R help - Aug 2008 - how to save an R object selectively?

Hi,
How to save an R object for example a matrix or vector and not all objects
created in a session (which is usually performed by save.image or
q("yes"))?

Best,

Carol



      
	[[alternative HTML version deleted]]

See ?save

On Thu, Aug 7, 2008 at 5:39 PM, carol white <wht_crl at yahoo.com>
wrote:
> Hi,
> How to save an R object for example a matrix or vector and not all objects
created in a session (which is usually performed by save.image or
q("yes"))?
>
> Best,
>
> Carol
>
>
>
>
>        [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Henrique Dallazuanna
Curitiba-Paran?-Brasil
25? 25' 40" S 49? 16' 22" O

On 8/08/2008, at 8:39 AM, carol white wrote:
> Hi,
> How to save an R object for example a matrix or vector and not all  
> objects created in a session (which is usually performed by  
> save.image or q("yes"))?
Two approaches:

(1) Clean up your workspace so that it contains only the objects you  
really want,
before doing q("yes").  Use rm(whatever you don't need). This is  
probably a Good Idea
anyhow.  Focuses the mind wondrously.

I've written myself a little interactive clean-up function:

	clean <- function ()
	{
     		lll <- ls(pos = 1)
     		for (xxx in lll) {
         	pmpt <- paste("remove ", xxx, "? ", sep =
"")
         	ans <- readline(pmpt)
         	if (ans == "y") remove(list = xxx, pos = 1)
     	}
     	invisible()
}

Then just type clean() and answer ``y'' for the objects that you want  
to get rid of.

(2) Save the object that you want to keep in a separate file.  E.g.:

	save(X,file="melvin")

Then, when you want to use the object X, in a new R session, you have to
either (a) load("melvin") or (b) attach("melvin").  Read the
help on
load()
and attach() to see the difference in effect if you are unfamiliar  
with these
functions.

You could also, instead of using save(), use dput:

	dput(X,"melvin")

Then you want ``X'' in a new R session you do

	X <- dget("melvin")

Note that the *name* ``X'' is not maintained by dput() --- only the  
``contents''
of X.  So unlike the usage of load() or attach() (applied to the  
results of save())
you need to *assign* the output of dget() to an object.  Which  
needn't necessarily
be called ``X''.  You could just as easily do

	clyde <- dget("melvin")

Then the ``contents'' of clyde will be the ``contents'' of what
you
used to call X.
Extra flexibility, more chance to confuse yourself! :-)

	cheers,

		Rolf

######################################################################
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}

On 11/08/2008, at 9:57 PM, carol white wrote:
> Thanks for your replies.
>
> So there is no way to add one or more objects to the existing  
> objects stored in .RData?
	Not really.  There is no ``append=TRUE'' argument for save().

	There is a workaround, of course.  There [almost] always is in R:

		attach(".RData",warn.conflicts=FALSE)
		assign("clyde",clyde,pos=2)
		savepos(2)
		detach(2)

	where ``clyde'' is the object you are trying to ``save
selectively''.

	***At this point*** the file .RData will contain the objects that  
were in it
	when you started your R session *and* ``clyde''.

	However, it could easily get mucked up.  In particular, if you were  
to answer
	``y'' to the question ``Save workspace image?'' when you quit,
then .RData
	will get overwritten, and will then contain whatever objects were in  
your
	global environment at the time you quit.

	Hence I would ***STRONGLY ADVISE YOU NOT TO PROCEED IN THIS MANNER***.

	You are trying to use .RData in a way in which it was not intended  
to be used,
	and consequently tying yourself in knots.

	If you want to save objects selectively, save them in a different  
file from .Rdata,
	and then load or append (or dget()) that file, as I suggested in my  
previous email.

	Use the facilities of R in the way that they were intended to be  
used, rather than
	trying to force your own inappropriate paradigm upon R.  You'll find  
it more efficient
	to do things R's way rather than trying to force R to do things your  
way.  R gets
	it right.  [Almost] always!

		cheers,

			Rolf Turner

P. S.  ``Almost always'' == ``except on a set of measure/probability  
zero''.  :-)

######################################################################
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}