Probing Connections

[youssefzaky]

It appears that a probe on a connection returns the output of the post object rather than just the signal from that connection. So if you have multiple inputs going into an object, the connection probes on each return the sum of all the inputs, not the signal going through each connection.

In the following example all the probes show the same value.

import nengo
import matplotlib.pyplot as plt

with nengo.Network() as model:
    input = nengo.Node(output=1)
    ens1 = nengo.Ensemble(dimensions=1, n_neurons=100)
    ens2 = nengo.Ensemble(dimensions=1, n_neurons=100)
    ens3 = nengo.Ensemble(dimensions=1, n_neurons=100, radius=1)
    nengo.Connection(input, ens1)
    nengo.Connection(input, ens2)
    c1 = nengo.Connection(ens1, ens3, transform=2)
    c2 = nengo.Connection(ens2, ens3, transform=-1)
    p1 = nengo.Probe(c1, synapse=0.05)
    p2 = nengo.Probe(c2, synapse=0.05)
    p3 = nengo.Probe(ens3, synapse=0.05)

s = nengo.Simulator(model)
s.run(1)
plt.plot(s.trange(), s.data[p1], label='conn1')
plt.plot(s.trange(), s.data[p2], label='conn2')
plt.plot(s.trange(), s.data[p3], label='ens')
plt.legend(loc='best')
plt.ylim(-2, 2)
plt.show()

[tbekolay]

Does doing

    p1 = nengo.Probe(c1, 'input', synapse=0.05)
    p2 = nengo.Probe(c2, 'input', synapse=0.05)

Give you what you want? By default, probing a connection probes the 'output' attribute.

[youssefzaky]

I thought probing the 'output' attribute on p1 and p2 would show values of 2 and -1 respectively. Probing 'input' attribute gives 100D values (the inputs from each presynaptic neuron?)

[hunse]

Does #974 do what you want?

[youssefzaky]

yup!