R help - Oct 2010 - Boundary correction for kernel density estimation

Dear R-users,

I have the following problem:

I would like to estimate the density curve for univariate data between 0
and 1. Unfortunately, the density function in the stats package is just
capable to cut the curve at a left and a right-most point. This
truncation would lead to an underestimation. An overspill of the bounded
support is unappropriate as well.

Do anyone knows a boundary correction method implmented in R? I did much
research but the correction methods I found are regarding survival or
spatial data.

Thanks a lot for any hint!
Cheers,
Katja

Look at the logspline package.  It uses a different method from what density
does, but it can take boundaries into account.

-- 
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
801.408.8111

> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Katja Hebestreit
> Sent: Monday, October 11, 2010 3:04 AM
> To: r-help at r-project.org
> Subject: [R] Boundary correction for kernel density estimation
> 
> Dear R-users,
> 
> I have the following problem:
> 
> I would like to estimate the density curve for univariate data between
> 0
> and 1. Unfortunately, the density function in the stats package is just
> capable to cut the curve at a left and a right-most point. This
> truncation would lead to an underestimation. An overspill of the
> bounded
> support is unappropriate as well.
> 
> Do anyone knows a boundary correction method implmented in R? I did
> much
> research but the correction methods I found are regarding survival or
> spatial data.
> 
> Thanks a lot for any hint!
> Cheers,
> Katja
> 
> ______________________________________________
> R-help at 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.