R help - Jan 2011 - ls.diag and lsfit very basic question

Hello,

Is the "std.dev" component of ls.diag( lsfit(x,y) ) the sample
standard
deviation of the residuals of the fit?

I have
  ls.diag(lsfit(xx,yy))$std.dev
different from
  sd(lsfit(xx,yy)$residuals)

where xx and yy are vectors of 5 elements.

Regards,

Hi,

Yes, that is basically the idea.  It is defined as:

stddev <- (colSums(as.matrix(resids^2))/(n - p))^0.5

Where n is the number of rows in the residual matrix and p is the rank
of the QR decomposition.  I believe the reason they are slightly
different is that the mean of the residuals is not necessarily exactly
0.

Cheers,

Josh

On Sat, Jan 22, 2011 at 2:54 PM, MM <finjulhich at gmail.com>
wrote:
> Hello,
>
> Is the "std.dev" component of ls.diag( lsfit(x,y) ) the sample
standard
> deviation of the residuals of the fit?
>
> I have
> ?ls.diag(lsfit(xx,yy))$std.dev
> different from
> ?sd(lsfit(xx,yy)$residuals)
>
> where xx and yy are vectors of 5 elements.
>
> Regards,
>
> ______________________________________________
> 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.
>


-- 
Joshua Wiley
Ph.D. Student, Health Psychology
University of California, Los Angeles
http://www.joshuawiley.com/

MM wrote:
> Hello,
> 
> Is the "std.dev" component of ls.diag( lsfit(x,y) ) the sample
standard
> deviation of the residuals of the fit?
> 
> I have
>   ls.diag(lsfit(xx,yy))$std.dev
> different from
>   sd(lsfit(xx,yy)$residuals)
> 
> where xx and yy are vectors of 5 elements.
Compare

     ls.diag(lsfit(xx,yy))$std.dev

with

     sd(resid(lsfit(xx,yy)))*sqrt(4/3)

and think about degrees of freedom.

Peter Ehlers
> 
> Regards,
> 
> ______________________________________________
> 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.