assemblies_v5_output-variability_multi.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "\n",
    "path = '/home/seq/pelenet'\n",
    "sys.path.append(path)\n",
    "os.chdir(path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Allow reload of objects\n",
    "%load_ext autoreload\n",
    "%autoreload\n",
    "\n",
    "from pelenet.utils import Utils\n",
    "from pelenet.experiments.assemblies import AssemblyExperiment\n",
    "\n",
    "# Official modules\n",
    "import numpy as np\n",
    "import nxsdk.api.n2a as nx\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "from time import time\n",
    "from copy import deepcopy\n",
    "import statsmodels.api as sm\n",
    "from nxsdk.logutils.nxlogging import LoggingLevel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Overwrite default parameters (pelenet/parameters/ and pelenet/experiments/random.py)\n",
    "parameters = {\n",
    "    'loihiLoggingLevel': LoggingLevel.ERROR,\n",
    "    # Experiment\n",
    "    'seed': 1,  # Random seed\n",
    "    'trials': 40, #20,  # Number of trials\n",
    "    'stepsPerTrial': 40, #50,  # Number of simulation steps for every trial\n",
    "    'isReset': True,  # Activate reset after every trial\n",
    "    # Network\n",
    "    'reservoirExSize': 400,  # Number of excitatory neurons\n",
    "    'reservoirConnPerNeuron': 35, #35 # Number of connections per neuron\n",
    "    # Noise\n",
    "    'isNoise': True,\n",
    "    'onlyExcitatory': True,  # Noisy input is only excitatory\n",
    "    'noiseMaxWeight': 120, #120,\n",
    "    # Neurons\n",
    "    'refractoryDelay': 2, # Refactory period\n",
    "    'voltageTau': 100, #100 # Voltage time constant\n",
    "    'currentTau': 6, #5 # Current time constant\n",
    "    'thresholdMant': 1200, #1200  # Spiking threshold for membrane potential\n",
    "    # Plasticity\n",
    "    'isLearningRule': True,\n",
    "    'learningEpoch': 4,\n",
    "    'learningRule': '2^-2*x1*y0 - 2^-2*y1*x0 + 2^-4*x1*y1*x0*y0 - 2^-3*y0*w*w', #'2^-2*x1*y0 - 2^-2*y1*x0 + 2^-4*x1*y1*y0 - 2^-3*y0*w*w',\n",
    "    #'learningRule': '2^-3*x1*y1*x0*y0 - 2^-3*y0*w*w',\n",
    "    #'learningRule': '2^-2*x1*y0 - 2^-2*y1*x0 + 2^-3*y0*w*w',\n",
    "    # Input\n",
    "    'inputIsVary': True,\n",
    "    'inputOffset': 10,\n",
    "    'inputVaryProbs': [0.1, 0.9],\n",
    "    'inputGenSpikeProb': 0.8,  # Probability of spikes for the spike generators\n",
    "    'inputNumTargetNeurons': 40,  # Number of neurons targeted by the spike generators\n",
    "    'inputSteps': 20, #30,  # Number of steps the input is active\n",
    "    # Probes\n",
    "    'isExSpikeProbe': True,  # Probe excitatory spikes\n",
    "    'isInSpikeProbe': True,  # Probe inhibitory spikes\n",
    "    'isWeightProbe': True  # Probe weight matrix at the end of the simulation\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_error(x, y, trials):\n",
    "    x_err = []\n",
    "    for i in range(trials):\n",
    "        # Estimate\n",
    "        model = sm.OLS(y, x[i].T)\n",
    "        params = model.fit().params\n",
    "\n",
    "        # Predict\n",
    "        ye = np.dot(x[i].T, params)\n",
    "        err = np.sqrt(np.sum(np.square(y-ye)))\n",
    "        x_err.append(err)\n",
    "    return x_err"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "supports = []\n",
    "\n",
    "a_y = np.sin(np.arange(20)/10*(np.pi))\n",
    "b_y = np.cos(np.arange(20)/10*(np.pi))\n",
    "\n",
    "def run_epoch(seed):\n",
    "    p = deepcopy(parameters)\n",
    "    print(seed)\n",
    "    p['seed'] = seed\n",
    "    \n",
    "    \"\"\"\n",
    "    Allocation\n",
    "    \"\"\"\n",
    "    # Initilizes the experiment and utils\n",
    "    exp = AssemblyExperiment(name='assemblies', parameters=p)\n",
    "    utils = Utils.instance()\n",
    "    \n",
    "    # Build network\n",
