Measures values are twice as big compare to corresponding fact table

The version used is SQL Server 2012
We have a CustomerAddressKey in fact tables with many partitions and Regular Dimension Usage between fact table and CustomerAddress dimension. As we do not process all partitions every day sometimes we encounter the issue when users see the old address for
a particular player.
We have played with 2 approaches: Referenced Dimension Usage and modification of CustomerAddress view when only current (latest) address is available.
After we build and process Referenced Dimension Usage version using CustomerAddress as Reference Dimension and Customer as Intermediate dimension and CustomerAddressKey and CurrentCustomerAddressKey.
Everything is OK besides that the measures value for any slice is twice as big as corresponding fact table value.
What did we do wrong?
serguei russky

There are a couple of ways that data can get duplicated in SSAS
One is incorrectly running multiple ProcessAdd operations. ProcessAdd always appends data. So if you run it twice with the same set of rows you end up with duplicate fact rows in your cube. The way to fix this is to run a ProcessFull on the affected partition/s.
The other common thing that can duplicate data is issues with your data and / or schema. The following is just psuedo code, the join conditions should be the same as you've specified in your DSV. If you run these two styles of SQL queries over your raw data
and they return different amounts then you probably have a problem with the way you have populated the data in your reference dimension table.
SELECT SUM( <measure column> )
FROM <fact table>
SELECT SUM( <measure column> )
FROM <fact table>
INNER JOIN <reference dim>
ON ...
INNER JOIN <CustomerAddress dimesion>
ON ...
http://darren.gosbell.com - please mark correct answers

Similar Messages

File sizes that are twice as big in PS CC after cropping compared to cs5

I'm getting file sizes that are twice as big when I resample an image in PS CC compared to what I was getting in cs5. Anyone know why and what I can do to correct it?

Wrong forum.
What file format are you saving in?
What compression/quality setting are you using if any?
Are you working in 8 or 16 bits?

How to calculate average if some of measure values are 0

Hi
I created a webI report on top of a Bex query, I have a more than 10 different measure in WebI report column on which i need to calculate average.
If I use the Average function provided with in WebI report it is calculating average for the measures which are having 0 values too, but i need to calculate average on a column only for those values which are not 0
Ex: Measure Quantity has values in the report like
100
100
0
100
100
Average function is calculating as 400/5
but I need 400/4, because one of the value is 0, so i need to avoid that count which is zero and should divide it by 4 but not 5.
One way is to custom create measure for all measures and make a if else condition and divided by that custom measure, but is there any other procedure?

i hope this can help
=Average([myMeasure]) Where ([myMesyre] <> 0)
good luck
Amr
before the 0 add <> not Equal too sign < and >
Edited by: Amr Salem on Jan 19, 2011 9:29 AM
Edited by: Amr Salem on Jan 19, 2011 9:30 AM

Measure Values are not shown in the pivot tables

Measures without aggregation (in rpd) are not shown in pivot table.
Im trying to add multiple columns in the fact table as measures,one which is summable had has the SUM function and three which are not summable.
The initial report will be as follows and I have been able to get the data displayed.
Year | Month| Company Name| SUM(C1)| C2|C3
However if I pivot the table as follows data in column C2 and C3 are not Displayed.
Columns
     Year
          Month
          Company
Rows
Measure Lablel
Measures
     SUM(C1)
               C2
               C3
Should we have an aggregation rule applued in C2 and C3 ?

Yes, all the column entries must have agg rule. You can specify one such as 'first' in the pivot table.

AWM adding very less rows in cube as compared to source fact table

AWM is adding just 4 lack rows in the cube. While there are above 10 billion rows in fact table. Even I have applied no aggregation rule at any dimension in cube. And number of rejected records is zero.I am confused what could be the possible reason

Take a look at the SQL query generated by OLAP during cube load.
You can look at it in AWM cube mapping, OR in the CUBE_BUILD_LOG table.
Maybe a SUM..GROUP BY ... is happening during cube load and that is why 1 billion rows are becoming 400,000 rows.
Are the numbers correct when you look at the data at top nodes?

