R help - May 2008 - Collection of lm()s

Hello,

I would like to create a subscriptable collection (presumably a list) 
of lm() models.

I have a data frame DX containing 6 groups of data.  The general idea 
is (NOT RUN) ...

for (i in 1:6)
{       DXS = subset( DX, <whatever>);
         LMX[ i] = lm( <formula>, data = DXS);
}

Now access model results by subscript ... e.g. coefficients( LMX[ 
2]).  Or would it be [[ 2]]?

I have experimented with various schemes, attempting to 
"pre-allocate" a list etc. without success.

Also, I assume that lm() does not make a copy of input data?  That 
is, if I want "predict( LMX[ i])", I would have to retain a copy the 
associated DXS subsets?

TIA,

Chip Barnaby



---------------------------------------------------------
Chip Barnaby                   cbarnaby at wrightsoft.com
Vice President of Research
Wrightsoft Corp.               781-862-8719 x118 voice
131 Hartwell Ave               781-861-2058 fax
Lexington, MA 02421         www.wrightsoft.com

On Mon, 12 May 2008, Chip Barnaby wrote:
> Hello,
>
> I would like to create a subscriptable collection (presumably a list) of
lm()
> models.
>
> I have a data frame DX containing 6 groups of data.  The general idea is
(NOT
> RUN) ...
>
> for (i in 1:6)
> {       DXS = subset( DX, <whatever>);
>        LMX[ i] = lm( <formula>, data = DXS);
> }
>
> Now access model results by subscript ... e.g. coefficients( LMX[ 2]).  Or 
> would it be [[ 2]]?
>
> I have experimented with various schemes, attempting to
"pre-allocate" a list
> etc. without success.
You want

LMX <- list(6)
...
LMX[[1]] <- lm( <formula>, data = DX, subset=<whatever>)

(No trailing ';' needed, [[]] not [] for list elements.)

> Also, I assume that lm() does not make a copy of input data?
It may.  This is controlled by the 'model' argument which defaults to
TRUE
(in recent versions of R: I think it once defaulted to FALSE).
> That is, if I 
> want "predict( LMX[ i])", I would have to retain a copy the
associated DXS
> subsets?
It does retain the fitted values:  see the 'Value' section of ?lm.
>
> TIA,
>
> Chip Barnaby
>
>
>
> ---------------------------------------------------------
> Chip Barnaby                   cbarnaby at wrightsoft.com
> Vice President of Research
> Wrightsoft Corp.               781-862-8719 x118 voice
> 131 Hartwell Ave               781-861-2058 fax
> Lexington, MA 02421         www.wrightsoft.com
>
> ______________________________________________
> 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.
>
-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

try something like the following:

x <- runif(1000, -3, 3)
y <- 2 + 3 * x + rnorm(1000)
f <- gl(10, 100)

lm.lis <- vector("list", 10)
for (i in 1:10) {
     lm.lis[[i]] <- lm(y ~ x, subset = f == as.character(i))
}

sapply(lm.lis, coef)
sapply(lm.lis, fitted)


I hope it helps.

Best,
Dimitris

----
Dimitris Rizopoulos
Biostatistical Centre
School of Public Health
Catholic University of Leuven

Address: Kapucijnenvoer 35, Leuven, Belgium
Tel: +32/(0)16/336899
Fax: +32/(0)16/337015
Web: http://med.kuleuven.be/biostat/
      http://www.student.kuleuven.be/~m0390867/dimitris.htm


Quoting Chip Barnaby <cbarnaby at wrightsoft.com>:
> Hello,
>
> I would like to create a subscriptable collection (presumably a list)
> of lm() models.
>
> I have a data frame DX containing 6 groups of data.  The general idea
> is (NOT RUN) ...
>
> for (i in 1:6)
> {       DXS = subset( DX, <whatever>);
>         LMX[ i] = lm( <formula>, data = DXS);
> }
>
> Now access model results by subscript ... e.g. coefficients( LMX[ 2]).
> Or would it be [[ 2]]?
>
> I have experimented with various schemes, attempting to
"pre-allocate"
> a list etc. without success.
>
> Also, I assume that lm() does not make a copy of input data?  That is,
> if I want "predict( LMX[ i])", I would have to retain a copy the
> associated DXS subsets?
>
> TIA,
>
> Chip Barnaby
>
>
>
> ---------------------------------------------------------
> Chip Barnaby                   cbarnaby at wrightsoft.com
> Vice President of Research
> Wrightsoft Corp.               781-862-8719 x118 voice
> 131 Hartwell Ave               781-861-2058 fax
> Lexington, MA 02421         www.wrightsoft.com
>
> ______________________________________________
> 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.


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm