R help - Jul 2007 - Error message when running lm() with na.action=NULL

Hi there,

I am trying to run a liner regression using lm with na.action = NULL, but I
am getting an error message. Any ideas as to why this may be happening?
Please see code and error message below:
> reg_test<-lm(yy~.,data=test,na.action=NULL)
Error in lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...) : 
        NA/NaN/Inf in foreign function call (arg 4)


This message and any attachment are confidential and may be privileged or
otherwise protected  from disclosure. If you are not the intended recipient,
please telephone or email the sender and delete this message and any attachment
from your system. If you are not the intended recipient you must not copy this
message or attachment or disclose the contents to any other person. Nothing
contained in the attached email shall be regarded as an offer to sell or as a
solicitation of an offer to buy any  services, funds or products or an
expression of any opinion or views of the firm or its employees. Nothing
contained in the attached email shall be deemed to be an advise of, or
recommendation by, the firm or its employees. No representation is made as to
accuracy, completeness, reliability or appropriateness of the information
contained in the attached email.


	[[alternative HTML version deleted]]

Sorry, I forgot to reply to all..... See Patrick suggestion below. I also
got another idea from Julian Faraway's book on R
> res <- rep(NA,153)
> res[as.numeric(row.names(na.omit(airquality)))] <- gl$res
where gl was a regression and airquality the data. 

Thanks,

Fabian

-----Original Message-----
From: Patrick Burns [mailto:pburns@pburns.seanet.com] 
Sent: Tuesday, July 31, 2007 12:55 PM
To: Garavito,Fabian
Subject: Re: [R] Error message when running lm() with na.action=NULL

Garavito,Fabian wrote:
> I was trying to get a vector of residuals where NA would be assigned 
> to the lm residual vector (in this case reg_test$residuals) whenever 
> NA was present in the sample. For instance,
>
> Obs     fitted  residual
> O1      f1              r1
> O2      f2              r2
> NA      NA              NA
>
> All I am getting is residual vectors for Non Missing observations 
> (i.e. the dimension of the residual vector could be different from 
> that of the response vector)
>
Makes sense.  I think you need to do it yourself.  Perhaps something
like (untested):

test.nona <- na.omit(test)
mod <- lm(y | ., data=test.nona)
allresids <- rep(NA, nrow(test))
allresids[-attr(test.nona, 'na.action')] <- residuals(mod)

Putting this into a function would probably be the way to go.
> Thank you very much for your help. I used to use R a few years ago, 
> but I have not used it in a long while. Moreover, I had been using SAS 
> in the last year and going back to R has been quite difficult. I was 
> actually looking for a good book in R. Any ideas?
>
'good' is questionable and not exactly about R, but S Poetry
is at least free.  Venables and Ripley is a common citation.
Peter Dalgaard's book is given good reviews.  There is a book
that is aimed more towards finance (which I'm assuming you
are in) but I've forgotten what it is nor have I seen it.  Searching
Amazon should probably show you a lot of your choices, and
I haven't heard of any books that generate aversion.

Pat
> Fabian
>
>
> -----Original Message-----
> From: Patrick Burns [mailto:pburns@pburns.seanet.com]
> Sent: Tuesday, July 31, 2007 11:39 AM
> To: Garavito,Fabian
> Subject: Re: [R] Error message when running lm() with na.action=NULL
>
> Doing
>
> na.action=NULL
>
> means that you need to get rid of missing values yourself
> from your data, which I'm guessing you haven't done.
>
>
> Patrick Burns
> patrick@burns-stat.com
> +44 (0)20 8525 0696
> http://www.burns-stat.com
> (home of S Poetry and "A Guide for the Unwilling S User")
>
> Garavito,Fabian wrote:
>
> >Hi there,
> >
> >I am trying to run a liner regression using lm with na.action = NULL, 
> but I
> >am getting an error message. Any ideas as to why this may be happening?
> >Please see code and error message below:
> >
> > 
> >
> >>reg_test<-lm(yy~.,data=test,na.action=NULL)
> >>   
> >>
> >Error in lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...)
:
> >        NA/NaN/Inf in foreign function call (arg 4)
> >
> >
> >This message and any attachment are confidential and may be 
> privileged or otherwise protected  from disclosure. If you are not the 
> intended recipient, please telephone or email the sender and delete 
> this message and any attachment from your system. If you are not the 
> intended recipient you must not copy this message or attachment or 
> disclose the contents to any other person. Nothing contained in the 
> attached email shall be regarded as an offer to sell or as a 
> solicitation of an offer to buy any  services, funds or products or an 
> expression of any opinion or views of the firm or its employees. 
> Nothing contained in the attached email shall be deemed to be an 
> advise of, or recommendation by, the firm or its employees. No 
> representation is made as to accuracy, completeness, reliability or 
> appropriateness of the information contained in the attached email.
>
> >
> >
> >       [[alternative HTML version deleted]]
> >
> >______________________________________________
> >R-help@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
> >and provide commented, minimal, self-contained, reproducible code.
> >
> >
> > 
> >
>
>
>
> This message and any attachment are confidential and may be privileged 
> or otherwise protected from disclosure. If you are not the intended 
> recipient, please telephone or email the sender and delete this 
> message and any attachment from your system. If you are not the 
> intended recipient you must not copy this message or attachment or 
> disclose the contents to any other person. Nothing contained in the 
> attached email shall be regarded as an offer to sell or as a 
> solicitation of an offer to buy any services, funds or products or an 
> expression of any opinion or views of the firm or its employees. 
> Nothing contained in the attached email shall be deemed to be an 
> advise of, or recommendation by, the firm or its employees. No 
> representation is made as to accuracy, completeness, reliability or 
> appropriateness of the information contained in the attached email.


	[[alternative HTML version deleted]]