Re: R or SAS?
- From: Richard Ulrich <Rich.Ulrich@xxxxxxxxxxx>
- Date: Sat, 23 Feb 2008 22:23:22 -0500
On Sat, 23 Feb 2008 16:42:03 -0600, David Winsemius
<doe_snot@xxxxxxxxxxx> wrote:
Francogrex <franco@xxxxxxxx> wrote in news:8022cc97-e247-4cf2-a11a-
be7c5675ff7f@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:
On Feb 22, 9:10 pm, Paige Miller <paige.mil...@xxxxxxxxx> wrote:
Pardon me for asking you a question ... but often times, the choice of
which one is better for your department depends on what you want to do
with the software. And you didn't really tell us what you want to do,
hence my question ... what do you want to do with the software?
Ok, I should have given a bit more details. Well it's mostly writing
programs for bayesian statistical analysis (eg MCMC) like for example
hierarchical bayesian logistic regression, or bayesian meta-analysis
of several studies... I can write such programs in R. In addition,
doing routine work for example statistical summaries of regular data
outputs (something that can be presented weakly to the board possible
with graphics)... Thanks
Shirley, you want summaries that ca be presented forcefully.
And especially if you want good graphics capacity, you will choose R.
Judging by the SAS newsgroup of 10 years ago,
*the* major application of SAS was writing reports
for corporations... Most of the questions were about
formatting or using printers.
For a "Programming language", I think I would have
assumed the superiority of R.
I know that SAS has some sort of programming language,
but I've always thought that the statistical forte of SAS,
versus R, was its Procedures that were integrated in a
conventional stat-pack form. And they are generally
well-known and of good repute. Another strong point, for some
users, was that it could read tapes and other files from almost
any format. A couple of problem with using SAS were that its
default-options are seemingly randomly chosen, and its
documentation is (or was) massive and difficult to use.
I've been a user of SPSS, including a little "programming" -
I would possibly be willing to develop a single program to
do an application in the SPSS programming language; but it
would be too slow and awkward to use as a programming
development language.
I don't know whether SAS has the same drawback, but I suspect
it would. I would be interesting in hearing about experiences.
--
Rich Ulrich
http://www.pitt.edu/~wpilib/index.html
.
- References:
- R or SAS?
- From: Francogrex
- Re: R or SAS?
- From: Paige Miller
- Re: R or SAS?
- From: Francogrex
- Re: R or SAS?
- From: David Winsemius
- R or SAS?
- Prev by Date: Re: What is the posterior distribution of a uniform prior lie on intervel
- Next by Date: Best fitting model
- Previous by thread: Re: R or SAS?
- Next by thread: Re: R or SAS?
- Index(es):
Relevant Pages
|
