question regarding fan trap....

[system]

Hi,
I have a universe with 3 tables say A,B,C and table A is security table,and table B is a fact table and I have joined these three tables based on the Dealer column, and there is a join for tables
A----->B complex join where i used @variable(‘BOUSER’) in the join condition and B----->C is an equi join and for table B Dealer I have selected the security table in the object definition to cause the use of specific join in the generated query, so that i’m not using any objects from the security table A, and I’m using the objects from table B and C.
Now my question is does this scenario forms a fan trap.

Thanks

---

santhosh_03 (BOB member since 2009-07-08)

[system]

Typically, no, but to be really able to tell one would need the cardinality of all involved joins.

---

Andreas (BOB member since 2002-06-20)

[system]

Thanks for your relply Andeas.

I’m bit confused so Just for clarification, in the scenario i have mentioned, i’m dragging both dimension and measure objects from table B and measure objects from table C and condition objects from table B (i,e in query filter) in a single query, if I’m not wrong the results will inflate if you have measure objects from two tables in a single query, to avoid this I can use a seperate query for table C measure objects with condition objects from table B.
or the other work around if i want to use all the objects in a single query i need to set the cardinality and insert the contexts so to avoid the inflated results.

Please clarify…

Thanks.

---

santhosh_03 (BOB member since 2009-07-08)

[system]

can some one please clarify my above scenario…

Thanks.

---

santhosh_03 (BOB member since 2009-07-08)

[system]

Hi,

I understand that you have 3 tables A, B and C: A — B — C
What are the cardinalities of joins A — B and B — C?
Which tables are dimensions and measures in a query coming from?
Can you post an example of a query?

---

Marek Chladny (BOB member since 2003-11-27)

[system]

Hi Marek,
Thanks for your reply,
The cardinality for A------>B is M---->M
and B------>C is M------>M
I’m taking both dimensions and measures from table B and a measure from table C.
I’m not using any objects from table A. but since table B Dealer(primary Key) I have selected the security table in the object definition so that its automatically showing the join in the genrated sql. Is this mean that the objects are coming from 3 tables.

Kindly suggest.
Thanks.

---

santhosh_03 (BOB member since 2009-07-08)

[system]

What is M---->M cardinality? Is it N:N or 1:N or 0:N?

In short, a fan trap occurs when you have a 1:N cardinality of a join between 2 tables and you have measures in a query from both of the tables (or at least from the table on the “1” end of the 1:N join).

Check more topics about a fan trap:

Problem of incompatible items...how to design?

Hi to everybody…i have 2 simple items in bo universe (2 aggregate indicators) , but when i put them together in a query, Bo think they are incompatible and so the result is this in the picture… he make 2 distinct queryes and then he join…but in pl/sql i usally do only a query, and i have the correct result…what i can do in the designer, to telle that the object are compatible, and that he can do it in only one query? Thanks to everybody!! [1.JPG] flavio20002 (BOB member since 2010-03-23)

Fan/Chasm Traps

Hi All, Yes I have searched QUITE a lot on this forum and I think I have covered most of the posts related to this topic. I saw a lot of posts exlaining how to solve them but haven’t come across any that describe the cause of traps. Here are somethings that I need help with: Fan Trap: D1 --< F1 --< F2 Fan trap will not exist if the fact tables above are at the same level of granularity no matter where the dimensions/measures are coming from. Is this a fair statement? If yes, then in real wo…

multiplying measure values =&gt; always fan/chasm trap??

Dear, I’m suffering a problem in my universe…I freshly created it and it’s based on an RDB model, not a star-scheme unfortunatelly. I defined some contexts and alias’ed some tables in order to provide clean operation of the universe. Today I started creating some reports and all went fine untill the measure (yes, ther’s only 1 numeric in the entire report) started multiplying its value from the moment I started adding more tables to the query… So in the beginning I started …

Duplicate records and miss match of values

I am designing a universe I am facing a issue where for one record I am fetching 7 records and getting huge values , I should get one record. I have checked the all the joins but could not solve the issue. In my date table I have dates from day,week and month is this some thing to do with it needful advice is helpful thanks in advance bobtom (BOB member since 2008-01-17)

fan trap

I have read the article from Dave on fan traps,A->B->C. Does it mean that if B and C have measures in them then it is a fan trap. Or does it have to be a specific on the type of the measures like for example may be total for orders in B and number of orders in C. hartford27 (BOB member since 2006-11-22)

joins

If we have three tables and the middle table gets a one to many join on one key from another table ie this table being the many side and its get a one to many join on an other key from an other table again this table being on the many side, what are the common issues. GC_hill (BOB member since 2007-08-29)

Fan trap issue

Hi All, We are facing a Fan Trap problem. We are having 3 tables which are refered in the report. The relation between these tables is as: CRT_CURRENCY ----< PRC_RATE-----< CAPACITY_DTL. The data in these tables is as: Table CRT_CURRENCY: [code:1:f8badc8b8f]CCY_CD LNG_DESC -------------------------- GBP Pound USD US Dollar[/code:1:f8badc8b8f] Table PRC_RATE: [code:1:f8badc8b8f]RATE_LINE_ID VER_NO CCY_CD TOT_PROMSD_REV -------------------------------------------------------------- 123…

---

Marek Chladny (BOB member since 2003-11-27)

[system]

Thanks for the information Marek,
Thats N---->N so if i use the measure objects from both tables which are in 1—>N cardinality, then the results will inflate here i can use the contexts and uncheck the Multiple sql statement for each measure in the universe or other work around is that i can take one more query with the objects coming from table C if no contexts are set and tick the multiple sql statement for each measure in universe…

Thanks.

---

santhosh_03 (BOB member since 2009-07-08)