Quick helper: sample a batch of subsets (for U-stat experiments)

Draws \(K\) subsets of size \(s\) uniformly at random from \(\{1,\ldots,n\}\). Within each subset, elements are sampled without replacement (all indices are distinct). Across rows, subsets are sampled independently (the same index may reappear in different rows).

Usage

cate_demo_subsets(n, K = 1000L, s = 8L, seed = NULL)

Arguments

n

Integer; number of observations.

K

Integer; number of subsets to sample.

s

Integer; subset size (order of the kernel).

seed

Optional integer seed for reproducibility.

Value

An integer matrix of size \(K \times s\). Each row contains \(s\) distinct integers in \(\{1,\ldots,n\}\) representing one subset.

Examples

subs <- cate_demo_subsets(n = 200, K = 1000, s = 8, seed = 1)
dim(subs)
#> [1] 1000    8