vkde: Variable Kernel Density Estimation
In kcucchi/vdke: Variable Kernel Density Estimation methods

Description Usage Arguments Details Value Examples

This is the description of the function vkde.

1	vkde(tdat, edat, bwtype, h, p = 1, verbose = F)

`tdat`	a vector containing the training dataset.
`edat`	a vector containing the evaluation data points.
`bwtype`	a string corresponding to the method for calculating bandwidths. Either `'fixed'`, `'balloon'` or `'sampleSmoothing'`.
`h`	the bandwidth(s) of the kernels. It should be a scalar when `'bwtype=fixed'`, a vector of length `length(edat)` for `'bwtype=balloon'`, a vector of length `length(tdat)` for `'bwtype=sampleSmoothing'`. It can be calculated from other functions in this package.

These are the details of the function.

The density at locations edat based on the training dataset tdat following the method specified by bwtype. The bandwidths h can be calculated from other functions in this package.

evalVect <- seq(from=-5,to=5,by=0.1)
trainingVect <- runif(5,min = -1,max = 2)
kernelVect <- vkde(tdat=trainingVect,edat=evalVect,'fixed',1)
plot(evalVect,kernelVect$kde,
     type='l',
     xlab = 'x',ylab='kernel density estimate')
axis(1,trainingVect,labels = NA,lwd.ticks = 2,col.ticks = 'red',tck=0.1)
for(i in 1:length(trainingVect)){
  lines(evalVect,kernelVect$individualKernels[i,],lty=2)
}