For the purposes of this tutorial we will be using the following XTAB program fragment:
*files input xt_data.txt *variables lastname/Last Name/1-10 firstname/First Name/11-20 age/Age/21-22 race/Race/23 sex/Sex/24 county/County/25-26 company/Company/27-36 phone//37-46 ac/Area Code/37-39 ssn/Social Security Number/47-55 salary/Salary/56-60 title/Job Title/61 class/Classification/62
The file xt_data.txt was generated utilizing the perl script xt_dgen.pl. It contains 1000 lines. It is to serve as our demonstration file.
First we will show a couple of file processing capabilities of XTAB:
Print the first 25 lines of the file:
*files input xt_data.txt limit 25 *instructions print data(Data is the default name of the input record.)
Write all Males to a file.
*files input xt_data.txt output file01 /v001/maledemo.txt *variables sex/Sex/24 *instructions select sex eq M write file01 data
Now we will show some cross tabulations.
A simple sex vrs race table:
*files input xt_data.txt *variables sex/Sex/24 race/Race/23 *tables table stub sex codes M,F spread race W,B,#(The # means everything else).
Now lets add some labels:
*files input xt_data.txt *variables sex/Sex/24 race/Race/23 *tables table Sex by Race Breakdown labels Men,Women stub sex codes M,F labels While, Black, Others spread race W,B,#Now lets add the same table, but just with supervisors.
*files input xt_data.txt *variables sex/Sex/24 race/Race/23 class/Classification/62 *tables table Sex by Race Breakdown labels Men,Women stub sex codes M,F labels While, Black, Others spread race W,B,# table Sex by Race Breakdown,,Supervisors select class eq S labels Men,Women stub sex codes M,F labels While, Black, Others spread race W,B,#(We can do multiple tables, of differing stubs and spreads, with differing selection criteria in a given run).
Now lets do a table of companys and salary ranges.*files input xt_data.txt *variables company/Company/27-36 salary/Salary/56-60 *tables table Companys and the Salary ranges they pay: stub company freq 20 spread salary values 0-30000,30001-65000,65001-99999In the above we may not know all the companys on the file so we say show me the first 20 unique ones. Since salary is numeric we use the key word values instead of codes.