hridge: Heteroskedastic Ridge Regression

Description Usage Arguments Value Author(s)

Description

Heteroskedastic Ridge Regression

Usage

1
2
hridge(y, X, lambda, predictor_weights = "varbased",
  optim_method = "Nelder-Mead")

Arguments

y

Vector of response variable

X

Matrix of predictor variables

lambda

Value of ridge (L2) regularization penalty

predictor_weights

Vector of length ncol(X) with weights for predictor variables. Defaults to "varbased", in which case weights are based on the variance of parameter estimates obtained through a set of univaraite regressions.

optim_method

Optimization algorithm, passed to optim()

Value

A list with parameter estimates, fitted values and multiple R-squared.

Author(s)

Michal Oleszak


MichalOleszak/momisc documentation built on May 29, 2019, 3:02 a.m.