Dimension values without data in a fact table

I have an ODS system and a Data warehouse system
I have a Sales fact table in the ODS system and I have these fields:
SALES
ID_CUSTOMER (PK),
ID_MODEL (PK),
ID_TIME (PK),
SALES,
QUANT_ART,
COST
Then in some records in the fields ID_Time or ID_Model or ID_Customer I donât have values (NULL) because in the transactional systems these record donât have values (NULL).
The users want to generate aggregate reports with the Sales table...
The question is:
I have to put a âdummyâ value in the dimensions Customer, Model and Time (for example â0â) and put this value in the fact table if the dimensions fields have NULL values????
Or I have to leave the NULL values?
What is the best choice? Why?

There's often some specific reason why these values don't exist, such as the record being a manual adjustment to sales (for example a journal voucher). In these cases it can be helpful to have a flag column to indicate this, so that when a user comes across a bunch of sales with a Store Name of "Unknown" or "Not Applicable" they can also look at the reason for this unusual entry.

Similar Messages

Display master data without data in the fact table

Characteristic 0PROJECT
Attribute Price
I want to show in the query all the prices including the projects that don't have registers in the fact table.
How do I do this?
Tnks.

I believe you are describing what SAP referes to as the Slow Moving Item scenario. Search SDN using that phrase and you'll get hits on documents and Notes that talk more about this. Here's something from an old How To
Slow Moving Item Scenario
You want to define a query that displays all products that have been purchased only
infrequently or not at all. In other words, the query is also display characteristic values for
which no transaction data or only low values exist for the selected period.
Procedure
In the Administrator Workbench;
1. Create a MultiProvider consisting of a revenue InfoCube, containing the InfoObject
Material (0MATERIAL), and the InfoObject 0MATERIAL. The InfoObject must be set as
an InfoProvider in InfoObject maintenance. In other words, you need to have assigned
the InfoObject to an InfoArea. (also refer to Tab Page: Master Data/texts [Ext.]).
In the BEx Analyzer:
2. Select your MultiProvider in the Query Designer.
3. Define a query that contains the InfoObject 1ROWCOUNT in the columns.
The InfoObject 1ROWCOUNT is contained in all flat InfoProviders, that is, in all
InfoObjects and ODS objects. It counts the number of records in the InfoProvider.
In this scenario, you can see from the row number display whether or nor values
from the InfoProvider InfoObject are really displayed.
4. Save the query and execute it. All values are now displayed, including those for materials
that were not purchased.
If you filter by time (0CALYEAR, for example), values from the InfoProvider
InfoObjects are not displayed since 0CALYEAR is not an attribute of
0MATERIAL. You can see this from the absence of values in the 1ROWCOUNT
column in the query. If you want to restrict by time, you need to proceed as
follows:
Constant Selection for the InfoObject
You need to set the constant selection for the 1ROWCOUNT key figure in order to be able to
set a filter by time in this query.
1. In the Query Designer, via the context menu for 1ROWCOUNT, choose Edit.
2. On the left hand half of the screen, under the data package dimension, select the
characteristic InfoProvider (0INFOPROV) and drag it into the right-hand screen area.
3. From the context menu for the InfoProvider, choose Restrict, and restrict across the
InfoProvider InfoObject.
4. Also from the context menu for the InfoProvider, choose the function Constant Selection.
5. Save the query and execute it. You can now also set a filter for a time characteristic, the
materials display remains as it was.
Displaying Slow Moving Items
SAP Online Help 05.11.02
MultiProviders 3.0B, Support Package 07 10
If you want to display a list of slow moving items, excluding products that are selling well, you
need to proceed as follows:
1. In the Query Designer, via the context menu for 1ROWCOUNT, choose Edit.
2. Via the context menu for InfoProvider, choose the function Display Empty Values. Also
select Constant Selection.
3. Save the query and execute it. The result is that the system displays the materials for
which there was no revenue.
Displaying Products with Small Revenues
If you want to display a list of products that have not been sold or have only been selling
badly, you need to proceed as follows:
1. Set constant selection as described above, but do not select the display empty values
function.
2. In the Query Designer, define a condition for the 0MATERIAL InfoObject. Specify a value
that is to be the upper limit for a bad sale.
3. Save the query and execute it. The result is that the system displays all materials that
have not been sold or have been selling badly.

