|Date: ||Tue, 28 Jan 1997 09:00:00 PST|
|Reply-To: ||Melvin Klassen <KLASSEN@UVVM.UVIC.CA>|
|Sender: ||"SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>|
|From: ||Melvin Klassen <KLASSEN@UVVM.UVIC.CA>|
|Subject: ||Re: SPSS and defining varaibles|
Mike Lane <mlane@NYIQ.NET> writes:
>The problem I'm having is one of data entry. I recently got some rather
>voluminous ASCII data files from the computer center where I work, and
>they have presented me with a problem. All the files are of the same
>format and layout, and all need the same column definitions to be brought
>into SPSS. I figured out how to replicate that part of the process using
>the paste command in the ascii import screen.
>What I cannot figure out is how to run or copy the same variable labels
>for each file without retyping them for each column of data for each
>file. It seems there MUST be a way to do this as it is so incredibly
>tedious to do it manually.
Back in the "dark-ages", i.e., when mainframes had less power than you
have on your desktop, SPSS users had to know the SPSS "syntax",
and they knew how to code the 'ADD FILES' command.
So, create a "reference" dataset, with all the variables defined,
and labelled correctly, and store exactly *ONE* observation of data
(all-missing-values if you wish) and save this as a system-file.
Then, for each "actual" data-file, read the data, and then use 'ADD FILES' to
merge the "reference" system-file with each data-file, and save the "active"
file as a system-file. As a result of this "concatenation" of observations,
the variable-labels and the value-labels from the "reference" file will be
retained, along with the "actual" data. You can even delete the "dummy" case.
Melvin (reaching for the Geritol) Klassen