www.neuron.yale.edu

I second Ted's point about the need for state-specific error tolerance scale multipliers. Unfortunately, the rxd state variables are not integrated into the VariableStepControl GUI's analysis at this time. You can, however, specify a species-specific atolscale. The rxd.Species constructor prototype ...

This is still a bit of a hack, but it should keep the GUI responsive and avoid the memory issue.

Before you do your h.run() (or equivalent), do:

Also: do the concentrations go negative with fixed-step integration?

Not directly. If you're not using mass-action, it's possible to write reactions that mathematically demand concentrations go negative. Can you tell us more about your reaction specification? Assuming some are not mass action, have you checked to make sure the reaction rates are not such that the con...

This windows-specific bug also seems to only occur if the simulation is run from the terminal or from a script. If you were to click the "Run" button on the GUI (e.g. in Tools - Run Control), the graphs should update etc during the simulation. That may or may not be useful to you.

(1) To get a concentration pointer in Python, use ._concentration_ref ; you can pass that to Vector.record . An example follows: from neuron import h, rxd from matplotlib import pyplot h.load_file('stdrun.hoc') dend = h.Section(name='dend') cyt = rxd.Region([dend], name='cyt', nrn_region='i') ca = r...

Point of clarification: are you saying it runs without memory problems if you don't use that workaround, but crashes when you do? How long does each simulation run (h.tstop) for, and how many simulations are you running? If you would like, I can take a look at your code to try to track down the memo...

Both of your difficulties are due to a Windows-specific bug that keeps NEURON from releasing the GIL during integration, which blocks both the NEURON GUI and Matplotlib threads from executing... thus the graphs never get updated. Thanks for bringing this to our attention. A workaround for now (besid...

Is this problem Spyder-specific, or does it also occur if you run the script from the regular Python prompt?

What operating system and NEURON version are you using?

The "from neuron import gui" is supposed to be all you need.

Yes but it's not currently anywhere near as clean as it should be. Diffusion between shells within a section can be implemented using MultiCompartmentReactions similarly to how one might do a leak between the ER and the cytosol. I don't have an example of this, but I'll try to make one tomorrow unle...

The leak current in the hh mechanism is il_hh not ileak_hh.

Alternatively, a more Pythonic syntax is to access it via the mechanism; e.g. seg.hh.il (or seg.hh._ref_il if you need the pointer).

In Spyder, if you store the value of the diam3d call and print it later, does that work? I'm wondering if you're seeing a return status and not the diameter itself.

A quick clarification:

The underlying issue was that vector.record returns a status (typically 1.0) and not the vector, so the original version was simply recording all the statuses and discarding the vectors.

This is puzzling because when you use h.load_file, the Python interpreter plays literally no role in loading the morphology (it's all done by the HOC interpreter instead). I was, unfortunately, not able to quickly reproduce your issue (details below), but I did have to modify the code slightly to ge...

pdb works for me with your script if you run it with python hh_example.py instead of using nrngui. Make sure the neuron libraries are on your PYTHONPATH. (There are subtle differences about how console input etc is handled when you launch with nrngui instead of python directly, that I suspect is re...