Skip to content
Prev 263955 / 398502 Next
Original Raw

Kernel Density Estimation at manually specified points

S Ellison 
May be a kludge, but it might be simpler to write your own density function for a few specified points.

For example

my.density <- function(x, bw = 'nrd0', at) {
    x<-na.omit(x)
    
    #####
    #Borrowed from density.default for compatibility
    
    if (is.character(bw)) {
        if (length(x) < 2) 
            stop("need at least 2 points to select a bandwidth automatically")
        bw <- switch(tolower(bw), nrd0 = bw.nrd0(x), nrd = bw.nrd(x), 
            ucv = bw.ucv(x), bcv = bw.bcv(x), sj = , `sj-ste` = bw.SJ(x, 
                method = "ste"), `sj-dpi` = bw.SJ(x, method = "dpi"), 
            stop("unknown bandwidth rule"))
    }
    ######
    at <- matrix(at, ncol=1)
    y <- apply(at, 1, FUN=function(a, x, bw) sum(dnorm(a, x, bw)/length(x)), x=x, bw=bw )
    return(list(x=at, y=y))

}

x<-rnorm(500)
plot(density(x))
at<-seq(-2, 2, 0.25)
points(my.density(x, at=at))
This email and any attachments are confidential. Any use...{{dropped:8}}

Thread (5 messages)

Carsten Harlaß Kernel Density Estimation at manually specified points Jun 26 David L Carlson Kernel Density Estimation at manually specified points Jun 27 Carsten Harlaß Kernel Density Estimation at manually specified points Jun 27 David Winsemius Kernel Density Estimation at manually specified points Jun 27 S Ellison Kernel Density Estimation at manually specified points Jun 28