自动生成标准的统计表格(配合ods)的macro（Automatic generation of standard statistical tables (along with ODS) macro）.docVIP

自动生成标准的统计表格(配合ods)的macro（Automatic generation of standard statistical tables (along with ODS) macro） Automatic generation of standard statistical tables (along with ODS) macro [url=mailto:wmqy2004@acrp]wmqy2004@acrp[/url] The program runs through sas9.13. %macro, comdts (J, K, dataname); %do i=j%to k; datanamei %end; %mend comdts; %macro, comparison (data=, vars=, vg1=, vg2=, vg3=, vg4=, class=, par=1, format=6.1, total=1, where=); Should be the name of a variable that contains the / * scanning macro parameter, and assign them to a series of macro variables. ODS select none; %let i=1; %do,%while (%scan (vars, i) ne);%let, vari=%scan (vars, i);%let, i=%eval (i+1);%end; Gets the number of variables should be / * * / %let vnum=%eval (i-1); Access to the specified number of packet level / * * / variable Proc freq data=data; tables class/out=freout; run; Data _null_; set freout end=last; if, last, then, call, symput ( lvnum _n_; compress ()); Run; / * level grouping variable number is less than 2 warning, and out of macro operation process. %if lvnum2%then%do; %put, Warning:, Level, number, of, class, variable, is,, less, than; %goto exitmacro; %end; / * * / data description %do i=1%to vnum; Proc, means, data=data, N, nmiss, mean, STD, min, Max, lclm,, uclm, median, maxdec=2; Output, out=tmpmeani, n=n, nmiss=nmiss, mean=mean, min=min, max=max, lclm=lclm, uclm=uclm, median=median, class, class;%if, where, NE,%then,%do;, std=std; where, where;%end; Var vari; Run; Data tmpmeani; set tmpmeani; Nnmiss=compress (put (n, 6) || ( ||left (put (nmiss, 6)) || )); Ms=put (mean, format) || + ||left (put (STD, format)); Minax=put (min, format) || ~ ||left (put (max, format)); Luclm=put (lclm, format) || ~ ||left (put (uclm, format)); Med=put (median, format); Run; %if lvnum=2%then%do; T-test / * * /; %if par=1%then%do; ODS output ttests=ttests equality=equality; Proc ttest data=data; VAR vari; class class; %if, where, NE,%then,%do; where where;%end; Data test2; merge ttests equality;