SpiNNaker/eg_testing-cell-models-network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Testing the action potential of cell models by using standard parameters\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fileName\n",
    "\n",
    "fileName = 'eg_testing-cell-models'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Detected PyNN version 0.9.4 and Neo version 0.6.1\n"
     ]
    }
   ],
   "source": [
    "# libs\n",
    "\n",
    "import pyNN.spiNNaker as sim\n",
    "from pyNN import space \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-17 09:53:30 INFO: Read cfg files: /home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spinn_front_end_common/interface/spinnaker.cfg, /home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spynnaker/pyNN/spynnaker.cfg, /home/spinnaker/.spynnaker.cfg\n",
      "2023-03-17 09:53:30 INFO: Will search these locations for binaries: /home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spinn_front_end_common/common_model_binaries : /home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spynnaker/pyNN/model_binaries\n",
      "2023-03-17 09:53:30 WARNING: /home/spinnaker/CNT-2023/SpiNNaker/reports has 36 old reports that have not been closed\n",
      "2023-03-17 09:53:30 WARNING: /home/spinnaker/CNT-2023/SpiNNaker/application_generated_data_files has 37 old reports that have not been closed\n",
      "2023-03-17 09:53:30 INFO: Setting time scale factor to 1.\n",
      "2023-03-17 09:53:30 INFO: Setting machine time step to 1000 micro-seconds.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['/home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spinn_front_end_common/interface/spinnaker.cfg', '/home/spinnaker/sPyNNaker/lib/python3.6/site-packages/spynnaker/pyNN/spynnaker.cfg', '/home/spinnaker/.spynnaker.cfg']\n"
     ]
    }
   ],
   "source": [
    "# simulation settings\n",
    "\n",
    "sim.setup(\n",
    "        timestep=1, # [ms]\n",
    "        min_delay=1, # [ms]\n",
    "        max_delay=100) # [ms]\n",
    "\n",
    "simtime = 250 # [ms+]\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## make the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define make_pop\n",
    "\n",
    "pops = {}\n",
    "\n",
    "def make_pop(cell_model):\n",
    "    pops = sim.Population(\n",
    "                            1, # one cell in each cell model\n",
    "                            cell_model,\n",
    "                            cellparams=None, # std pars used\n",
    "                            structure = space.Line(dx=1.0, x0=0.0, y=0.0, z=0.0),\n",
    "                            initial_values=None,\n",
    "                            label=str(cell_model),\n",
    "                            constraints=None,\n",
    "                            additional_parameters=None,\n",
    "                            )\n",
    "    pops.record('v')\n",
    "    \n",
    "    return pops\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute pops\n",
    "\n",
    "for idx, value in enumerate([\n",
    "                            sim.IF_cond_exp, \n",
    "                            sim.IF_curr_alpha,\n",
    "                            sim.IF_curr_exp,\n",
    "                            sim.Izhikevich,\n",
    "                            sim.extra_models.IF_curr_dual_exp,\n",
    "                            sim.extra_models.IFCondExpStoc,\n",
    "                            sim.extra_models.IFCurDelta,\n",
    "                            sim.extra_models.IFCurrExpCa2Adaptive,\n",
    "                            sim.extra_models.Izhikevich_cond]):\n",
    "\n",
    "    pops['model_' + str(idx)] = make_pop(value) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['model_0', 'model_1', 'model_2', 'model_3', 'model_4', 'model_5', 'model_6', 'model_7', 'model_8'])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check pops\n",
    "\n",
    "pops.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## make the thalamic input stimulus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['model_0', 'model_1', 'model_2', 'model_3', 'model_4', 'model_5', 'model_6', 'model_7', 'model_8', 'thalamus'])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# a sequence of ten spikes with regularly spaced spike times\n",
    "\n",
    "pops['thalamus'] = sim.Population(\n",
    "                            1, # one thalamic cell for each cell model\n",
    "                            sim.SpikeSourceArray(np.arange(50,60)), # ten sequential spike times\n",
    "                            structure = space.Line(dx=1.0, x0=0.0, y=0.0, z=0.0),\n",
    "                            )    \n",
    "\n",
    "pops.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## make the projections"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys([('thalamus', 'model_0'), ('thalamus', 'model_1'), ('thalamus', 'model_2'), ('thalamus', 'model_3'), ('thalamus', 'model_4'), ('thalamus', 'model_5'), ('thalamus', 'model_6'), ('thalamus', 'model_7'), ('thalamus', 'model_8')])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "projs = {}\n",
