Database Reference
In-Depth Information
Understanding dimensions
After the data source view has been created, it is time to really get into the multidi-
mensional modeling by creating the first dimension. This is the part where you really
need to start thinking about how the data will look to an end user. Up until this task,
nothing that you have done will actually be seen directly by the end user when brows-
ing the cube. But the dimensions that you create are fundamental to the usability of
your cube.
The dimension is used when slicing the measures in the fact table. In the example
that we are using here, we have the five dimensions with different attributes on them.
When we add dimensions to the database, they can be shared by cubes in the data-
base. Dimensions are objects that belong to the Analysis Services database and you
can and should reuse them if you have several cubes or fact tables in your data ware-
house.
A dimension can also be connected to the fact table several times, this is called a
role-playing dimension. You have already worked with such a dimension and it is the
Date
dimension.
A dimension consists of attributes, all dimensions need at least one and that is the
attribute that joins it to the fact table. This attribute is called the key attribute. In addi-
tion to this attribute, other attributes can also be defined. Consider the
Date
dimen-
sion, in this case the attribute that joins it to the fact is the
date
. The
Date
dimension
also contains other attributes such as
year
,
quarter
,
month
, and
week
. An attrib-
ute contains the individual attribute members. Again consider the
Date
dimension,
the key attribute is
date
and the attribute member is the individual date, for the
year
attribute the individual year is the member.
Tip
Moreinformation about dimensionscanbefound at
http://msdn.microsoft.com/en-
Information about attributes and hierarchies can be found at
ht-
Search WWH ::
Custom Search