[R] Three most useful R package

[kMan]

(1) - nlme, lattice, stats
(2) - a usable large-file/out of memory regression package that abstracts
"all" the details of connections & etc from the user, accept perhaps the
initial function call, so I don't have to actually know anything about the
file I'm opening, how big it is, how many lines of data, how much data my
system can load into memory at once without paging or crashing R, etc, but
will still give me parameter estimates for multiple categorical and
continuous predictors on a TB of data in less than a half hour, and can work
with something more interesting than a matrix.

Sincerely,
KeithC.

-----Original Message-----
From: Ralf B [mailto:ralf.bierig at gmail.com] 
Sent: Tuesday, March 02, 2010 1:14 PM
To: r-help at r-project.org
Subject: [R] Three most useful R package

Hi R-fans,

I would like put out a question to all R users on this list and hope it will
create some feedback and discussion.

1) What are your 3 most useful R package? and

2) What R package do you still miss and why do you think it would make a
useful addition?

Pulling answers together for these questions will serve as a guide for new
users and help people who just want to get a hint where to look first. Happy
replying!

Best,
Ralf