/***************************************
 Combining SAS datasets
 filename: Exercise5.sas
****************************************/
OPTIONS FORMCHAR="|----|+|---+=|-/\<>*";
title;
options pageno=1;
options formdlim = " ";

/*Stack datasets to add observations*/
libname sasdata2 "C:\Users\kwelch\Desktop\sasdata2";

title "Tecumseh Wave 1 Data";
proc contents data=sasdata2.wave1 varnum;
run;
title "Tecumseh Wave 2 Data";
proc contents data=sasdata2.wave2 varnum;
run;
title "Tecumseh Wave 3 Data";
proc contents data=sasdata2.wave3 varnum;
run;

title "Tecumseh Wave 1 Data";
proc means data=sasdata2.wave1;
run;
title "Tecumseh Wave 2 Data";
proc means data=sasdata2.wave2;
run;
title "Tecumseh Wave 3 Data";
proc means data=sasdata2.wave3;
run;

title "Tecumseh Wave 1 Data";
proc print data=sasdata2.wave1(obs=25);
run;

data sasdata2.allwaves;
  set sasdata2.wave1 sasdata2.wave2 sasdata2.wave3;
run;

title "All Waves";
proc means data=sasdata2.allwaves;
  class round;
run;

/*Merge datasets to add variables*/
title "Autism Demog Data";
proc contents data=sasdata2.autism_demog varnum;
run;
title "Autism Socialization Data";
proc contents data=sasdata2.autism_socialization varnum;
run;

proc sort data=sasdata2.autism_demog;
  by newinit;
run;
proc sort data=sasdata2.autism_socialization;
  by newinit;
run;

data sasdata2.autism_combined;
   merge sasdata2.autism_demog
         sasdata2.autism_socialization;
   by newinit;
   format bestviq bestnviq 5.0;
run;

data autism_demog;
   set sasdata2.autism_demog;
run;

/*Don't do this!!!!!!!!*/
*data sasdata2.autism_demog;
*run;

/*Note: we actually don't need to sort the dataset
        again, because we sorted it earlier, but it
        won't hurt to include the sort command*/
proc sort data=sasdata2.autism_socialization;
  by newinit;
run;

/*We use the mean function mean(vsae) to get the mean of
     the variable, and the new variable we create is
     called mean_vsae. 
  We use the n function n(vsae) to get the number of values
     for vsae, and the new variable we create is called 
     n_vsae*/
proc means data=sasdata2.autism_socialization noprint;
  by newinit;
  output out=meandata mean(vsae)= vsae_mean
                      mean(vsst)= vsst_mean
                      n(vsae)   = vsae_n
					  n(vsst)   = vsst_n;
run;

/*Note: the summary dataset will have one obs per child,
  and since there are 213 children, will have 213 observations*/
/*Note: The variable _freq_ tells you how many obs
        there were for each child.
        The variable _type_ is zero for all obs and can
        be dropped from your dataset*/

title "Summary Data Set";
proc print data=meandata;
run;