Date Dimension

In (almost) every Data Warehouse you will have a Date dimension. 

As the content of this dimension is predictable, you will usually generate it instead of loading it from some source. KNIME has the 'Create Date&Time Range' node just for this.

One thing you have to decide beforehand is up to what 'level' you will need to create the Date dimension. When you look at our Airline example data, you will see that in the source data we have Year, Quarter and Month, so the lowest level is Month. This is called the 'grain' of our Fact table. So we will create our Date dimension down to this level.

Also, as seen in previous examples, we will be adding a Surrogate key to join to our Facts table.

Below you have a screenshot of the Workflow and of the resulting table SCD_date:

Remember, in order to use this dimension, we have to modify the loading of our Facts table. We will see this on the next page