Multiclass_deep_net.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import gzip\n",
    "import pickle\n",
    "from urllib.request import urlretrieve\n",
    "import zipfile\n",
    "\n",
    "import nengo\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "#from nengo_dl.compat import tf_shape\n",
    "\n",
    "import nengo_dl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000\n",
      "50000\n",
      "(50000, 784)\n",
      "<class 'numpy.ndarray'>\n",
      "[[0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.328125   0.72265625\n",
      "  0.62109375 0.58984375 0.234375   0.140625   0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.8671875  0.9921875  0.9921875  0.9921875\n",
      "  0.9921875  0.94140625 0.7734375  0.7734375  0.7734375  0.7734375\n",
      "  0.7734375  0.7734375  0.7734375  0.7734375  0.6640625  0.203125\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.26171875 0.4453125  0.28125    0.4453125  0.63671875 0.88671875\n",
      "  0.9921875  0.87890625 0.9921875  0.9921875  0.9921875  0.9765625\n",
      "  0.89453125 0.9921875  0.9921875  0.546875   0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.06640625 0.2578125  0.0546875\n",
      "  0.26171875 0.26171875 0.26171875 0.23046875 0.08203125 0.921875\n",
      "  0.9921875  0.4140625  0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.32421875 0.98828125 0.81640625 0.0703125\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.0859375\n",
      "  0.91015625 0.99609375 0.32421875 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.50390625 0.9921875  0.9296875\n",
      "  0.171875   0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.23046875 0.97265625 0.9921875  0.2421875  0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.51953125 0.9921875\n",
      "  0.73046875 0.01953125 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.03515625 0.80078125 0.96875    0.2265625  0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.4921875\n",
      "  0.9921875  0.7109375  0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.29296875 0.98046875 0.9375     0.22265625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.07421875\n",
      "  0.86328125 0.9921875  0.6484375  0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.01171875 0.79296875 0.9921875  0.85546875\n",
      "  0.13671875 0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.1484375  0.9921875  0.9921875  0.30078125 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.12109375 0.875      0.9921875\n",
      "  0.44921875 0.00390625 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.51953125 0.9921875  0.9921875  0.203125   0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.23828125 0.9453125\n",
      "  0.9921875  0.9921875  0.203125   0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.47265625 0.9921875  0.9921875  0.85546875\n",
      "  0.15625    0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.47265625 0.9921875  0.80859375 0.0703125  0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]\n",
      " [0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.453125   0.48828125\n",
      "  0.66796875 0.99609375 0.99609375 0.5859375  0.36328125 0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.66015625 0.98828125 0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.98828125 0.8515625  0.1171875  0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.66015625 0.98828125\n",
      "  0.98828125 0.98828125 0.83203125 0.5546875  0.6875     0.98828125\n",
      "  0.98828125 0.4765625  0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.203125   0.9765625  0.98828125 0.8203125  0.125\n",
      "  0.046875   0.         0.0234375  0.8046875  0.98828125 0.546875\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.30078125\n",
      "  0.98046875 0.8203125  0.09765625 0.         0.         0.\n",
      "  0.4765625  0.96875    0.98828125 0.25390625 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.12109375 0.0703125\n",
      "  0.         0.         0.         0.         0.81640625 0.98828125\n",
      "  0.98828125 0.25390625 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.45703125 0.96484375 0.98828125 0.7734375  0.0390625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.296875   0.96484375\n",
