Assuming your 40 variables are consecutively numbered in some way (here I have assumed var1-var40), the following (untested) code should be about right...
gen cons=0
gen maxcons=0
forvalues i=1/40 {
replace cons=cons+1 if !missing(var`i')
replace cons=0 if missing(var`i')
replace maxcons=cons if cons>maxcons
}
gen byte cons20=maxcons>=20
The final variable (cons20) is the 1/0 variable you are after (maxcons will contain the longest run of consecutive non-missing variables).
Hope this helps
David
-----Original Message-----
From: Alejandro Delafuente
[mailto:[email protected]]
Sent: 10 June 2005 16:23
To: [email protected]
Subject: st: Specification for a 1/0 variable with consecutive
non-missing observations
Hi statalisters,
I have a time series for 40 years (40 continuos variables) and am trying to
create a 1/0 variable that indicates 1 (for each row) if at least 20 variables
in the series have consecutive non-missing values and 0 if this condition is
not accomplished. I do have more than 20 non-missing values for most of the
variables per row, but there are gaps (ie, missing values) between them. Thus I
want to know how many fulfill the condition of 20 non-missing values in a road.
Any clues as to how can I accomplish this with programming?
*
* 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/
*
* 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/
