[R] fun.aggregate=mean in reshape

[Charilaos Skiadas]

On Feb 12, 2008, at 1:31 PM, [Ricardo Rodriguez] Your XEN ICT Team  
wrote:

> Sundar Dorai-Raj wrote:
>
> Could it we advisable that cast, melt or whatever function we deal  
> with
> throws an more informative error message when this kind of conflicts
> occur? I am guessing this is a pretty frequent situation, so if it is
> not done there must be a number of good reasons!
>
> Just my two cents!
>

I think this is appropriate here:

 > library(fortunes)
 > fortune("dog")

Firstly, don't call your matrix 'matrix'. Would you call your dog  
'dog'? Anyway, it might clash
with the function 'matrix'.
    -- Barry Rowlingson
       R-help (October 2004)


in other words, it is advisable that you don't call anything "mean".  
The error seems to be almost as clear as it could in this case:

Error in get(as.character(FUN), mode = "function", envir = envir) :
   variable "fun" of mode "function" was not found

It tells you that a call to get was attempted looking for a variable  
of mode "function", and such a variable was not found. The problem is  
of course that the call tells you the variable is named "fun", while  
you expected it to be named "mean". But it alerts to the fact that  
the program was expecting a function in a place where it was given  
something else, and there is only one place where you provide it a  
function.
Also, the fact that the call works fine without the fun.aggregate, or  
with a different fun.aggregate, prompts to the problem being the word  
"mean".
Then typing "mean" followed by return would show you, instead of the  
function body for  the function mean, something quite different.
I have to admit that a cursory look at the code for cast does not  
immediately tell me where "fun" shows up, however, so I am slightly  
mystified by the error.

> Thanks for your help,
>
> Ricardo
>
> -- 
> Ricardo Rodríguez
> Your XEN ICT Team

Haris Skiadas
Department of Mathematics and Computer Science
Hanover College