[Numpy-discussion] Annoyance of memap rray with multiprocessing.Pool.applay_async

[Robert Kern]

On Sat, Apr 3, 2010 at 14:29, Nadav Horesh <nadavh at visionsense.com> wrote:
>
> The following script generate the following error on every loop iteration in the function average:
>
> Exception AttributeError: AttributeError("'NoneType' object has no attribute 'tell'",) in <bound method memmap.__del__ of memmap(xxxxxx)> ignored
>
> where xxxxxx is a scalar (the array sum).
>
> I get this error with numpy1.4 on a linux64 (dual core) machine. A winXP/Pentium4 with 2GB Ram could not run it since it explode the memory.
>
>  Any idea what is the origin on the error (my interset is the linux box)?
>

memmap instances don't pickle. Don't pass them as arguments to
apply_async() or related functions. Instead, pass the filename and
other arguments to memmap() and reconstruct the arrays in each
process.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco