Repaired a problem with the ./metrics createPR where all NaN slices were being encountered. Needed to transpose the numpconc numpy array so it organized into columns then also throw in a check during testing that catches all zero outputs (sub-threshold spiking) and finds the output neuron with the highest input.
Repaired a problem with the ./metrics createPR where all NaN slices were being encountered. Needed to transpose the numpconc numpy array so it organized into columns then also throw in a check during testing that catches all zero outputs (sub-threshold spiking) and finds the output neuron with the highest input.