C++ implementation of the Iteratively Re-Weighted Least Squares algorithm (IRLS) for generalized linear models (GLM) IRLS is free software, you can redistribute it and/or modify it under the terms of the GNU General Public License. The GNU General Public License does not permit this software to be redistributed in proprietary programs. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. ** Install and usage prerequisite: GNU Scientific Library (GSL) and C++ compiler (for instance g++) To test, run `R --vanilla < test.R`. This will: - simulate data; - run the built-in glm() function (output shown on stdout); - save data in files (test_y.dat, test_X.dat and test_offset.dat); - compile c++ code (test.cpp) and run IRLS (output also shown on stdout). ** Extension to other GLM To add a new link function, create a class mimicking "LogLink". The new class should inherit from "LinkFunc" (base class) and implement all required methods. ** References Dobson, A. J. An Introduction to Generalized Linear Models (Chapman and Hall/CRC, 2002), second edn. URL http://www.worldcat.org/isbn/1584881658. McCullagh, P. & Nelder, J. A. Generalized Linear Models. Monographs on Statistics & Applied Probability (Chapman & Hall/CRC, London, United Kingdom, 1989), second edition edn. URL http://www.worldcat.org/isbn/0412317605.