Note that this is not an answer to the question as asked.
-regexm("Tom and Tim", "Tim")- is 1.
You could use -regexm(<whatever>, "^Tim ")- to catch the first word.
It is a matter of taste, but it seems simpler to me to reach for regular
expression functions only when simpler functions such -substr()- and
-strpos()- don't yield easier answers.
In fact -word()- is also available for the purpose.
Nick
[email protected]
John Wagai
To get to your final answer use the regex command:
gen seq = regexm(newvar, "Tim")
NB use the newvar generated by Rich below
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Richard
Goldstein
Sent: 17 July 2008 16:04
To: [email protected]
Subject: Re: st: Wildcard for string variables
Sorry about the typo:
gen byte newvar=substr(sequence,1,4)=="Tim "
Richard Goldstein wrote:
> I don't see any need for a wildcard; how about
>
> gen byte newvar=substr(sequence,1,4)=="Tim"
>
> Rich
>
> Hugh Colaco wrote:
>> Hi Statalisters,
>>
>> I am not sure how the string wildcard works. For example, I have a
>> string variable called "sequence" that has the following
>> observations:-
>>
>> 1. "Tim and Tom and Jane"
>> 2. "Tim and Jane and Tom"
>> 3. "Tom and Jane and Harry"
>> 4. "Jane and Jill"
>>
>> I want to generate a new variable that equals one if Tim is the first
>> name mentioned, and zero otherwise.
>>
*
* 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/
