Archives of the TeradataForum
Message Posted: Fri, 19 Jul 2002 @ 13:45:25 GMT
| Subj: | | Attributes at a lower level of granularity than metrics. |
| |
| From: | | Jim Downey |
Does anyone have situations where they have an attribute which is at a lower grain than the metric? For example, we have a metric
(premium) which is calculated at a policy level. Since each policy might have multiple drivers, we are faced with how to report the
premium metric with driver attributes. How much premium do we get for Male drivers? For those of you who know insurance, this is
not the actual problem, but I have rolled the description up to a level that I think more people would understand.
Some of the choices are:
Do not allow metrics to be reported with a metric at a lower level of granularity. You can't report gender with
premium.
Allow the metric to product join if that is what is requested.
sum(premium)
Force the join so only one attribute is available. You can report on one gender but not both at the same time.
Allocate a percentage of the metric across the multiple attributes.
sum(premium)*%male +sum(premium)*%female.
I know that the correct answer is "What the business wants." but I am interested in how others handle this problem. What are the
pro's and con's to any approach. Does anyone know published references how to handle this. What would be considered "Best
Practice".
Thanks
Jim
| 
