Skip to contents

Nonparametric regression singleton score from observed outcome \(Y\) (order-1 pure score; no weights)

Source: R/score-npr.R
score_np_y.Rd

Constructs a pure-score kernel for nonparametric regression that uses only the observed response \(Y\). For a singleton subset \(S=\{i\}\) the kernel returns exactly \(h(S) = Y_i\). This score is order-1 only (s = 1). It performs no nearest-neighbor ordering, kernel weighting, or L-stat aggregation. Any weighting/averaging of subset evaluations is handled outside the score layer (e.g., by rank-/distance-weight modules).

Usage

score_np_y(x0, x_cols, s = 1, y_col = NULL, y = NULL)

Arguments

x0

Numeric length-\(p\) target point. Used only to check that length(x0) matches length(x_cols); it is not used in the kernel.

x_cols

Integer or character indices of the \(X\) columns in data, used for dimension checks with x0.

s

Intended subset size (kernel order). Must be 1; otherwise an error is thrown.

y_col

Optional name of the column in data that contains the response \(Y\). If NULL, supply y.

y

Optional numeric vector (length \(n\)) of \(Y\) values to use when y_col is NULL.

Value

An ustat_score object with:

  • $name: a descriptive name,

  • $order: always 1L,

  • $setup: stores g <- Y in the context,

  • $eval/$eval_batch: return \(Y_i\) for each singleton subset.

Details

The arguments x0, x_cols, and s are accepted for API consistency and basic dimension checks; only \(Y\) is used to construct the score. This factory enforces s = 1.

Examples

set.seed(1)
n <- 50
p <- 2
dat <- data.frame(X1 = rnorm(n), X2 = rnorm(n))
dat$Y <- with(dat, X1 - 2 * X2 + rnorm(n))

sc <- score_np_y(
  x0 = c(0, 0),
  x_cols = c("X1", "X2"),
  s = 1L,
  y_col = "Y"
)
ctx <- score_setup(sc, dat)
subs <- matrix(seq_len(n), ncol = 1L) # singleton subsets
score_eval_batch(sc, subs, dat, ctx) # returns dat$Y
#>  [1] -2.043032249  1.449811984 -2.428789644  4.012035767 -3.191124276
#>  [6] -3.013980912  1.938579481  3.736768187 -0.179472545  1.646896901
#> [11] -3.927190807  0.006678512 -0.568437067 -2.921400558  2.404096592
#> [16] -0.815326137  3.273734126 -2.266386815  1.008902850 -3.928652502
#> [21] -0.537999148  3.545067988 -1.361467132 -0.300712963  3.026901807
#> [26]  0.073645096  0.657223836 -1.510597259 -1.308493182  1.272713180
#> [31]  2.556177458 -0.421324983 -1.437006189  1.474934478 -2.258394071
#> [36] -2.617345129 -2.821465755  0.020774546 -0.292107029  0.172081389
#> [41] -0.993842960 -1.492513980 -3.288814292 -1.307294502 -4.978342709
#> [46] -2.575287009  5.004932925  1.932459373  1.050578487  0.187303465