ImpShrinkage: Improved Shrinkage Estimations for Multiple Linear Regression

Overview

ImpShrinkage is an R package that provides a broad collection of classical and modern shrinkage estimators for multiple linear regression.

These include:

• Unrestricted estimator
• Restricted estimator
• Preliminary test estimator
• Improved preliminary test estimator
• Stein estimator
• Positive-rule Stein estimator

Installation

Install the package from CRAN:

install.packages("ImpShrinkage")

Or install the development version from GitHub:

# install.packages("devtools")
devtools::install_github("mnrzrad/ImpShrinkage")

Quick Start

library(ImpShrinkage)

# Simulated data
set.seed(123)
X <- matrix(rnorm(100 * 5), ncol = 5)
beta <- c(1, 0.5, 0, 0, 0)
y <- X %*% beta + rnorm(100)

# Apply unrestricted and Stein estimators
fit_unres <- shrinkage_unrestricted(X, y)
fit_stein <- shrinkage_stein(X, y)

print(fit_unres)
print(fit_stein)

Authors

• Mina Norouzirad : Lead Developer, Conceptualization
• Danial Mazarei: Developer
• Ricardo Moura : Testing, Website and Integration

Documentation

For detailed documentation, please refer to:

• CRAN Documentation
• GitHub Repository

Citation

If you use PSinference in your research, please cite:

@Manual{ImpShrinkage2023,
  title = {ImpShrinkage: Improved Shrinkage Estimations for Multiple Linear Regression},
  author = {Danial Mazarei and Ricardo Moura},
  year = {2023},
  note = {R package},
  url = {https://cran.r-project.org/package=ImpShrinkage}
}

License

GPL (>= 2)

Contributing

Contributions are welcome! Please see our GitHub repository