      "  0.98828125 0.90234375 0.24609375 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.5        0.98828125 0.98828125 0.5625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.6875\n",
      "  0.9609375  0.98828125 0.62109375 0.046875   0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.09765625 0.9140625  0.98828125 0.91015625\n",
      "  0.13671875 0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.7734375  0.98828125 0.98828125 0.55078125 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.3046875  0.96875    0.98828125\n",
      "  0.73828125 0.046875   0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.07421875 0.78125    0.98828125 0.98828125 0.55078125 0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.5234375  0.98828125\n",
      "  0.98828125 0.67578125 0.046875   0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.96875    0.98828125 0.98828125 0.09765625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.96875    0.98828125 0.98828125 0.16796875 0.078125   0.078125\n",
      "  0.078125   0.078125   0.01953125 0.         0.01953125 0.078125\n",
      "  0.078125   0.14453125 0.5859375  0.5859375  0.5859375  0.57421875\n",
      "  0.0390625  0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.96875    0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.98828125 0.98828125 0.98828125\n",
      "  0.65625    0.55859375 0.6484375  0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.98828125 0.48046875 0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.6796875  0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.98828125 0.97265625 0.96484375 0.96484375 0.66015625\n",
      "  0.45703125 0.45703125 0.22265625 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.4609375  0.48046875 0.48046875 0.48046875 0.6484375\n",
      "  0.98828125 0.98828125 0.98828125 0.60546875 0.48046875 0.48046875\n",
      "  0.16015625 0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "urlretrieve(\"http://deeplearning.net/data/mnist/mnist.pkl.gz\",\n",
    "            \"mnist.pkl.gz\")\n",
    "with gzip.open(\"mnist.pkl.gz\") as f:\n",
    "    train_data, _, test_data = pickle.load(f, encoding=\"latin1\")\n",
    "train_data = list(train_data)\n",
    "test_data = list(test_data)\n",
    "print(len(test_data[1]))\n",
    "print(len(train_data[0]))\n",
    "print(train_data[0].shape)\n",
    "print(type(train_data[1]))\n",
    "print(test_data[0][0:2])\n",
    "for data in (train_data, test_data):\n",
    "    one_hot = np.zeros((data[0].shape[0], 10))\n",
    "    one_hot[np.arange(data[0].shape[0]), data[1]] = 1\n",
    "    data[1] = one_hot\n",
    "\n",
    "for i in range(3):\n",
    "    plt.figure()\n",
    "    plt.imshow(np.reshape(train_data[0][i], (28, 28)),\n",
    "               cmap=\"gray\")\n",
    "    plt.axis('off')\n",
    "    plt.title(str(np.argmax(train_data[1][i])));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.]], dtype=float32),\n",
       " array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "        [1., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0., ..., 0., 1., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 1., 0.]])]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.01171875 0.0703125  0.0703125  0.0703125\n",
      "  0.4921875  0.53125    0.68359375 0.1015625  0.6484375  0.99609375\n",
      "  0.96484375 0.49609375 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.1171875  0.140625   0.3671875  0.6015625\n",
      "  0.6640625  0.98828125 0.98828125 0.98828125 0.98828125 0.98828125\n",
      "  0.87890625 0.671875   0.98828125 0.9453125  0.76171875 0.25\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.19140625\n",
      "  0.9296875  0.98828125 0.98828125 0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.98046875 0.36328125 0.3203125\n",
      "  0.3203125  0.21875    0.15234375 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.0703125  0.85546875 0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.98828125 0.7734375  0.7109375\n",
      "  0.96484375 0.94140625 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.3125     0.609375   0.41796875 0.98828125\n",
      "  0.98828125 0.80078125 0.04296875 0.         0.16796875 0.6015625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.0546875  0.00390625 0.6015625  0.98828125 0.3515625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.54296875 0.98828125 0.7421875  0.0078125  0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.04296875\n",
