R devel - Jul 2001 - predict.lm(...., se=T), with 1-column model matrix (PR#1018)

#       r-bugs@r-project.org

The problem occurs when the model matrix has a single
column.
> elastic <- data.frame(stretch=c(46,54,48,50,44,42,52),
   distance=c(183,217,189,208,178,150,249))
> elastic.lm <- lm(distance ~ -1 + stretch, data=elastic)
> predict(elastic.lm,se=T)
Error in XRinv^2 %*% rep(res.var, p) : non-conformable arguments

The fix is to replace 
  XRinv <- qr.Q(object$qr)[, p1]
with
  XRinv <- qr.Q(object$qr)[, p1, drop=FALSE]

The statement is five lines above XRinv^2 %*% rep(res.var, p)

John Maindonald.

--please do not edit the information below--

Version:
 platform = i386-pc-mingw32
 arch = x86
 os = Win32
 system = x86, Win32
 status = 
 major = 1
 minor = 3.0
 year = 2001
 month = 06
 day = 22
 language = R

Windows ME 4.90 (build 3000)  

Search Path:
 .GlobalEnv, package:ctest, Autoloads, package:base

John Maindonald               email : john.maindonald@anu.edu.au        
Statistical Consulting Unit,  phone : (6125)3473        
c/o CMA, SMS,                 fax   : (6125)5549  
John Dedman Mathematical Sciences Building
Australian National University
Canberra ACT 0200
Australia

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

john.maindonald@anu.edu.au writes:
> #       r-bugs@r-project.org
> 
> The problem occurs when the model matrix has a single
> column.
> 
> > elastic <- data.frame(stretch=c(46,54,48,50,44,42,52),
>    distance=c(183,217,189,208,178,150,249))
> > elastic.lm <- lm(distance ~ -1 + stretch, data=elastic)
> 
> > predict(elastic.lm,se=T)
> Error in XRinv^2 %*% rep(res.var, p) : non-conformable arguments
> 
> The fix is to replace 
>   XRinv <- qr.Q(object$qr)[, p1]
> with
>   XRinv <- qr.Q(object$qr)[, p1, drop=FALSE]
> 
> The statement is five lines above XRinv^2 %*% rep(res.var, p)
> 
Ick! Thanks.

For a workaround, notice that
> predict(elastic.lm,se=T,new=elastic)
$fit
       1        2        3        4        5        6        7 
188.8946 221.7458 197.1074 205.3202 180.6818 172.4690 213.5330 

$se.fit
[1] 6.423642 7.540797 6.702930 6.982219 6.144353 5.865064 7.261508

$df
[1] 6

$residual.scale
[1] 17.79576


-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._