English

Anne-Lindsay Beall 0
WUSS 2011: By the numbers

Ready for some fun facts? These tidbits were all shared during the Western Users of SAS Software (WUSS) conference held in San Francisco this week:    1.    How many million dollars did a simple Excel spreadsheet error cost TransAlta?  Answer: $24 million  2.    What percentage of mobile phone users older

Rick Wicklin 0
Optimizing? Two hints for specifying derivatives

I previously wrote about using SAS/IML for nonlinear optimization, and demonstrated optimization by maximizing a likelihood function. Many well-known optimization algorithms require derivative information during the optimization, including the conjugate gradient method (implemented in the NLPCG subroutine) and the Newton-Raphson method (implemented in the NLPNRA method). You should specify analytic