I have a 3-way interaction model as follows:   Y = A + B + C + A*B + A*C + B*C + A*B*C   A is a dummy and B and C are centred continuous variables.   I am mainly interested in the parameters for BC and B. The interpretation of these parameters depends on the inclusion of AB*C and all 2-way interactions. I am using SAS and PROC GLMSELECT with selection=none, thus (IMHO) forcing the interactions to be included?    For some data the p-values for the aforementioned interactions are above 0.05. Does that mean they are not ‘included’ and thus the interpretation of the aforementioned parameters changes?   Any feedback would be very much appreciated. ... View more

I tend to use a lot of dynamic (explicit) sql in poc sql. Something like this dynamic WHERE CLAUSE works fine :   WHERE TableAlias.ColumnName IN (%bquote('&MacroVariable.')) However for some reason, this snippet of a SQL query:   SELECT &MacroVariableContainingColumnName. FROM ...   does not work. I get the error message:   Table does not contain MacroVariableContainingColumnName   So the actual content of    (e.g. Column1) is not used.     Just to clarify, I want to generate the dynamic SQL:   select C1   from something along those lines:   %let MacroVariableContainingColumnName = C1; using proc sql; connect to TERADATA (OVERRIDE_RESP_LEN=YES SERVER=DBC AUTHDOMAIN="Bla" MODE=TERADATA); create table X.Y as Select * from connection to teradata ( SELECT &MacroVariableContainingColumnName. ...   Any ideas? ... View more

I can access this stored process on our server:   Sample: Hello World DATA Step−generated output using PUT statements.   I presume this comes with the default SAS implementation. Is anyone aware of a tutorial which describe the underlying source code of this stored procedure? I would like to replicate this stored process. Any pointers would be very much appreciated. Many thanks! ... View more

Let us say I have a dataset like so:   ID X1 X2 1 1 1 2 1 0   Here Id indicates something to be modeled. The 0 and 1 in columns X1 and X2 indicate whether the indepedent varible should be used in the model (0=do not include, 1=include). So the model code to be produced would be:   PROC GLMSELECT data=TrainingData; model DependentVariable = X1 X2 / selection = stepwise(select=SL choose=AICC); run;   and     PROC GLMSELECT data=TrainingData; model DependentVariable = X1 / selection = stepwise(select=SL choose=AICC); run respectivley. Ideally I would like to invoke a macro given the Id. The Id is then used to write and execute the modeling code. Btw the relvantTrainingData given Id can be easily initialised from anothe dataset. Any ideas how to achieve the above? Thanks!     ... View more