{ "cells": [ { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "%matplotlib inline" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Tsodyks facilitating example\n\nThis scripts simulates two neurons. One is driven with dc-input and\nconnected to the other one with a facilitating Tsodyks synapse. The\nmembrane potential trace of the second neuron is recorded.\n\nThis example reproduces figure 1B of [1]_\nThis example is analog to ``tsodyks_depressing.py``, except that\ndifferent synapse parameters are used. Here, a small facilitation\nparameter ``U`` causes a slow saturation of the synaptic efficacy\n(Eq. 2.2), enabling a facilitating behavior.\n\n## References\n\n.. [1] Tsodyks M, Pawelzik K, Markram H (1998). Neural networks with dynamic synapses. Neural\n computation, http://dx.doi.org/10.1162/089976698300017502\n\n## See Also\n\n:doc:`tsodyks_depressing`\n\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "First, we import all necessary modules for simulation and plotting.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "import nest\nimport nest.voltage_trace\nfrom numpy import exp" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Second, the simulation parameters are assigned to variables. The neuron\nand synapse parameters are stored into a dictionary.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "h = 0.1 # simulation step size (ms)\nTau = 40. # membrane time constant\nTheta = 15. # threshold\nE_L = 0. # reset potential of membrane potential\nR = 1. # membrane resistance (GOhm)\nC = Tau / R # Tau (ms)/R in NEST units\nTauR = 2. # refractory time\nTau_psc = 1.5 # time constant of PSC (= Tau_inact)\nTau_rec = 130. # recovery time\nTau_fac = 530. # facilitation time\nU = 0.03 # facilitation parameter U\nA = 1540. # PSC weight in pA\nf = 20. / 1000. # frequency in Hz converted to 1/ms\nTend = 1200. # simulation time\nTIstart = 50. # start time of dc\nTIend = 1050. # end time of dc\nI0 = Theta * C / Tau / (1 - exp(-(1 / f - TauR) / Tau)) # dc amplitude\n\nneuron_param = {\"tau_m\": Tau,\n \"t_ref\": TauR,\n \"tau_syn_ex\": Tau_psc,\n \"tau_syn_in\": Tau_psc,\n \"C_m\": C,\n \"V_reset\": E_L,\n \"E_L\": E_L,\n \"V_m\": E_L,\n \"V_th\": Theta}\n\nsyn_param = {\"tau_psc\": Tau_psc,\n \"tau_rec\": Tau_rec,\n \"tau_fac\": Tau_fac,\n \"U\": U,\n \"delay\": 0.1,\n \"weight\": A,\n \"u\": 0.0,\n \"x\": 1.0}" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Third, we reset the kernel and set the resolution using ``SetKernelStatus``.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "nest.ResetKernel()\nnest.SetKernelStatus({\"resolution\": h})" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Fourth, the nodes are created using ``Create``. We store the returned\nhandles in variables for later reference.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "neurons = nest.Create(\"iaf_psc_exp\", 2)\ndc_gen = nest.Create(\"dc_generator\")\nvolts = nest.Create(\"voltmeter\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Fifth, the ``iaf_psc_exp`` neurons, the ``dc_generator`` and the ``voltmeter``\nare configured using ``SetStatus``, which expects a list of node handles and\na parameter dictionary or a list of parameter dictionaries.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "neurons.set(neuron_param)\ndc_gen.set(amplitude=I0, start=TIstart, stop=TIend)\nvolts.set(label=\"voltmeter\", interval=1.)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Sixth, the ``dc_generator`` is connected to the first neuron\n(`neurons[0]`) and the `voltmeter` is connected to the second neuron\n(`neurons[1]`). The command `Connect` has different variants. Plain\n``Connect`` just takes the handles of pre- and post-synaptic nodes and\nuses the default values for weight and delay. Note that the connection\ndirection for the ``voltmeter`` reflects the signal flow in the simulation\nkernel, because it observes the neuron instead of receiving events from it.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "nest.Connect(dc_gen, neurons[0])\nnest.Connect(volts, neurons[1])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Seventh, the first neuron (`neurons[0]`) is connected to the second\nneuron (`neurons[1]`). The command ``CopyModel`` copies the\n``tsodyks_synapse`` model to the new name ``syn`` with parameters\n``syn_param``. The manually defined model ``syn`` is used in the\nconnection routine via the ``syn_spec`` parameter.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "nest.CopyModel(\"tsodyks_synapse\", \"syn\", syn_param)\nnest.Connect(neurons[0], neurons[1], syn_spec=\"syn\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, we simulate the configuration using the command ``Simulate``,\nwhere the simulation time `Tend` is passed as the argument. We plot the\ntarget neuron's membrane potential as function of time.\n\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": false }, "outputs": [], "source": [ "nest.Simulate(Tend)\nnest.voltage_trace.from_device(volts)\nnest.voltage_trace.show()" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.6.12" } }, "nbformat": 4, "nbformat_minor": 0 }