Best modelling packages in R? In R I use a lot the packages plyr, stringr, Hmisc and ggplot2. Each of these packages take the base code and make functions that are more intuitive and easier to work with. Each of these packages tend to focus on the data munging and exploration side of data analysis. Are there any packages that make working with linear models easier and more intuitive?
 A: There are a number of very handy packages. I'll mention a few.
MASS (which comes with R) has many useful things in it, which I find quite valuable, especially for generalized linear models, but some parts also with normal linear models.
car I use all the time now; it's filled with handy functions for regression models.
rms has already been mentioned, but I also use a number of the helpful functions in
Hmisc, some of which are handy when modelling.
A: It's difficult to fully answer your question without known what you'd like to do with the linear models. But if you just want to run a straight forward regression...
Consider rms, also from Frank Harrell. You'd be looking for lm in that package, although there is a lot more in that. You could use glm(), as part of base.
A: Thomas Yee's package Vector Generalized Linear and Additive Models (VGAM) is an incredibly versatile framework for fitting a wide variety of models. It is moreover accompanied by a comprehensive book. It does, however, lack many of the useful features that 'rms' provide for visualization, residuals, prediction, calibration, and more.
