Summary statistics for bootstrap samples

DESCRIPTION:

Compute means, products, sums, or variances for samples

USAGE:

indexMeans(x, indices, group = NULL) 
indexProducts(x, indices) 
indexSums(x, indices) 
indexVars(x, indices) 

REQUIRED ARGUMENTS:

x
numeric vector or matrix with n rows and p columns. The columns represent variables, each consisting of n observations.
indices
numeric vector or matrix with m rows and B columns , containing integer values between 1 and n. B is the number of bootstrap samples and m is the number of observations per sample.

OPTIONAL ARGUMENTS:

group
vector of length equal to number of rows of x, dividing x into groups. If supplied, then the return values correspond to the sum (across groups) of group means.

VALUE:

B x p matrix, with i,jth element equal to
mean(x[indices[,i],j])
(or product, sum, or variance).

SEE ALSO:

, .

EXAMPLES:

data <- cancer.vet[,c("age","survival")] 
obj <- bootstrap(data, mean, B=300, save.indices=T) 
indexMeans(data, obj$indices) 
 
# Linear approximation for a robust location estimate 
set.seed(1); x <- rcauchy(40) 
obj <- bootstrap(x, location.m, B=200, save.indices=T) 
L <- resampGetL(obj) 
plot(x, L)               # outliers have less influence 
plot(indexMeans(L, obj$indices), obj$replicates)