Overview
This R package estimates the inflation and variance of the MLE from a high-dimensional binary regression model. The package supports two methods:
High-dimensional theory (which provides an exact asymptotic distribution assuming the covariates are multivariate Gaussian or sub-Gaussian).
Resized bootstrap methods (which approximates the MLE distribution but does not make distributional assumptions of the covariates)
Getting started
- You can find the package vignette under the Articles tab (https://zq00.github.io/glmhd/articles/my-vignette.html).
- You can read more about the theory of high dimensional logistic MLE, and the methods used in the package here.
- You can read more about the resized bootstrap method here.
Installation
You can install the package using
install.packages("devtools")
devtools::install_github("zq00/glmhd")To install with vignettes, please run
devtools::install_github("zq00/glmhd", build_vignettes = TRUE)Note that the vignette takes about 20 min to knit, so feel free to download the Rmarkdown file and run code line by line.
Function documentation
You can find the function documentations under the Reference tab. To get started, you can take a look at the function adjust_glm, which computes the adjusted coefficient and standard error estimates.
Source code
The source code is located at the Github -> R folder.
Feedback
If you encounter error or would like to provide feedback, please use Github -> Issues to reach us. Thank you!