kalekiu/easyesn

Accessing reservoir states

pfdominey opened this issue · 2 comments

pfdominey commented

Greetings,
Thanks for the great system. I am interested in studying the activity of the reservoir units during the time-course of the processing of the input sequence. After training the reservoir we then expose it to a test inputDataTesting

prediction = esn.predict(inputDataTesting)

After this execution, is it possible to have access to the data structure that contains the activations of the reservoir units during the processing of the input? This would be a matrix of NxM where N is the number of reservoir units, and M is the number of time steps during the simulation.

This would be used to analyse the temporal dynamics of the reservoir activity.

Thanks for your help

PF Dominey

(ps - I hope this is an appropriate question for this issue forum - please excuse me if this is out of line)

zimmerrol commented

Hi,
this is not possible at the moment. But you can easily add this function to the library! Just modify the last line of the predict function` and change it from

return Y.T

to

return Y.T, X

Does this solve your problem?

pfdominey commented

Sure does! Thanks: