Multivalued attribute
Hi I have read a lot about this modeling technique but either I didn't understand what I read either this is not the solution suitable to my needs. Here it is. I am dealing with sales of products in stores. Each product can be assigned to a

data as an attribute on a dimension or a key on the fact table
Hi, I am in the process of trying to model a mortgage (origination/servicing) data mart. For requirements all I was given at this point is a spreadsheet of attributes. My plan was to try and pull out dimensions and metrics from the list and then have a

What to do when a single attribute rolls up to multiple categories?
I'm working with healthcare claims data, and developing a new data warehouse. A single pharmacy claim line will have a drug identifier called a NDC (national drug code). A single NDC may belong to one or more drug categories. Given the one-to-many

dimension attribute denormalisation in fact table
Hello, in our data warehouse, we have a fact table containing about 100 million rows per month joined to a customer dimension containing about 10 million customers. The customer dimension contains 30 attributes. Two of these attributes are very

Same attribute in multiple dimensions or Create new dimension?
Hi all - I'm struggling with this concept. I have 2 dimensions: Customer and Products. Both share a common single attribute that I'll just call "Owner" for now. If a customer purchased a product, the owner would be the same for both. I

Modeling Descriptive Flex Fields (Attribute Columns) on an Order Line
For an order line that contains 15 generic attribute columns, how would one model all the miscellaneous data elements (i.e Vendor SKU, Customer Part Number, RMA Product Condition, etc) to make it available to end users as part of the order fact

Multivalue attributes
I am working on a model for use by the fund raising unit of our university. They want to be able to do some CRM type reporting on our alumni / donor data. In creating a constituent dimension I have found that there are a large number of attributes tha

Multi valued dimension attributes, weighting factor and cognos
Hi, we have several attributes in our product and customer dimension which might by multi valued. We modelled these as multi valued dimension attributes using bridge tables and weighting factors therein. Can anybody reference a best practice

Flexiblity To Add Attributes To A Dimension And Not Process The Whole Cube/s
Hello, We are in the process of designing a DW for Games and have run into a business problem : Business users want to add attributes to a dimension and decide its level in the hierarchy on the fly. The new attributes should be availabe in reports

Dimension Attribute that has 13.5million members
The DW and cubes I design all use the Kimball method which I find works really well for me. I'm using SSSAS 2005 standard. However, I have come across a problem with a new cube I have just completed. It's quite basic in that there is 1 fact table and

ETL as a source for some attributes of a dimension
Hi all, is it allowed to programmatically add value on a row of a dimension by the etl process ? For instance : Source contains only books : NK | NAME 142 | invisible Etl source the DW dimension 'Product' and programmatically add a category :

attributes depending on more than 1 dimension (2 dimensions)
suppose you have - a discount code attribute - a status attribute which depend on the combination of the dimension customer & the dimension product. You want to be able to relate these attributes to a fact table invoice details. What are the

Dimension with multivalue attributes
I can't seem to wrap my brain around this one, need some help. Data model is simplified for the example given here: fact: Interview dimension: applicant The grain is obviously interviews, however the users want to be able to report interview

Adding new attributes to a dimension
We're an Oracle 10g warehouse site and still get nervous when adding new attributes to our ageing dimensions. The issue stems from not being able to reorder columns in Oracle. So your new attributes either sit at the bottom of the table (not good) or yo

Transitive Dependency
There are many dimensional tables in our organzation has got transitive dependency in them and we are forced to adopt different SCD types like type 1 and type 2 I will explain my scenario with an example(see table below) ,the Item dimension has got