Dimension Attributes: Introduction and Overview, Part IV - Page 2

Overview of the Attribute Properties

As we noted in previous articles of this subseries, Analysis Services exposes many properties that determine how dimensions and dimension attributes function. We can review the properties for our selected attribute, Geography Key, within our sample UDM, by taking the following steps.

The Properties pane appears for the Geography Key attribute. (The Properties pane likely appeared when we selected the Geography Key attribute dimension within the Dimensions pane, by default, below the Solution Explorer. The design environment can, of course, be customized in many ways to accommodate our local development needs.)

We can, at this stage, see the thirty DimensionAttribute properties for the Geography Key attribute within the Properties pane. We examined the first eleven properties, the members of the Advanced properties group, in Dimension Attributes: Introduction and Overview, Part I. We noted that these properties include the following:

• AttributeHierarchyDisplayFolder
• AttributeHierarchyEnabled
• AttributeHierarchyOptimizedState
• AttributeHierarchyVisible
• DefaultMember
• DiscretizationBucketCount
• DiscretizationMethod
• EstimatedCount
• IsAggregatable
• OrderBy
• OrderByAttribute

The five Basic properties, which appear underneath the Advanced properties group, and which we reviewed in Part II, include the following:

• Description
• ID
• Name
• Type
• Usage

The Misc group, which we explored in Part III, comes next, and includes the following four properties:

• AttributeHierarchyOrdered
• GroupingBehavior
• InstanceSelection
• MemberNamesUnique

Beneath the Misc group in the Properties pane lies the Parent-Child group, which we examine in this article, and which includes the following five properties:

• MembersWithData
• MembersWithDataCaption
• NamingTemplate
• RootMemberIf
• UnaryOperatorColumn

Finally, the five Source properties (which we explore in the next article of this subseries), appearing underneath the Parent-Child properties group, include the following:

• CustomRollupColumn
• CustomRollupPropertiesColumn
• KeyColumns
• NameColumn
• ValueColumn

The Properties pane for the Geography Key attribute, with the Parent-Child properties group (which we will examine in the practice session that follows) expanded, appears as shown in Illustration 6.

Having finished our overview of the Misc attribute properties in Part III, we will examine the attribute properties group that next appears within the Properties pane for Geography Key.

Examine Attribute Properties: Parent-Child Properties

The Parent-Child properties group appears just underneath the Misc properties within the Properties pane. In this section, we will discuss the purpose of each property within the Parent-Child group for our chosen attribute example. As we did for the other properties groups, we will discuss / examine, in most cases, possible settings with which we can come into contact within the context of each property.

Each of the property values within the Parent-Child properties group is relevant only when the Usage property (a member of the Basic properties group we discuss in the section we dedicate to the subject above) for the attribute under consideration is set to Parent, meaning that a parent-child hierarchy has been defined. When the Usage property is set to Regular or Key, this setting is ignored by Analysis Services.

Parent-Child Property: MembersWithData

The MembersWithData property specifies the manner of treatment of fact data that is associated with non-leaf members. In effect, the value selected for the property (there are two options) dictates whether data members for non-leaf members in the parent attribute are displayed.

The two selection options include NonLeafDataHidden and NonLeafDataVisible, as we can see in the illustration above.

Parent-Child Property: MembersWithDataCaption

MembersWithDataCaption affords us a means of specifying, if we desire to do so, a string. The purpose of the string is to serve as a template, to be used by parent attributes, upon which captions, for system-generated data members within the parent attribute, can be based.

Parent-Child Property: NamingTemplate

The NamingTemplate property allows us to specify the manner in which levels are named in a parent-child hierarchy constructed from the parent attribute. This is accomplished via a Level Naming Template we can access at the property level, as we shall see in the next step.

The Level Naming Template dialog appears, as depicted in Illustration 9.

The Level Naming Template affords us a means of specifying the level names displayed to information consumers as they browse the cube that houses the dimension displayed in the dialog box title bar.

