R help - Jul 2010 - strange error : isS4(x) in gamm function (mgcv package). Variable in data-frame not recognized???

Dear all,

I run a gamm with following call :

result <- try(gamm(values~ s( VM )+s( RH )+s( TT )+s( PP
)+RF+weekend+s(day)+s(julday) ,correlation=corCAR1(form=~ day|month
),data=tmp) )"

with mgcv version 1.6.2

No stress about the data, the error is not data-related. I get :

Error in isS4(x) : object 'VM' not found

What so? I did define the dataframe to be used, and the dataframe
contains a variable VM :
> str(tmp)
'data.frame':   4014 obs. of  12 variables:
 $ values : num  73.45 105.45 74.45 41.45 -4.55 ...
 $ dates  :Class 'Date'  num [1:4014] 9862 9863 9864 9865 9866 ...
 $ year   : num  -5.65 -5.65 -5.65 -5.65 -5.65 ...
 $ day    : num  -178 -177 -176 -175 -174 ...
 $ month  : Factor w/ 156 levels
"1996-April","1996-August",..: 17 17
17 17 17 17 17 17 17 17 ...
 $ julday : num  -2241 -2240 -2239 -2238 -2237 ...
 $ weekend: num  -0.289 -0.289 -0.289 0.711 0.711 ...
 $ VM     : num  0.139 -1.451 0.349 0.839 -0.611 ...
 $ RH     : num  55.2 61.4 59.8 64.1 60.7 ...
 $ TT     : num  -23.4 -23.6 -19.5 -16.1 -15.3 ...
 $ PP     : num  6.17 4.27 -4.93 -9.23 -2.63 ...
 $ RF     : Ord.factor w/ 3 levels
"None"<"<2.5mm"<..: 1 1 1 1 1 1 1 1 1 1 ...
 - attr(*, "means")= num

Any idea what I'm missing here?

Cheers
Joris


-- 
Joris Meys
Statistical consultant

Ghent University
Faculty of Bioscience Engineering
Department of Applied mathematics, biometrics and process control

tel : +32 9 264 59 87
Joris.Meys at Ugent.be
-------------------------------
Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php

Dear all,

it gets even more weird. After restarting R, the code I used works
just fine. The call is generated in a function that I debugged using
browser(). Problem is solved, but I have no clue whatsoever how that
error came about. It must have something to do with namespaces, but
the origin is dark. I tried to regenerate the error, but didn't
succeed.

Somebody an idea as to where I have to look for a cause?

Cheers
Joris

On Wed, Jul 28, 2010 at 1:16 PM, Joris Meys <jorismeys at gmail.com>
wrote:
> Dear all,
>
> I run a gamm with following call :
>
> result <- try(gamm(values~ s( VM )+s( RH )+s( TT )+s( PP
> )+RF+weekend+s(day)+s(julday) ,correlation=corCAR1(form=~ day|month
> ),data=tmp) )"
>
> with mgcv version 1.6.2
>
> No stress about the data, the error is not data-related. I get :
>
> Error in isS4(x) : object 'VM' not found
>
> What so? I did define the dataframe to be used, and the dataframe
> contains a variable VM :
>
>> str(tmp)
> 'data.frame': ? 4014 obs. of ?12 variables:
> ?$ values : num ?73.45 105.45 74.45 41.45 -4.55 ...
> ?$ dates ?:Class 'Date' ?num [1:4014] 9862 9863 9864 9865 9866 ...
> ?$ year ? : num ?-5.65 -5.65 -5.65 -5.65 -5.65 ...
> ?$ day ? ?: num ?-178 -177 -176 -175 -174 ...
> ?$ month ?: Factor w/ 156 levels
"1996-April","1996-August",..: 17 17
> 17 17 17 17 17 17 17 17 ...
> ?$ julday : num ?-2241 -2240 -2239 -2238 -2237 ...
> ?$ weekend: num ?-0.289 -0.289 -0.289 0.711 0.711 ...
> ?$ VM ? ? : num ?0.139 -1.451 0.349 0.839 -0.611 ...
> ?$ RH ? ? : num ?55.2 61.4 59.8 64.1 60.7 ...
> ?$ TT ? ? : num ?-23.4 -23.6 -19.5 -16.1 -15.3 ...
> ?$ PP ? ? : num ?6.17 4.27 -4.93 -9.23 -2.63 ...
> ?$ RF ? ? : Ord.factor w/ 3 levels
"None"<"<2.5mm"<..: 1 1 1 1 1 1 1 1 1 1 ...
> ?- attr(*, "means")= num
>
> Any idea what I'm missing here?
>
> Cheers
> Joris
>
>
> --
> Joris Meys
> Statistical consultant
>
> Ghent University
> Faculty of Bioscience Engineering
> Department of Applied mathematics, biometrics and process control
>
> tel : +32 9 264 59 87
> Joris.Meys at Ugent.be
> -------------------------------
> Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
>


-- 
Joris Meys
Statistical consultant

Ghent University
Faculty of Bioscience Engineering
Department of Applied mathematics, biometrics and process control

tel : +32 9 264 59 87
Joris.Meys at Ugent.be
-------------------------------
Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php