Effective coupling

[zbarni]

In the function computing the effective coupling in meanfield_calcs.py on line 971

def effective_coupling_strength(tau_m, tau_s, tau_r, V_0_rel, V_th_rel, J,  mean_input, std_input):
...
    for pre in np.arange(dim):
        for post in np.arange(dim):
            w_ecs[post][pre] = aux_calcs.d_nu_d_nu_in_fb(
                tau_m, tau_s, tau_r, V_th_rel, V_0_rel, J[post][pre],
                mean_input[pre], std_input[pre])[1] # linear (mu) contribution

shouldn't the mean and std of the input depend on the postsynaptic neuron instead of the presynaptic one, i.e., mean_input[pre], std_input[pre] be replaced with mean_input[post], std_input[post]? I think that as it currently is implemented, the resulting entry w_ij does not correspond to the definitions used in the following papers Senk et al., eq. 56, Helias et al., eq. A3, where w_ij depends on the stationary firing rate of the postsynaptic neuron i.

[zbarni]

Just wondering if this issue was followed upon?

[moritzlayer]

It's still on my ToDo List, but we need to finish the review of the tests first, before changing any code. Hopefully, Johanna can have a look at the tests after she's back from holidays.