Biological Neuron Parameters

[cpehle]

Right now we don't provide biologically plausible neuron parameters (that is parameters that have realistic units and magnitude), instead we normalise most values to be between 0 and 1 and identify reset and leak voltages for neurons. It would be good to introduce biologically plausible neuron parameter ranges based on for example on the cell models of PyNN. One immediate problem is to determine appropriate weight initialisations

[cpehle]

Some plausible starting point would be for example:

lif_default_parameters = LIFParameters(
    v_leak = -65.0,
    v_th = -50.0,
    v_reset = -65.0,
    tau_syn_inv = 1 / 5,
    tau_mem_inv = 1 / 20
)

[Jegp]

Can this be closed?