A hierarchy filter is a filter type that shows hierarchical data in an expandable tree view. Good examples of data that can be used in a hierarchy filter are date and time information, or geographical data such as continents/countries/cities. Data connections to external data sources often contain hierarchies that are predefined in the external source, which can be analyzed using hierarchy filters. The examples below illustrate two uses of hierarchy filters.
Hierarchy filter from a date column
In this example, a column with various dates is first shown as a check box filter. This column is used to create a hierarchy filter, grouping the dates into Year/Quarter/Month, instantly providing a better overview.
In this case, you go from a linear series of dates that you can only manipulate on the most granular level, to a view that groups the dates from the column into a hierarchy you decide. You can select to show the dates as Year/Quarter/Month or Year/Month/Day or Year/Month/Week/Day or any other configuration you want, when setting up the hierarchy.
Hierarchy filter from different but related columns
With date information, you often only need the data from one single column to create a filter that is hierarchically structured. However, for other data it might not be so easy for Spotfire to interpret the data as hierarchical. In this case you can create a hierarchy filter by combining several columns into a fixed hierarchy. Such data can be geographical data, or perhaps product information such as Market category/Product category/Product.
In this example, we see three separate columns (Continent, Country, and City) originally displayed as three separate filters. These are then combined into one hierarchy filter whose structure is easier to work with.
Note: The separate filters for the three columns will not be removed. However, if you are working with the hierarchy in the Filters panel, you can manually hide them, if you wish. Just right-click, and select Hide filter for each filter you want to hide, or use the Organize Filters dialog to hide multiple filters at once.
In order to combine many columns into a hierarchy filter, there must be an inherent hierarchy in the data. The data in the example above looks like this:
Hierarchy filter from an external data source
When the data in your analysis comes from a data connection to an external data source, there may be predefined, fixed hierarchies available. If the data comes from a cube data source, with large hierarchies containing many different levels and categories, then the hierarchy filter may have a hard time loading all of the possible values. In that case, you can configure the data connection so a certain hierarchy should Load Values on Request only. See Working with Large OLAP Hierarchies in Spotfire for more information on how to do this.
When the connection has been configured this way, the hierarchy filter cannot display the search field that is otherwise available. You may also see progress indications for each category you expand, as new values are fetched from the external source.
Hierarchy filters from external cube sources will also include an (All) check box and have the setting to indicate that a category is only partially selected turned on by default. These settings are available for all hierarchy filters, in the Filtering Scheme Properties dialog.