6 messages · Duncan Murdoch, Greg Snow, wangguojie2006 +1 more
Hi, guys, I'm using kernel "density" estimation. But how can I return to a density estimation for a fixed point? For example, b<-runif(1000,0,1) f<-density(b) How can I get the value of density(b) at b=0.5? Your help is extremely appreciated. Thanks. Jay
View this message in context: http://r.789695.n4.nabble.com/help-in-density-estimation-tp2552264p2552264.html Sent from the R help mailing list archive at Nabble.com.
b<-runif(1000,0,1) f<-density(b)
f is a list of things, including x values where the density is computed, and y values for the density there. So you could do it by linear interpolation using approx or approxfun. For example > b <- runif(1000,0,1) > flist <- density(b) > f <- approxfun(flist$x, flist$y) > f(0.2) [1] 0.9717893 > f(-1) [1] NA If you don't like the NA for an out-of-range argument, then choose something different from the default for the "rule" argument to approxfun. Duncan Murdoch
On 23/09/2010 11:42 AM, wangguojie2006 wrote:
b<-runif(1000,0,1) f<-density(b)
f is a list of things, including x values where the density is computed, and y values for the density there. So you could do it by linear interpolation using approx or approxfun. For example
> b <- runif(1000,0,1) > flist <- density(b) > f <- approxfun(flist$x, flist$y) > f(0.2)
[1] 0.9717893
> f(-1)
[1] NA If you don't like the NA for an out-of-range argument, then choose something different from the default for the "rule" argument to approxfun. Duncan Murdoch
Or, perhaps more transparently (and more explicitly modifiable):
b<-runif(1000,0,1)
f <- density(b, from=0, to=1, n=512)
plot(f$x, f$y, type="l", col="blue",
xlim=c(0,1), ylim=c(0,1.5)) ## Plot the density estimate
x0 <- 0.5 ## Target value of x
i0 <- max(which(f$x <= x0))
i1 <- min(which(f$x > x0))
u0 <- f$x[i0] ; v0 <- f$y[i1]
u1 <- f$x[i1] ; v1 <- f$y[i1]
y0 <- v0 + (v1-v0)*(x0-u0)/(u1-u0) ## Linear interpolation
points(x0, y0, pch="+", col="red") ## Add interpolated point
Ted.
--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 23-Sep-10 Time: 18:05:41
------------------------------ XFMail ------------------------------
You could do: b <- runif(10000, 0, 1) tmp <- density(b, from=0.5, to=0.5, n=1) tmp$y As one direct approach. You could also look at the logspline package for an alternative for density estimation that provides you with a density function (and also allows for finite domains).
Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.snow at imail.org 801.408.8111 > -----Original Message----- > From: r-help-bounces at r-project.org [mailto:r-help-bounces at r- > project.org] On Behalf Of wangguojie2006 > Sent: Thursday, September 23, 2010 9:43 AM > To: r-help at r-project.org > Subject: [R] help in density estimation > > > Hi, guys, > > I'm using kernel "density" estimation. But how can I return to a > density > estimation for a fixed point? > > For example, > > b<-runif(1000,0,1) > f<-density(b) > > How can I get the value of density(b) at b=0.5? > > Your help is extremely appreciated. Thanks. > > Jay > -- > View this message in context: http://r.789695.n4.nabble.com/help-in- > density-estimation-tp2552264p2552264.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting- > guide.html > and provide commented, minimal, self-contained, reproducible code.
You guys are really good. Thanks.
View this message in context: http://r.789695.n4.nabble.com/help-in-density-estimation-tp2552264p2552484.html Sent from the R help mailing list archive at Nabble.com.
There was a typo error in my code below. See the inserted correction.
On 23-Sep-10 16:52:09, Duncan Murdoch wrote:
On 23/09/2010 11:42 AM, wangguojie2006 wrote:
b<-runif(1000,0,1) f<-density(b)
f is a list of things, including x values where the density is computed, and y values for the density there. So you could do it by linear interpolation using approx or approxfun. For example
> b <- runif(1000,0,1) > flist <- density(b) > f <- approxfun(flist$x, flist$y) > f(0.2)
[1] 0.9717893
> f(-1)
[1] NA If you don't like the NA for an out-of-range argument, then choose something different from the default for the "rule" argument to approxfun. Duncan Murdoch
Or, perhaps more transparently (and more explicitly modifiable):
b<-runif(1000,0,1)
f <- density(b, from=0, to=1, n=512)
plot(f$x, f$y, type="l", col="blue",
xlim=c(0,1), ylim=c(0,1.5)) ## Plot the density estimate
x0 <- 0.5 ## Target value of x
i0 <- max(which(f$x <= x0))
i1 <- min(which(f$x > x0))
##u0 <- f$x[i0] ; v0 <- f$y[i1] ## WRONG!
u0 <- f$x[i0] ; v0 <- f$y[i0] ## Correction
u1 <- f$x[i1] ; v1 <- f$y[i1]
y0 <- v0 + (v1-v0)*(x0-u0)/(u1-u0) ## Linear interpolation
points(x0, y0, pch="+", col="red") ## Add interpolated point
Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk> Fax-to-email: +44 (0)870 094 0861 Date: 23-Sep-10 Time: 18:05:41 ------------------------------ XFMail ------------------------------
-------------------------------------------------------------------- E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk> Fax-to-email: +44 (0)870 094 0861 Date: 23-Sep-10 Time: 20:38:53 ------------------------------ XFMail ------------------------------