Skip to content

profile likelihood

3 messages · Ben Bolker, plocq

#
Hi,

I try to use the function profile() of the SpatialExtremes' package to
obtain the profile likelihood of parameters for an extreme values fit based
on Poisson process :

fit<-fpot(data, threshold, model="pp", npp=365).

But when I call "profile(fit)", I obtain the following error (even if I
precise others arguments of the function) :

[1] "profiling loc"
Erreur dans nlpot(p[1], p[2], ...) :
argument(s) inutilis?(s) (loc = 9.58435562072)


I don't understand what is unused and why. From where does this error come
from? If someone could help me it would be nice...!

--
View this message in context: http://r.789695.n4.nabble.com/profile-likelihood-tp4174074p4174074.html
Sent from the R help mailing list archive at Nabble.com.
1 day later
#
plocq <less.is.mo_ <at> hotmail.com> writes:
We need a reproducible example, please.

  fpot() is in the evd package, not in the SpatialExtremes package
(you didn't say explicitly that it was, but this would be helpful
information).

  This minimal example works:

library(evd)
uvdata <- rgpd(100, loc = 0, scale = 1.1, shape = 0.2)
M1 <- fpot(uvdata, 1)
library(SpatialExtremes)
profile(M1)

  So you need to provide us with a minimal example that doesn't work.
#
Thank you for your response. In fact, it seems not to be adapted to the
Poisson model... :

library(evd)
uvdata <- rgpd(100, loc = 0, scale = 1.1, shape = 0.2)
M1 <- fpot(uvdata, 1, model="pp")
library(SpatialExtremes)
profile(M1) 

--
View this message in context: http://r.789695.n4.nabble.com/profile-likelihood-tp4174074p4179315.html
Sent from the R help mailing list archive at Nabble.com.