Implict fact column issue

HI ,
I have two fact and two dimension tables.
D1 is joined to F1 and D2 joined to F2 and i have set as implicit column from F1. I am trying to develop a report from D1 and D2 . We are not getting proper results.
But D1 is not joined to F2 and D2 is not joined to F1. An y idea on this.
Regards,
Malli

Hi,
Create an alias of the dimension tables and create a dummy fact table with
select 1=1 from dual;
and join the three tables.
But in the dimensions, there should be relating data in the two tables.
Try and let me know further
Regards,
MuRam

Similar Messages

Implicit Fact Column being ignored

Hi,
I have a Business model containing two fact tables with a shared sets of conformed dimensions. Dim1 and Dim2 are both mapped to Fact1 and Fact2.
When I select attributes from Dim1 and Dim2 (without facts), OBIEE picks Fact2 to pull up the records. Since I want OBIEE to use Fact1, I've setup an Implicit Fact Column using a measure of Fact1.
After restarting all services, OBIEE still picks Fact2 (probably because it has less -conformed- dimensions). When I check the physical query, the Implict Fact Column is NOT included in the select statement so OBIEE is basically ignoring it.
Is there any config-setting which influences how OBIEE activates the Implicit Fact Column?
OBIEE-version = 10.1.3.4.1
Thanks!

Well.... not really suggestions... but I did some more google research and came out with some comments...Pls. note that they are not mine comments... just copy/paste from a couple of blogs.
"This column is generally the *lowest granularity column* of the fact table like Row_wid etc"
"It all boils down to the granularity of the data present in the fact tables.
If a fact contains data at a more granular level then always that will be preferred over the other fact.
The 'Implicit fact column' comes into picture when you want to enforce a fact as the default path when used in a report."
"For dimension-only queries across multiple dimensions, Oracle BI Server picks the most economical fact table source
based on t*he number and levels of joined dimensions*."
Antonio
BExpert, Brazil

Using a Fact Column in Prompt

Hi,
I have requirement to use a Fact Column (Eg: Control flag) to be used in the Column Prompt. What is the best way of using it such that query performance will improve. If I use it directly it might have scan through my entire fact table which will have a performance impact.
Is there a best way to do it?
Thanks,
GJ

Hi GJ,
The "proper" way to model this in the RPD is to add the column on to a dimension vs. leaving it on the fact table. i.e. the underlying physical fact table becomes a source for both a logical dimension and the logical fact. OBIEE smart enough to sort everything out.
Having said that - the real performance issue could be because, in a prompt, it would have to scan a potentially huge fact table to get the distinct values of the control flag. If that's the case, then you should either 1) build a real dimension table to make it faster (this would be preferable), or 2) add a bitmap index to the control flag column on the fact table, which would also greatly improve the performance of the "select distinct control flag from fact table" query.
Hope this helps!
Scott

Implicit Fact column creating problem to the reports.

Hi Friends,
I have two reports from same subject area from two different facts F1 & F2.
I have created one implicit column, as it is unable to judge from which fact it needs to build the query. So, it is genarting query from fact F1.
I have created one logical column in the fact and marked it as Implict column in presnenation layer.( as we the process)
But the problem is, for both the reports it is using same implicit column and it is generating the query with only one fact F1.
It is excluding the other fact F2.
One report it shoud use Fact1 and other report it should use Fact2.
PLease hep me in regarding this,
Thanks & Regards,
Kalyan J V

Kallu wrote:
Hi Friends,
I have two reports from same subject area from two different facts F1 & F2.
I have created one implicit column, as it is unable to judge from which fact it needs to build the query. So, it is genarting query from fact F1.
I have created one logical column in the fact and marked it as Implict column in presnenation layer.( as we the process)
But the problem is, for both the reports it is using same implicit column and it is generating the query with only one fact F1.
It is excluding the other fact F2.
One report it shoud use Fact1 and other report it should use Fact2.
PLease hep me in regarding this,
Thanks & Regards,
Kalyan J VFrom my understanding, implicit fact column's are set a subject area level and not dimension/fact level. So the fact table you have set for implicit fact will be used in both the reports. Why dont you just drag the second fact into the report and hide it, if you are not trying to use the fact 1 that is setup as implicit fact?

Fact columns in dimension table

In OBIEE, Can we use a fact column in related dimension table in case this column is just an description column? Yup I know that use fact columns in fact and dimension in dimension.. But what exactly will be issue if we do this? I observed it in one model.. No data connection was there to test..

