merging datafiles with string variables of different length

[Frans Marcelissen-3]

Hello,
I had an annoying problem with merging datafiles/adding cases with "add files" (spss 17.0). Because both files were imported from excel, the string variables with the same name in both files have different length. In that case "add file" is fails, the variables in both files must must be exactly of the same type. So one solution is to use "alter type" to make both the samen, but if there are hundreds of variables....
One simple solution would be to use "apply dictionary" to synchronise oth, but this also fails if one variable in the dictionary file has shorter length than the target variable.
Of course it is possible to write a python program for this problemen, but it would be much more simple if add file would have an option to ignore unequal length. Or is there another solution?
 
Actually I have to merge about 10 files. A solution for the problem is build a alter type statement and to repeat that for each file. Then another problem arises: not all variables are present in each file. So a "alter type" statement with a long list of variables fails if one variable does not exist. Is there a solution for this?????

---

Haal meer uit je Hotmail met Internet Explorer 8. Download nu

[Oliver, Richard]

Use the ALL keyword, as in:

 

ALTER TYPE ALL (A=A200).

 

Specify a string length that is sufficient for all string variables in the file.

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of FransiePansie Kevertje
Sent: Tuesday, June 16, 2009 10:06 AM
To: [hidden email]
Subject: merging datafiles with string variables of different length

 

Hello,
I had an annoying problem with merging datafiles/adding cases with "add files" (spss 17.0). Because both files were imported from excel, the string variables with the same name in both files have different length. In that case "add file" is fails, the variables in both files must must be exactly of the same type. So one solution is to use "alter type" to make both the samen, but if there are hundreds of variables....
One simple solution would be to use "apply dictionary" to synchronise oth, but this also fails if one variable in the dictionary file has shorter length than the target variable.
Of course it is possible to write a python program for this problemen, but it would be much more simple if add file would have an option to ignore unequal length. Or is there another solution?
 
Actually I have to merge about 10 files. A solution for the problem is build a alter type statement and to repeat that for each file. Then another problem arises: not all variables are present in each file. So a "alter type" statement with a long list of variables fails if one variable does not exist. Is there a solution for this?????

---

Haal meer uit je Hotmail met Internet Explorer 8. Download nu