| Remarks: |
Network-wise the parrot neuron behaves like other neuron models
regarding connections and communication. While the number of
outgoing spikes equals that of incoming ones, the weigth of the
outgoing spikes solely depends on the weigth of outgoing connections.
A Poisson generator that would send multiple spikes during a single
time step due to a high rate will send single spikes with
multiple synaptic strength instead, for effiacy reasons.
This can be realized because of the way devices are implemented
in the threaded environment. A parrot neuron on the other
hand always emits single spikes. Hence, in a situation where for
example a poisson generator with a high firing rate is connected
to a parrot neuron, the communication cost associated with outgoing
spikes is much bigger for the latter.
Please note that this node is capable of sending precise spike times
to target nodes (on-grid spike time plus offset). If this node is
connected to a spike_detector, the property "precise_times" of the
spike_detector has to be set to true in order to record the offsets
in addition to the on-grid spike times.
|