[Numpy-discussion] memory usage of numpy arrays

[Robert Kern]

On Thu, Mar 4, 2010 at 19:10, kaby <kackvogel at gmail.com> wrote:
>
> Hi.
> I am using numpy arrays and when constructing an array I get a  "cannot
> allocate memory for thread-local data: ABORT"
> The array i'm constructing is
> zeros((numVars, 2, numVars, 2), dtype=float) Where numVars is at about 2000.
>
> I was expecting the memory usage to be
> 2000*2000*2*2*8Bytes=128.000.000Bytes=128MBytes
>
> So why is that happening? What am I missing?

It's not having a problem allocating the memory itself. There are some
things under the covers that use thread-local storage, and numpy is
apparently not able to do those things. Can you show us a complete,
minimal, self-contained script that fails? Can you please run that
script and copy-and-paste the full traceback? Can you tell us what
platform you are on and how you built numpy? If you did not build
numpy yourself, please tell us exactly where you got your binaries
from (the URL to the package itself is necessary). Thanks.

-- 
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