[R] r error

[Marc Schwartz]

On Tue, 2006-10-24 at 11:54 -0400, Stanfield, Les (MNR) wrote:
> I hope you can help me.  I am new to "R" and am working through the
> example of quantile regression analysis in the paper by Roger Koenker
> (quantile regression in R: a Vignette.  I got to page 6 and in trying to
> execute the following code I received the error below. I have searched
> on the web and while I have found similar error messages, I have not
> been able to ascertain the root cause.  Any assistance would be
> appreciated.  Basically, is this a bug in the program or in the coding,
> or does this function not work in the versions that I have downloaded.
> I have downloaded: R 2.4.0 and quantreg 4.0.1 and SparseM.  Am I missing
> something?
> 
>  
> 
> > xx <- x -mean(x)
> 
> > fit1 <- summary(rq(y ~ xx, tau = 2:98/100))
> 
> Error in model.frame(formula, rownames, variables, varnames, extras,
> extranames,  : 
> 
>         variable lengths differ (found for 'xx')


The error is indicating that the number of elements in 'y' is not the
same as in 'xx':

> y <- rnorm(50)
> xx <- rnorm(49)
> lm(y ~ xx)
Error in model.frame(formula, rownames, variables, varnames, extras,
extranames,  :
        variable lengths differ (found for 'xx')


Thus, check the nature of 'y' and 'xx':

> str(y)
 num [1:50] -0.902  1.002 -1.064  2.630  2.193 ...

> str(xx)
 num [1:49]  1.590 -2.354  0.166  0.332 -0.547 ...


And that should reveal the source of the error.

HTH,

Marc Schwartz