In article <113DA521-771E-4E99-9873-570EC3EA1B62 (AT) microsoft (DOT) com>,
KitenutDave (AT) discussions (DOT) microsoft.com says...
Quote:
I understand that it is probably not optimal design, but our source data has
some "holes" in it. That is, certain dimensions are not populated in every
record. For text fields I can update null fields to a value such as 'N/A'
during the load into the fact table. For numeric fields, I can't use 'N/A'
unless I change the datatype to text. Are there any consequences if I do this?
The usual pattern for dealing with this sort of issue is to add a record
|
to your dimension table to catch these 'unknown' records. I often use an
ID of 0 or -1 across all my dimensions. When inserting records into the
fact table I then COALESEC() null values to the 'unknown' ID.
--
Regards
Darren Gosbell [MCSD]
Blog: http://www.geekswithblogs.net/darrengosbell