[Eeglablist] pop_resample out of memory?

Andreas Widmann widmann at uni-leipzig.de
Fri May 19 01:20:58 PDT 2017


Hi Agnieszka,

m should be 67 using the default parameters for the anti-alias filter when downsampling 512 to 256 Hz. Nothing to get out of memory, actually. Do you use custom cutoff or transition band width for the anti-alias filter (cf or df)?

Could you please make a file available (e.g. Dropbox; the bugtracker is still offline; btw, any news on this?) affected by the problem? Could you please send minimal sample code (relevant output of eegh in case working with the GUI) and information on OS, EEGLAB and MATLAB versions?

Best,
Andreas

> Am 17.05.2017 um 10:08 schrieb Agnieszka Zuberer <azuberer at googlemail.com>:
> 
> Dear eeglab-community,
> 
> While trying to down-sample my continuous data set of only 3 min recording with an original sampling rate of 512 to 256, pop_resample() tells me I am out of memory. I tried to resample with and without previous low pass filtering of 45 Hz.
> 
> 
> this is my memory capacity
> 
> Maximum possible array:      2901 MB (3.042e+09 bytes) *
> Memory available for all arrays:      2901 MB (3.042e+09 bytes) *
> Memory used by MATLAB:      6901 MB (7.236e+09 bytes)
> Physical Memory (RAM):      7888 MB (8.271e+09 bytes)
> 
> *  Limited by System Memory (physical + swap file) available.
> 
> 
> 
> The code stucks exactly here (bold) in the windows function
> 
> % Kaiser
> function w = kaiser(m, a)
>     if nargin < 2 || isempty(a)
>         a = 0.5;
>     end
>     m = [-1:2 / (m - 1):1]';
>     w = besseli(0, a * sqrt(1 - m.^2)) / besseli(0, a);
> end
> 
> 
> Any help would be greatly appreciated.
> 
> 
> Kind regards
> 
> Agnieszka
> _______________________________________________
> 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



More information about the eeglablist mailing list