How we will know that dimension size is more than the fact table size?

how we will know that dimension size is more than the fact table size?

Hi,
Let us assume that we are going to take Division and distribution channel in a dimension and assume we have 20 distinct values for Division in R/3 and 30 Distinct values for Distribution channel .So Maximum, we can get 20 * 30 records in dimension table and we can take rough estimation of records in the cube by observing the raw data in source system.
With rgds,
Anil Kumar Sharma .P

Update data automatically in fact table in Data Warehouse

Hi,
I'm working on the creation of a data warehouse that include different data source like SQL Server performance (more than one), Active Directory users, Server performance (more than one), Exchange server mailboxes. The problem is that performance data change
frequently (like CPU and Memory), so my question is how to update data in fact table every 5 seconds automatically with SSIS.
Thank you for any advice

I'm assuming you have already figured out how to capture the data e.g. Powershell, extended events, MDW etc. and just need to know what dimensions or fact tables do you need.
You need to decide how often you are going to capture this data and based on that you will have dimensions with appropriate grain. Don't try to cram everything in the same fact table if it not of the same granularity. Also, separate process usually
have separate fact tables.
In addition to the Date dimension, you will need a Time dimension with a grain of 1 second (or maybe 5 seconds if that is when you get your data) then run the SSIS every 5 seconds to capture and append that data in the fact table.
- Aalamjeet Rangi | (Blog)

Load data in a fact table

Hello,
I have implemented SCD2 dimension and mapping executing works fine.
Now I have question about loading data in a fact table.
How do I need to use OWB (maybe JOINER operator - Join condition - between dimensions and source table) in case of:
- update on source table
- delete on source table
I think the most simple is insert on source table. It is probably to_char(source_transaction_date,'dd.mm.yyyy') = to_char(sysdate,'dd.mm.yyyy'), if I load once a day..
What is the procedure for fact table mapping to handle updates and deletes on source table?
Regards

Some discussions in previous forums should help you
http://forums.sdn.sap.com/thread.jspa?threadID=2019448
http://forums.sdn.sap.com/thread.jspa?threadID=1908902
In the SAP tutorial, you can see a sample example of making fact tables.
http://help.sap.com/businessobject/product_guides/boexir32SP1/en/xi321_ds_tutorial_en.pdf
Arun

How can we delete the data in e-fact table.

how can we delete the data in e-fact table.

hii,
You cannot delete the request individually but you can one of the following:
1. Do a selective deletion from the cube. RSA1 -> Cube -> Contents -> selective deletion.
2. Delete all the data in the cube and then reconstruct only the required request ids. This would work only if you have the PSA available for all the requests.
3. Reverse posting is another possibility.
hope it helps,
partha

Alternate fact table without disturbing the existing fact table in EIS

I have a question in EIS, how to load or work with alternate fact table, without disturbing the existing fact table, pls provide me with some procedure and steps
thank u in advance
sunil

Yes you can have multiple models and metaoutlines load one cube. I've done it before. The trick is you ahve one "master" metaoutline that you put all the measures in (even if they are manually added. You use that to build the cube. Then you can have multiple metaoutlines build different measures within the cube. The idea is you have to have them update different intersections or the last one in will be what you get. I have a client that gets sales info from one source, shipment info from a different relational source . in all I have 10 differnt models and metaoutlines that build one cube. you just lump the load statements into one CBS file. It works great (or as great as EIS works). The biggest issue is timing to make sure all the sources are available when the load is going to be doe. Of course you could build different parts at different times.

Confirmed Dimensions. OBIEE Not able to pull data from two fact tables.