Hi Hari,
Why can't you create two tables in BMM one is as FACt and DIM and place all description columns in dim and measures in fact?
this way you would achieve Star Schema in BMM
http://www.varanasisaichand.com/2012/04/fact-and-dimension-from-single-source.html (2nd approach)
thanks,
Saichand

Filter on Fact columns at report level- obiee 11g

Hello,
The error I am getting in my report is : Incorrectly defined logical table source (for fact table SUMM) does not contain mapping for [LEDGERFACT.CD, LEDGERFACT.YEARID].
The Model(Physical and BMM Layer) is as follows: DIM_LF_B is joined to 2 fact tables LEDGERFACT and SUMMARY
DIM_RPT_PARAM -> DIM_LF_B -> LEDGERFACT
DIM_RPT_PARAM -> DIM_LF_B -> SUMMARY
Note that the aggregation is not set for any of the Fact columns. (I am aware that all measures have to be set some aggregation type , but Client restrictions as not to edit the rpd :( )
Now on the report front, i have pulled following columns:
>> Col name: Description ||| Forumla: DIM_RPT_PARAM.ITEM_DESC
>> Col name: Actual Amt ||| Formula: case when SUMMARY.MONTHID=3 and SUMMARY.YEARID=2012 then sum(SUMM.AMT_1) else 0 end
>> Col name: Plan Amt ||| Formula: case when LEDGERFACT.YEARID=2012 and LEDGERFACT.CD=300 then sum(LEDGERFACT.AMT_1) else 0 end
I am not able to understand as to why the BI server is throwing the error.
I tried removing the case condition and just applying the sum(as below), and the report comes up.
Col name: Description ||| Forumla: DIM_RPT_PARAM.ITEM_DESC
Col name: Actual Amt ||| Formula: sum(SUMM.AMT_1)
Col name: Plan Amt ||| Formula: sum(LEDGERFACT.AMT_1) Also for the following combination the report works
1)
>> Col name: Description ||| Forumla: DIM_RPT_PARAM.ITEM_DESC
>> Col name: Actual Amt ||| Formula: sum(SUMM.AMT_1)
>> Col name: Plan Amt ||| Formula: case when LEDGERFACT.YEARID=2012 and LEDGERFACT.CD=300 then sum(LEDGERFACT.AMT_1) else 0 end
2)
>> Col name: Description ||| Forumla: DIM_RPT_PARAM.ITEM_DESC
>> Col name: Actual Amt ||| Formula: case when SUMMARY.MONTHID=3 and SUMMARY.YEARID=2012 then sum(SUMM.AMT_1) else 0 end
>> Col name: Plan Amt ||| Formula: sum(LEDGERFACT.AMT_1)
Any help appreciated.
Thnks,
Shruti
Edited by: 970002 on Mar 28, 2013 7:18 AM

Hi Shruthi,
To be clear you have one table 'DIM_LF_B' and its separated out as FACT and DIM in BMM layer with respective formulas ? If yes check this http://www.varanasisaichand.com/2012/04/fact-and-dimension-from-single-source.html
let me know if it is different
Thanks,
Saichand

When we should go for implicit fact column

hi all,
In which scenario we have to go for implicit fact column....
let say we are having dimensions D1 and D2 and D3.. and two Facts F1and F2
all dimensions are connected to both facts F1 and F2,
once you drag the columns from D1 and D2... the report will come based on facts F1 and F2.... so if we want to restrict the report .. i mean report should come based one Fact table let us say
F1 that time only we have to go for implicit fact column .....
or any other scenarios ... are there for implicit fact column
Thanks

Hi ,
have a look http://obiee.nl/?tag=dimensional-modelling&paged=2 you willl get a good idea.
Hope this may helpful for you
Thanks,
Ananth

How to create Implicit Fact column ?and how to use it ?

Hi all,
How to create Implicit Fact column ?
and how to use it ?
Thanks in advance

When you create a request with only dimension columns, the Oracle BI engine has to choose a fact table to join the dimensions to each other. If you have multiple fact tables in your BMM layer, the Oracle BI engine will use the most economical fact table, depending on the number of dimensions of the fact table.
In some cases you do not want to use this fact table, but you want to force the Oracle BI engine to use the fact table you want.
In that case you should create a fact (or select an existing fact) of the fact table you want to use for the dimension only query. In the presentation catalog of the subject area you should select this fact as the implicit fact. This fact will always be used in the query, but it won't be visible in your report. An example of an implicit fact is the count of the primary key (id) of the fact table.

