Plotting figures after individual sim runs in a batch simulation

[Maliha A]

Is there no way of plotting figures after each run in a batch simulation? 

Or is the only option to read batch data from individual simulation files at the very end of the full batch run?

Let's say for example I would like to plot the average membrane potential of a population after each simulation. I've been playing around with the model in Tutorial 8. I felt optimistic and tried adding lines of code to the tut8_init.py file after 

sim.create(netParams, simConfig)
sim.simulate()
sim.analyze()

like so:

stimtimevec = sim.allSimData.t
VavgM=np.zeros(len(sim.allSimData.t))
for i in range(20,40):
    arrayname = f'cell_{i}'
    VavgM = np.add(VavgM,sim.allSimData.V_soma[arrayname])
    VavgM = (1/20)*VavgM

plt.figure(figsize=(10,6))
plt.plot(stimtimevec,VavgM)
plt.xlabel("Time (ms)")
plt.ylabel("Voltage (mV)")
plt.rcParams['font.size'] = 16

But that didn't work... any suggestions would be greatly appreciated.

Thank you
Maliha

[J C]

I'm not certain how to do this during a batch run using the current NetPyNE batch tools.
if this code works on a single run, then it could possibly be modified to work in a batch simulation.
First, any display is usually disabled during batch runs, so doing something to save the figure would be more appropriate.
`plt.savefig()`
the second issue is that we would need to provide a unique name for the figure as opposed to something like "results.png" or else all of the simulations would clobber each other.
to provide a string to the `plt.savefig()` that would be unique for each run, you could generate a string based on some `cfg` variable that changes between simulations in the batch...
if you look at the `batchCell` example, the generated `_cfg.json` files might show some useful unique identifiers that can be referenced.
The `simConfig` dictionary is essentially what the `cfg` that is being imported in the `init.py` file is...
At the bottom, one string includes `"simLabel"` so you could theoretically do
`plt.savefig("{}.png".format(cfg.simLabel))`
to generate a name unique to each run (and therefore a unique png)

[J C]

```

(base) ➜  batchNa git:(development) ✗ cat batchNa_0_1_cfg.json | grep "simLabel" -B 5 -A 5
        "seeds": {
            "conn": 4321,
            "loc": 4321,
            "stim": 1234
        },
        "simLabel": "batchNa_0_1", <---- this value
        "tau1NMDA": 15,
        "timestampFilename": false,
        "timing": true,
        "tstop": 1000.0,
        "use_fast_imem": false,
```

what I don't know is if there are issues with the matplotlib library that would cause the figures from the analysis scripts to conflict with the any custom plot function in your init.py

[Maliha A]

Thank you for the suggestion! I gave it a try and unfortunately it still didn't work.. From my understanding it seems to not execute any lines written after sim.analyze(). I tested this using the following lines:

randomname = 3
randomname2 = 4
with open('tut8_output.txt', "a") as tut8_output_txt:
    l1 = '\n [%.3f,' % randomname
    l2 = '%.3f],' % randomname2
    tut8_output_txt.writelines([l1,l2])

I expected that each time the init.py is called for each run, these lines would also get executed but that's not the case. Hmm..

[J C]

Can you clarify, nothing executes after `sim.analyze()` when running it as a batch job or as a single job?
Can you link me a repository of your code? 

Thanks,

James 

[Maliha A]

Hi James, 

I realized I had errors in my code! So i am indeed able to execute lines after sim.analyze() as a batch job :) 

Thanks

Maliha