***********************************
*          Dr. Baokun Li          *
*      Creating New Variables     *
*         June 28th, 2007         *
***********************************;

*定義３個新變量  ;
DATA tdata;
INFILE  'd:\data\tomhs4.dat' ;
INPUT ptid $ educ sbp12;

* 這樣的話，缺失值被變換為2;
if educ < 7 then grad1 = 2 ; else
if educ >=7 then grad1 = 1 ;

* 下面的兩種方法結果一樣且是正確的;
if educ < 7 and educ ne . then grad2 = 2; else
if educ >=7 then grad2 = 1;

* IN 是SAS的一個特殊函數;
if educ IN(1,2,3,4,5,6) then grad3 = 2; else
if educ IN(7,8,9) then grad3 = 1;

*象下面的新變量定義，如果發生錯誤就慘了 ;
if sbp12 = .   then sbp12c = . ; else
if sbp12 < 120 then sbp12c = 1 ; else
if sbp12 < 140 then sbp12c = 2 ; else
if sbp12 >=140 then sbp12c = 3 ;

run;

PROC FORMAT;
 VALUE grad   1-6 = '< College Graduate' 
              7-9 = 'College Graduate';

 VALUE sbpcat LOW - 119 = '< 120' 
              120-139 = '120-139' 
              140-HIGH = '140+' ;
run;
*Proc Freq 是常用的檢查新變量是否定義得對的方法   ;

PROC FREQ DATA=tdata; 
 TABLES grad1 grad2 grad3 educ ;
 FORMAT educ grad.;
RUN;

PROC FREQ DATA=tdata; 
 TABLES sbp12c sbp12;
 FORMAT sbp12 sbpcat.;
RUN;


quit;