[R] Calculation of Durbin-Watson p-value

[Achim Zeileis]

On 19 May 2005 04:31:54 -0000 Ramesh Kolluru wrote:

> Sir,  
> I am unable to get the source code for Durbin-Watson test, as I want
> to calculate the p-value for Durbin Watson statistic using
> interpolation method. I sent this mail to r-help, but it was rejected,

Perhaps you should have tried to figure out why it was rejected instead
of sending it directly to BDR... Please *do* read the posting guide at
  http://www.R-project.org/posting-guide.html
on how to ask for help on the R mailing lists.

As for your mail (which does not ask any question, and surely no
specific one): which source code are you referring to? There are at
least two implementations in R (dwtest in package lmtest, durbin.watson
in package car), the source code of both is available in the usual ways.
But maybe you're referring to some other implementation?

Concerning the p values: dwtest() interfaces the pan/gradsol algorithm
which computes the null distribution from a linear combination of
chi-squared variables and also implements a normal approximation.
durbin.watson() computes p values by bootstrapping. So I don't see the
need for implementing an interpolation method (although I have to admit
that it is not clear to me what exactly this means in this context).
Z

> please suggest me some way. I will be highly greatful to you.
>
> Thanks in advance
> Ramesh
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html
>