Speedglm is a fast version of glm()
SL.speedglm(Y, X, newX, family, obsWeights, maxit = 25, k = 2, ...)
Outcome variable
Training dataframe
Test dataframe
Gaussian or binomial
Observation-level weights
Maximum number of iterations before stopping.
numeric, the penalty per parameter to be used; the default k = 2 is the classical AIC.
Any remaining arguments, not used.
Enea, M. A. R. C. O. (2013). Fitting linear models and generalized linear models with large data sets in R. Statistical Methods for the Analysis of Large Datasets: book of short papers, 411-414.
predict.SL.speedglm
speedglm
predict.speedglm