Hierarchy Structures

Typical hierarchy structure

A typical hierarchy is organized in a tree-like structure with one value at the top of the tree and multiple values branching out from the top.

The following illustration shows a typical hierarchy.

Parent and child values

The relationships between the various values are referred to as though they were part of a family tree. Each dimension value in a hierarchy (except the top-most value) has a parent value, which is the value directly above it in the structure. The dimension values directly below a parent are called its children.

Sibling values

Each parent value can have any number of children. Dimension values that have the same parent are called siblings.

The following illustration shows siblings in a hierarchy.

Descendant and ancestor values

Each child value can also have children of its own, and so on. Multiple levels of dimension values that roll up to a common value are called descendants of that value, and the value itself is known as the ancestor of the descendent values.

Leaf values

In a hierarchy, any dimension value that has no children is called a leaf dimension value.

Leaf dimension values are the only dimension values for which you can input data for a hierarchy. The data for each non-leaf dimension value is consolidated from its children.

Example: Hierarchy structure

This example illustrates a sample hierarchy for the Organization dimension for a fictitious company, US Global Computers. In this diagram:

• All of the divisions (East, Central, West) at the bottom of the hierarchy are leaf organizations, as are the values Support and Training.

• Data for the Support, Consulting and Training divisions will roll up into Sales data, and Resource, Manufacturing, Sales, and Finance data will roll up into corporate data for US Global Computers.

The following diagram illustrates a sample hierarchy that is based on the Organization dimension.

Using dimension values in multiple hierarchies

Financial Analyzer supports multiple hierarchies. This means that you can specify any number of hierarchies for rolling up data. For example, you might want to define separate hierarchies for examining management and functional data.

In most hierarchies, any dimension value that is not the top-most can have only one parent. However, if that dimension value belongs to more than one hierarchy, it can have a different parent in each hierarchy.

Example: Using dimension values in multiple hierarchies

In this example, there are two hierarchies, each representing a different type of computer system.

• The Sentinel Standard hierarchy contains the dimension values Monitor and System Software

• The Sentinel Multimedia hierarchy contains the dimension value Fax/Modem, as well as Monitor and System Software

  
  

  In the Selector, these would appear as separate product hierarchies that you can choose from to use in a report or worksheet:

  
  

Multiple top-level parents in the same hierarchy

You can create a hierarchy that has more than one top-level parent. Specifying a hierarchy with more than one top-level value can be useful when working with time-related dimensions.

Example: Multiple top-level parents in the same hierarchy

In this example, 1999 and 2000 are top level parents in the Application Time hierarchy.

In the Selector, 1999 and 2000 appear as top level parents in the Application Time hierarchy. Notice that these values are siblings, and there are no higher level values in the hierarchy.