dpkde
PURPOSE
Penalised density on hyperplane: Projection index for MDH
SYNOPSIS
function [df, dkde] = dpkde(x, proj, h, alpha, eta, epsilon)
DESCRIPTION
Penalised density on hyperplane: Projection index for MDH
[DF, DKDE] = DPKDE(X, PROJ, H, ALPHA, ETA, EPSILON)
Inputs:
(X): location of split along the projected data
(PROJ): Univariate projection of dataset (X*v)
(H): bandwidth parameter
(ALPHA): range over which minimisers of 1D density are sought
(ETA): Term in penalty function controlling maximum distance between minimisers of the
kde and the penalised density integral (recommended value: 0.01)
(EPSILON): Term in penalty function controlling smoothness (recommended value: 1)
Output:
(DF): derivative of one-dimensional penalised density on hyperplane w.r.t. split point (x)
(DKDE): derivative of one-dimensional kernel density estimator w.r.t. split point (x)CROSS-REFERENCE INFORMATION
This function calls:- f_md Penalised density of optimal hyperplane with normal vector (V)
Generated on Tue 17-Jul-2018 18:58:09 by m2html © 2005