CKNNRLD.tune: Tune CKNNRLD Model with Automatic Cluster Selection

In CKNNRLD: Clustering-Based K-Nearest Neighbor Regression for Longitudinal Data

Tune CKNNRLD Model with Automatic Cluster Selection

Description

Automatically selects the best number of clusters (C) using internal validation criteria, then tunes the CKNNRLD model within each cluster via cross-validation.

Usage

CKNNRLD.tune(
  y,
  x,
  nfolds = 10,
  folds = NULL,
  seed = NULL,
  A = 10,
  C_range = 2:6,
  cluster_method = "kml"
)

Arguments

y	Matrix of longitudinal outcomes (subjects x timepoints).
x	Matrix of predictor variables (subjects x features).
nfolds	Number of folds for cross-validation (default = 10).
folds	Optional list of pre-specified fold indices.
seed	Random seed for reproducibility.
A	Maximum number of neighbors to evaluate (searches from 2 to A, default = 10).
C_range	Range of cluster numbers to evaluate (default = 2:6).
cluster_method	Clustering method to use. Currently supports "kml" (default).

Value

A list containing:

best_c	Optimal number of clusters
cluster_results	Tuning results for each cluster (from KNNRLD.tune)
cluster_sizes	Size of each cluster
cluster_assignments	Cluster labels for each subject
criteria	Data frame of quality criteria used for selecting best C

Examples

set.seed(123)
n <- 30
T <- 3
d <- 2
x <- matrix(runif(n * d), nrow = n)
y <- matrix(rnorm(n * T), nrow = n)
tune_result <- CKNNRLD.tune(
  y = y,
  x = x,
  nfolds = 3,
  A = 4,
  C_range = 2:3
)
print(tune_result$best_c)

CKNNRLD documentation built on May 29, 2026, 1:06 a.m.