Consistent behavior of sampling time for multimeters and spike_recorders for co-simulation

[dionperd]

In the context of tvb-multiscale cosimulation, it is possible to perform a NEST -> TVB update either

• via a population spike trains -> mean population rate conversion, or
• via a population membrane potentials -> mean population membrane potential averaging,

after a fixed interval of cosimulation equal to a number of TVB simulation steps, after which the two simulators have to synchronize by exchanging data.

Spike recorders, used for the first interface, will send spike times within the full proper interval, i.e., within the times [interval_start + resolution, interval_stop].

Instead, multimeters, used for the second kind of interface, need to be simulated for an extra NEST resolution step or NEST min delay step, for the values to be recorded by the multimeter, i.e., one has to simulate until interval_stop + resolution/min_delay, in order to get the value with time stamp "interval_stop".

I am not sure if there is a solution for this problem, but I would still like to raise an issue!

[github-actions[bot]]

Issue automatically marked stale!

[heplesser]

This would likely require significant changes, but we keep it in the backlog.

[github-actions[bot]]

Issue automatically marked stale!