Filtering based on Fact Columns makes the result different?

Dear All,
I have an analysis report for Publisher, there are several columns, for example,
Col A = ROUND(Col B * Col C / Col D, 0) Whereas Col B, Col C and Col D are integers
Col E = ROUND(Col F * Col C / Col D, 0) Whereas Col F, Col C and Col D are integers
Filter: Col G is not null, Whereas Col G is fact column, Integer
I have the case is that
Col B = 120, Col C = 4, Col D = 14, Col F = 10
If Filter is not applied:
Col A = Round(120 * 4/14, 0) = 34
Col E = Round (120 * 10/14, 0) = 86
When Filter is applied,
Col A = Round (Floor(120 * 4/14), 0) = 34
Col E = Round(Floor(120 * 10/14), 0) = 85
C

If you want to avoid the floor function Use (120*10/14.0)
If all the values in a division are integers, OBIEE uses a Floor function automatically.
SO CAST either the numerator or the denominator to DOUBLE and then run your calculations, the floor function will not be applied.

OBIEE - Answers - Grand total of a fact column

I need to determine if a fact column (lets say X) is within a certain range of the total of X. I want to know how we can get the total of X in each record of the answers resultset. The column X has been defined as an aggregated column in the repository.
The answers request is : Dim A, Dim B, Fact X and I need to calculate whether Fact X is less than the average of Total of X.
Any help in this regard is highly apreciated
Thanks
Shiva
[email protected]

avg = 1.00 * sum(x)/max(rsum(1))
then use case statement

Multi fact columns to Single fact column

Hi,
We have fact table which stored the data in monthly wise fact columns.
Ex:
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
100
200
150
250
223
1212
171
12123
31123
112
2113
1123
150
223
222
142
1354
1567
452
763
41733
441
1211
1213
333
222
55
256
455
445
752
4752
45214
114
8122
4555
My requirement, We have convert all month fact columns into a single month fact column.
How to achive this. Kindly Let me know.

It rather seems that you don't understand his input or the concept of LTSs.
You create 12 LTS, each using the same physical table as a source.
Then you create one single column called "Monthly Value"
Then you map the "Jan" column from the first LTS into the column
Then "Feb" from the second LTS
etc.
Bob's your uncle

Can we make a fact column to be displayed as a single ROW in obiee 11g

Hi,
I have a requirement recently, where i need to show a fact column which shows the total value, as a single row in the main report.
Total
1000
1000
1000
1000
Changed to..
Total
1000
Can we achieve this in OBIEE.
Please share your inputs. Much Appreciate it !
Thanks,
Ramya

Shailja,
Yes you can. Follow these steps:
1> Create the LOV (which you can create from another table, as you mention). and test it.
2> Now you create one ADF Table on the page.
3> Select the Table in staructure window > select af:column(the column where you want put you lov) > Inside the af:column, Replace the Uicomponent by the LOV(which you created in step-1)
4> Save & Run
--Neelmani Jaiswal
http://www.jneelmani.blogspot.com

Query on filter on fact column

Query on filter on fact column
IN OBIA AR Transactions subject Area
the fact is filtered on (RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='REV' OR RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='UNEARN') in the Informatica ETL
(RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='REV' OR RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='UNEARN')
I need to bring in the Tax information within the subject area from the table above and extend the Filter to include 'Tax'
(RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='REV' OR RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='UNEARN')
and
(RA_CUST_TRX_LINE_GL_DIST_ALL.ACCOUNT_CLASS='TAX')- which comes from Account_class
will this amendment affect the existing subject area and in turn produce higher report numbers
secondly
if i would like to understand how i would try and work out the report
AR amount split by tax
Invoice Number Account class AR Amount
001 REV 100
001 Tax 20
to produce the above report i have to remove the filter on the ETL
but make sure that when the user DOES NOT hits the Column Account class in analysis then by default put a filter on the fact column source which keeps only 'REV' or 'UNEARN'
thanks

The answer lies in GL revenue analysis subject area

Multiple fact column should show the value in different row.

