Indicator Interactions

Overview

Indicator interactions present dashboards with which a dashboard user can interact, enabling the user to explore data directly from within the dashboard.
When you add two or more indicators to a dashboard, you can choose to have one indicator interact with one or more other indicators. The first indicator (the source) influences a second indicator (the target). You specify this influence by mapping a source parameter to a target parameter.
Note: You can specify more than one target parameter.

Example

Here is an example of a dashboard with two indicators. There is no interaction set up between them, so as a dashboard user explores data in one indicator, the other indicator remains static.
two indicators (without interactions) in a dashboard
A dashboard administrator then modifies the indicators in the dashboard designer. The dashboard administrator sets up an interaction so that the source parameter Make in the Average City MPG indicator is mapped to the target parameter Make in the Car Detail Table indicator, and then selects the brush (highlight) interaction.
When a dashboard user clicks a car make in the Average City MPG indicator, here is what happens:
  1. The parameter Make in the Average City MPG indicator is set to the car make that the dashboard user clicked.
  2. The parameter Make in the Car Detail Table indicator is also set to the car make that the dashboard user clicked, because the parameter Make in the Average City MPG indicator is mapped to the parameter Make in the Car Detail Table indicator.
  3. The interaction is then processed, which results in the car make being brushed (highlighted) in the Car Detail Table indicator.
In the dashboard designer, this indicator interaction is shown by an arrow from the Average City MPG indicator to the Car Detail Table indicator
set up of the interaction between two indicators
To display this arrow, the dashboard administrator selects the Show interactions check box from the dashboard's Properties pane.
Show interactions check box
Later, as a dashboard user explores data in the modified Average City MPG indicator, the car make clicked by the user highlights the car make data in the Car Detail Table indicator.
highlighted data in the target indicator

Parameter Mapping

For maximum flexibility, the source parameter and the target parameter are not required to have the same name; they merely need to reference the same type of data. For example, the source indicator can have a data column named REGION_ID and the target indicator can have similar data contained in a data column named SALES_REGION_ID. By mapping REGION_ID to SALES_REGION_ID, you link together the two data columns.
Further, the data can be contained in different data sources and different types of data sources. For example, a source indicator can have indicator data from an SQL query that maps to a target indicator that has indicator data from a stored process.
There are three types of interactions:
  • brush
  • client-side filter
  • server-side filter
Note: If you define an indicator with both a hyperlink and an interaction, the hyperlink takes precedence over the interaction when the hyperlink opens in the same window as the interaction. For example, if you create a dashboard that opens a report and that interacts with another indicator, when the dashboard user clicks the indicator, the report opens in the current window. If the hyperlink opens in a new window, then the dashboard user sees the interaction in the current window and the hyperlink in a new window.

Brush Interaction

A brush interaction specifies that when the dashboard user selects a specific value in the source indicator, related values in the target indicator are highlighted.
Note: You can specify more than one target indicator.
Any type of indicator data can be used for a brush interaction.
All display types can be the source for a brush interaction, except dynamic prompt, dynamic text, and the interactive displays.
All display types can be the target for a brush interaction, except KPI, dynamic prompt, dynamic text, and the interactive displays.

Client-Side Filter

A client-side filter interaction specifies that when the dashboard user selects a specific value in the source indicator, data in the target indicator is filtered, based on the selection.
Any type of indicator data can be used for a client-side filter interaction.
Note: Do not specify a default value for the mapped parameters that severely limits the amount of data returned by the indicator data. This limited data, when filtered further by the client-side filter, can reduce the amount of data to few, or no, matching rows.
All display types can be the source for a client-side filter interaction, except dynamic text and the interactive displays.
All display types can be the target for a client-side filter interaction, except KPI and the interactive displays.
A client-side filter operates on data that is stored on the dashboard user's computer, so a client-side filter can operate against any target parameter, and a client-side filter can be very responsive. However, because a client-side filter requires that a target indicator first load all data on to the dashboard user's computer, load times might be long, and there is an upper limit to how much data can be used.

Server-Side Filter

Overview

A server-side filter is conceptually the same as a client-side filter, except that the parameters are put into the query and the query is sent to the middle tier (the server) with the new parameters. Generally, a server-side filter is not as responsive as a client-side filter because of the time required to communicate with the middle tier, and the target parameters can be only parameters that are understood by the query.
Unlike a client-side filter, which can be based on a target parameter that does not have a prompt defined in the target indicator data, a server-side filter can be based only on a target parameter that is defined in the target indicator data.
The only indicator data that can be used with a server-side filter is an information map or a stored process. In both cases, you must supply a default value for a prompt.
All display types can be the source for a server-side filter interaction, except dynamic text and the interactive displays.
All display types can be the target for a server-side filter interaction, except dynamic text and the interactive displays. The only indicator data that can be the target for a server-side filter is a relational information map and a stored process. In both cases, you must supply a default value for a prompt. An information map must also have at least one filter defined, and this filter must be a mapped parameter.
CAUTION:
Do not change the default value of a prompt in another application while SAS BI Dashboard is running.
Log off from SAS BI Dashboard, change the value, and then log in again.

Date Formats in Information Maps

When using a prompt based on a date with an information map, ensure that you pass values to the prompt in a format that the prompt accepts.
Here are guidelines for each type of date:
  • DAY
    You can pass values such as August 04, 2010, Aug 04, 2010, or 04AUG2010. No other format is accepted.
    Note: The easiest way to ensure that the proper date format is accepted is to use a prompt with a format of DATE9 or MMDDYY10.
  • WEEK
    You can pass a value such as Week 31 2010.
    Note: There is no SAS format that can create such as value. You have to create a custom data set or use SQL to create a data column with the correct format. Then, pass this data column as the source in the interaction or link.
  • MONTH
    You can pass a value such as Jan 2009. No other format is accepted.
    Note: There is no SAS format that can create such as value. You have to create a custom data set or use SQL to create a data column with the correct format. Then, pass this data column as the source in the interaction or link.
  • QUARTER
    You can pass a value such as 3rd quarter 2010.
    Note: There is no SAS format that can create such as value. You have to create a custom data set or use SQL to create a data column with the correct format. Then, pass this data column as the source in the interaction or link.
  • YEAR
    You can pass a value such as 2010.

Chaining Filters

You can chain filters, such as indicator A that interacts with indicator B, which in turn interacts with indicator C.
Note: Although SAS BI Dashboard does not prevent you from defining a recursive filter interaction, it is not recommended. A recursive filter interaction might not provide the user with the expected result and might make further filtering of the data impossible.