One other important point in "selling" R is that its use results in
analysts doing a better job. I find that many SAS users for example
routinely assume that all covariable effects are linear because it is
messy to do otherwise in SAS. S has a natural modeling language for
flexible nonlinear effects (primarily because predictors can be
automatically-generated matrices containing basis functions) in addition
to computing dummy variables "secretly" and computing product terms for
interactions on the fly.
I prefer open source software for many reasons, but a major reason that
R is superior for statistical analysis and graphics today is its
functionality and extendibility. This leads to better analyses.
Frank Harrell
______________________________________________
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