    "    exp.build()\n",
    "\n",
    "    # Get spectral radius\n",
    "    #utils.getSpectralRadius(exp.net.initialWeights)\n",
    "    \n",
    "    # Run network\n",
    "    exp.run()\n",
    "    \n",
    "    supportMasks = utils.getSupportMasksForAllTrials(exp.net.initialWeights.exex, exp.net.trainedWeightsExex)\n",
    "    \n",
    "    #0: A, 1: B, 2: others\n",
    "    a_masks = np.array([ supportMasks[i, 0] for i in range(exp.net.p.trials) ])\n",
    "    b_masks = np.array([ supportMasks[i, 1] for i in range(exp.net.p.trials) ])\n",
    "    \n",
    "    backgroundNeurons = exp.net.exSpikeTrains[80:]\n",
    "    \n",
    "    a_x = [ backgroundNeurons[a_masks[i],40+i*60:(i+1)*60] for i in range(exp.net.p.trials) ]\n",
    "    b_x = [ backgroundNeurons[b_masks[i],40+i*60:(i+1)*60] for i in range(exp.net.p.trials) ]\n",
    "    \n",
    "    err_al = np.array([calc_error(a_x, a_y, exp.net.p.trials), calc_error(b_x, b_y, exp.net.p.trials)])\n",
    "    \n",
    "    \"\"\"\n",
    "    Reallocation\n",
    "    \"\"\"\n",
    "    # New parameters\n",
    "    newParameters = {\n",
    "        **p,\n",
    "        'trials': 40,\n",
    "        'inputVaryProbs': [0.9, 0.1]\n",
    "    }\n",
    "    \n",
    "    # Initilizes a second experiment\n",
    "    exp2 = AssemblyExperiment(name='assemblies-retrain', parameters=newParameters)\n",
    "\n",
    "    # Build network\n",
    "    weights = deepcopy(exp.net.initialWeights)\n",
    "    weights.exex = exp.net.trainedWeightsExex[-1]\n",
    "    exp2.build(exp.net.initialMasks, weights)\n",
    "\n",
    "    # Run network\n",
    "    exp2.run()\n",
    "    \n",
    "    # Get support masks\n",
    "    supportMasks2 = utils.getSupportMasksForAllTrials(exp2.net.initialWeights.exex, exp2.net.trainedWeightsExex)\n",
    "    \n",
    "    #0: A, 1: B, 2: others\n",
    "    a_masks = np.array([ supportMasks2[i, 0] for i in range(exp2.net.p.trials) ])\n",
    "    b_masks = np.array([ supportMasks2[i, 1] for i in range(exp2.net.p.trials) ])\n",
    "    \n",
    "    backgroundNeurons = exp2.net.exSpikeTrains[80:]\n",
    "    \n",
    "    a_x = [ backgroundNeurons[a_masks[i],40+i*60:(i+1)*60] for i in range(exp2.net.p.trials) ]\n",
    "    b_x = [ backgroundNeurons[b_masks[i],40+i*60:(i+1)*60] for i in range(exp2.net.p.trials) ]\n",
    "    \n",
    "    err_re = np.array([calc_error(a_x, a_y, exp2.net.p.trials), calc_error(b_x, b_y, exp2.net.p.trials)])\n",
    "    \n",
    "    return np.array([err_al, err_re])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1   Processing timeseries... .. \n",
      "2   Processing timeseries... .. \n",
      "3   Processing timeseries... .. \n",
      "4   Processing timeseries... .. \n",
      "5   Processing timeseries... .. \n",
      "6   Processing timeseries... .. \n",
      "7   Processing timeseries... .. \n",
      "8   Processing timeseries... .. \n",
      "9   Processing timeseries... .. \n",
      "    Processing timeseries... .. \r"
     ]
    }
   ],
   "source": [
    "errs = np.array([run_epoch(i) for i in range(10)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "err_means = np.mean(errs, axis=0)\n",
    "err_al = err_means[0]\n",
    "err_re = err_means[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,3))\n",
    "plt.plot(err_al[0], label='A')\n",
    "plt.plot(err_al[1], label='B')\n",
    "plt.xlim(0,39)\n",
    "plt.ylim(-0.1,2.5)\n",
    "plt.xlabel('trials')\n",
    "plt.ylabel('RMSE')\n",
    "plt.legend()\n",
    "pl = plt.show()\n",
    "\n",
    "plt.figure(figsize=(10,3))\n",
    "plt.plot(err_re[0], label='A')\n",
    "plt.plot(err_re[1], label='B')\n",
    "plt.xlim(0,39)\n",
    "plt.ylim(-0.1,2.5)\n",
    "plt.xlabel('trials')\n",
    "plt.ylabel('RMSE')\n",
    "plt.legend()\n",
    "pl = plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}