Does anyone know where to find the formula SAS uses for calculating the standard error when summing the fixed model intercept and the random intercept effects for each level of the random variable in a mixed model?  It's not sqrt(SE1**2 + SE2**2 - 2*Cov12), not to mention that covariance between the fixed intercept and the random intercepts does not exist--fixed intercept is, well, fixed.   I can get the answer using the ESTIMATE statement but would like to do this utilizing ODS OUTPUT SolutionR dataset.         ESTIMATE 'Desired Effect' Intercept 1 | Intercept 1 / subject 1 cl;   Thanks, Haris ... View more

I know how to embed individual SAS code links in SEG.  Is there a way to tell SEG to embed all SAS code at once?   Conversely, is there a way to Save As all embeded SAS code into separate SAS code snippets?  I know I can save the entire SEG project as one giant SAS code module and I can save each code module as a SAS code.  But can I point SEG to a folder (in my case it's Windows) and have it use the existing module names to create separate SAS code files with *.sas extensions?   What I am trying to do is to import multiple SAS programs into SEG.  These are all parts of the code system used to run annual reports.  I need to make changes to those modules, run the new report based on the previous report's code and then save the updated code to a new location preserving the structure of the previous report.  I really don't want to merge all those multiple SAS code modules into a single giant SAS code module.   Also, I thought there was an OPTION for changing what happens when SAS code is opened in SEG.  By default, external SAS code is linked.  I thought I remembered an option to automatically embed rather than link the code.  Can's seem to find that option in SEG 7.11.  Am I imagning things?   Thanks in advance. ... View more

I was very eager to try LASSO variable selection for LOGISTIC regression but can't seem to find the answer to standardization question.  Variables selected with STEPWISE and LASSO are completely different.  I was under impression that LASSO would select a robust subset of STEPWISE choice but that is not what is happening.  LASSO does not select any dichotomous class variables which in my dataset are all coded 0/1 and tend to have larger raw beta coefficients.  I wonder if it's a matter of scale?   1. Does SAS do that automatically, or do I need to standardize my variables before I run HPGENSELECT? 2. What about polynomials--do they need to be standardized as well? 3. Do predictors need to be centered?   Thank you. ... View more