      "  0.7421875  0.98828125 0.2734375  0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.13671875 0.94140625\n",
      "  0.87890625 0.625      0.421875   0.00390625 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.31640625 0.9375     0.98828125\n",
      "  0.98828125 0.46484375 0.09765625 0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.17578125 0.7265625  0.98828125 0.98828125\n",
      "  0.5859375  0.10546875 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.0625     0.36328125 0.984375   0.98828125 0.73046875\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.97265625 0.98828125 0.97265625 0.25       0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.1796875  0.5078125  0.71484375 0.98828125\n",
      "  0.98828125 0.80859375 0.0078125  0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.15234375 0.578125\n",
      "  0.89453125 0.98828125 0.98828125 0.98828125 0.9765625  0.7109375\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.09375    0.4453125  0.86328125 0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.78515625 0.3046875  0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.08984375 0.2578125  0.83203125 0.98828125\n",
      "  0.98828125 0.98828125 0.98828125 0.7734375  0.31640625 0.0078125\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.0703125  0.66796875\n",
      "  0.85546875 0.98828125 0.98828125 0.98828125 0.98828125 0.76171875\n",
      "  0.3125     0.03515625 0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.21484375 0.671875   0.8828125  0.98828125 0.98828125 0.98828125\n",
      "  0.98828125 0.953125   0.51953125 0.04296875 0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.53125    0.98828125\n",
      "  0.98828125 0.98828125 0.828125   0.52734375 0.515625   0.0625\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.         0.         0.\n",
      "  0.         0.         0.         0.        ]]\n"
     ]
    }
   ],
   "source": [
    "print(train_data[0][0:1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "with nengo.Network() as net:\n",
    "    # set some default parameters for the neurons that will make\n",
    "    # the training progress more smoothly\n",
    "    net.config[nengo.Ensemble].max_rates = nengo.dists.Choice([100])\n",
    "    net.config[nengo.Ensemble].intercepts = nengo.dists.Choice([0])\n",
    "    neuron_type = nengo.LIF(amplitude=0.01)\n",
    "    #print(neuron_type.probeable)\n",
    "    #spikes = nengo.Probe(neuron_type,'spikes')\n",
    "    # we'll make all the nengo objects in the network\n",
    "    # non-trainable. we could train them if we wanted, but they don't\n",
    "    # add any representational power. note that this doesn't affect\n",
    "    # the internal components of tensornodes, which will always be\n",
    "    # trainable or non-trainable depending on the code written in\n",
    "    # the tensornode.\n",
    "    nengo_dl.configure_settings(trainable=True)\n",
    "    # the input node that will be used to feed in input images\n",
    "    inp = nengo.Node([0] * 28 * 28)\n",
    "    temp = nengo.Probe(inp)\n",
    "    # add the first convolutional layer\n",
    "    nengo_a = nengo.Ensemble(784, 1, neuron_type)\n",
    "    conn_first = nengo.Connection(inp, nengo_a.neurons,solver=nengo.solvers.LstsqL2(weights=True))\n",
    "\n",
    "    #x,conn1 = nengo_dl.tensor_layer(inp, tf.layers.dense,return_conn=True,units=784)\n",
    "    #print(\"blah\")\n",
    "    #print(type(x))\n",
    "    #print(x.tensor_func)\n",
    "    #print(x.size_in)\n",
    "    #print(x.size_out)\n",
    "    #x,conn = nengo_dl.tensor_layer(inp, tf.layers.dense,return_conn=True,units=784)\n",
    "    #conn = nengo.Connection(inp, x)\n",
    "    #net.config[conn].trainable = True\n",
    "    # apply the neural nonlinearity\n",
    "    #x = nengo_dl.tensor_layer(x, neuron_type)\n",
    "    \n",
    "    #conv\n",
    "    #x,conn2 = nengo_dl.tensor_layer(\n",
    "       # x, tf.layers.dense,return_conn=True, units=256)\n",
    "    \n",
    "    #print(x.size_in)\n",
    "    #print(x.size_out)\n",
    "\n",
    "    # apply the neural nonlinearity\n",
    "    #solver=nengo.solvers.LstsqL2(weights=True)\n",
    "    x,conn10 = nengo_dl.tensor_layer(x, neuron_type,return_conn=True, )\n",
    "    conn10.solver = nengo.solvers.LstsqL2(weights=True)\n",
    "    print(\"blah again\")\n",
    "    print(type(x))\n",
    "    print(conn10.solver)\n",
    "\n",
    "    # add another convolutional layer\n",
    "    x,conn3 = nengo_dl.tensor_layer(\n",
    "        x, tf.layers.dense, return_conn=True, units=64)\n",
    "    x = nengo_dl.tensor_layer(x, neuron_type)\n",
    "\n",
    " \n",
    "\n",
    "    # another convolutional layer\n",
    "    x,conn4 = nengo_dl.tensor_layer(\n",
    "        x, tf.layers.dense, return_conn=True, units=16)\n",
    "    x = nengo_dl.tensor_layer(x, neuron_type)\n",
    "\n",
    " \n",
    "\n",
    "    # linear readout\n",
    "    x,conn5 = nengo_dl.tensor_layer(x, tf.layers.dense, return_conn=True, units=10)\n",
    "\n",
    "    # we'll create two different output probes, one with a filter\n",
    "    # (for when we're simulating the network over time and\n",
    "    # accumulating spikes), and one without (for when we're\n",
    "    # training the network using a rate-based approximation)\n",
    "    print(x.probeable)\n",
    "    out_p = nengo.Probe(x)\n",
    "    out_p_filt = nengo.Probe(x, synapse=0.1)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "n_in = 784\n",
    "n_hidden = 64\n",
    "minibatch_size = 50\n",
    "with nengo.Network() as auto_net:\n",
    "    # input\n",
    "    nengo_a = nengo.Node(np.zeros(n_in))\n",
    "    print(nengo_a.size_out)\n",
    "\n",
    "    # first layer\n",
    "    nengo_b = nengo.Ensemble(\n",
    "        n_hidden, 1, neuron_type=nengo.RectifiedLinear())\n",
    "    print(nengo_b.size_in)\n",
    "    nengo.Connection(\n",
    "        nengo_a, nengo_b.neurons, transform=nengo_dl.dists.Glorot())\n",
    "\n",
    "    # second layer\n",
    "    nengo_c = nengo.Ensemble(\n",
    "        n_in, 1, neuron_type=nengo.RectifiedLinear())\n",
    "    nengo.Connection(\n",
    "        nengo_b.neurons, nengo_c.neurons,\n",
    "        transform=nengo_dl.dists.Glorot())\n",
    "\n",
    "    # probes are used to collect data from the network\n",
    "    p_c = nengo.Probe(nengo_c.neurons)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "with nengo.Network() as net:\n",
    "    # set some default parameters for the neurons that will make\n",
    "    # the training progress more smoothly\n",
    "    net.config[nengo.Ensemble].max_rates = nengo.dists.Choice([100])\n",
    "    net.config[nengo.Ensemble].intercepts = nengo.dists.Choice([0])\n",
    "    neuron_type = nengo.LIF(amplitude=0.01)\n",
    "\n",
    "    nengo_dl.configure_settings(trainable=True)\n",
    "    # the input node that will be used to feed in input images\n",
    "    inp = nengo.Node(np.zeros(784))\n",
    "    print(inp.size_out)\n",
    "    temp = nengo.Probe(inp)\n",
    "    # add the first layer\n",
    "    nengo_a = nengo.Ensemble(784, 1, neuron_type = neuron_type)\n",
    "    nengo.Connection(inp, nengo_a.neurons, transform=nengo_dl.dists.Glorot())\n",
    "    \n",
    "    nengo_b = nengo.Ensemble(256, 1, neuron_type = neuron_type)\n",
    "    print(nengo_b.size_in)\n",
    "    nengo.Connection(nengo_a, nengo_b, solver=nengo.solvers.LstsqL2(weights=True),transform=nengo_dl.dists.Glorot())\n",
    "    \n",
    "    nengo_c = nengo.Ensemble(64, 1, neuron_type = neuron_type)\n",
    "    nengo.Connection(nengo_b, nengo_c,solver=nengo.solvers.LstsqL2(weights=True),transform=nengo_dl.dists.Glorot())\n",
    "    \n",
    "    nengo_d = nengo.Ensemble(16, 1, neuron_type = neuron_type)\n",
    "    nengo.Connection(nengo_c, nengo_d,solver=nengo.solvers.LstsqL2(weights=True), transform=nengo_dl.dists.Glorot())\n",
    "    \n",
    "    # add another convolutional lay\n",
    "    #nengo_e = nengo.Ensemble(10, 1, neuron_type = neuron_type)\n",
    "    #nengo.Connection(nengo_d, nengo_e,solver=nengo.solvers.LstsqL2(weights=True),transform=nengo_dl.dists.Glorot(), label=\"conn_fifth\")\n",
    "\n",
    "\n",
    "    # we'll create two different output probes, one with a filter\n",
    "    # (for when we're simulating the network over time and\n",
    "    # accumulating spikes), and one without (for when we're\n",
    "    # training the network using a rate-based approximation)\n",
