Extending the Power to Know - Update on the SAS and ESRI Integration

Yesterday we were given a briefing by Jim Metcalf, director of SAS technology platform product management.He provided an overview of the GIS capability inside the SAS BI Server and the SAS BI Enterprise (EBI) Server.These two products came on the market March 1, and were launched at the SAS user group meeting on April 11.Metcalf first highlighted that SAS has had a relationship with ESRI since 2002, when the SAS Bridge for ESRI was released.This new capability is also based on the very solid ESRI relationship.

The SAS Web OLAP Viewer for JAVA 2.1 that resides within SAS' BI servers, is GIS enabled.In October of this year, other viewers SAS uses will also be GIS enabled.The company will also release a public API that allows users to build their own applications based on this integration.Currently the company is marketing this capability to SAS9 users.

This new integrated functionality is web-based with a zero-download (e.g.you don't need a plug-in) client.The basic architecture looks like this.

Source: SAS Institute.(Click for larger image)

A primary part of the integration is the ability to create a tiered functionality for various levels and skill sets within the enterprise. While SAS has had the "bridge" to ESRI since 2002, this new combination allows for almost seamless integration and bi-directional movement of data from the SAS Business Intelligence environment to the ESRI platform.From the user's perspective, this is done through the SAS Web OLAP Viewer for JAVA 2.1, with the ties to the two servers.

Key functionality from within the SAS Viewer includes the following.

Web-based zero-download client that delivers OLAP query and reporting (e.g.a "map-enabled cube")
Integrates with SAS BI and EBI Servers
Associates cube structures with map layers
GIS-enabled through integration with ESRI's ArcGIS 9.0
Synchronizes OLAP and GIS map data exploration
Pan and zoom
Drill-up/drill-down
Expands/collapses
Select and identify region
Filter by value, filter by region
Filter and rank
Create calculated measures
Bookmark-save and return

The maps and SAS table can be seen on the same user interface, which is slick.However, there are several features that are also very slick.First is a tile map showing the concentrations of data.This allows the user to see, in a grid format, the concentrations of data in terms of grid sizes from the base geography.For example, if you have data that span a state, the grid will represent only that state and determine the size of the individual cell in the grid by the data value.If you have more discriminate data within a state, such as county, or ZIP, the state will have sub-cells that are sized relative to the data value.

Source: SAS Institute.(Click for larger image)

A second representation of data that is so obvious that it has been almost totally overlooked by others is the thematic cells of a data table.This is a very cool way to represent data in a thematic table using the same symbology as on a map.

Source: SAS Institute.(Click for larger image)

The ability to move seamlessly from data, statistically processed data and visualization both in maps and in other graphics, moves closer to the goal of extending the insight to what is actually happening in the enterprise.As these tools and some of the other things that SAS intends to bring into play appear, they will definitely extend the power to know.The objective of this integration is clearly to form a bi-directional link between BI and spatial data.