Stimulating currents that are not step-shaped

[bgmeulem]

Hiya

Is it possible to stimulate a cell with current injections that are not step current shaped? I'm trying to simulate a bi-exponential (or epsp-shaped) current to test out Jaxley for more complicated intrinsic physiology, such as paired injections and active dendritic physiology.
Current shape should be something like:

def epsp_current(x, tau_rise, tau_decay, imax, onset):
    tpeak = tau_rise* tau_decay* np.log(tau_rise/ v) / (tau_rise- tau_decay)
    adjust = 1 / ((1 - np.exp(-tpeak / tau_rise)) - (1 - np.exp(-tpeak / tau_decay)))
    amp = adjust * imax
    if x < onset:
        return 0
    else:
        a0 = 1 - np.exp(-(x - onset) / tau_rise)
        a1 = 1 - np.exp(-(x - onset) / tau_decay)
        return amp * (a0 - a1)

Image of this shape below.

I couldn't find anything in the source code that supports this. I was looking at datapoint_to_step_currents that could be hacked into giving individual currents, each of length dt, but it doesn't support passing different delays.
Will I simply need to pass a whole array of dt-length currents, each evaluated at a different point on the epsp shape to simulate this shape?

[jnsbck]

Hiya, any jax array can be used as a stimulus, just make sure that the time-steps line up with those of your simulation :)

some_stim = jnp.array([...])
comp = jx.Comp()
comp.stimulate(some_stim)

Jonas

[bgmeulem]

Cool, that worked, thanks for the quick repsonse!