Python kernel dying when running jupyter notebooks with docker

[AABBASI89]

I have started exploring this tool to detect CS recorded in rats. I was using the tutorial notebook to understand how to execute the steps of this algorithm but my kernel keeps on dying as soon as I launch run training cell.

Also on a side note, I was wondering if I can just trail a network using the data provided by you and then use that trained network to label CS on my dataset. Do you think this will be possible?

Also I don't have raw wide band signal. I only have high pass spyking data sampled at 25 kHz and low pass LFP data sampled at 1 kHz. Is it absolutely necessary to have both spyking data and LFP data have the same sampling frequency?

[jobellet]

I have started exploring this tool to detect CS recorded in rats. I was using the tutorial notebook to understand how to execute the steps of this algorithm but my kernel keeps on dying as soon as I launch run training cell.

Maybe your computer is not powerful enough. I would suggest that you use the colab option.

Also on a side note, I was wondering if I can just trail a network using the data provided by you and then use that trained network to label CS on my dataset. Do you think this will be possible?

Yes, it can work if your data looks similar to the one provided with our paper. You should anyway always double-check the output of the algorithm to see if it works.

Also I don't have raw wide band signal. I only have high pass spyking data sampled at 25 kHz and low pass LFP data sampled at 1 kHz. Is it absolutely necessary to have both spyking data and LFP data have the same sampling frequency?

Yes it is absolutely necessary to have the LFP and spiking data at the same sampling frequency. The LFP is quite informative about CSs occurrence and human experts use it to accurately detect complex spikes. I guess our algorithm does a better job when the LFP is provided, However, I have put an example in the tutorial showing that our algorithm can also deal with the spyking data only.