If the names are as in `c(Mons)', you
could go
tokenize "`c(Mons)'"
forval i = 1/12 {
replace x = m`i' if index(Y, "``i''")
}
Nick
[email protected]
Joseph Coveney
> FUKUGAWA Nobuya wrote:
>
> Suppose we have string variable Y, and we would like to replace
> variable X when variable Y contains certain characters such as
> "Dec".
> In other words, we want to write;
> replace X="m12" if Y (contains "Dec")
> replace X="m11" if Y (contains "Nov")
> replace X="m10" if Y (contains "Oct") .
> How can I write the latter part of command?
> Please note Y can be Dec123abc, 456Novpqr, and zzz789Oct.
>
\> --------------------------------------------------------------
> ------------------
>
> Rafal Raciborski responded with an answer that uses Stata's
> new regular
> expression functions. If you have an earlier release, then:
>
> replace X = "m12" if index(Y, "Dec")
> replace X = "m11" if index(Y, "Nov")
> replace X = "m10" if index(Y, "Oct")
>
> Note that in Stata 9, you can use -strpos()- instead of -index()-:
>
> replace X = "m12" if strpos(Y, "Dec")
> replace X = "m11" if strpos(Y, "Nov")
> replace X = "m10" if strpos(Y, "Oct")
>
> You can get cute, too, if you want to:
>
> foreach month in `c(Mons)' {
> replace X = "m" + ///
> string(month(date("2000-`month'-01", "ymd")), "%02.0f") ///
> if strpos(Y, "`month")
> }
>
> This last snippet helps avoid tedium if you need to replace X
> for all twelve
> months, but would be overkill for just three.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
