R help - Jun 2012 - POT Package

Hi,

I have a problem in fitting GPD distribution.
i generate random numbers from gpd distribution from specific parameters
using pot packege then i used fitgpd function to estimete the parameters.The
estimated parameters are not matched with the given parameters i.e.from
which i generate random numbers.I think estimated parameters should be
matched with the given parameters.Also suggest me another good packege for
GPD.


for e.g.
k2<-rgpd(100,loc=100000,scale=229678.21 ,shape=0.41)

fitgpd(k2,100000,"mle")$param
       scale        shape 
4.271966e+05 2.092887e-01 






--
View this message in context:
http://r.789695.n4.nabble.com/POT-Package-tp4632055.html
Sent from the R help mailing list archive at Nabble.com.

Hi,

You assumed 

scale=229678.21 and shape=0.41

and the procedure estimated

\hat{scale}=427196.6 and \hat{shape}=0.2092887

I think these estimates are natural, since 
1) you use a simulated generated data with n=100, and 2) the true value your
scale parameter is very large.

Hope this helps
Best
Ozgur

--
View this message in context:
http://r.789695.n4.nabble.com/POT-Package-tp4632055p4632065.html
Sent from the R help mailing list archive at Nabble.com.