
  #2  
Old   
Bill Cheng [MSFT]
 
Posts: n/a

Default RE: Problem with Calc Cell and Hidden Measures - 08-21-2003 , 03:43 AM






Hi Dave,

I have reproduced the issue using the steps you provided. I suggest that
you open an incident with Microsoft Product Support Services via
http://support.microsoft.com/default...EN-US;CNTACTMS, so that
we can have dedicated engineer to assist you. If you would like our help in
creating the incident, please also let us know.


Bill Cheng
Microsoft Online Partner Support

Get Secure! - www.microsoft.com/security
This posting is provided "as is" with no warranties and confers no rights.
--------------------
Quote:
Content-Class: urn:content-classes:message
From: "Dave Fackler" <davef (AT) intellinet (DOT) com
Sender: "Dave Fackler" <davef (AT) intellinet (DOT) com
Subject: Problem with Calc Cell and Hidden Measures
Date: Wed, 20 Aug 2003 06:26:05 -0700
Lines: 56
Message-ID: <0ceb01c3671e$9bc25570$a601280a (AT) phx (DOT) gbl
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Newsreader: Microsoft CDO for Windows 2000
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4910.0300
Thread-Index: AcNnHpvCGxC1R1uRTd+3GDaeJQT8kA==
Newsgroups: microsoft.public.sqlserver.olap
Path: cpmsftngxa06.phx.gbl
Xref: cpmsftngxa06.phx.gbl microsoft.public.sqlserver.olap:41931
NNTP-Posting-Host: TK2MSFTNGXA14 10.40.1.166
X-Tomcat-NG: microsoft.public.sqlserver.olap

I believe I've uncovered a bug with calculated cells, but
would like comments or suggestions before I open a support
call with Microsoft. Here's the situation.

I have a cube where all of the base measures are hidden
and calculated members in the [Measures] dimension expose
the data to the end user in various ways. I've added a
calc cell to the cube in order to handle some calculations
within a dimension. Since I want the calc cell to resolve
before the calculated members do, I have it's solve order
set lower than all of the calculated members.

The problem is that in this situation, the cells affected
by the calc cell do not show any data. It appears to be
some conflict between the calc cell and the hidden
measures. If I make a measure visible, then data appears
for any calculated member (measure) based on the visible
base measure.

You should be able to recreate this problem by doing the
following:

1. In the [Sales] cube in [Foodmart 2000], set the
Visible property for all of the base measures to false.
Save and process the cube. You should now only have the
two calculated members ([Profit] and [Sales Average])
showing in the [Measures] dimension.

2. Set the solve order for both of the calculated members
to something higher than 0. Save the cube.

3. Add a calc cell to the cube. For the calculation
subcube definition, use the descendants of the [Time].
[1998] member (i.e., DESCENDANTS([Time].&[1998])). For
the calculation value, use the following formula to make
[1998] values 10% higher than their corresponding [1997]
values: ParallelPeriod([Year]) * 1.1. Name the calc cell
soemthing like "Forecast 1998". Save the cube. The
default solve order for the calc cell will be 0 -- which
should now be lower than the calculated members.

Now browse the cube and put the [Time] dimension on rows
and the [Measures] dimension on columns. You should see
data for 1997, but not for 1998. If you make the base
measures visible again, the forecasted numbers for 1998
will appear.

Can anyone out there try this and confirm that it happens
the same for them? Want to make sure I'm not going
crazy. Also, if anyone has any ideas on what the issue
might be or work-arounds (or an explanation that what I'm
trying to do is just bad or wrong), please chime in!!

Thanks (and sorry for the long post)...
Dave Fackler




Reply With Quote
Reply




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off



Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.