LPALS: compute directions using principal asymmetric least squares

Description Usage Arguments Value

View source: R/LPALS.R

Description

compute directions using principal asymmetric least squares

Usage

1
LPALS(X, Y, lambda = 1, tau = seq(0.1, 0.9, by = 0.1))

Arguments

X

A design or model matrix which cannot include categorical variables

Y

A continuous response variable

lambda

A tuning parameter which must be greater than 0

tau

The quantile levels which must be at least two

Value

Returns the directions and their corresponding eigenvalues, and the coefficients of the expectile estimates


ansoale/Pals documentation built on Oct. 5, 2020, 12:02 a.m.