Aggregate Dimension Based on Fact?

Date Dimension at Various Grains

Recently our data model work group has been discussing expanding/changing our Date Dimension to more of a time dimension to encompass smart keys to faciliate the Month and Quarterly fact tables, and still use the new time Dimension surrgoate key.

Aggregate value

Hello everyone, I was wondering if it's acceptable to calculate an aggregate_value by aggregating a measurement from the fact_table and by subtracting from a value located in another dimension.

Building aggregate fact tables from staging

Hi, Is it acceptable to create an aggregate fact table from the staging tables after you have cleaned out the data in staging? I know usually you're supposed to create aggregate fact tables from the fact/dimension tables, however can it be done straigh

Why not snowflake if the aggregate dimension is needed for a different fact table?

I am in a bit of a dilemma in that if let's say we have a Product dimension needed for the Sales fact table but our forecast fact table is at the ProductCategory level. The Kimball toolkit suggests to still keep the Product Catergory attributes in the

rollup fact table

If creating a roll up fact table would you consider using something other than the surrogate key for your date dimension join? I would not but my colleagues think we should be using month_id for example if we are creating a monthly roll up table from a

2 fact at different granuality demanding different level of a dimension hierarchy.

Hi, Here is what i am trying to design. There is a transaction level fact table say Fact1.. One of the dimensions for this fact table is Dim1, which has hierarchy as L1>L1>R1.. leaf to root. The Leaf node (L1) of the dimension is Foreign key in the

Overlapping Dates in a dimension

Hi, I am wondering if someone could advise me on how to deal with this design problem. I have a promotion dimension keyed to a sales fact table. Initially, we thought that only one promotion could be active at one time for a sale, but it is actually th

Wide and large Dimension or Survey Factless Fact Table

Hi I would like some advice on this particular point. We receive responses to surveys containing hundreds of questions. Those surveys are submitted many times to about ten thousands people we follow. Their answers could change in time. Usually the

Sparse dimensions, how to handle?

I've purchased the book "The Microsoft Data Warehouse Toolkit" and have reviewed it in order to build out a data warehouse solution I am currently designing. I am stuck in a spot that I don't seem to find any examples of , or am not abl

Modelling Dimension relations over time

Hi, We are planning to model a star schema based data warehouse with the following challenges: 1- The fact table is dimensioned across sales rep postion (a logical designation), geography, product, sales and atime 2- The sales reps have different

Dimensions Advent Calendar.... Ho Ho Ho.

I thought it could be a good idea for the following 24 days, to open together on each day a door of our.... DIMENSIONS ADVENT CALENDAR. So feel free to post and open a door, if you would like too... I will be brave to open the first. I´m

Random strange facts

It is possible to sail all the way around the world at 60 degrees S latitude without hitting any land. The average coal power plant emits 100 times as much radiation as the average nuclear plant. The easternmost point in the United States is in

Regarding Factless Fact Tables

Some time ago I was asked to build the BI platform for a road safety organization. The warehouse was built of three primary fact tables, each with a collection of associated dimensions. I am now trying to branch out into ETL and Warehouse design, and from