[R-sig-finance] R vs. S-PLUS

[David Kane]

My Newletters Etc. writes:
 > I'm wondering if you might be willing to share some specifics about
 > your use of R.  

Of course. I am a real R evangalist.

 > For example, what types of analysis have you done
 > with R?  

I work in quantitive global equity modeling. How much is a share of
IBM worth? For me, the power in R is not so much that the statistical
tools are fancier than what one find in SAS or Stata or whatever ---
although this is often the case --- but that the programming language
is richer and the production tools (especially packages and test
cases) are so easy to use.

 > Have you used pre-defined "packages" or have you "rolled
 > your own?"  

I use all sorts of R packages but, for the actual financial analysis
parts, have had to role my own. It is on my to-do list to more fully
explore things like Rmetrics. I am unaware of any packages devoted to
the sort of stuff that I need to do regularly. As an example, I am
today calculating growth rates of various sorts. Which company has the
fastest growing sales? The answer to this question depends on all
sorts of sticky points that reasonable people can disagree about.

 > Do you have any pointers about how to most effectively
 > approach learning R?  Hopefully these questions are sufficient
 > to give you a feel for the direction of my inquiry.

The standard R documentation is a fine place to begin. I would start
by reading An Introduction to R cover to cover (while doing the
exercises) and then, depending on your level of computer experience,
going on to Writing R Extensions.

I encourage you to use R, especially in academics. To the extent that
you believe, as I do, that research should be public and replication
easy (see: http://gking.harvard.edu/replrepl/replrepl.html), R
provides the perfect tool.


-- 
David Kane
Kane Capital Management