R help - Oct 2008 - How to calculate an appropiate threshold of a GPD distribution.

Dear R-users

I would like to fit the GPD distribution using the "gpd" function from
"evir" package. I need to pick an appropiate threshold. I know there
are
some useful plots to assist me in this choice like meplot or mrlplot. Do you
know if there is any technique to compute or calculate an exact threshold?.

Regards
Borja

	[[alternative HTML version deleted]]