The Joy of SAS Enterprise Guide

SAS Enterprise Guide is hard to describe, difficult to pigeonhole, a very powerful tool that is the “Excel” of SAS, and the closest thing there is to an Integrated Development Environment (IDE) for SAS.   Since the release of SAS 9, Enterprise Guide is the new face of SAS for general purpose use in a variety of situations.   Enterprise Guide is now included with desktop SAS for Windows, sold with almost every BI Server sale, bundled with many of the SAS Solutions, and loved by users at many companies I have visited.   It does have shortcomings, but it is  capable of a breadth of  work that no other tool is capable of performing (at least that I have ever seen!)   Simply stated, it leverages many of the old and new strengths of SAS in an easy to use  desktop application.

I will venture to  offer a brief overview of the capabilities  with SAS Enterprise Guide:

1) Easy access to import data from almost any source accessible via your PC: Excel, Access, CSV, ODBC, OLE DB, and even Exchange mailbox data can be imported to SAS with a few simple clicks in Enterprise Guide.

2) Enterprise Guide provides users with very powerful data management capabilities.   Frequently used data management capabilities include a sophisticated query builder, data sampling, ranking, transposing, and even creating and editing data.   I have seen many customers use Enterprise Guide as a “poor person’s” data warehousing toolbox, but I typically discourage such practices.   I do encourage the use of Enterprise Guide as a data staging ground for reporting and analytics.   SAS also encourages this by bundling it with Enterprise Miner (it also can access scoring models from Enterprise Miner, but that falls under analytics below!)

3) Data reporting was the weakest link in Enterprise Guide prior to SAS 9 and EG 4.1.   However,  EG now leverages the new SAS Report Model and has some pretty nice capabilities in the report authoring area.   Reporting tasks can vary from simple listings, data summaries, and sophisticated pivot table like functionality via the Summary Tables task.   More important, with EG 4.1 you can format reports via a point and click interface and maintain all of your formatting when you rerun the project at a later time.   Finally, you can combine multiple pieces of your project output (graphs, summary tables, and analytics) into one report and publish it for use on the web or in Microsoft Office!

4) Graphics- while SAS has many basic and sophisticated graphics available to users, it was traditionally very difficult to get what you wanted without wading through a thousand page manual.   Graphs that I used to spend hours or days writing in SAS code can now be made in minutes with Enterprise Guide.   Every chart comes with a multitude of options to tweak it just so.   A new experimental download with EG 4.1 add the ability to create and explore data interactively through charts.   While not as sophisticated as products like Tableau, it has the advantage of accessing and aggregating the data on the SAS Server, allowing this experimental task to use millions or even billions of rows of data as the input to the charts.

5) Analytics- point and click your way through basic statistics like t-tests, regression, and correlation to advanced methods for forecasting, quality control, and data mining.   Most of these tasks also bundle in nice diagnostic graphs to understand the analysis and present it to non-technical audiences.   Logistic regression in EG is a poor persons data mining (notice a trend here?) but Enterprise Miner is the complete package when it comes to data mining.   Forecasting in EG is very useful and powerful, but SAS Forecast Server is the complete package (and it integrates very nicely with Microsoft Office via the SAS Add-In for Microsoft Office.)

6) SAS Programming- if you are an experienced SAS programmer, you can combine the power of your experience with the point and click capabilities of EG in areas that you might be unfamiliar with.   In fact, you can use EG and never access the wizards provided in it, but why would you limit yourself?   I have often said that becoming an expert at SAS programming takes around 5 years, but mastering EG takes about 1-2 years.   More important, EG allows new programmers to learn the SAS syntax to complex areas (like graphs and procedures like Mixed) in short order.   Programming in SAS continues to be relevant, but focuses on advanced areas like data step, macros, and data access optimization.

I have personally visited many large companies that have a similar story of Enterprise Guide adoption and widespread growth of SAS usage.   In the “old days”, with just a handful of SAS programmers, SAS was perceived as the province of the SAS “priesthood.”   Since it is hard to find and/or train SAS programmers, someone brought in Enterprise Guide typically for people experienced with Excel or competitor’s business intelligence tools- such as Brio, Cognos, or Crystal Reports.  

These people were frequently seeking more power than their tools provided, but wanted a  quick learning curve- faster than the path to  becoming an expert SAS programmer.   After a year or two of SAS and Enterprise Guide, other business teams were very interested in accessing the power of SAS, so they would add Enterprise Guide to their team.   Eventually, someone would buy the BI Server to get their Excel or PowerPoint junkies connected with SAS via the Add-In for Microsoft Office.   Work in EG can be easily published to these Add-In for Microsoft Office users.   In just a few years the number of SAS users would skyrocket from 10 or 20 to 500 or even 5,000!

More Joy of Enterprise Guide  is forthcoming  in part 2…   🙂