Jian Zhang:
Easiest way to do that is to use the -normalden- function like this:
sysuse auto, clear
capture program drop try1
program try1
args lnf xb ln_sigma
quietly replace `lnf'=ln(normalden($ML_y1, `xb',
exp(`ln_sigma')))
end
ml model lf try1 (price=mpg foreign) ()
ml maximize
If Stata has a function for the density you want to use preprogrammed
than using it will generally result in more readable code and less
bugs.
HTH,
Maarten
--- Jian Zhang <[email protected]> wrote:
> I am using ml command to do maximum likelihood estimation with normal
> distribution of error terms. But the the result returned with an
> error message.
> Here is the code I wrote:
>
> capture program drop try1
> program try1
> args lnf theta1 theta2
> quietly replace
> `lnf'=-1/2*ln(`theta2')-1/2*($ML_y1-`theta1')
> end
>
> ml model lf try1 (year=i firm) ()
> ml maximize
>
> But the result shows as follows :
> initial: log likelihood = -<inf> (could not be evaluated)
> feasible: log likelihood = -97165.343
> rescale: log likelihood = -60002.996
> rescale eq: log likelihood = -60002.996
> could not calculate numerical derivatives
> discontinuous region with missing values encountered
> r(430);
>
> end of do-file
> r(430);
>
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
___________________________________________________________
Inbox full of spam? Get leading spam protection and 1GB storage with All New Yahoo! Mail. http://uk.docs.yahoo.com/nowyoucan.html
*
* 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/