    "\n",
    "for i in range(9):\n",
    "    projs['thalamus', 'model_' + str(i)] = sim.Projection(\n",
    "                                    pops['thalamus'],\n",
    "                                    pops['model_' + str(i)],\n",
    "                                    sim.OneToOneConnector(),\n",
    "                                    synapse_type=sim.StaticSynapse(weight=0.1, delay=1.0),\n",
    "                                    receptor_type = 'excitatory',\n",
    "                                    space = space.Space(axes = 'x'),\n",
    "                                    label=None,\n",
    "                                )\n",
    "\n",
    "projs.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## run the simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-17 09:53:36 INFO: Simulating for 250 1.0ms timesteps using a hardware timestep of 1000us\n",
      "2023-03-17 09:53:36 INFO: Starting execution process\n",
      "2023-03-17 09:53:39 INFO: Time 0:00:03.250143 taken by SpallocMaxMachineGenerator\n",
      "Pre allocating resources for Extra Monitor support vertices\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:53:48 INFO: Time 0:00:08.747342 taken by PreAllocateResourcesForExtraMonitorSupport\n",
      "Partitioning graph vertices\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Partitioning graph edges\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:53:53 INFO: Time 0:00:05.118075 taken by PartitionAndPlacePartitioner\n",
      "Created spalloc job 7076690\n",
      "2023-03-17 09:53:53 INFO: Created spalloc job 7076690\n",
      "Waiting for board power commands to complete.\n",
      "2023-03-17 09:53:53 INFO: Waiting for board power commands to complete.\n",
      "2023-03-17 09:54:03 INFO: Time 0:00:09.626833 taken by SpallocAllocator\n",
      "2023-03-17 09:54:03 INFO: Creating transceiver for 10.11.193.17\n",
      "2023-03-17 09:54:03 INFO: Working out if machine is booted\n",
      "2023-03-17 09:54:07 INFO: Attempting to boot machine\n",
      "2023-03-17 09:54:12 INFO: Found board with version [Version: SC&MP 3.2.5 at SpiNNaker:0:0:0 (built Thu Aug  1 08:15:06 2019)]\n",
      "2023-03-17 09:54:12 INFO: Machine communication successful\n",
      "2023-03-17 09:54:12 INFO: Detected a machine on IP address 10.11.193.17 which has 855 cores and 120.0 links\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:09.632306 taken by MachineGenerator\n",
      "Generating partitioner report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.008259 taken by PartitionerReport\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.000501 taken by NetworkSpecificationReport\n",
      "Allocating virtual identifiers\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.059663 taken by MallocBasedChipIDAllocator\n",
      "Inserting extra monitors into graphs\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.031482 taken by InsertExtraMonitorVerticesToGraphs\n",
      "Writing the board chip report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.005517 taken by BoardChipReport\n",
      "Filtering edges\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.032844 taken by GraphEdgeFilter\n",
      "Placing graph vertices\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.037043 taken by OneToOnePlacer\n",
      "Inserting edges between vertices which require FR speed up functionality.\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.067855 taken by InsertEdgesToExtraMonitorFunctionality\n",
      "Generating routing tables for data in system processes\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:12 INFO: Time 0:00:00.006977 taken by DataInMulticastRoutingGenerator\n",
      "Generating fixed router routes\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.016080 taken by FixedRouteRouter\n",
      "Generating placement report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Generating placement by core report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.060453 taken by PlacerReportWithApplicationGraph\n",
      "Routing\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.029002 taken by NerRoute\n",
      "Discovering tags\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Allocating tags\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.044808 taken by BasicTagAllocator\n",
      "Reporting Tags\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.011773 taken by TagReport\n",
      "Getting number of keys required by each edge using application graph\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.034209 taken by EdgeToNKeysMapper\n",
      "Getting constraints for application graph\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.030650 taken by ProcessPartitionConstraints\n",
      "Allocating routing keys\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.010214 taken by MallocBasedRoutingInfoAllocator\n",
