[R] The profile log-likelihood problem

[Duncan Mackay]

Hi

Without knowing anything about the data and the warning messages, however
the p.vec value seems a little strange.
try 
p.vec= seq(1.35, 1.75, by = 0.1) 

It could be that you have run out of memory to do 50 sets of calculations

If ok with the result then repeat with a narrower focus. Most probably
quicker


Regards

Duncan

Duncan Mackay
Department of Agronomy and Soil Science
University of New England
Armidale NSW 2351
Email: home: mackay at northnet.com.au


-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On
Behalf Of smart hendsome
Sent: Thursday, 21 November 2013 12:43
To: r-help at r-project.org
Cc: roslinazairimah at ump.edu.my
Subject: [R] The profile log-likelihood problem

Hi, Im still new to R and I have a problem regarding the code below. I don't
know how to solve that. I want to compute profile log-likelihood using the
rainfall data. I use the code below from Dunn:



"
tweedie.profile(Amount~Year*Month,p.vec=seq(1.35,1.75,length=50),method="ser
ies",do.ci=TRUE) "

I got the message below when I run the code above. Anyone know why the p.vec
taken from 1.35 to 1.75? Can someone help me? 


"1.35 1.358163 1.366327 1.374490 1.382653 1.390816 1.398980 1.407143
1.415306 1.423469 1.431633 1.439796 1.447959 1.456122 1.464286 1.472449
1.480612 1.488776 1.496939 1.505102 1.513265 1.521429 1.529592 1.537755
1.545918 1.554082 1.562245 1.570408 1.578571 1.586735 1.594898 1.603061
1.611224 1.619388 1.627551 1.635714 1.643878 1.652041 1.660204 1.668367
1.676531 1.684694 1.692857 1.701020 1.709184 1.717347 1.725510 1.733673
1.741837 1.75 ..................................................Done.
  No valid values of the likelihood computed: smooth aborted
   Consider trying another value for the input  method.
Error in if ((xi.max > 0) & (xi.max < 1)) { : 
  missing value where TRUE/FALSE needed
In addition: There were 50 or more warnings (use warnings() to see the first
50)"

Thanks for yourattention and help.

Regard,
Zuhri
	[[alternative HTML version deleted]]