Hi Experts,
I have a very simple set up of Star Schema with two fact tables and 1 dimension. Both fact tables joined to the dimension at the same level.
When i pull a column from both fact tables and the dimension table in OBIEE, it has to create simple SQL like below:
select FACT1.column1,
Fact2.Column1,
Dim.Column1
from FACT1, FACT2, DIM
where FACT1.ID = DIM.ID and FACT2.ID = DIM.ID
but instead it creating a query in a very complex way:
select case when D1.c2 is not null then D1.c2 when D2.c2 is not null then D2.c2 end as c2,
     D1.c1 as c3,
     D2.c1 as c4
from
     (select FACT1.Column1 as c1,
               DIM.Column1 as c2
          from
               DIM T1287863,
               FACT1 T1287945
   where (DIM.ID = FACT1.ID)
       ) D1 full outer join (
        select FACT2.Column1 as c1,
               DIM.Column1 as c2
          from
               DIM,
               FACT2
          where ( DIM.ID = FACT2.ID)
     ) D2 On isnull(D1.c2 , '1') = isnull(D2.c2 , '1') and isnull(D1.c2 , '2') = isnull(D2.c2 , '2')
I even tried setting the levels for both the fact tables and it still creates the query in avove way. Any thoughts on this will be vary helpful.

Subramanian,
see below the code we're using for the RFM.
on the ct_containers table i'm passing a line, and its getting updated after the call.
on the ct_errors table i just want to receive the errors and i only receive the line, we add manually there ('Serious error with validation code').
kr, achim
FUNCTION zbapi_ra_validations .
*"*"Local Interface:
*" IMPORTING
*"     VALUE(IS_RA_SCREEN) TYPE ZBAPI_S_RA_SCREEN
*" CHANGING
*"     VALUE(CT_ERRORS) TYPE ZRA_T_ERRORS
*"     VALUE(CT_CONTAINERS) TYPE ZRA_T_CONT_IP
DATA:
    lo_badi_handle TYPE REF TO zra_validation_rule,
    ls_error       TYPE zra_s_error.
GET BADI lo_badi_handle.
TRY.
      CALL BADI lo_badi_handle->validate_rules
        EXPORTING
          is_screen_flds = is_ra_screen
        CHANGING
          ct_containers = ct_containers
          ct_errors      = ct_errors.
    CATCH zcx_ra.
      ls_error-message = 'Serious error with validation code'.
      APPEND ls_error TO ct_errors.
ENDTRY.
ENDFUNCTION.
if i call this rfm in SE37 the ct_errors table is populated with all errors and the manually created line.
Message was edited by: Achim Hauck

Creating Time Dimension from date columns in fact tables.

I remember watching a demo from a BI Tool a couple years ago, wich I swear was OBIEE, and the presentator stated it was possible to create a Time Dimension in the admin tool, based on a date column in other table.
Can you guys tell me if there's such functionality in OBIEE?
If so, how could I achieve that?!
Thanks in advance!
Marcos

hi,
You are trying to make Fact table as Dim table ???
Fact table has some dim columns??
We can do this by making a fact table as dim table ,create a dim hierarchy on that table
Year level :Extract(year from fact_date_column)
Month and year level:CAST (Extract(month from fact_date_column) As CHAR(5) ) || CAST (Extract(yearfrom fact_date_column) As CHAR(5) )
Like this
But,be careful while doing this make sure that all joins and content levels are good
As per my knowledge this is not a good way,Experts can add some words lets see!!!!!! :-)
thanks,
saichand.v

Filters on non-conforming dimensions when fetching data from multiple facts

Hi,
I would like to confirm my understanding about how OBI fetches data from multiple facts. I have set up the BMM layer. Through Answers I am getting data from multiple facts as expected. However, is it necessary to base filters only on the conforming dimensions? If I do so, I am getting the correct data. But if I filter data on non-conforming dimensions, then I am getting more data which I am not expecting.
Suppose, Fact1, Fact2 and Fact3 share a CommonDim dimension. If I have a filter on CommonDim, then the filter is used on the three fact tables. However, if I filter data based on Fact1's DimTwo which is not shared by Fact2 and Fact3 and I have measurements from all the fact tables, then from the SQL generated, it seems that OBI is fecting all data from Fact2 and Fact3 and data from Fact1 where DimTwo = 'SomeValue'. Then it seems to be merging the data from the three result sets.
Is this expected, esp. a scenario where a filter is based on a non-conforming dimension (DimTwo in the example)? Have I missed anything? What do I need to do, if I want to return rows from Fact1 based on DimTwo's filter and then if there is data on Fact2 and Fact3 based on the value in CommonDim corresponding to DimTwo's filter?
Please let me know.
Thanks and regards,
Manoj.
PS.: I have got a link now which is http://obibb.wordpress.com/2010/06/06/multiple-fact-reporting-on-non-conforming-dimensions-part-ii/
Edited by: mandix on 25-Oct-2011 12:01