    "    x, conn = nengo_dl.tensor_layer(nengo_d, tf.layers.dense, return_conn=True, units=10)\n",
    "    print(x.probeable)\n",
    "    out_p = nengo.Probe(x)\n",
    "    out_p_filt = nengo.Probe(x, synapse=0.1)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "784\n",
      "<class 'nengo.ensemble.Neurons'>\n",
      "256\n",
      "WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/nengo_dl/tensor_node.py:325: dense (from tensorflow.python.layers.core) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use keras.layers.dense instead.\n",
      "WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n",
      "('output',)\n"
     ]
    }
   ],
   "source": [
    "with nengo.Network() as net:\n",
    "    # set some default parameters for the neurons that will make\n",
    "    # the training progress more smoothly\n",
    "    net.config[nengo.Ensemble].max_rates = nengo.dists.Choice([100])\n",
    "    net.config[nengo.Ensemble].intercepts = nengo.dists.Choice([0])\n",
    "    neuron_type = nengo.LIF(amplitude=0.01)\n",
    "\n",
    "    nengo_dl.configure_settings(trainable=True)\n",
    "    # the input node that will be used to feed in input images\n",
    "    inp = nengo.Node(np.zeros(784))\n",
    "    \n",
    "    print(inp.size_out)\n",
    "    temp = nengo.Probe(inp)\n",
    "    # add the first layer\n",
    "    x, conn1 = nengo_dl.tensor_layer(inp, neuron_type, transform=nengo_dl.dists.Glorot(),shape_in=(784,), return_conn=True)\n",
    "    print(type(x))\n",
    "    A_spikes = nengo.Probe(x)\n",
    "    x, conn2 = nengo_dl.tensor_layer(x, neuron_type, transform=nengo_dl.dists.Glorot(),shape_in=(256,), return_conn=True)\n",
    "    print(conn2.size_out)\n",
    "    x, conn3 = nengo_dl.tensor_layer(x, neuron_type, transform=nengo_dl.dists.Glorot(),shape_in=(64,), return_conn=True)\n",
    "    \n",
    "    x, conn4 = nengo_dl.tensor_layer(x, neuron_type, transform=nengo_dl.dists.Glorot(),shape_in=(16,), return_conn=True)\n",
    "    \n",
    "    #x, conn5 = nengo_dl.tensor_layer(x, neuron_type, transform=nengo_dl.dists.Glorot(),shape_in=(10,), return_conn=True)\n",
    "    \n",
    "    x,conn_l = nengo_dl.tensor_layer(x, tf.layers.dense, units=10, return_conn=True)\n",
    "\n",
    "\n",
    "    # we'll create two different output probes, one with a filter\n",
    "    # (for when we're simulating the network over time and\n",
    "    # accumulating spikes), and one without (for when we're\n",
    "    # training the network using a rate-based approximation)\n",
    "    print(x.probeable)\n",
    "    out_p = nengo.Probe(x)\n",
    "    out_p_filt = nengo.Probe(x, synapse=0.1)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Build finished in 0:00:00                                                      \n",
      "|#                Optimizing graph: creating signals                  | 0:00:00"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/nengo_dl/simulator.py:131: UserWarning: No GPU support detected. It is recommended that you install tensorflow-gpu (`pip install tensorflow-gpu`).\n",
      "  \"No GPU support detected. It is recommended that you \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization finished in 0:00:00                                               \n",
      "Construction finished in 0:00:01                                               ##################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################| ETA: 0:00:00\n"
     ]
    }
   ],
   "source": [
    "minibatch_size = 200\n",
    "sim = nengo_dl.Simulator(net, minibatch_size=minibatch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(400, 30, 784)\n"
     ]
    }
   ],
   "source": [
    "# add the single timestep to the training data\n",
    "train_data = {inp: train_data[0][:, None, :],\n",
    "              out_p: train_data[1][:, None, :]}\n",
    "\n",
    "# when testing our network with spiking neurons we will need to run it\n",
    "# over time, so we repeat the input/target data for a number of\n",
    "# timesteps. we're also going to reduce the number of test images, just\n",
    "# to speed up this example.\n",
    "n_steps = 30\n",
    "test_data = {\n",
    "    inp: np.tile(test_data[0][:minibatch_size*2, None, :],\n",
    "                 (1, n_steps, 1)),\n",
    "    out_p_filt: np.tile(test_data[1][:minibatch_size*2, None, :],\n",
