Compare 2 Data Files

* This is a general routine for comparing 2 files, for example pre- and post- cleaning; 
* or data entry by different operators; or partial data delivery compared to final delivery and so on.

* For each variable it lists all the differences.
* It produces similar output to the stata routine cf3.
* cheers Simon

/* Compare two files and lists all differences */ 
/* Simon Freidin 09/05/2006 */ 
/* Cases are matched by ID */ 
/* Output is variable by variable. */ 
/* Typical output: */ 
/* ID var1_old var1_new */ 
/* */ 
/* 32747 3 2 */ 
/* 32676 4 5 */ 
/* 36472 2 3 */ 
/* 47699 5 4 */

set printback=no mprint=no length=none.

file handle oldfile /name='substitute path and filename'. 
file handle newfile /name='substitute path and filename'.

define checkvar (!pos=!tokens(1)). 
match files file=oldfile /rename= (!1=!concat(!1,'_old')) /in=inold 
  /file=newfile /rename= (!1=!concat(!1,'_new')) /in=innew 
  /keep=!concat(!1,'_old') !concat(!1,'_new') ID 
  /by ID. 
sel if inold and innew. 
/* select cases in both files */ 
display labels var=!concat(!1,'_new'). 
sel if !concat(!1,'_old') ne !concat(!1,'_new'). 
list /ID !concat(!1,'_old') !concat(!1,'_new'). 
!enddefine.

/* drop ID and variables not in new file here */ 
/* drop any other variables you don't want to compare */ 
get file=oldfile /drop= ID . 
n of cases 1. 
oms /destination viewer=no. 
flip. 
omsend. 
write outfile="c:\temp\checkvar.sps"/"checkvar " case_lbl ".". 
exe. 
include "c:\temp\checkvar.sps".