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

Re: st: rescaling a variable by the maximum value question

 From Nick Cox To statalist@hsphsun2.harvard.edu Subject Re: st: rescaling a variable by the maximum value question Date Sat, 13 Oct 2012 10:23:40 +0100

```Justina is correct. You can create a new variable, but there is no
need to do that.

su age, meanonly
gen age2 = age / r(max)

-su, meanonly- is more efficient when you just want a maximum. (The
name -meanonly- is not indicative here.) That's unimportant in one-off
cases, but a detail to note if doing this repeatedly.

SJ-7-3  st0135  . . . . . . . . . . . Stata tip 50: Efficient use of summarize
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
Q3/07   SJ 7(3):438--439                                 (no commands)
tip on using the meanonly option of summarize to more
quickly determine the mean and other statistics of a
large dataset

What Jay is missing is that -max()- will in his context not return a
scalar that is the maximum of a variable. In fact  Stata functions
typically do not that, return a scalar that summarizes a variable.
-max()- wants two or more arguments and will return the result of
comparing them, as it were rowwise. If -max(x, y)- were calculated
from two variables -x- and -y-, then there is scope for a new variable
to be produced.

Nick

On Sat, Oct 13, 2012 at 9:02 AM, Justina Fischer <JAVFischer@gmx.de> wrote:
> you can generate a variable that takes on the maximum value
>
> e.g.
>
>  su age
> gen var = `r(max)'
>
> gen newvar = oldvar/var
>
>

"JVerkuilen (Gmail)" <jvverkuilen@gmail.com>

>> I'm sure I'm simply overlooking something easy but I want to make a
>> new variable that is rescaled by dividing by the maximum value of the
>> old one. Old variable: 1, 2, 3, 4. New variable should be 0.25, 0.5,
>> 0.75, 1.
>>
>> The syntax
>>
>>      gen newvar = oldvar/max(oldvar)
>>
>> does not work and I understand why it doesn't, because max doesn't
>> work that way. Is there a straightforward way to do this? (Yes I could
>> simply
>>
>>      summarize oldvar
>>
>> and pick out the max by hand and rescale. I'd like to avoid that, if
>> possible. ;)
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/
```