Hi,
When we have only Dimesion source data,
Can we update only dimensional table without updating the fact table using Slowly Changing Dimension Stage?
Or is it mandatory to load fact table too?
Updating Dimension Table using SCD stage
Moderators: chulett, rschirm, roy
Updating Dimension Table using SCD stage
I never let school to interfere in my education
-
- Participant
- Posts: 54607
- Joined: Wed Oct 23, 2002 10:52 pm
- Location: Sydney, Australia
- Contact:
Your "fact table" can be the file NUL (which will eat, and discard, all the rows you append to it). However, be really, really certain that what you say is true. Factless dimensions are extremely rare things, and suggest that something has been overlooked in the design phase - what are they dimensions for? "Factless fact tables" certainly are valid constructs, perhaps to record that some combination of dimension values has occurred without their being any form of measure associated with the combination, but "factless dimensions" just rings alarm bells for me.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
-
- Participant
- Posts: 54607
- Joined: Wed Oct 23, 2002 10:52 pm
- Location: Sydney, Australia
- Contact:
Typically that is ALL that a factless fact table contains. A set of foreign keys to dimensions; the combination of those foreign keys effectively forms the primary key of the fact table.
A factless fact table records that some combination of dimension values occurred. There are no measures. For example a combination of time, employee and customer dimensions might record the fact that the employee visited the customer - this might be a VISITS fact table. No other information is required. Counts of visits grouped by time, employee or customer can be created in reporting tools.
A factless fact table records that some combination of dimension values occurred. There are no measures. For example a combination of time, employee and customer dimensions might record the fact that the employee visited the customer - this might be a VISITS fact table. No other information is required. Counts of visits grouped by time, employee or customer can be created in reporting tools.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
-
- Charter Member
- Posts: 193
- Joined: Tue Sep 05, 2006 8:01 pm
- Location: Australia
Re: Updating Dimension Table using SCD stage
[quote="basav_ds"]Hi,
When we have only Dimesion source data,
Can we update only dimensional table without updating the fact table using Slowly Changing Dimension Stage?
Or is it mandatory to load fact table too?[/quote
You can just load your dimension tables first but make sure you have proper timestamping (e.g valid time and/or transaction time) in your dimension tables. When you are ready to load you fact table then you need to make sure your fact record links to the "correct" dimension record. Do some research on temporal data in data warehouses.
When we have only Dimesion source data,
Can we update only dimensional table without updating the fact table using Slowly Changing Dimension Stage?
Or is it mandatory to load fact table too?[/quote
You can just load your dimension tables first but make sure you have proper timestamping (e.g valid time and/or transaction time) in your dimension tables. When you are ready to load you fact table then you need to make sure your fact record links to the "correct" dimension record. Do some research on temporal data in data warehouses.