Bookmark and Share

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


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

RE: st: wildcard in numeric variable


From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   RE: st: wildcard in numeric variable
Date   Fri, 1 Jun 2012 13:11:30 +0100

I am found of 42 jokes and managed to sneak "fortitude" into the Stata Journal. 

(If your reaction is ???, see http://en.wikipedia.org/wiki/42_(number) although it won't do anything for your next paper, program, proposal, project or poem.) 

Nick 
[email protected] 


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Colin Campbell
Sent: 01 June 2012 12:54
To: [email protected]
Subject: Re: st: wildcard in numeric variable

Thanks,

An answer that stops short of 42 but works perfectly nonetheless!

Colin



----- Original Message -----
From: Nick Cox <[email protected]>
To: "'[email protected]'" <[email protected]>
Cc: 
Sent: Friday, 1 June 2012, 13:34
Subject: RE: st: wildcard in numeric variable

The second argument of the -string()- function if you need it should be a _numeric_ display format. That's because you are specifying how a number should be displayed. Try something like 

... if substr(string(numvar, "%20.0f", 1, 4) == "9999"

for "any suitable value of 20". (I think this was a Salomon Bochner joke.) 

Nick 
[email protected] 


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Colin Campbell
Sent: 01 June 2012 12:27
To: [email protected]
Subject: Re: st: wildcard in numeric variable

Thank you Nick,

the solution

... if substr(string(numvar), 1, 4) == "9999"

looks great and nearly works a treat. But, this solution will not replace those with 8 or more digits.
I assume this has to do with the str format %#s but am not sure how to specify this correctly.

I tried the following:

    replace cv =. if substr(string(cv, "%12s"), 1, 4) == "9999"

but this returns "(0 real changes made)". Any other ideas?

Colin



----- Original Message -----
From: Nick Cox <[email protected]>
To: [email protected]
Cc: 
Sent: Friday, 1 June 2012, 10:47
Subject: Re: st: wildcard in numeric variable

Wildcards don't apply to numeric values.

... if substr(string(numvar), 1, 4) == "9999"

is the simplest solution that occurs to me. No need for new variables,
-tostring- or any regex machinery.

Nick

On Fri, Jun 1, 2012 at 9:40 AM, Colin Campbell
<[email protected]> wrote:
>
> Dear Statalist,
>
> I have a dataset in which a numeric variable cv contains values that need to be coded as missing.
> There are a variety of numeric values 9999, 99999, 999999 etc that need to be recoded.
> While I could replace each value individually, I wondered if there is a way to use wildcards with numeric variables such that e.g.
>
>
>     replace numvar =. if numvar == 9999*
>
> would work. I have a workaround using string variables as follows
>
>     tostring cv, gen(cv_str)
>     gen cv_mis = regexm(cv_str, "9999")
>     replace cv =. if cv_mis ==1
>     drop cv_mis cv_str
>
> Is there a more elegant way to achieve this without resorting to strings?

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?searchhttp://www.stata.com/support/statalist/faqhttp://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*  http://www.stata.com/help.cgi?searchhttp://www.stata.com/support/statalist/faqhttp://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?searchhttp://www.stata.com/support/statalist/faqhttp://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


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