R help - Jun 2012 - Need help in using "OPTIM / OPTIMIZATION" function

Dear rehena,

You can use optim() with method="Brent", or optimize() directly, for
your
case.

But for you case, a popular approach is Newton-Raphson root finding
algorithm. You can use newton function under my package OOmisc.

You can check whether the global maximum or local maximum is found by
replacing the solution in the second derivative of the log-likelihood
function and if its value at the solution is less than 0, you can be sure
that global maximum is secured.

Hope this helps
Ozgur
 

--
View this message in context:
http://r.789695.n4.nabble.com/Need-help-in-using-OPTIM-OPTIMIZATION-function-tp4633796p4633921.html
Sent from the R help mailing list archive at Nabble.com.

rehena wrote
> 
> Hi Ozgur,
> 
> Many thanks for your help. 
> 
> May be I am not very strong enough in the OPTIMIZATION field. Can I check
> with you whether second derivative of the function at global max will be
> less than 0 and will be greater than 0 at local max? 
> 
> Thanks in advance. 
> 
> Regards,
>  rehena
> 

Dear rehena,

If you want to check whether the second derivative is negative by OOmisc,
the answer is no.

You an derive the derivatives by hand and code in R. 

Ozgur

--
View this message in context:
http://r.789695.n4.nabble.com/Need-help-in-using-OPTIM-OPTIMIZATION-function-tp4633796p4633924.html
Sent from the R help mailing list archive at Nabble.com.