[SciPy-User] Problem with handling big matrices with Windows

[Antonelli Maria Rosaria]

Ok, thank you.
I will try splitting and keep you posted.
Thank you very much for your help !
Best, 
Rosa

On 6/11/14 12:23 PM, "Matthieu Brucher" <matthieu.brucher at gmail.com> wrote:

>You can use whatever tool you want, in the end you will have to use
>directly on undirectly this kind of tools, and this means that you
>won't be able to create your array as is. If you split it in half, you
>may have the same issue in the end, as it seems that you want to
>create it, and then assign another array just as big to this one. Try
>splitting your algorithm in smaller pieces, so that you may even be
>able to parallelize it.
>
>Cheers,
>
>2014-06-11 12:16 GMT+02:00 Antonelli Maria Rosaria
><maria-rosaria.antonelli at curie.fr>:
>> Thank you very much !
>> I just wanted to be sure that there are not other solution.
>> Do Blaze say anything to you ?
>> Otherwise I will split my matrix by two !
>> Best,
>> Rosa
>>
>> On 6/11/14 12:13 PM, "Matthieu Brucher" <matthieu.brucher at gmail.com>
>>wrote:
>>
>>>You will need to construct this array bits by bits, not in one go. No
>>>other solution, as it would be even worse in terms of memory usage.
>>>
>>>2014-06-11 11:12 GMT+01:00 Antonelli Maria Rosaria
>>><maria-rosaria.antonelli at curie.fr>:
>>>> Yes, thank you,.
>>>> But later I will have to assign a matrix to float to fpŠ
>>>> Best,
>>>> Rosa
>>>>
>>>> From: Da¼id <davidmenhur at gmail.com>
>>>> Reply-To: SciPy Users List <scipy-user at scipy.org>
>>>> Date: Wednesday, June 11, 2014 12:07 PM
>>>> To: SciPy Users List <scipy-user at scipy.org>
>>>> Subject: Re: [SciPy-User] Problem with handling big matrices with
>>>>Windows
>>>>
>>>> On 11 June 2014 12:01, Antonelli Maria Rosaria
>>>> <maria-rosaria.antonelli at curie.fr> wrote:
>>>>>
>>>>> I can np.memmap of that size, but when I ask to assign a zeros matrix
>>>>>of
>>>>> the same size to the np.memmap, it gives the same error :
>>>>> fp = np.memmap(filename, dtype='float, mode ='w+', shape(80, 100,
>>>>>384,
>>>>> 285))
>>>>> fp = zeros((80, 100, 384, 285)
>>>>
>>>>
>>>> You are not assigning values, but just replacing the memap object by a
>>>>new
>>>> matrix. You can zero the matrix by doing:
>>>>
>>>> fp = np.memap(...)
>>>> fp[:] = 0
>>>>
>>>> _______________________________________________
>>>> SciPy-User mailing list
>>>> SciPy-User at scipy.org
>>>> http://mail.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>
>>>
>>>
>>>--
>>>Information System Engineer, Ph.D.
>>>Blog: http://matt.eifelle.com
>>>LinkedIn: http://www.linkedin.com/in/matthieubrucher
>>>Music band: http://liliejay.com/
>>>_______________________________________________
>>>SciPy-User mailing list
>>>SciPy-User at scipy.org
>>>http://mail.scipy.org/mailman/listinfo/scipy-user
>>
>> _______________________________________________
>> SciPy-User mailing list
>> SciPy-User at scipy.org
>> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>
>
>-- 
>Information System Engineer, Ph.D.
>Blog: http://matt.eifelle.com
>LinkedIn: http://www.linkedin.com/in/matthieubrucher
>Music band: http://liliejay.com/
>_______________________________________________
>SciPy-User mailing list
>SciPy-User at scipy.org
>http://mail.scipy.org/mailman/listinfo/scipy-user