Dear all,
I’ve been having trouble with the nl command in stata. I’m trying to estimate the following function
Crh= b0 + b1*triwk + b2*(max(triwk-k1,0))
The code is given below
gen trisp1=max(triwk -14, 0)
regress crh triwk trisp1
global b0= _b[_cons]
global b1= _b[triwk ]
global b2= _b[trisp1]
capture program drop nlspline
program define nlspline
version 8.2
if "`1' " == "?" {
global S_1 "b0 b1 b2 k1"
global b0= $b0
global b1= $b1
global b2= $b2
global k1=14
exit
}
replace `1' = $b0 + $b1*triwk + $b2*(max(triwk-$k1,0))
end
nl spline crh
When I run the code I get the error:
variable ? not found
nlsplinec refused query, rc=111
Initially I was trying to estimate this with an additional knot point and my data set had 1158 observations of which I use only 42 by using conditioning commands. At that time stata gave me an error “triwk-invalid name”.
I tried updating stata to version 8.2, I also tried simplifying the function as well as changing the variable name from ‘triwk’ as well as restarting stata. I still got the same error message. For the above code, I created a new data set with only the observations that I need.
The format I used for the code was from Mr. Maarten Buis’s message on Wednesday 5th October with subject: Re: Explaining Interaction terms. In this mail Maarten gave an example code which works perfectly on my computer!, but on adapting it to my variables it doesn’t work. I checked my variables couldn’t find anything amiss with them.
Thank you for your help.
Leny
*
* 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/
