bw_CV_fun: CV optimal bandwidth

In svjaco/lpreba: Local Polynomial Regression with Option for Explicit Boundary Adjustment

Description

Selection of the bandwidth that yields the smallest leave-one-out cross-validation (LOOCV) error over a supplied grid.

Usage

bw_CV_fun(
  X,
  Y,
  kernel = epanechnikov,
  bw_grid,
  degree = 1L,
  kernel_left = epanechnikov_left,
  boundary_left = NA,
  boundary_right = NA,
  boundary_adjustment = FALSE,
  plot = TRUE
)

Arguments

X	Data for the regressor (vector).
Y	Data for the regressand (vector).
kernel	Kernel (function). Default is epanechnikov.
bw_grid	Bandwidth grid to find global minimum of CV error (vector).
degree	Degree of the locally fitted polynomial (integer). Default is 1L.
kernel_left	Left boundary kernels (function). Default is epanechnikov_left.
boundary_left	Lower boundary of the support of X (scalar). Default is NA.
boundary_right	Upper boundary of the support of X (scalar). Default is NA
boundary_adjustment	Explicit boundary adjustment (boolean). Default is FALSE.
plot	Plot CV error over bandwidth grid (boolean). Default is TRUE.

Value

CV optimal bandwidth (scalar) and plot of CV errors if plot == TRUE.

See Also

CV_error_fun

Examples

m_fun <- function(x) {sin(2*pi*x)} # True regression function
n <- 100 # Sample size
X <- seq(0, 1, length.out = n) # Data for the regressor
m_X <- m_fun(X) # True values of regression function
epsilon <- rnorm(n, sd = 0.25) # Error term
Y <- m_X + epsilon # Data for the regressand
bw_grid <- seq(0.05, 0.2, length.out = 200)

# Local polynomial estimator
bw_CV_fun(X = X, Y = Y, kernel = epanechnikov, bw_grid = bw_grid, degree = 1L,
          kernel_left = epanechnikov_left,
          boundary_left = NA, boundary_right = NA,
          boundary_adjustment = FALSE,
          plot = TRUE)

# Boundary-adjusted Nadaraya-Watson estimator
bw_CV_fun(X = X, Y = Y, kernel = epanechnikov, bw_grid = bw_grid, degree = 0L,
          kernel_left = epanechnikov_left,
          boundary_left = 0, boundary_right = 1,
          boundary_adjustment = TRUE,
          plot = TRUE)

svjaco/lpreba documentation built on March 4, 2022, 12:42 a.m.