怎么用SAS删除重复的变量？
数据中有一些变量重复值达到90%，想把多余的删掉，怎么操作 %macrovar_namelist(data=,tarvar=,dsor=);%letlib=%upcase(%scan(&data.,1,'.'));
%letdname=%upcase(%scan(&data.,2,'.'));
%globalvar_list var_num;
proc sql ;
create table &dsor.as
select name
from sashelp.VCOLUMN
where left(libname)="&lib."and left(memname)="&dname."and lowcase(name)^=lowcase("&tarvar.");
quit;
%mend;
%macrotest(data,tarvar,data_result,data_drop,rate);
proc datasets lib=work;
delete base;
run;
data base;
length variable$100.;
run;
%var_namelist(data=&data.,/*coltype=num,*/tarvar=&tarvar.,dsor=aa);
data _null_;
set aa;
call symput(compress("var"||left(_n_)),compress(name));
call symput(compress("n"),compress(_n_));
run;
%put&n.;
%doi=1%to&n.;
%put&&var&i.;
proc freq data=&data.(keep=&&var&i.) noprint;
tables &&var&i./out=PERCENT_&&var&i.;
/*(keep=PERCENT)*/
run;
proc sql;
select max(PERCENT) into: max_percent from
PERCENT_&&var&i.;
quit;
%if&max_percent>&rate.%then%do;
data next;
variable="&&var&i.";
run;
proc append base=base data=next force;
run;
%end;
proc datasets lib=work noprint;
delete PERCENT_&&var&i.;
run;
%end;
data base;
set base(where=(variable^=''));
run;
proc transpose data=base out=base1(drop=_name_);
id variable;
run;
/*这步是删除单一变量超过90的重复值的缺失值的可以按照这个写下*/
proc sql noprint;
select name into :var_list separated by' '
from sashelp.VCOLUMN
where upcase(left(libname))="WORK"and UPCASE(left(memname))="BASE1";
quit;
%PUT&var_num1.;
data &data_result.;
set &data.;
drop &var_list.;
run;
data &data_drop.;
set &data.;
keep &tarvar.&var_list.;
run;
%mend;
页:
[1]