Hello experts,
We have a requirement in this way. I have to build a report with three fields: date, Fact :No of jobcard open, No of jobcard closed. Report should come in this way..
Date
1 2 3 4 5 6 7 8 9 10
No of jobcard open 30 45 46 67 83 28 23 45 01 92
No of Jobcard closed 23 40 25 29 70 20 10 20 01 89
I am new in this field, please help me in this regard.
Regards
Niraj
Edited by: user13412398 on Jan 19, 2011 7:46 PM

Hi Niraj,
Just pull the columns into your criteria request and from compound layout select pivot view so all your columns will be in place in pivot table.
Then you pull your columns accordingly seeing the pivot structure i.e. placing date at the top and in measure section place the fact column and in rows section place the jobcard closed.Folow these and you would get the result.
Follow this etiquette http://forums.oracle.com/forums/ann.jspa?annID=939
UPDATED POST
Place both those columns in measure and the date column as you need at the top.Just play around with the pivot view by placing here and there and check results.
So you will get idea and that will fulfill your requirement.
Hope answers your question.
Cheers,
KK
Edited by: Kranthi on Jan 19, 2011 9:34 PM

Implicit Fact Column - am I thinking about this the right way?

My exploration of the Implicit Fact Column is below. Let me know if you think this is the intended usage.
When the BI Server sees a request where one Dimension is being constrained by another, it has to pick a fact table that contains references to these two dimensions. The "Implicit Fact Column", which is set in the Presentation Layer of the RPD is used to guide the BI Server towards one or more Fact tables that can be used to satisfy the request. In the absence of the Implicit Fact Column, I noticed that the BI Server was choosing random fact tables so I went exploring how this feature worked. I set the Implicit Fact column in my Presentation layer and then went looking to see how the queries were generated.
For my example, my two dimensions are Program and Channel. I want to see the list of Channels available for a Program. Without the Implicit Fact Column, the queries are of the form:
select distinct T289.CHANNEL_NAME as c1,
T36001.PROGRAM_NAME as c2
from
DIM_PROGRAMS T36001,
DIM_CHANNELS T289,
SomeRandomFactTable T41210
where ( T289.DIM_CHANNEL_KEY = T41210.DIM_CHANNEL_KEY and T36001.DIM_PROGRAM_KEY = T41210.DIM_PROGRAM_KEY )
order by c1, c2
The nice part about this is that there is only one set of Program / Channel combinations returned. The downside is that the Fact table that is chosen to resolve the combinations can be any Fact table under the purview of the BI Server. It doesn't even have to be in the same subject area where the Program and Channel were chosen. When I set the Implicit Fact Column, the queries are of the form:
select distinct T289.CHANNEL_NAME as c1,
T36001.PROGRAM_NAME as c2,
T35832.DIM_SITE_KEY /* this is the implicit fact column */
from
DIM_PROGRAMS T36001,
DIM_CHANNELS T289,
AFactTableMappedToFactColumn T35832
where ( T289.DIM_CHANNEL_KEY = T35832.DIM_CHANNEL_KEY and T35832.DIM_PROGRAM_KEY = T36001.DIM_PROGRAM_KEY )
order by c1, c2
I was happy that I could predict what Fact table was being used, but now the results are wrong as I get a duplicate set of records for every DIM_SITE_KEY / Channel / Program combination. I decided to play a little Algebra trick. Knowing that the DIM_SITE_KEY was a numeric value, I defined the Implicit fact column as DIM_SITE_KEY / DIM_SITE_KEY which will always result in a value of 1. Now I get queries that look like:
select distinct T289.CHANNEL_NAME as c1,
T36001.PROGRAM_NAME as c2,
T35832.DIM_SITE_KEY / nullif( T35832.DIM_SITE_KEY, 0) as c3
from
DIM_PROGRAMS T36001,
DIM_CHANNELS T289,
AFactTableMappedToFactColumn T35832
where ( T289.DIM_CHANNEL_KEY = T35832.DIM_CHANNEL_KEY and T35832.DIM_PROGRAM_KEY = T36001.DIM_PROGRAM_KEY )
order by c1, c2
Since DIM_SITE_KEY / DIM_SITE_KEY is always equal to 1, I only get one set of Program / Channel combinations. I get the added bonus of knowing which set of Fact tables are going to be used to satisfy the request.

Perfect Analysis, But one important note is when using implicit fact table, Always select measure( Which has some aggregation) as a implicit fact column. So that will allows the query will eliminate duplicate rows by doing a group by operation.
Other note is make sure performance is good. When we don't have a good model for fact tables, when the tables contains large set of data, make sure performance is good by doing indexes or caching to load the prompts.
- Madan

Implict fact column issue

Similar Messages

Maybe you are looking for