R help - Aug 2008 - Polygon issue in lattice & Bivariate density estimation

Dear all,

I am trying to set a specific bandwidth for a bivariate kernel density
estimation and then plot it in lattice: managed all that except that the
plot appears to have an issue regards the setting of the polygon and as a
result I end up with horizontal lines disecting my plotting region. If I
lower the bandwidth excessively then the issue resolves itself, but this
defeats the object as I wish and need h=0.8481.

Any suggestions would be much appreciated. Thank you

Here is the code I have used:
library(lattice)
library(KernSmooth)
#####
CAPEL2=read.csv("CAPEL2.csv",header=T,row.names=1)
#####
panel.density2d <-
    function(x, y, nlevels = 10, bandwidth, ...)
{
    k2d <- bkde2D(na.omit(cbind(x, y)), bandwidth = bandwidth)
    cl <-
        contourLines(x = k2d$x1, y = k2d$x2, z = k2d$fhat,
                     nlevels = nlevels)
    lapply(cl, panel.polygon, ...)
}
#####
xyplot(PCO2~PCO1,data=CAPEL2,panel = panel.density2d,bandwidth=0.8481)
http://www.nabble.com/file/p18893107/CAPEL2.csv CAPEL2.csv 
-- 
View this message in context:
http://www.nabble.com/Polygon-issue-in-lattice---Bivariate-density-estimation-tp18893107p18893107.html
Sent from the R help mailing list archive at Nabble.com.

On Fri, Aug 8, 2008 at 7:40 AM, Andrewjohnclose <a.j.close at ncl.ac.uk>
wrote:
>
> Dear all,
>
> I am trying to set a specific bandwidth for a bivariate kernel density
> estimation and then plot it in lattice: managed all that except that the
> plot appears to have an issue regards the setting of the polygon and as a
> result I end up with horizontal lines disecting my plotting region. If I
> lower the bandwidth excessively then the issue resolves itself, but this
> defeats the object as I wish and need h=0.8481.
>
> Any suggestions would be much appreciated. Thank you
>
> Here is the code I have used:
> library(lattice)
> library(KernSmooth)
> #####
> CAPEL2=read.csv("CAPEL2.csv",header=T,row.names=1)
> #####
> panel.density2d <-
>    function(x, y, nlevels = 10, bandwidth, ...)
> {
>    k2d <- bkde2D(na.omit(cbind(x, y)), bandwidth = bandwidth)
>    cl <-
>        contourLines(x = k2d$x1, y = k2d$x2, z = k2d$fhat,
>                     nlevels = nlevels)
>    lapply(cl, panel.polygon, ...)
> }
Without looking at this in detail, I would suggest using

    lapply(cl, panel.lines, ...)

instead. Polygons are drawn as closed, whereas contourLines does not
necessarily give you closed contours.

-Deepayan
> #####
> xyplot(PCO2~PCO1,data=CAPEL2,panel = panel.density2d,bandwidth=0.8481)
> http://www.nabble.com/file/p18893107/CAPEL2.csv CAPEL2.csv
> --
> View this message in context:
http://www.nabble.com/Polygon-issue-in-lattice---Bivariate-density-estimation-tp18893107p18893107.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.
>