Question GEVFIT in fExtremes

[Brommundt Jürgen]

Hello,
Doing some extreme value evaluations of wind speed data I found your very useful R-Package fExtremes. Since I started doing my evaluations in MatLab I had a chance to compare the results between the results of the MatLab gevfit and your gevfit routine. Because we need the data for the design of buildings I am interested in extremes of given return intervalls and the uncertainty, estimated from the confidence intervalls.
Matlab's gevfit estimates the parameters of the GEV and provides another set of parameters for the limits of the confidence intervalls. From these I estimated the design values using the inverse of the GEV. From fExtremes I used the function gevrlevelPlot with the same confidence intervall (c=0.9) to get the uncertainty bands.
As you can see from the attached Excel-File the MLM-fits of the GEV agree very well from both programs. Unfortunately the confidence bands do not agree. I tried to find out the reason from the documentation, but unfortunately did not succeed. I would appreciate, if you could be so kind, to give me a hint, what a possible reason might be??
Thanks a lot in advance!

Kind regards,
Jürgen Brommundt


 <<windstatistik_fExtremes.xls>> 


Jürgen Brommundt
Dr.-Ing., Projektingenieur
AF-Colenco AG, Grundwasserschutz und Entsorgung
Täfernstrasse 26 | CH-5405 Baden | Schweiz
Direkt: +41 (0)56 483 15 35 | Fax: +41 (0)56 483 18 46
juergen.brommundt at afconsult.com | www.af-colenco.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/rmetrics-core/attachments/20080520/8bf029c0/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: windstatistik_fExtremes.xls
Type: application/vnd.ms-excel
Size: 25088 bytes
Desc: windstatistik_fExtremes.xls
URL: <https://stat.ethz.ch/pipermail/rmetrics-core/attachments/20080520/8bf029c0/attachment-0001.xls>