      "Generating Routing info report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.023285 taken by routingInfoReports\n",
      "Generating routing tables\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.023280 taken by BasicRoutingTableGenerator\n",
      "Finding executable start types\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.039755 taken by LocateExecutableStartType\n",
      "Initialising buffers\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.026652 taken by BufferManagerCreator\n",
      "Generating data specifications\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.087902 taken by SpynnakerDataSpecificationWriter\n",
      "Preparing Routing Tables\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.010287 taken by RoutingSetup\n",
      "Finding binaries\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:13 INFO: Time 0:00:00.056558 taken by GraphBinaryGatherer\n",
      "Running routing table compression on chip\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:18 INFO: Time 0:00:04.530218 taken by MundyOnChipRouterCompression\n",
      "Generating Router table report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:18 INFO: Time 0:00:00.009055 taken by unCompressedRoutingTableReports\n",
      "loading fixed routes\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:18 INFO: Time 0:00:00.045122 taken by LoadFixedRoutes\n",
      "Executing data specifications and loading data for system vertices\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:18 INFO: Time 0:00:00.152977 taken by HostExecuteSystemDataSpecification\n",
      "Loading system executables onto the machine\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:25 INFO: Time 0:00:06.735600 taken by LoadSystemExecutableImages\n",
      "Clearing tags\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Loading Tags\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:25 INFO: Time 0:00:00.020384 taken by TagsLoader\n",
      "Writing data\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:25 INFO: Time 0:00:00.028284 taken by WriteMemoryIOData\n",
      "Executing data specifications and loading data for application vertices\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:27 INFO: Time 0:00:02.178132 taken by HostExecuteApplicationDataSpecification\n",
      "Expanding Synapses\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:27 INFO: Time 0:00:00.560172 taken by SynapseExpander\n",
      "Writing fixed route report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:54:27 INFO: Time 0:00:00.046834 taken by FixedRouteFromMachineReport\n",
      "Loading executables onto the machine\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:34.555231 taken by LoadApplicationExecutableImages\n",
      "Reading Routing Tables from Machine\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.017314 taken by RoutingTableFromMachineReport\n",
      "Generating compressed router table report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.009548 taken by compressedRoutingTableReports\n",
      "Generating comparison of router table report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.005806 taken by comparisonOfRoutingTablesReport\n",
      "2023-03-17 09:55:02 INFO: Running for 1 steps for a total of 250.0ms\n",
      "2023-03-17 09:55:02 INFO: Run 1 of 1\n",
      "Generating SDRAM usage report\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.045593 taken by SdramUsageReportPerChip\n",
      "Updating run time\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.026028 taken by ChipRuntimeUpdater\n",
      "2023-03-17 09:55:02 INFO: creating live event connection database in /home/spinnaker/CNT-2023/SpiNNaker/application_generated_data_files/2023-03-17-09-53-30-911445/run_1/input_output_database.db\n",
      "Creating graph description database\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.050071 taken by DatabaseInterface\n",
      "2023-03-17 09:55:02 INFO: ** Notifying external sources that the database is ready for reading **\n",
      "2023-03-17 09:55:02 INFO: Time 0:00:00.000976 taken by NotificationProtocol\n",
      "2023-03-17 09:55:02 INFO: *** Running simulation... *** \n",
      "Loading buffers\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:02 INFO: ** Awaiting for a response from an external source to state its ready for the simulation to start **\n",
      "2023-03-17 09:55:02 INFO: ** Sending start / resume message to external sources to state the simulation has started or resumed. **\n",
      "2023-03-17 09:55:02 INFO: ** Awaiting for a response from an external source to state its ready for the simulation to start **\n",
      "2023-03-17 09:55:02 INFO: Application started; waiting 0.35s for it to stop\n",
      "2023-03-17 09:55:03 INFO: ** Sending pause / stop message to external sources to state the simulation has been paused or stopped. **\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.443675 taken by ApplicationRunner\n",
