* This files identifies some of the errors in the 
* longitudinal match between 1982 and 1992 CM datafiles
***********************************************************

options compress=no;
libname in 'D:/cm/';
libname un 'D:/enrico/new/data';
libname on 'D:/enrico/imported/data';

data tmp; set in.merged2;

*numeric-to-character conversion;
ppn_tmp = put(ppn,$10.);
drop ppn;

* plant exists in 1992 but not in 1982;
if county82 =. and county92 ^=. then miss82=1; else miss82=0;
*plant exists in 1982 but not in 1992;
if county82 ^=. and county92 =. then miss92=1; else miss92=0;

* phisical location has changed. This is impossible;
if county82 ^= county92 then err1=1; else err1=0;
if   smsa82 ^=   smsa92 then err2=1; else err2=0;

*smsa code is either missing or 9999;
if smsa82 =9999 or smsa82=. then rural82=1; else rural82=0; 
if smsa92 =9999 or smsa92=. then rural92=1; else rural92=0; 

/*
proc freq; tables miss82*miss92;
proc freq; tables err1*err2;
where (miss82=0 and miss92=0);
proc freq; table rural82*rural92;
where (miss82=0 and miss92=0 and err1=0 and err2=0);
*/
run;

data in.errors; set tmp;
ppn = ppn_tmp;
keep ppn miss82 miss92 err1 err2 rural82 rural92;
proc means;
run;