Database Reference
In-Depth Information
Junk Dimensions
You may discover that you end up with a number of dimension tables that hold nothing more than the key
column, name column and just a few rows. In a situation where there are a large number of these small tables,
your data warehouse can become quite cluttered. One way to simplify things is to create a junk dimension.
Although the name sounds a bit silly, a junk dimension can be quite useful.
With a junk dimension, you can bind a bunch of small dimensional tables into a single table. By taking the
combination of possible values from these smaller tables and storing the combined values in a single table, you
create a junk dimension table.
For example, if the values of one dimension table were “go to lunch” and “go home” and another dimension
table had the values “yes” and “no,” then the possible combinations would be as follows:
•
Go to Lunch, Yes
•
Go to Lunch, No
•
Go Home, Yes
•
Go Home, No
If you build a junk dimension table to hold these values, you will need an ID to link to the fact table and,
in this case, two other columns to hold a combination of values. Figure
4-17
shows an example of taking two
Figure 4-17.
Converting to a junk dimension