      "Extracting buffers from the last run\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.135654 taken by BufferExtractor\n",
      "Getting provenance data from machine graph\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting provenance data from application graph\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.108725 taken by GraphProvenanceGatherer\n",
      "Getting provenance data\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.043658 taken by PlacementsProvenanceGatherer\n",
      "Getting Router Provenance\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.052247 taken by RouterProvenanceGatherer\n",
      "Getting profile data\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 INFO: Time 0:00:00.042001 taken by ProfileDataGatherer\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "250.0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sim.run(simtime) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## save results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_base.IFCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_base.IFCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_base.IFCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_base.IFCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_base.IFCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_alpha.IFCurrAlpha'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "2023-03-17 09:55:03 WARNING: Warning getting data on a whole population when selective recording is active will result in only the requested neurons being returned in numerical order and without repeats.\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_alpha.IFCurrAlpha'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_alpha.IFCurrAlpha'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_alpha.IFCurrAlpha'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_alpha.IFCurrAlpha'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_base.IFCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_base.IFCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_base.IFCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_base.IFCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_base.IFCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_curr_exp_base.IzkCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_curr_exp_base.IzkCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_curr_exp_base.IzkCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_curr_exp_base.IzkCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_curr_exp_base.IzkCurrExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_dual_exp_base.IFCurrDualExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_dual_exp_base.IFCurrDualExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_dual_exp_base.IFCurrDualExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_dual_exp_base.IFCurrDualExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_dual_exp_base.IFCurrDualExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_stoc.IFCondExpStoc'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_stoc.IFCondExpStoc'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_stoc.IFCondExpStoc'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_stoc.IFCondExpStoc'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_cond_exp_stoc.IFCondExpStoc'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_delta.IFCurrDelta'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_delta.IFCurrDelta'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_delta.IFCurrDelta'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_delta.IFCurrDelta'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_delta.IFCurrDelta'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_ca2_adaptive.IFCurrExpCa2Adaptive'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_ca2_adaptive.IFCurrExpCa2Adaptive'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_ca2_adaptive.IFCurrExpCa2Adaptive'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_ca2_adaptive.IFCurrExpCa2Adaptive'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.if_curr_exp_ca2_adaptive.IFCurrExpCa2Adaptive'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_cond_exp_base.IzkCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_cond_exp_base.IzkCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_cond_exp_base.IzkCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_cond_exp_base.IzkCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n",
      "Getting v for <class 'spynnaker.pyNN.models.neuron.builds.izk_cond_exp_base.IzkCondExpBase'>\n",
      "|0%                          50%                         100%|\n",
      " ============================================================\n"
     ]
    }
   ],
   "source": [
    "# save the results\n",
    "\n",
    "outputs = {}\n",
    "\n",
    "for i in range(9):\n",
    "    outputs['model_' + str(i)] = pops['model_' + str(i)].get_data()\n",
    "    for recording in ['v', 'gsyn_inh', 'gsyn_exc', 'spikes']:\n",
