How about this approach?
1. run -sureg- to fit the regressions separately on x1 and x2. Apply
-constraint- first to get equal intercepts.
2. Use b0 + b1*X1 where b1*x1 < b2*x2; otherwise use b2*x2.
Steve
On Fri, Mar 5, 2010 at 2:14 PM, Austin Nichols <austinnichols@gmail.com> wrote:
> Sebastian van Baal <s.vanbaal@arcor.de> :
> This seems likely to be problematic no matter what you do--typically
> the objective function should be differentiable in the parameters in
> these kinds of problems. What is the theory that drives this
> specification? Is there an alternative parameterization that is
> differentiable?
>
> On Fri, Mar 5, 2010 at 10:51 AM, Sebastian van Baal <s.vanbaal@arcor.de> wrote:
>> Dear Stata Users:
>>
>> I attempt to fit a regression model that has a minimum function on the
>> right-hand side. My current working solution is to use nonlinear least
>> squares:
>>
>> . nl ( y = {b0} + min({b1}*x1 , {b2}*x2) )
>>
>> Being admittedly unexperienced with nonlinear least squares, I am unsure if
>> this is correct. I have not seen an application of nonlinear least squares
>> (or any other form of regression analysis) involving a minimum function.
>> Would you say that is the correct way to proceed?
>>
>> Advance thanks for any hints you can give me.
>>
>> Sebastian van Baal
>> Cologne, Germany
> *
> * 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/
>
--
Steven Samuels
sjsamuels@gmail.com
18 Cantine's Island
Saugerties NY 12477
USA
845-246-0774
*
* 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/
