under certain conditions, model.matrix appears to lack one column (PR#646)

Dear R Team,

# Summary of the problem: setting contrasts as 

   > contrasts(g) <- contr.treatment 

or > contrasts(g) <- matrix(c(1,-1,0),ncol=1)
(i.e. without quotes around `contr.treatment' or `contr.sum', etc.)
and fitting an lm model without an intercept results in a model matrix
that lacks one column.

Date: Mon, 28 Aug 2000 07:28:19 +0100 (BST)
From: Prof Brian D Ripley <ripley@stats.ox.ac.uk>
To: rnassar@duke.edu
Cc: r-devel@stat.math.ethz.ch, R-bugs@biostat.ku.dk
Subject: Re: [Rd] under certain conditions, model.matrix appears to lack one column (PR#646)

On Sun, 27 Aug 2000 rnassar@duke.edu wrote:

Dear R Team,

# Summary of the problem: setting contrasts as 

   > contrasts(g) <- contr.treatment 

or > contrasts(g) <- matrix(c(1,-1,0),ncol=1)
(i.e. without quotes around `contr.treatment' or `contr.sum', etc.)
and fitting an lm model without an intercept results in a model matrix
that lacks one column.

Right, but first is not allowed by (?contrasts)

     ctr: either a matrix whose columns give coefficients for contrasts
          in the levels of `x', or the (quoted) name of a function
          which computes such matrices.

and the second is arguably correct: you chose to use only two columns.
However, the behaviour is different in the development version, 
which does not code x in models such as this, and so does not
have the problem.

That is not precisely what S does, but seems a sane way of interpreting the
contradictory information given (which is that you don't want an intercept
and that you only want two columns for the coding matrix for x!  That
applies to the function case too: there the coding matrix is evaluated
before knowing if contrast are required, that is using argument contrasts =
FALSE).

-- 
Brian D. Ripley,                  ripley@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

Dear Professor Ripley,

Thank you very much for your kind explanation.  If I may lamely say
something in my defence, even as I apologize for my error: I mistook the
sentence "the (quoted) name of a function" to mean "optionally quoted"
because of the parentheses surrounding "quoted", and was encouraged by the
fact that R did not complain when I gave the unquoted form and that it did
implement a change (although not the desired one), and (even more lamely)
the fact that none of the examples on the help page show the function
quoted, although I now realize that in the example the function has an
argument, and is not given simply the name of the function. 

Again, thank you for taking the time to reply and point out my mistake.

Rashid Nassar


On Mon, 28 Aug 2000, Prof Brian D Ripley wrote:

Date: Mon, 28 Aug 2000 07:28:19 +0100 (BST)
From: Prof Brian D Ripley <ripley@stats.ox.ac.uk>
To: rnassar@duke.edu
Cc: r-devel@stat.math.ethz.ch, R-bugs@biostat.ku.dk
Subject: Re: [Rd] under certain conditions, model.matrix appears to lack one column (PR#646)

On Sun, 27 Aug 2000 rnassar@duke.edu wrote:

Dear R Team,

# Summary of the problem: setting contrasts as 

   > contrasts(g) <- contr.treatment 

or > contrasts(g) <- matrix(c(1,-1,0),ncol=1)
(i.e. without quotes around `contr.treatment' or `contr.sum', etc.)
and fitting an lm model without an intercept results in a model matrix
that lacks one column.

Right, but first is not allowed by (?contrasts)

     ctr: either a matrix whose columns give coefficients for contrasts
          in the levels of `x', or the (quoted) name of a function
          which computes such matrices.

and the second is arguably correct: you chose to use only two columns.
However, the behaviour is different in the development version, 
which does not code x in models such as this, and so does not
have the problem.

That is not precisely what S does, but seems a sane way of interpreting the
contradictory information given (which is that you don't want an intercept
and that you only want two columns for the coding matrix for x!  That
applies to the function case too: there the coding matrix is evaluated
before knowing if contrast are required, that is using argument contrasts =
FALSE).

-- 
Brian D. Ripley,                  ripley@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