Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Scaling of variables

From   Nick Cox <>
Subject   Re: st: Scaling of variables
Date   Tue, 7 Jun 2011 09:15:13 +0100

foreach v of var <put your varlist here> {
     gen `v'_at = `v' / at

A varlist is always a list of variable names, but possibly including
wildcards such as foo? or bar*.


On Tue, Jun 7, 2011 at 9:05 AM, Achmed Aldai <> wrote:
> Hi Maarten
> sorry I have to ask again about the loops:
> What would be my lname and do I use in or of list? I have like 30 variables that have to be scaled by at.
> If I use of what would be the name of my varlist or can I just set a random name for that?
> Concerning the macros, I would do it with ds [varlist], I guess the name from above and then say var/at?
> Thanks a lot for your help
> -------- Original-Nachricht --------
>> Datum: Tue, 7 Jun 2011 09:21:34,
>  +0200
>> Von: Maarten Buis <>
>> An:
>> Betreff: Re: st: Scaling of variables
>> --- On Tue, Jun 7, 2011 at 9:10 AM, Achmed Aldai wrote:
>> > I have to do an analysis with balance sheet data of several firms
>> (gvkey) and want to scale some of the variables by total assets(at). E.g. total
>> liablities, earnings, and so on. Is there a way that I can do this with one
>> command and not do it for each variable since I have around 40 and that
>> would take a lot of time. Still I dont have to scale all of them by total
>> assets. The company (gvkey) and industry specifications for example should stay
>> unchanged.
>> You can loop over variables, see -help foreach-. The trick is than is
>> to create a macro containing the appropriate variables, and feed that
>> macro to -foreach-. For the creation of that macro I often use -ds-.

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index