extracellular times-recording unnecessary

[cjayb]

The fix in #427 also includes a removal of the arbitrary one-sample-mismatch between the length of the potential and dipole traces. The latter are one short because extra_scatter_gather is only called after a solver step. The solution was to set the initial state to 0 mV (which I suppose is debatable).

Presently _ExtracellularArrayBuilder creates an independent copy of h.t for every array attached to the network:

self._nrn_times = h.Vector().record(h._ref_t)

This is redundant, since h._ref_t is already recorded in network_builder.py:_simulate_single_trial(). One simple solution is to let arr._times = sim_data[0]['times'] in the below (parallel_backends.py:_gather_trial_data())

        # extracellular array
        for arr_name, arr in net.rec_arrays.items():
            # voltages is a n_trials x n_contacts x n_samples array
            arr._data.append(sim_data[idx]['rec_data'][arr_name])
            arr._times = sim_data[idx]['rec_times'][arr_name]

Do I get a +1 for this? And in the bigger picture, do we want keep having times embedded in each of the various output objects we create (dpl, rec_arr, CellResponse)?

[rythorpe]

Sounds good to me. Idk why we need redundant copies of times passed around if they are all the same.

[jasmainak]

agree there should only be one "master" times. Not sure if the initial state should be included? I think having times in the different objects is fine for convenience with plotting etc? We just make sure it's set only in _gather_trial_data so they are all equal. Could even have a test that checks for this

[jasmainak]

0.2 milestone @cjayb ?

[cjayb]

We just make sure it's set only in _gather_trial_data so they are all equal.

Yeah this seems like the easiest approach that's still convenient (plotting). I hesitate because it causes an unnecessary copy, but times isn't something to worry about, I guess.

Not sure if the initial state should be included?

True, especially since our initial state is somewhat poorly defined.

0.2 milestone @cjayb ?

I have a suspicion this might balloon a bit, possibly interact with the pending discussion on initial 'transients' etc. Let's leave this Issue open for now and see what that other topic churns up?

[jasmainak]

sounds good, seems a bit independent of the 'transients' conversation to me but I let you decide

[raj1701]

Can I work on this issue @cjayb @jasmainak @rythorpe ?

[cjayb]

Hey Raj, I'm afraid I'm quite useless here as I haven't looked at the code since last year :( I don't know the relevance of this Issue, but it is labeled as a good first issue, @jasmainak @rythorpe ?

[jasmainak]

One issue at a time @raj1701 ... let's first see your other PRs merged