An R-package for fitting glm's with high-dimensional k-way fixed effects.
This is a preliminary R-package based on the working paper "Fast and Feasible Estimation of Generalized Linear Models with High-Dimensional k-way Fixed Effects" (Stammann, 2018), https://arxiv.org/abs/1707.01815. A special iteratively re-weighted least squares pseudo-demeaning algorithm is implemented, such that the estimation of glm's with high-dimensional fixed effects becomes feasible.
This version of the package is able to estimate poisson and logit models with high-dimensional k-way fixed effects. In near future the package will be extended to probit models and robust/clustered standard errors.
If you have any suggestions for improvements, feel free to contact me (Amrei.Stammann@hhu.de).
The package can be installed by devtools::install_github("amrei-stammann/alpaca", build_vignettes = TRUE).