LS: Least Squares Estimation for Grouped Data with Ridge...

In DLMRMV: Distributed Linear Regression Models with Response Missing Variables

Least Squares Estimation for Grouped Data with Ridge Regularization

Description

This function implements the least squares estimation for grouped data, supporting ridge regression regularization. It can handle missing data and returns regression coefficients and the sum of squared residuals for each group.

Usage

LS(d, yidx, Xidx, n, lam = 0.005)

Arguments

d	A data frame containing dependent and independent variables.
yidx	The column index of the dependent variable.
Xidx	The column indices of the independent variables.
n	A vector of starting indices for the groups.
lam	Regularization parameter for ridge regression, default is 0.005.

Value

A list containing the following elements:

beta	A matrix of regression coefficients for each group.
SSE	The sum of squared residuals for each group.
df	The sample size for each group.
gram	The Gram matrix for each group.
cgram	The Cholesky decomposition result for each group.
comm	An unused variable (reserved for future expansion).

Examples

# Example data
set.seed(123)
n <- 1000
p <- 5
d <- list(all = cbind(rnorm(n), matrix(rnorm(n*p), ncol=p)))

# Call the LS function
result <- LS(d, yidx = 1, Xidx = 2:(p + 1), n = c(1, 300, 600, 1000))

# View the results
print(result$beta)  # Regression coefficients
print(result$SSE)   # Sum of squared residuals

DLMRMV documentation built on April 12, 2025, 1:25 a.m.