    "                        (1, n_steps, 1))}\n",
    "print(test_data[inp].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def objective(outputs, targets):\n",
    "    return tf.nn.softmax_cross_entropy_with_logits_v2(\n",
    "        logits=outputs, labels=targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt = tf.train.RMSPropOptimizer(learning_rate=0.001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|####################Calculating loss (100%)###################| ETA:  00:00:00WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use standard file APIs to check for files with this prefix.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use standard file APIs to check for files with this prefix.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculation finished in 0:00:12                                                \n",
      "error before training: 89.00%\n"
     ]
    }
   ],
   "source": [
    "def classification_error(outputs, targets):\n",
    "    return 100 * tf.reduce_mean(\n",
    "        tf.cast(tf.not_equal(tf.argmax(outputs[:, -1], axis=-1),\n",
    "                             tf.argmax(targets[:, -1], axis=-1)),\n",
    "                tf.float32))\n",
    "\n",
    "print(\"error before training: %.2f%%\" % sim.loss(\n",
    "    test_data, {out_p_filt: classification_error}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|                   Training (0%)                  | ETA:  --:--:-- (loss: ---)"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/nengo_dl/simulator.py:575: UserWarning: Simulator was created with tensorboard=False; ignoring requested summaries\n",
      "  warnings.warn(\"Simulator was created with tensorboard=False; \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|                   Training (0%)                  | ETA:  --:--:-- (loss: ---)WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/array_grad.py:425: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/array_grad.py:425: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training finished in 0:12:04 (loss: 5.7278)                                    \n"
     ]
    }
   ],
   "source": [
    "do_training = True\n",
    "#summ_list = []\n",
    "if do_training:\n",
    "    # run training\n",
    "    sim.train(train_data, opt, objective={out_p: objective},n_epochs=10,summaries=[conn_l])\n",
    "    #print(sim.probeable)\n",
    "    # save the parameters to file\n",
    "    sim.save_params(\"./mnist_params\")\n",
    "#else:\n",
    "    # download pretrained weights\n",
    "    #urlretrieve(\n",
    "      #  \"https://drive.google.com/uc?export=download&id=1u9JyNuRxQDUcFgkRnI1qfJVFMdnGRsjI\",\n",
    "       # \"mnist_params.zip\")\n",
    "    #with zipfile.ZipFile(\"mnist_params.zip\") as f:\n",
    "        #f.extractall()\n",
    "\n",
    "    # load parameters\n",
    "    #sim.load_params(\"./mnist_params\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculation finished in 0:00:03                                                \n",
      "error after training: 3.00%\n"
     ]
    }
   ],
   "source": [
    "print(\"error after training: %.2f%%\" % sim.loss(\n",
    "    test_data, {out_p_filt: classification_error}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulation finished in 0:00:02                                                 \n",
      "oooo\n",
      "(200, 30, 784)\n",
      "(784,)\n",
      "[0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.328125   0.72265625\n",
      " 0.62109375 0.58984375 0.234375   0.140625   0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.8671875  0.9921875  0.9921875  0.9921875\n",
      " 0.9921875  0.94140625 0.7734375  0.7734375  0.7734375  0.7734375\n",
      " 0.7734375  0.7734375  0.7734375  0.7734375  0.6640625  0.203125\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.26171875 0.4453125  0.28125    0.4453125  0.63671875 0.88671875\n",
      " 0.9921875  0.87890625 0.9921875  0.9921875  0.9921875  0.9765625\n",
      " 0.89453125 0.9921875  0.9921875  0.546875   0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.06640625 0.2578125  0.0546875\n",
      " 0.26171875 0.26171875 0.26171875 0.23046875 0.08203125 0.921875\n",
      " 0.9921875  0.4140625  0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.32421875 0.98828125 0.81640625 0.0703125\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.0859375\n",