Because the sample attribute with which we are working is not a member of a parent-child dimension, the dialog box that we see is not typical. A parent-child dimension always contains a single metadata level (excluding the “All” level, if any), which typically produces multiple displayed levels. Because the number of displayed levels is not always known when the dimension is created or edited, and because this number can change – if and when the data in the dimension table is updated – the Level Naming Template dialog box is made available to allow us to specify the names applied to the displayed levels.

NOTE: For more information on parent-child dimensions, together with more details of the property settings that support these dimensions and their member attributes, see other articles of my Introduction to MSSQL Server Analysis Services series at Database Journal.

Parent-Child Property: RootMemberIf

The value we select for the RootMemberIf property specifies the criteria by which members of the highest level (excluding the “All” level) are identified.

The four selection options include the following:

• ParentIsBlankSelfOrMissing: (Default) Only members that meet one or more of the conditions described for ParentIsBlank, ParentIsSelf, or ParentIsMissing are treated as root members.
• ParentIsBlank: Only members with a null, a zero, or an empty string in the key column or columns are treated as root members.
• ParentIsSelf: Only members with themselves as parents are treated as root members.
• ParentIsMissing: Only members with parents that cannot be found are treated as root members.

The behavior of the RootMemberIf property in determining how the root or topmost members of a parent-child hierarchy are identified, is, therefore, dependent upon which of the selections above is made. The default, as noted above, is ParentIsBlankSelfOrMissing.

Parent-Child Property: UnaryOperatorColumn

The value we supply to UnaryOperatorColumn is used to specify the column within the underlying data source that provides unary operators.

The two selection options that are available are “None” and “New.”

The Object Binding dialog appears. As most of us are well aware, we use the Object Binding dialog box in Business Intelligence Development Studio to define bindings between the property of an Analysis Services object and a table / column in a data source view. The Object Binding dialog box can be called by selecting (new) from the drop-down list for the value of the following properties of an Analysis Services object in the Properties window of Business Intelligence Development Studio:

• NameColumn;
• ValueColumn;
• CustomRollupColumn;
• CustomRollupPropertiesColumn;
• UnaryOperatorColumn (where we currently rest).

Using the Object Binding dialog, we select Binding type, Source table and Source column as appropriate to our needs, and then save our changes. Because we are not dealing with a Parent-Child dimension as our practice example, we will simply dismiss the dialog via the Cancel button, as we have done in similar situations before.

Having completed our review of the Parent-Child attribute properties, we will conclude this part of our examination of attribute properties. We will extend our introductory examination of dimension attributes, specifically continuing our discussion with the Source group of properties, within the next article of this series.

NOTE: Please consider saving the project we have created to this point for use in subsequent related articles of this subseries. Doing so will allow us to avoid the need to repeat the preparation process we have undertaken, initially, to provide a practice environment.

Conclusion

In this, the fourth part of a multi-part article introducing dimension attributes, we continued our current subseries focusing upon dimensional model components, with an objective of discussing the associated concepts, and of providing hands-on exposure to the properties supporting each. We reviewed our initial introduction to the dimensional model and summarized its role in meeting the primary objectives of business intelligence. Next, we provided a brief review of dimension attributes in general.

We overviewed many of the general characteristics and purposes of attributes, including their names, and the names of the groups within which each is classified. We then continued our focus upon the properties underlying them, based upon the examination of a representative attribute within our sample cube. In this article, we extended our discussion beyond the Advanced, Basic and Misc groups of properties, which we began in Part I, and continued in
Part II and Part III, respectively, and examined the attribute properties belonging to the Parent-Child group, including what they define and support, and how we can manage them. We will continue our examination of attribute properties, this time for those that constitute the membership of the Source properties group, in the next part of this article.

» See All Articles by Columnist William E. Pearson, III

Discuss this article in the MSSQL Server 2000 Analysis Services and MDX Topics Forum.