[R] Problems with Cochrane-Orcutt procedures

[John Fox]

Dear Tolga,

I'm afraid that I don't see an error. (I expect in any event that the
Cochrane-Orcott and Prais estimators are now only of historical interest.)

Regards,
 John

------------------------------
John Fox, Professor
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
web: socserv.mcmaster.ca/jfox


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On
> Behalf Of tolga.i.uzuner at jpmorgan.com
> Sent: June-17-08 11:13 AM
> To: r-help at r-project.org
> Subject: [R] Problems with Cochrane-Orcutt procedures
> 
> Hi John,
> 
> Hi Folks/Prof. Fox,
> 
> I found some code John Fox had written sometime back on the
Cochrane-Orcutt
> and Prais procedures here:
> https://stat.ethz.ch/pipermail/r-help/2002-January/017774.html
> 
> I thought I would try it out and get the following errors below. Was
> wondering if anyone had any immediate opinions why this might be ?
> 
> The linear model is the object regrCMSlm .
> 
> Thanks,
> Tolga
> 
> > regrCMSlm
> 
> Call:
> lm(formula = regrCMS[, 1] ~ regrCMS[, 2])
> 
> Coefficients:
>  (Intercept)  regrCMS[, 2]
>      25.7067       -0.3409
> 
> > summary(regrCMSlm)
> 
> Call:
> lm(formula = regrCMS[, 1] ~ regrCMS[, 2])
> 
> Residuals:
> 09/20/07 11/28/07 02/01/08 04/09/08 06/16/08
>  10.0593   0.3588  -1.1459   0.1340  -9.8520
> 
> Coefficients:
>              Estimate Std. Error t value Pr(>|t|)
> (Intercept)  25.70673    0.85300   30.14   <2e-16 ***
> regrCMS[, 2] -0.34092    0.02205  -15.46   <2e-16 ***
> ---
> Signif. codes:  0 b