R help - Dec 2002 - Crawley's book on S-Plus and one strangeness

Hi,

I have got to my hands an excellent book by Michael J. Crawley
``Statistical Computing: An Introduction to Data Analysis using
S-Plus'' (John Wiley & Sons, Ltd, ISBN 0-471-56040-5). Its beauty
for me is in the fact, that it is more of ``An Introduction to
Data Analysis'' than ``using S-Plus'', but I guess that it may
be
of interest for many others.

Most of the examples in the book are however taken from S-Plus
and using datasets provided with it. Is there anywhere a copy of
these datasets available for R?

And one small question aside: I was very much surprised (in this
book as well as on this list) how many times people use
sqrt(var(x)) when what they want to say (IMHO) is sd(x). Is it
just a macho way to show that I understand more complicated
things, or is there any real difference between the two?

Have a nice day,

	Matej

-- 
Matej Cepl, matej at ceplovi.cz,
Finger: 89EF 4BC6 288A BF43 1BAB  25C3 E09F EF25 D964 84AC
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
Science is meaningless because it gives no answer to our
question, the only question important to us: ``What shall we do
and how shall we live?''
    -- Lev Nikolaevich Tolstoy

matej at ceplovi.cz (Matej Cepl) writes:
> I have got to my hands an excellent book by Michael J. Crawley
> ``Statistical Computing: An Introduction to Data Analysis using
> S-Plus'' (John Wiley & Sons, Ltd, ISBN 0-471-56040-5). Its
beauty
> for me is in the fact, that it is more of ``An Introduction to
> Data Analysis'' than ``using S-Plus'', but I guess that it
may be
> of interest for many others.
> Most of the examples in the book are however taken from S-Plus
> and using datasets provided with it. Is there anywhere a copy of
> these datasets available for R?
Can you be more specific?  Which datasets?
> And one small question aside: I was very much surprised (in this
> book as well as on this list) how many times people use
> sqrt(var(x)) when what they want to say (IMHO) is sd(x). Is it
> just a macho way to show that I understand more complicated
> things, or is there any real difference between the two?
The var function was available in S long before the sd function was
introduced and many 'old-timers' instinctively use sqrt(var(x)) rather
than sd(x).  The sd function ends up calling sqrt(var(x, na.rm na.rm)) when
argument x is a vector.

On Mon, 2 Dec 2002, Matej Cepl wrote:
> I have got to my hands an excellent book by Michael J. Crawley
> ``Statistical Computing: An Introduction to Data Analysis using
> S-Plus'' (John Wiley & Sons, Ltd, ISBN 0-471-56040-5). Its
beauty
> for me is in the fact, that it is more of ``An Introduction to
> Data Analysis'' than ``using S-Plus'', but I guess that it
may be
> of interest for many others.
> Most of the examples in the book are however taken from S-Plus
> and using datasets provided with it. Is there anywhere a copy of
> these datasets available for R?
They are not in S-PLUS.  They are Crawley's own, but he does not
say where they are.  Recently I managed to find them via the publisher's
web site, which links to

http://www.bio.ic.ac.uk/research/mjcraw/statcomp/
> And one small question aside: I was very much surprised (in this
> book as well as on this list) how many times people use
> sqrt(var(x)) when what they want to say (IMHO) is sd(x). Is it
> just a macho way to show that I understand more complicated
> things, or is there any real difference between the two?
sd(x) does not exist in S.  There is stdev(x) in more recent versions of
S-PLUS.  You need a better introductory guide!

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595