    "        pops['model_' + str(i)].write_data(fileName + '_' + str(recording) + '.pkl')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## recover results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make the recover results function\n",
    "\n",
    "def recover_results(outputs):\n",
    "    results = {}\n",
    "    for key in outputs.keys(): \n",
    "        \n",
    "        # to get voltage and conductances\n",
    "        for analogsignal in outputs[key].segments[0].analogsignals:\n",
    "            print(analogsignal.name)\n",
    "            results[key, analogsignal.name] = analogsignal\n",
    "\n",
    "        # to get spikes\n",
    "        results[key, 'spikes'] = outputs[key].segments[0].spiketrains\n",
    "    return results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "v\n",
      "v\n",
      "v\n",
      "v\n",
      "v\n",
      "v\n",
      "v\n",
      "v\n",
      "v\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "dict_keys([('model_0', 'v'), ('model_0', 'spikes'), ('model_1', 'v'), ('model_1', 'spikes'), ('model_2', 'v'), ('model_2', 'spikes'), ('model_3', 'v'), ('model_3', 'spikes'), ('model_4', 'v'), ('model_4', 'spikes'), ('model_5', 'v'), ('model_5', 'spikes'), ('model_6', 'v'), ('model_6', 'spikes'), ('model_7', 'v'), ('model_7', 'spikes'), ('model_8', 'v'), ('model_8', 'spikes')])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# recover results\n",
    "\n",
    "results = recover_results(outputs)\n",
    "results.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## check the voltage signature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x504 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "fig, axes = plt.subplots(3, 3, sharex=True, sharey=False, figsize=(11,7))\n",
    "fig.tight_layout(pad=3)\n",
    "fig.suptitle('voltage signature in cell models')\n",
    "\n",
    "axes_list = fig.axes\n",
    "idx = 0\n",
    "for i in range(9):\n",
    "    im = axes_list[i].plot(results['model_' + str(i), 'v'], label=str('model_' + str(i)))\n",
    "    #axes_list[i].set_xticklabels([])\n",
    "    axes_list[i].legend()\n",
    "    axes_list[i].grid(which='both')\n",
    "    if i==0 or i==3 or i==6:\n",
    "        axes_list[i].set_ylabel('[mv]')\n",
    "    if i==6 or i==7 or i==8:\n",
    "        axes_list[i].set_xlabel('[ms]')\n"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Plot legend:\n",
    "    \n",
    "model_0 = sim.IF_cond_exp, \n",
    "model_1 = sim.IF_curr_alpha\n",
    "model_2 = sim.IF_curr_exp\n",
    "model_3 = sim.Izhikevich\n",
    "model_4 = sim.extra_models.IF_curr_dual_exp\n",
    "model_5 = sim.extra_models.IFCondExpStoc\n",
    "model_6 = sim.extra_models.IFCurDelta\n",
    "model_7 = sim.extra_models.IFCurrExpCa2Adaptive\n",
    "model_8 = sim.extra_models.Izhikevich_cond"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# end simulation\n",
    "\n",
    "sim.end()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task 1: on the action potential of cell models\n",
    "\n",
    "- Have a look of the plot, you should observe the different voltage signatures having similarities as well as differences. <br>\n",
    "  Describe them or try making a classification by adopting criteria indicated in this bottom image<br>\n",
    "  \n",
    "  e.g., \n",
    "      - depolarization timing\n",
    "      - shape of depolarization\n",
    "      - action potential timing\n",
    "      - shape of repolarization\n",
    "      - presence/absence of refractory period\n",
    "      - resting state timing\n",
    "      \n",
    "      \n",
    "- This simulation has been done with standard parameters. <br>\n",
    "  Choose a parameter shared from all the models and tune it. <br>\n",
    "  Observe how the voltage trace of the action potential changes. <br>\n",
    "  To explore the default paramters, use e.g., `sim.IF_cond_exp.default_parameters` <br>\n",
    "  Change them by defining a new dictionary of pars for each model call e.g., pars[model_x]<br>\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](https://upload.wikimedia.org/wikipedia/commons/4/4a/Action_potential.svg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task 2: on the network of different cell models\n",
    "\n",
    "- The previous results are related to single behaviour of the given cell models.<br>\n",
    "  Try to make a population of 25 cells for each model and record their voltage.<br>\n",
    "  The input should be design as a sequence of ten input spikes separeted 150 ms for each model.<br>\n",
    "  E.g., the first model receives input at 50-59ms, the second at 200-209ms, the third at 350-359ms etc.<br>\n",
    "  Plot the results in a single plot showing the all traces that should appear sequentally displayed.<br>\n",
    "  Set the simtime at 1500ms."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sPyNNaker",
   "language": "python",
   "name": "spynnaker"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}