Thanks for your reply. The advantage of nonparametric statistics is that you do not need to specify the distribution because the estimator (e.g. kernel estimator) converges to the true density as the number of observations converges to infinite. The mode in this case is the maximum of this estimator as a *function*. Therefore it actually makes a lot of sense and it is an activate research area in statistics. Please check these concepts before replying.
> Date: Fri, 20 Jan 2012 12:05:10 -0800 > Subject: Re: [R] Estimation of the mode > From: gunter.ber...@gene.com > To: chikwi...@hotmail.com > CC: r-help@r-project.org > > You realize, I trust, that "the mode" of a continuous distribution has > no meaning without prior specification of the distribution, which, for > a fitted density estimate would mean specification of the fitting > parameters (bandwidth, etc.) at a minimum. > > So perhaps you need to rethink what you are trying to do and/or > perhaps get some advice from a local statistical resource. > > -- Bert > > On Fri, Jan 20, 2012 at 8:27 AM, Javier xyz <chikwi...@hotmail.com> wrote: > > > > Hi all, > > > > I am trying to estimate the mode of a 4-dimensional nonparametric density > > estimator (any) using a sample of size n=10,000. I have tried using the > > package 'ks' and 'np' but they are extremely slow; this is related to the > > estimation of the bandwidth matrix. I also checked the package 'modeest' > > but it contains only methods for univariate distributions. I am only > > interested in the mode. Any suggestion? > > > > Thanks in advance. > > > > Kind regards. > > > > [[alternative HTML version deleted]] > > > > ______________________________________________ > > R-help@r-project.org mailing list > > https://stat.ethz.ch/mailman/listinfo/r-help > > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > > and provide commented, minimal, self-contained, reproducible code. > > > > -- > > Bert Gunter > Genentech Nonclinical Biostatistics > > Internal Contact Info: > Phone: 467-7374 > Website: > http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