Pivot Table: measure values are not shown

Good afternoon!
I'm trying to show data in pivot table. All the columns are in single logical table. I have achieved next result:
Image: !http://pics.livejournal.com/whitish/pic/0000hgz1!
But measures in cells are'not displayed.
In physical data source (xls file) my data stored as:
Image: !http://pics.livejournal.com/whitish/pic/0000kctt!
Could someone please help me to display my measures in cells?
Thanks in advance,
Alex.

Hi.
Maybe is aggregation rule on Fuzzy Clussification measure set to None (more option/aggregation rule), if yes leave it to default.
Regards
Goran
http://108obiee.blogspot.com

SEM Measure Builder - dataSource values are not appearing in measures

Dear Experts
we recently upgraded our SEM system from 4.0 to 6.0. I am getting few problems in measure builder catelouge entries. when i checked the measures in measure catalouge the datasource tab is blank. Can any one please tell me how do I get the results in datasource tab. Measure values are not appearing in DataSource tab in measure catalouge.
Thanks
Pramod

Yes, all the column entries must have agg rule. You can specify one such as 'first' in the pivot table.

How to convert a measure value to null when the actual measure value is 0

Hello Gurus,
I have year, products, measure 1, measure 2, measure 3. I am added three products into one category by using bin in the analysis i am getting zeros for some of the measure cells when it should be null values, because the actual measure values are null for indivudual product. How can i convert that 0's to null. Please advise.
Thanks,
RC

To answer your immediate question, you can write this:
CASE WHEN table.column = 0 THEN NULL ELSE table.column END
Remember though, that any aggregations done on a record that is NULL will not yield a sum. That is why the usual case is to convert NULLs to 0s. But that may not be a concern for what you are doing. Good luck!

Display -ve Measure Values in Horizontal Bar Chart on right hand side or in bar

In OBIEE I have an issue where the negative measure values in a horizontal bar chart are displayed on the left and overlap with the Y axis values, how do I force these values to be displayed on the right or in the bar itself ? The measure values are set to always on, hoverover cannot be the solution

Hi,
It should not be overlapped. If you are using horizontal bar then 0 will come in center. It should show structure like:
|
|
|
|
|
|-------------------------------------------------
-4 -2 0 2 4
May I know in your report, how it is looking and what is your requirement.
Thanks

Hiding measure values line chart

Hie Guys,
I have a line chart with measure object "X" and dimension "Month". Now, the measure values are going to be zero for all previous months. I do not want those zero values to be displayed in the line chart. Is there a way to do this in Webi? I tried to create a variable to substitute Measure object X.
Variable = If (Measure Object =0) Then "" Else Measure Object. This did not work as the chart only takes measures and the variable is text. Anybody have any suggestions? Thanks.

Jeewan, Consider this as sample data.
Now create three variables....
Month 1 = [M1] Where ([M1]<>0)
Month 2 = [M2] Where ([M2]<>0)
Month 3 = [M3] Where ([M3]<>0)
Now your data will be
Right Click on it. Turn into Line Chart. Finally your output is as you required.
Regards,
G Sujitha

How to hide rows if all values are Zero (0) horizontally?

Hi,
I have got a requirement to hide rows from the results if all values are equal to zero horizontally.
Example:
Region   M1    M2   M3
East     0      0    3
West    10     20    0
South   -5      0    5
North    0      0    0In the above example only North Region row should be hidden from the results becasue all the measures values are zero.
Can someone please let me know the logic to achieve this?
Thank in advance.

Hi Karthik,
Your answer was spot on(you are awarded points) but not my question. My client uses some X erp, which removes all zero value records when he clicks to see results. He wanted the same thing for Answers (he does not want to add any filters). Is there anyway to achive that, like using javascript or something?
Thanks in advance.

