DNN weights (asymptotic full-U), distance-based ranking

The asymptotic DNN rule is evaluated on the full rank grid and is therefore typically treated as a full-support approximation rather than as a hard finite-support cutoff.

Usage

weights_dnn_asymp(
  X,
  X0 = NULL,
  x0 = NULL,
  distance = "l2",
  k,
  s,
  normalize = TRUE,
  ctx = NULL,
  return = c("W")
)

Arguments

X

numeric matrix (n x p)

X0

numeric matrix (m x p); mutually exclusive with x0

x0

numeric length-p single target (convenience for m=1)

distance

distance spec or registered name (e.g., "l2")

k

integer; neighbors used in each s-subset (1 <= k <= s)

s

integer; U-statistic order / subset size

normalize

logical; renormalize columns to sum 1

ctx

optional context

return

character vector of extras to include; any of c("W","ord","ranks","dx")

Value

A weights_result object (same structure as weights_dnn_exact()).

Examples

set.seed(1)
x_mat <- matrix(rnorm(20), 10, 2)
x0_mat <- matrix(rnorm(4), 2, 2)
out <- weights_dnn_asymp(x_mat, x0_mat, k = 1L, s = 3L, normalize = TRUE)
dim(out$W)
#> [1] 10  2