REMINDER: The next meeting of the Michigan SAS Users Group will be held on Thursday, November 6 from 2:00-4:30 PM at Team Detroit in Dearborn, MI. There is no charge for attending this meeting! Here are the meeting details:
Date: Thursday, November 6, 2008 Time: 2:00 - 4:30 PM Place: Team Detroit 550 Town Center Drive Dearborn, MI 48126-2755 Cost: Free! Agenda:Using SAS® to Control and Automate a Multi SAS Program Process- Patrick Halpin, dunnhumby USA
Often times a project is comprised of many SAS® programs that need to run in a given order. Additionally, the SAS programs are dependent on previous SAS programs completing. One way to accomplish and automate this task is to use "Done" files and the SLEEP command in SAS. "Done" files are files that are created when a SAS program finishes. SAS programs dependent on previous jobs finishing look for those "Done" files to be created. Once the "Done" files are created the next SAS job in the process will start.
Using SAS to Assess Normality and Find the Optimal Transform- Brandy Sinco, University of MichiganMy presentation will cover the user-friendly features of Proc Univariate to assess normality: normality diagnostic statistics, histogram with normal curve super-imposed, and the qq-plot. I will also discuss whether to standardize the data, how SAS computes skewness and kurtosis, and interpretation of q-q plots. Then, I will show how to find the optimum transform for normality by using the Box-Cox feature in Proc TransReg.
One-Step Change from Baseline Calculations and other DOW-Loop Tricks- Nancy Brucken, i3 StatprobeChange from baseline is a common measure of safety and/or efficacy in clinical trials. The traditional way of calculating changes from baseline in a vertically-structured dataset requires multiple DATA steps, and thus several passes through the data. This paper demonstrates how change from baseline calculations can be performed with a single pass through the data, through use of the Dorfman-Whitlock DO- (DOW-) Loop. It also looks at how similar logic may be used to overcome a limitation of PROC TRANSPOSE, and create a single record per BY statement in an output dataset using a single pass through the input dataset when transposing multiple variables simultaneously.
We hope to see you there!
