I recommend adding a dummy key according to NULL. The key can be unique at
its level by itself or by being concatenated with its parent key. (Here,
OrganizationID?) Or you can make the key have the same value as its parent
depending on the business situation.
FYI, it's better to avoid multiple parallel joins between the dimension and
fact tables, although Analysis Services supports it without problems.
Ohjoo Kwon
"gNM" <gedukas99 (AT) hotmail (DOT) com> wrote
Quote:
I have some ORG dimension with several levels (Country, Organization,
Operator). Fact table contains OrganisationId and OperatorId.
OrganisationId
in Fact table always NOT NULL while OperatorId sometimes contains NULL
values. I join facts with dimension by OrganisationId and OperatorId. In
case OperatorId is NULL I miss facts. How can I get these facts at such
cases only by OrganisationId, while in other cases I get them by
OperatorId? |