      " 0.91015625 0.99609375 0.32421875 0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.50390625 0.9921875  0.9296875\n",
      " 0.171875   0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.23046875 0.97265625 0.9921875  0.2421875  0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.51953125 0.9921875\n",
      " 0.73046875 0.01953125 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.03515625 0.80078125 0.96875    0.2265625  0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.4921875\n",
      " 0.9921875  0.7109375  0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.29296875 0.98046875 0.9375     0.22265625\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.07421875\n",
      " 0.86328125 0.9921875  0.6484375  0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.01171875 0.79296875 0.9921875  0.85546875\n",
      " 0.13671875 0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.1484375  0.9921875  0.9921875  0.30078125 0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.12109375 0.875      0.9921875\n",
      " 0.44921875 0.00390625 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.51953125 0.9921875  0.9921875  0.203125   0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.23828125 0.9453125\n",
      " 0.9921875  0.9921875  0.203125   0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.47265625 0.9921875  0.9921875  0.85546875\n",
      " 0.15625    0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.47265625 0.9921875  0.80859375 0.0703125  0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.        ]\n",
      "(30, 10)\n",
      "[[ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00\n",
      "   0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00\n",
      "   0.00000000e+00  0.00000000e+00]\n",
      " [-1.60276875e-04 -5.17056789e-04  3.86100175e-04 -8.58501124e-04\n",
      "  -2.41348636e-04 -3.80599755e-04 -7.15419534e-04  5.59254957e-04\n",
      "   9.22847656e-04  5.76934486e-04]\n",
      " [-3.18958977e-04 -1.02896872e-03  7.68358586e-04 -1.70846004e-03\n",
      "  -4.80295799e-04 -7.57412519e-04 -1.42372050e-03  1.11294526e-03\n",
      "   1.83651282e-03  1.14812842e-03]\n",
      " [-4.76062181e-04 -1.53578713e-03  1.14681351e-03 -2.54996167e-03\n",
      "  -7.16865412e-04 -1.13047590e-03 -2.12497381e-03  1.66112627e-03\n",
      "   2.74108676e-03  1.71363889e-03]\n",
      " [-6.31602190e-04 -2.03756243e-03  1.52150262e-03 -3.38309025e-03\n",
      "  -9.51081107e-04 -1.49982725e-03 -2.81924941e-03  2.20385264e-03\n",
      "   3.63666005e-03  2.27352232e-03]\n",
      " [-7.85594515e-04 -2.53434526e-03  1.89246354e-03 -4.20792913e-03\n",
      "  -1.18296640e-03 -1.86550349e-03 -3.50661715e-03  2.74117896e-03\n",
      "   4.52332245e-03  2.82783504e-03]\n",
      " [-9.38054640e-04 -3.02618486e-03  2.25973339e-03 -5.02456073e-03\n",
      "  -1.41254428e-03 -2.22754106e-03 -4.18714527e-03  3.27315880e-03\n",
      "   5.40116243e-03  3.37663223e-03]\n",
      " [-1.08899770e-03 -3.51313059e-03  2.62334896e-03 -5.83306653e-03\n",
      "  -1.63983786e-03 -2.58597638e-03 -4.86090221e-03  3.79984546e-03\n",
      "   6.27026754e-03  3.91996885e-03]\n",
      " [-3.41612637e-01 -9.34340898e-03  4.78102528e-02 -1.47572413e-01\n",
      "  -1.61891002e-02 -2.32324675e-01 -1.85696647e-01  3.26650351e-01\n",
      "  -5.16599901e-02 -2.85138506e-02]\n",
      " [-3.87107641e-01 -1.29621223e-01 -5.88811599e-02 -2.36843079e-01\n",
      "  -1.19775403e-02 -2.50062823e-01 -1.61184475e-01  3.12979192e-01\n",
      "  -2.43161004e-02 -4.51021828e-03]\n",
      " [-3.83416146e-01 -1.28848523e-01 -5.79091832e-02 -2.35344946e-01\n",
      "  -1.20997103e-02 -2.47955263e-01 -1.60296082e-01  3.10424238e-01\n",
      "  -2.31513046e-02 -3.88840656e-03]\n",
      " [-6.21901333e-01 -2.37596646e-01  2.04829797e-02 -2.55396694e-01\n",
      "  -7.69269615e-02 -4.13024217e-01 -7.25301355e-02  4.22803521e-01\n",
      "  -4.19940613e-02 -2.44447216e-01]\n",
      " [-5.67248642e-01 -1.62706584e-01  7.87632391e-02 -2.62741804e-01\n",
      "  -3.80913876e-02 -4.49259102e-01 -2.05812663e-01  5.02350092e-01\n",
      "   4.25246134e-02 -2.02584833e-01]\n",
      " [-5.61764717e-01 -1.61604688e-01  7.83656314e-02 -2.60986000e-01\n",
