R help - Oct 2010 - C code called from R: reallocate arrays passed as arguments?

Hello,

I used to use a (somehow complicated) code in C and I have tried to adapt it
to be called from R without making an excessive amount of changes. I am
afraid I cannot post the code here, but I am going to try to describe the
problem as accurately as possible.

I have some arguments that are most of the cases required for internal C
operations only; but as they are needed a few times as arguments, they are
still passed as arguments. Another argument is a "switch" (with value
0 or
1) that states whether these "arguments" are really needed as
arguments or
not. When such switch is 0 (says "no"), the C code "feels
free" to
manipulate these "arguments", and in particular it tries to modify its
extension (reallocate).

To be clear:

I define the "argument" in R: values1 <- double(size)
values1 is passed as an argument in ".C".
If the switch is set to 0 (says that "values1" is not a real
argument), the
C function tries to reallocate the pseudo-argument values1:

double *tmp2 = realloc(values1, new_size);
        if(tmp2 != NULL){
            values1 = tmp2;
        }
        else{
            Rprintf("\nFATAL ERROR: Could not allocate memory for the
array\n");
            exit(2);
        }

But then, when calling the function ".C", I get *** error for object
...:
pointer being reallocated was not allocated

Is there any workaround for this?

Thank you very much in advance,

David
Can you please Cc to me any replies, just in case I may miss any of them
among the whole amount of emails :-) ?

	[[alternative HTML version deleted]]

On 06/10/2010 10:53 AM, David wrote:
> Hello,
>
> I used to use a (somehow complicated) code in C and I have tried to adapt
it
> to be called from R without making an excessive amount of changes. I am
> afraid I cannot post the code here, but I am going to try to describe the
> problem as accurately as possible.
>
> I have some arguments that are most of the cases required for internal C
> operations only; but as they are needed a few times as arguments, they are
> still passed as arguments. Another argument is a "switch" (with
value 0 or
> 1) that states whether these "arguments" are really needed as
arguments or
> not. When such switch is 0 (says "no"), the C code "feels
free" to
> manipulate these "arguments", and in particular it tries to
modify its
> extension (reallocate).
You should be using .Call rather than .C if you need to modify 
allocations.  You should treat arrays passed via .C as fixed size.

If you want to stick with the .C convention, then I think you'll need 
two calls:  one to return the required sizes, and a second one with new 
vectors allocated at those sizes.  It's often easier to learn how .Call 
works.

Duncan Murdoch
> To be clear:
>
> I define the "argument" in R: values1<- double(size)
> values1 is passed as an argument in ".C".
> If the switch is set to 0 (says that "values1" is not a real
argument), the
> C function tries to reallocate the pseudo-argument values1:
>
> double *tmp2 = realloc(values1, new_size);
>          if(tmp2 != NULL){
>              values1 = tmp2;
>          }
>          else{
>              Rprintf("\nFATAL ERROR: Could not allocate memory for the
> array\n");
>              exit(2);
>          }
>
> But then, when calling the function ".C", I get *** error for
object ...:
> pointer being reallocated was not allocated
>
> Is there any workaround for this?
>
> Thank you very much in advance,
>
> David
> Can you please Cc to me any replies, just in case I may miss any of them
> among the whole amount of emails :-) ?
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> 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.