Kernel: Kernel density estimation.

In lakerwsl/WUNT: Covariate balancing by uniform transformer

Description

This function calculates the kernel density estimates of x with kernel choice kernel and bandwidth h. If the number of covariates is large, higher order kernels are recommended.

Of limited interest to most users; function Kernel() would typically be called by some other functions.

Usage

Kernel(x,h,kernel='gaussian')

Arguments

x	Data matrix or data vector to calculate the kernel estimate.
h	The bandwidth to be used.
kernel	Type of kernel. kernel can be chosen from 'gaussian','gaussian4','gaussian6','gaussian8','gaussian10','epanechnikov','uniform','triangular','triweight','tricube','biweight',cosine','silverman'

Value

A vector of the estimated density values.

Examples

Kernel(x=cbind(runif(10),rnorm(10)),h=1/10,kernel='gaussian')

lakerwsl/WUNT documentation built on Sept. 13, 2020, 12:15 a.m.