Hi
The problem you describe is similar to mine (I wrote several threads ago).
That is, if there is a non-conformed dimension in a report then the column from the fact table which is not connected with those non-conformed dimensions containes zeros or blanks.
Was this problem solved? I didn't quite catch one of previuos messages... If you can please repeat it more detailed.
What I tried to do is to set Total level in the Content tab (LTS properties). In this example - I could set Total level to JOBS in the plan type dimension. But it didn't help.
So if you know what to do then please describe it here...

Unexpected results getting data from two fact tables through conformed dim

Hi all,
We are getting an unexpected behaviour in our OBIEE 10.1.3.3.3. We have this scenario:
We have {color:#0000ff}2 fact tables{color}{color:#000000} called F1 and F2. F1 has one measure, f1m1 and F2 has another one, f2m1.
We have {color:#0000ff}4 conformed dimensions{color}, called D1, D2, D3, Date.
When we are requesting for individual fact tables, we are getting:
date d1 d2 d3 f1m1
dt1 - x - y - z - m1
dt1 - x - y - z' - m2
date d1 d2 d3 f2m1
dt1 - x - y - z - m3
dt1 - x - y - z'' - m4
But, trying to obtain a compare scenario, we are getting
date d1 d2 d3 f1m1 f2m1
dt1 x y z m1 m4
Instead of
date d1 d2 d3 f1m1 f2m1
dt1 x y z m1 m3
Looking at query log, we have catched the reason. That's why BI Server is using to solve this request using ROW_COUNT() to join SAWITH0 and SAWITH1 in SAWITH2 result set. So, the order may not be the same in the results sets in every fact table. More or less, generated query is like:
WITH
SAWITH0 AS
(select ....
from F1),
SAWITH1 AS
(select ...
from F2),
SAWITH2 AS
select from (select ...
ROW_NUMBER() OVER PARTITION (....) c10
from SAWITH0.d1 full outer join SAWITH1.d1 ....) D1
{color:#ff0000}where (D1.c10 = 1){color}
select SAWITH2. ....
from SAWITH2
order by c1..c10
The problems seems to be that BI server is ordering the result sets SAWITH0 and SAWITH1 and getting row number to join this results sets, but this is not getting the correct result.
Any ideas?
TIA
Javier
{color}
Edited by: jirazazábal on Mar 13, 2009 2:46 PM

I have done a logical fact table with two fact table source on it.
The Sql performed against the database was this one.
-------------------- Sending query to database named PRODS_AIX (id: <<153418>>):
WITH
SAWITH0 AS (select sum(T21296.CONSUMERS_SALES_EURO) as c1,
     T21309.DIVISION_CODE as c2
from
     DIVISION T21309,
     C_CONSUMERS_SALES T21296
where ( T21296.DIVISION = T21309.DIMENSION_KEY )
group by T21309.DIVISION_CODE),
SAWITH1 AS (select sum(T21356.ORDER_VALUE) as c1,
     T21309.DIVISION_CODE as c2
from
     DIVISION T21309,
     DWH_SALES_ORDER_OVERVIEW T21356
where ( T21309.DIMENSION_KEY = T21356.DIVISION_KEY )
group by T21309.DIVISION_CODE)
select distinct case when SAWITH0.c2 is not null then SAWITH0.c2 when SAWITH1.c2 is not null then SAWITH1.c2 end as c1,
     SAWITH0.c1 as c2,
     SAWITH1.c1 as c3
from
     SAWITH0 full outer join SAWITH1 On nvl(SAWITH0.c2 , 'q') = nvl(SAWITH1.c2 , 'q') and nvl(SAWITH0.c2 , 'z') = nvl(SAWITH1.c2 , 'z')
order by c1As you can see one select (SAWITH0) for the first fact table C_CONSUMERS_SALES and one select for the second fact table DWH_SALES_ORDER_OVERVIEW (SAWITH1 ) and the two statement are joined with a full outer join.
I ask me why you have the three select (SAWITH0,SAWITH1 and SAWITH2). Can you please paste the complete SQL performed ?
Can you tell us also which SQL is performed if you select only the columns from one fact table and not for the other ?
Regards
Nico
http://gerardnico.com

Pulling values from Variables for my Fact Table

Hi All,
The Load_Date in my Fact table is currently, current_timestamp-1, as the data is one day old when it is being loaded. I want to get that date from a variable because if I need to rerun with an earlier date like 20 or 30 days before the current date, I do not want to touch the mapping instead modify the variable and run the package. Please advise on how to do this?
Thanks for your time and help.

