[R] lm with predictor matrix

Liaw, Andy andy_liaw at merck.com
Tue May 11 03:48:30 CEST 2004


Try:

x = matrix(runif(30), 10, 3)
newx = matrix(runif(15), 5, 3)
y = rnorm(10)

fit = lm(y ~ x)
newyhat = predict(fit, newdata=data.frame(x=I(newx)))

HTH,
Andy

> From: Claus Gwiggner
> 
> Hello,
>   
>   can someone give me an example of how to fit a linear model 
> where the
>   response is a matrix, please? (other than a loop over lm calls)
>   The ?lm points me to model.matrix but it is not clear to me 
> how to use a
>   model matrix in an lm call.
>  
>   fit<-lm(headmatrix ~ tailmatrix) causes problems as soon as I
>  use predict(fit,newframe). The variables in tailmatrix are 
> not found in
>  newframe.
>  
>   Thanks,
>   
>   Claus




More information about the R-help mailing list