How to update measure values

I have road sections modelled with x,y and m values. The m values are evenly spread along the shape. The road topography is at times mountainous, and so these interpolated measure values are not a good representation of the actual measures.
I now have actual driven measure values every 10m and want to update the geometry spatial array with these. I therefore have about 1 million measures to update and have two questions:
1. What is the SQL to update just the measures of an SDO_ORDINATE_ARRAY?
2. What is the best way to update this many records?

Phil,
1. There is no simple SQL statement to update an M value. But it would be a rather simple thing to create a procedure to do so, imputing a geometry, the vertex to modify, and the new value, getting the revised geometry back.
2. If you are going to update multiple vertices's per geometry, it might be best to send the procedure a vertex number / value pair list, and have it update each geometry as a whole.
A good place to start is to look at the 3D to 2D conversion routine:
Re: 3-D conversion to 2-D
Bryan

Why runstate.sequencefile.data.seq.mysequence.mystep.result.numeric doesent contain a measurement value when i use this in a post uut callback?

Hello
I want to write a csv file witch contains all measurement values and limits from my machine. So i try to give the values in post uut callback to a vi. The limits are working and the csv contains all my limit values but the measurement values are all 0. I used runstate.sequencefile.data.seq.mysequence.mystep.result.numeric
I would not like to do it with FileGlobals because they are much values.
Thanks for any advices for this problem.
Solved!
Go to Solution.

There's probably a better way to accomplish what you want. But to make your life easier I'll give you the hacky solution:
Use: Runstate.ProcessModelClient.Data.Seq.MyStep.Result.Numeric
The funny thing is the report should contain all the information you want. I would just create an engine to convert it to your csv look. If you are using 2012 then this seems like a perfect candidate for a plugin.
Hope this helps,
jigg
CTA, CLA
teststandhelp.com
~Will work for kudos and/or BBQ~

Restricting the Prompt Values based on Fact table data

Hi,
We need dashboard prompts in OBIEE reports that will fetch data from dimension tables. Our all dimensions are conformed dimension having joined with multiple fact tables.
Because prompts are showing data from dimension table, it is showing all dimension data and few data might not present in the particular fact table that we have used in the report.
For Example:
We have one Calendar dimension table that has 20 years data that means all dates. Now we have two fact tables Fact1 and Fact2.
In Fact1 we have records related to only 2007 Years data. And In Fact2 We have 2008 years data.
Now We are creating two separate reports in two separate page in one dashboard where the prompt is same. That is Calendar date prompt.
So In report1, The prompt will be Calendare and Result will be from Fact1. But as the calendar prompt from Calendar dimension, it showing all 20 years data though only 2007 years data is valie.
Similarly for Report2, It is showing all calendar dates in the prompt where Fact2 is having 2008 year data.
I want to show only 2007 years dates in the prompt for Report1 and 2008 years dates for Report2. That means I want to restrict prompt values based on the fact.
It will be great if someone can suggest on this.
Regards,
SS

Hi, I have a similar problem.
I have two dimensions: time and customer.
The two dimensions are joined into two Fact Table (Order, Forecast).
I created a dashboard prompt with two selections: the first is the Year (time dimension) and the second is the Region (customer dimension) with the option "cosntraint."
When I select a Year, not all Regions are shown, but only those that are present in one of two fact table for the year chosen. Which fact table?
What criteria will limit the values of the Region?
Thanks

BO Webi report hierarchy with measure values showing more(almost double value) compare to BW Bex report

Hi,
In our BO Webi report hierarchy with measure values showing more(almost double value) compare to BW Bex report. Can any one please help on this.
Is it BW problem or BO problem?
I checked in some other threads but it's not given solution.
Thanks,
Manjunatha

Hi,
Is it BW problem or BO problem? : BO
is it causing problem with hierarchly data only ? without hierarchies data is matching or not?
Post same in SAP BusinessObjects Web Intelligence .
Thanks.

Measures values are twice as big compare to corresponding fact table

Similar Messages

Maybe you are looking for