Hi there, during my work I have to use kernel smoothing methods for multivariate ordinal data. The R-package "KernSmooth" unfortunately includes only a version for continous scaled variables. Does anybody know whether there exists also a version for ordinal data? Thanks for help! --
Meike Gebel <meike.gebel at gmx.de> writes:> Hi there, > during my work I have to use kernel smoothing methods for multivariate > ordinal data. > The R-package "KernSmooth" unfortunately includes only a version for > continous scaled variables. > Does anybody know whether there exists also a version for ordinal data? > Thanks for help!Er, do you have a reference for what such a beast should be doing? Kernel smoothing methods generally rely quite heavily on having a continuous distribution (for density estimates) or having a continuous x distribution and something on the y axis that allows taking weighted avereages (for nonparametric regression type problems). -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907
They do exist: the term has a somewhat different meaning for categorical data. Titerington, D. M. (1980) Technometrics 22, 259-268 might be a good start. On 17 Jun 2003, Peter Dalgaard BSA wrote:> Meike Gebel <meike.gebel at gmx.de> writes: > > > Hi there, > > during my work I have to use kernel smoothing methods for multivariate > > ordinal data. > > The R-package "KernSmooth" unfortunately includes only a version for > > continous scaled variables. > > Does anybody know whether there exists also a version for ordinal data? > > Thanks for help! > > Er, do you have a reference for what such a beast should be doing? > Kernel smoothing methods generally rely quite heavily on having a > continuous distribution (for density estimates) or having a continuous > x distribution and something on the y axis that allows taking weighted > avereages (for nonparametric regression type problems). > >-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595