marginals1d: marginals1d
In twDEMC: parallel DEMC

Description Usage Arguments Details Value Author(s) See Also Examples

integrate over all others but 1 column

1
2
3

marginals1d(x,
 vars = 2, intCol = 1, n = round(nrow(x)/12), dimnames = .colNames(x, 
    vars), fAgg = mean, ...)

`x`	numeric matrix
`vars`	index (integer or string): column to define integration boxes
`intCol`	index (integer or string): column to average
`n`	number of groups, defaults to about 10 observations per group
`dimnames`	dimension names of the result array
`fAgg`	function applied over the subsets of x[,intCol]
`...`	further arguments to fAgg

There are several methods to get calculate marginal densities for multivariate sample.

integrate over all others but 1 column: this method
integrate over all others but 2 column: marginals2d
integrate over all others but 3 column: marginals3d

numeric vector with each item representing application of fAgg to x[,intCol] and dimnames equal to means of the classes

Thomas Wutzler

twDEMCBlockInt

data(twdemcEx1)
sample <- stackChains(twdemcEx1)
#res <- marginals1d(sample,vars="kO",n=20)
res <- marginals1d(sample,vars="b",n=20)
plot( res ~ as.numeric(names(res)) )