You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Mar 20, 2024. It is now read-only.
Repository navigation
This repository was archived by the owner on Mar 20, 2024. It is now read-only.
Are you loading the data in Matlab (load_open_ephys_binary.m) or Python (Python3/Binary.py)? Either way, the data will be loaded as a matrix with dimensions of samples x channels. Selecting one value in the second dimension (e.g. data[:,1] in Python) will return data for a specific channel.
Would be nice to have a Matlab reader with optional loading of the selected channels if a data size is too huge and a memory is not enough. Has anyone solved this problem?
For big datasets the MATLAB reader has an optional 'mmap' option that loads the file in memory mapped mode. In this mode the file is not loaded to memory in one go, but chunks are loaded and unloaded automatically by MATLAB as needed. In this case, it would be possible to load the data in this mode and just copy the required channels into a different array.
See help load_open_ephys_binary for more info on this mode.
Activity
jsiegle commentedon Jan 2, 2021
Are you loading the data in Matlab (
load_open_ephys_binary.m) or Python (Python3/Binary.py)? Either way, the data will be loaded as a matrix with dimensions of samples x channels. Selecting one value in the second dimension (e.g.data[:,1]in Python) will return data for a specific channel.Alkohost commentedon Mar 1, 2021
Would be nice to have a Matlab reader with optional loading of the selected channels if a data size is too huge and a memory is not enough. Has anyone solved this problem?
aacuevas commentedon Mar 1, 2021
For big datasets the MATLAB reader has an optional 'mmap' option that loads the file in memory mapped mode. In this mode the file is not loaded to memory in one go, but chunks are loaded and unloaded automatically by MATLAB as needed. In this case, it would be possible to load the data in this mode and just copy the required channels into a different array.
See
help load_open_ephys_binaryfor more info on this mode.Alkohost commentedon Mar 1, 2021
Thanks, this is a good way.