LSL + Timeflux as a potential experimental infrastructure.
- install conda environment with
conda env create -f environment-mac.yml
(tested on mac), orconda env create -f environment-win.yml
(tested on win10) - (for windows, if timeflux throws an error on pytables, reinstall pytables from conda
conda install -c conda-forge pytables
)
- activate the conda environment
conda activate pylsl
- start LSL stream by running
python send_data.py
, which generates random 8-channel EEG data - in another terminal, start LSL stream by running
python send_marker.py
, which generates random markers in another outlet - in another terminal, start timeflux by running
timeflux -d test_timeflux.yaml
(d for debug)- it will receive the data from LSL
- do some basic sigproc (by adding 1 to the signal)
- stream to a monitor (visible in a browser tab in the following link)
- save the data to a new subdirectory called
./data
in hdf5 format
- timeflux allows creating individual plugins using their template
- some sigproc and ML tools already available as exisiting plugins
- make brain product and vive eye tracking data play well with it