Defining Analysis Services Dimensions - SQL Server Analysis Services 2012 Cube Development - page 57

ively, you could also supply MDX to find the last month for which the sales amount is
greater than zero.
The IsAggregatable property controls whether the attribute hierarchy has the All
level. The default value is "true". If you set this attribute's value to false, then the "All"
level will not exist. Furthermore, since there is no "All" level, you should define an
explicit default member to be used by queries which do not explicitly reference this
hierarchy.
The DiscretizationMethod property allows grouping attribute members so that
you don't have to show the full list but rather a few groups that represent the entire
population of members. For example, the customer dimension has the yearly in-
come attribute. As you might guess, each company could have millions of custom-
ers with many different income levels, so breaking down sales by each salary value
would produce an impractical result set. Instead, we need to come up with a few in-
come brackets to identify product sales' trends for each salary range. Of course, you
can define your own way of categorizing income brackets. Alternatively, you could
let Analysis Services do it for you—simply choose one of the predefined discretiz-
ation algorithms: automatic, equal areas or clusters. The DiscretizationBuck-
etCount property lets you specify the number of buckets you would like the cube
users to see while browsing this attribute.
The AttributeHierarchyOrdered property controls whether the hierarchy is
ordered. The default value is true. If a particular attribute is not used for querying
very often, you can save some processing time by setting this property to false; in
this case the OrderBy property is ignored and member values aren't sorted.