Displaying 1 result from an estimated 1 matches for "bivn".
Did you mean:
bin
2003 Sep 01
0
Re: Plotting bivariate normal distributions.
You'll find that it is a lot easier to do it in R:
# lets first simulate a bivariate normal sample
library(MASS)
bivn <- mvrnorm(1000, mu = c(0, 0), Sigma = matrix(c(1, .5, .5, 1), 2))
# now we do a kernel density estimate
bivn.kde <- kde2d(bivn[,1], bivn[,2], n = 50)
# now plot your results
contour(bivn.kde)
image(bivn.kde)
persp(bivn.kde, phi = 45, theta = 30)
# fancy contour with image
image(bivn.kde);...