As Bhabani said, create a project variable with the code you would like to run to generate the earlier date. For example, "select current_timestamp - 30". Then, add this variable to the Package before the Interface, setting it to refresh. In the Interface, change the mapping from what it is now (current_timestamp - 1) to the variable (#VARIABLE_NAME). The variable will now be refreshed and the value added to the mapping.
If you want to change it to "current_timestamp - 20", just update the variable.
Regards,
Michael Rainey

How to fetch data from multiple fact tables from a normalized schema?

Hello everybody,
I am working on a normalized schema to build my repository. I have categorized the tables into dimensions and facts. I have 3 fact tables in my schema and they have a 1:m:1 relationship i.e. if I have table A, B, and C; A has 1:m relationship with B and C also has 1:m relationship with B. How can I use measures from these tables to create a star schema. If A = Sales, B = Transaction, C = Payment_Amount
Sales(Sales ID, Amount, Tax pct, ...)
Transaction(Transaction ID, Sales ID, Payment ID,Transaction Amt, ...)
Payment_Amount(Payment ID, Check ID, Payment Made, ...)
Please give me some direction to pursue.
Thank you!
D

Hi dcole,
Go through this link for snowflake schema http://gerardnico.com/wiki/datamodeling/snowflake_
http://www.rittmanmead.com/2007/06/19/obiee-data-modeling-tips-1-integrating-1-1-and-1-many-source-tables/
I suppose it should work with snowflake schema...Im interested in this topic please let me know what is your approach every now and then.
By,
KK

Why Negative values are entering in to fact table.

Dear All,
After Data File uploaded. The Fact table is showing Negative values.
Can any body give solution.
And in DataManager->Financial Process>FX Restatament (Is running Succussfully)
But Currency is not getting translated.
Please help.
Thanks,
Satish.

The reason the values are negative may be caused by 2 items. The first is the default method by which BPC store information; BPC is designed to store values based on the natural sign. So, if a Revenue account is positive in the Data File, and the Property "ACCTTYPE" is INC, the value is stored as a negative number. When an expense value is loaded to an account with "EXP", the value is stored as a positive. LEQ are Negative and AST are positive. (CREDITS and DEBITS storage)
BPC then uses tables and measures to report the account based on type correctly in the Excel interface. This is how BPC is designed.
There is also the CREDITPOSITIVE = YES / NO that is part of the Data Transformation instructions, and if this is setup incorrectly, you may store the reverse signs from the data file that is loaded. So this may need to be reviewed as well.
As for the FXTranslation, I assume you have rates in the RATE cube to use in the translation process. You will need to verify that the RATETYPE property is filled in for the accounts you wish to translate? DId you process the FX script logic and the RATE script logic prior to running FX.
Hope this helps.
Edited by: Petar Daniel on Dec 10, 2008 2:20 PM

Decode the values (Re-group Dimension Values) without ETL

Hello guys, i have a question which is partly triggered by me not wanting to modify the default ETLs.
I have values in a dimension table coming as:
Region A
Region B
Region C
Region D
Region E
However, i am hoping to re-org the hierarchy as below:
Region New A
Region New D
I know this can be done in ETL but is there any place else that this can be done? Possibly in the business layer? Is there a place in the business layer where we can decode the values and re-group them?
Essentially, there is a new org structure in our company where we are grouping (consolidating) old regions into new and renaming the values.
Thanks in advance for your help,

In the logical column in the BMM you have two options to write expression:
1. In general tab/use existing logical columns as the source
2. In data type tab/logical table source/edit/logical column and write expression using physical tables and columns
In both cases you can write your mapping expression whether it is a case statement or something else.
Regards
Goran
http://108obiee.blogspot.com

Dimension values without data in a fact table

Similar Messages

Maybe you are looking for