Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



Re: OBRE algorithm
Posted:
Dec 5, 2012 1:41 PM


Hi Koen! Just done. Not much docs yet
https://github.com/moiseevigor/obre
Entry point for procedures https://github.com/moiseevigor/obre/blob/master/matlab/obre.m which contains some info how to execute the code.
Later I'll provide the data snippets and examples how to execute all of this.
Please be patient, it is quite a "programmer design".
Please ask, open issues, request features at https://github.com/moiseevigor/obre/issues
Please Contribute, Star, Watch, Fork and Pull request.
If you appreciate the code and months of hacking tip me please at https://www.gittip.com/moiseevigor/ Have fun coding! Igor
"Koen" wrote in message <k9n1d0$m2o$1@newscl01ah.mathworks.com>... > Hi Igor, > Have you been able to open the repository yet and share the procedures? I am dying to get that OBRE implemented. > Thanks. > Kind regards, > > Koen > > "Moiseev Igor" <moiseev.igor@gmail.com> wrote in message <k99um7$lth$1@newscl01ah.mathworks.com>... > > Hi everybody, > > we've managed to convert the Fortran procedures of OBRE (from the article Dupuis et al) into Matlab and it seems to produce correct results, but still remember this is a partial machine conversion, so it is full of non matlabstyle code. > > > > As I've noted the magior interest to this algorithm, I'll provide to open a new Github repository for this code, free to access and contribute. > > > > I'll let you know the repo link in a few days. > > > > All the best, > > Igor. > > > > "Koen" wrote in message <k97fks$mlh$1@newscl01ah.mathworks.com>... > > > Hi Jorge, > > > > > > How was your luck? Managed to implement your OBRE in Matlab? > > > > > > Kind regards, > > > Koen > > > > > > "Jorge Lara" <jorge.lara.alvarez@gmail.com> wrote in message <k01rpe$68p$1@newscl01ah.mathworks.com>... > > > > Hi Igor, > > > > > > > > I see that it has been a year. But I am going to implement OBRE in matlab. > > > > > > > > Did you have any luck before? > > > > > > > > "Moiseev Igor" <moiseev.igor@gmail.com> wrote in message <ihrbe7$brg$1@fred.mathworks.com>... > > > > > Hi Simon, I see it is three years past, but in any case we're going to implement the OBRE algorithm for extreme data using Matlab. The implementation is based on the article > > > > > > > > > > "Exceedances over High Thresholds A Guide to Threshold Selection" by D.J. Dupuis, Department of Engineering Mathematics, DalTech Dalhousie University, Halifax Nova Scotia > > > > > Canada > > > > > > > > > > Are you still interested? Did you got any info about the problem that time? > > > > > > > > > > Thank you in advance and all the best! > > > > > Igor. > > > > > > > > > > "Simon " <shchan.nospam@northwatercapital.com> wrote in message <fpurag$ll$1@fred.mathworks.com>... > > > > > > Has anyone implemented an Optimal Bias Robust Estimator for > > > > > > fitting the Generalized Pareto Distribution or know a link > > > > > > to get one? Thanks.



