dimensional fact model problem
4 posters
Page 1 of 1
dimensional fact model problem
Hallo!
I am unexperienced in building a DWH and I hope you can help me
At https://i.servimg.com/u/f76/18/23/84/53/unbena10.png you can see my (simple) ERM.
The tables A and B are my Dimensions. But what should I do with C?
C has a n:m relationship with B and 1:n relationship with A: a A belongs to ONE C, but a C has serveral A's. For a fact, the C joined over A is identical with the C joined over B. In other words, one fact has exactly one C! It does not matter which "way" (A or B) you use. The relationship between A and C as well as B and C are not important for me. I need A, B and C for aggregation&filtering.
Consequently my intention is to handle C as third Dimension > "unlink" C from A and B. What do you think about it?
Originally I tried to develop the dimensional fact model using the data driven approach described by Golfarelli et. al. but I didn't find a solution for my construct. I run into the problem that it seems C is a hierarchy instead of dimension. But I think it is nonsense to add the same hierarchy to the dimensions A and B. Why does the data driven approach not working for me what i'am doing wrong?
Do you know any literature which copes with such a construct?
Thank you!
I am unexperienced in building a DWH and I hope you can help me
At https://i.servimg.com/u/f76/18/23/84/53/unbena10.png you can see my (simple) ERM.
The tables A and B are my Dimensions. But what should I do with C?
C has a n:m relationship with B and 1:n relationship with A: a A belongs to ONE C, but a C has serveral A's. For a fact, the C joined over A is identical with the C joined over B. In other words, one fact has exactly one C! It does not matter which "way" (A or B) you use. The relationship between A and C as well as B and C are not important for me. I need A, B and C for aggregation&filtering.
Consequently my intention is to handle C as third Dimension > "unlink" C from A and B. What do you think about it?
Originally I tried to develop the dimensional fact model using the data driven approach described by Golfarelli et. al. but I didn't find a solution for my construct. I run into the problem that it seems C is a hierarchy instead of dimension. But I think it is nonsense to add the same hierarchy to the dimensions A and B. Why does the data driven approach not working for me what i'am doing wrong?
Do you know any literature which copes with such a construct?
Thank you!
ben_1990 Posts : 2
Join date : 20130411
Re: dimensional fact model problem
Hi there,
To handle many to many relationships i recommend a relationship table (Bridge)
To handle many to many relationships i recommend a relationship table (Bridge)
Re: dimensional fact model problem
Relate C to the fact. There is only one C that would satisfy the situation for a particular fact row, correct?
Facts are held in context with its related dimensions, C being one of them. Relationships A has with C and B has with C are immaterial to the fact. If you want to track the states of relationships between A, B, and C, you use another fact (or facts).
Facts are held in context with its related dimensions, C being one of them. Relationships A has with C and B has with C are immaterial to the fact. If you want to track the states of relationships between A, B, and C, you use another fact (or facts).
Re: dimensional fact model problem
If you did care about the relationships between A, B, and C, you could handle via a factless fact table.
BoxesAndLines Posts : 1212
Join date : 20090203
Location : USA
Similar topics
» Purchasing Fact Model problem
» Single fact table vs multiple fact tables  what is the right way in a dimensional model?
» General Ledger Fact Table: Dimensional Model
» Dimensional Model: Connecting dimension to fact table using two approaches
» Rule based algorithm to convert an ER model to a dimensional model
» Single fact table vs multiple fact tables  what is the right way in a dimensional model?
» General Ledger Fact Table: Dimensional Model
» Dimensional Model: Connecting dimension to fact table using two approaches
» Rule based algorithm to convert an ER model to a dimensional model
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum

