Calcium channel can lead to NaN

[michaeldeistler]

I found that, if only a calcium channel is inserted, then the simulation can fail.

import matplotlib.pyplot as plt
import jaxley as jx
from jaxley_mech.channels.fm97 import Ca


# If decreased below 0.0001 the simulation will succeed.
i_amp = 0.001

dt = 0.025
t_max = 200.0
i_delay = 1.0
i_dur = 20.0

comp = jx.Compartment()
comp.insert(Ca())
comp.record()
comp.stimulate(jx.step_current(i_delay, i_dur, i_amp, dt, t_max))
v = jx.integrate(comp)

fig, ax = plt.subplots(1, 1, figsize=(5, 3))
_ = ax.plot(v.T)
_ = ax.set_ylim([-80, 150])

Do you know if this is a bug or if this is just how the differential equation for calcium is set up?

[huangziwei]

likely a bug, I will check it tomorrow.

UPDATE:

@michaeldeistler It's a bug. Ca concentration in the compartment went negative, which shouldn't happen... fixed in the latest commit.