06-02-2016 03:04 PM
I have a number of variables that were log transformed, and I wanted to use these new variables to do ANOVA.
I first created a macro variable to store all the log transformed variables names and then used a do loop to repeat the anova analysis.
The problem is, my last log variables was not included in the anova analysis. I checked the values of the macro variables "collist" and "n", which were both correct.
What could be the problem?
%macro loop_var(data); proc contents data=&data short out=test; run; proc sql; select name into :collist separated by ' ' from test where name like 'log_%'; %let n=&sqlobs; quit; ods graphics off; %do i=1 %to &n; %let currentvar=%scan(&collist,&i); proc glm data=&data ; class type; model ¤tvar=type; lsmeans type /pdiff stderr cl ADJUST=SCHEFFE; %end; %mend; %loop_var(mydata);
06-02-2016 03:14 PM
What values do you have for collist and n?
%put collist=&collist i=&i currentvar=¤tvar n=&n;
Immediately before your PROC GLM statement. That should help debug.
Need further help from the community? Please ask a new question.