[R] Problem with gam() after R update

[David Winsemius]

On Aug 1, 2011, at 5:01 AM, Przemek Jura wrote:

> Dear group,
> I experience some problems with gam() function after R update to  
> version 2.13.1
> The function in both gam and mgcv packages stopped to work. Before,  
> with the same code I used, everything was fine.

Reports like this often turn out to be inaccurate because either the  
(not offered) code  was not the "same" or the (also not offered) data  
was different. Did you reinstall these packages? How? How many  
versions "up" was the update? sessionInfo()?

> The function from gam package yields following warning:
>
> Residual degrees of freedom are negative or zero.  This occurs when  
> the sum of the parametric and nonparametric degrees of freedom  
> exceeds the number of observations.  The model is probably too  
> complex for the amount of data available

That certainly looks like an informative error message. What do you  
want us to do about it?

>
> while gam() from mgcv crashes R.

A  report of a real "crash" should go to the package maintainer with a  
lot more detail than you have provided above.

>
> Did I miss something?

Perhaps reading the Posting Guide?

>
> Thank you in advance.
>
> PJ
> ______________________________________________
> 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.

David Winsemius, MD
West Hartford, CT