[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: new reshape command compatibility issue

From   "Martin Weiss" <[email protected]>
To   <[email protected]>
Subject   st: RE: new reshape command compatibility issue
Date   Mon, 25 Aug 2008 17:18:17 +0200

Would a -version- statement not be easier?

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of James Muller
Sent: Monday, August 25, 2008 5:06 PM
To: [email protected]
Subject: st: new reshape command compatibility issue

Hi list,

A new major version of reshape was released, but it imposes a restriction on
the maximum length of variables it can reshape (at least from wide->long).
Specifically, reshape fails on variables with names longer than 20

I realized this when it caused a large (a couple of thousand do-files
involved) system to fail...

There's a quick solution, but it sucks. The old reshape command is still
available as the command reshape_10. The solution is to replace all reshape
command calls with reshape_10. It is a shame 'reshape' is also a word in

Under the hood, reshape obtains details of the variables via the 'char'
system (?), which creates variables describing variables from your dataset,
prefixing them... That is, the dataset variable name plus the prefix can
violate Stata's 32-character variable name limit.

I ask, is there the possibility of thes very important Stata command,
being modified so that it can accept variable named according to the general
convention in Stata?

As I say, there is the option of using reshape_10, but, well, that's a lazy


*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index