[Eeglablist] SIFT Memory Issue

Makoto Miyakoshi mmiyakoshi at ucsd.edu
Thu Jul 17 09:14:38 PDT 2014


How many frequencies and latencies do you have in your time-frequency
matrix?

Makoto


On Thu, Jul 17, 2014 at 6:28 AM, Mostafa Rz <mostafa.rouzbahani at gmail.com>
wrote:

> I used 64bit and 8gig RAM but it doesn't work again.
>
>
> On Thursday, July 17, 2014, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:
>
>> Dear Mostafa,
>>
>> Two solutions.
>> 1. You use less resolution in time and/or frequency to save memory.
>> 2. You use 64bit OS and > 8 GB RAM.
>>
>> Makoto
>>
>>
>> On Tue, Jul 15, 2014 at 7:41 AM, Mostafa Rz <mostafa.rouzbahani at gmail.com
>> > wrote:
>>
>>> Hi
>>> when i want to use SIFT options i got memory errors.
>>> My RAM is 4 GIG and i could load data and preprocess and fit a modell to it. But
>>> after that i don't have any free memory to use other options.
>>> For produce it first I load sample data which download it fromhttp://sccn.ucsd.edu/wiki/SIFT. I load only one of them and it use about 80MB
>>> of my RAM.
>>> 2)After that from SIFT toolbox I select preprocessing. It increase my RAM usage
>>> to 220MB.
>>> 3)I select fit AMVAR MODEL from model fiiting and validatin and i decrease the
>>> model order to 3 for less usage of RAM. After this process its take 2010MB of
>>> my RAM.
>>> 4)Estimation of the connectivity is my next step.It's increase the usage a bit
>>> more.
>>> 5)After that I couldn't do anything else. I try to use visualization tab for
>>> creating brain movie or plotting time-freq grid but my RAM starting to overflow
>>> and I get the out of memory error.
>>> I have a question. Is there any way to use my hard disk for saving data after
>>> some process to free my  RAM or any task which free my RAM?
>>>
>>> I have to mention using windows 7 Page File is decrease the process speed very much.
>>>
>>> Thanks for response
>>>
>>>
>>> _______________________________________________
>>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>>> To unsubscribe, send an empty email to
>>> eeglablist-unsubscribe at sccn.ucsd.edu
>>> For digest mode, send an email with the subject "set digest mime" to
>>> eeglablist-request at sccn.ucsd.edu
>>>
>>
>>
>>
>> --
>> Makoto Miyakoshi
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>>
>


-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20140717/c6084817/attachment.html>


More information about the eeglablist mailing list