      "  -3.79537232e-02 -4.45169508e-01 -2.04480216e-01  4.97910857e-01\n",
      "   4.30243351e-02 -1.99992150e-01]\n",
      " [-6.40913188e-01 -8.23504403e-02 -4.10035998e-03 -3.52489471e-01\n",
      "  -1.42361850e-01 -3.88001323e-01 -3.07117790e-01  5.18696904e-01\n",
      "  -1.14241652e-02 -1.50086462e-01]\n",
      " [-7.37369478e-01 -9.38041508e-02 -4.42715026e-02 -3.24995667e-01\n",
      "  -3.30092192e-01 -4.18179274e-01 -3.67856026e-01  4.86342937e-01\n",
      "   1.38554201e-01 -1.75883591e-01]\n",
      " [-9.67826426e-01  3.41677964e-02 -2.84506768e-01 -4.88567919e-01\n",
      "  -2.01176316e-01 -5.14637589e-01 -5.34803569e-01  6.70999169e-01\n",
      "   3.62091362e-02 -1.69593960e-01]\n",
      " [-8.96075368e-01  8.03521127e-02 -2.72661239e-01 -4.67430830e-01\n",
      "  -3.16030920e-01 -4.34513807e-01 -5.00388086e-01  5.65836906e-01\n",
      "   1.03801176e-01 -2.89340258e-01]\n",
      " [-9.38229144e-01  1.19569033e-01 -1.84322119e-01 -5.38535476e-01\n",
      "  -3.77609462e-01 -5.81970930e-01 -6.32098436e-01  7.13666201e-01\n",
      "   2.41796255e-01 -2.35642478e-01]\n",
      " [-1.06232619e+00  2.14151740e-02 -2.32396841e-01 -6.57500088e-01\n",
      "  -5.30726850e-01 -6.54027402e-01 -8.50868821e-01  6.69509172e-01\n",
      "   5.97586781e-02 -6.74789697e-02]\n",
      " [-1.16504860e+00  2.28763130e-02 -2.43277073e-01 -6.39909029e-01\n",
      "  -4.19812411e-01 -7.83852100e-01 -7.56453097e-01  7.79317319e-01\n",
      "  -1.04997233e-02 -2.27412313e-01]\n",
      " [-1.25315094e+00 -1.03778597e-02 -2.13328287e-01 -7.00269639e-01\n",
      "  -5.18669784e-01 -8.87869358e-01 -7.52326787e-01  8.41828227e-01\n",
      "   4.54546660e-02 -2.13183910e-01]\n",
      " [-1.32378614e+00  3.00268698e-02 -1.94390580e-01 -6.73600197e-01\n",
      "  -3.44672978e-01 -8.72303665e-01 -9.48439658e-01  9.21173036e-01\n",
      "  -2.77070329e-02 -8.51938576e-02]\n",
      " [-1.24849319e+00  7.62523860e-02 -1.83441713e-01 -6.50622010e-01\n",
      "  -4.58099753e-01 -7.88621068e-01 -9.09908414e-01  8.13521504e-01\n",
      "   4.05209847e-02 -2.05779970e-01]\n",
      " [-1.36724043e+00  4.74035740e-04 -3.01358640e-01 -8.80591393e-01\n",
      "  -5.09393036e-01 -8.07172179e-01 -9.15340662e-01  8.94941270e-01\n",
      "  -2.21386515e-02 -2.55985558e-01]\n",
      " [-1.55995488e+00  1.69500992e-01 -3.91765535e-01 -8.68920028e-01\n",
      "  -4.25876260e-01 -9.85178590e-01 -1.06575513e+00  1.05628276e+00\n",
      "  -4.77841683e-03 -1.84680179e-01]\n",
      " [-1.63634777e+00  4.10611182e-02 -3.32243323e-01 -8.45839918e-01\n",
      "  -3.73053133e-01 -9.70907748e-01 -1.08276904e+00  1.07466900e+00\n",
      "  -7.51738995e-02 -1.51636317e-01]\n",
      " [-1.56929910e+00  8.03663433e-02 -2.01907530e-01 -8.99772346e-01\n",
      "  -2.86628574e-01 -1.06105912e+00 -1.13919318e+00  1.22248375e+00\n",
      "  -2.44681425e-02 -2.34009236e-01]\n",
      " [-1.66927898e+00  1.14053056e-01 -2.81636983e-01 -8.27308297e-01\n",
      "  -2.22785413e-01 -1.12736642e+00 -1.10871494e+00  1.25203538e+00\n",
      "  -5.97454160e-02 -2.68973768e-01]\n",
      " [-1.73740757e+00  1.90564454e-01 -3.60520899e-01 -9.13176775e-01\n",
      "  -3.25354427e-01 -1.06341028e+00 -1.20235527e+00  1.26531768e+00\n",
      "  -1.13171339e-01 -2.18381703e-01]]\n",
      "we are here\n"
     ]
    },
    {
     "ename": "NameError",
     "evalue": "name 'conn5' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-12-4a18755c9272>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     47\u001b[0m     \u001b[0;31m#print(sim.data[conn2].weights)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     48\u001b[0m     \u001b[0;31m#print(sim.data[conn3].weights)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 49\u001b[0;31m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mconn5\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mweights\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     50\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m     \u001b[0;31m#print(sim.data[ens].bias)  # bias values\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'conn5' is not defined"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#import pandas as pd\n",
    "#pd.set_option('display.height', 1000)\n",
    "#pd.set_option('display.max_rows', 500)\n",
    "#pd.set_option('display.max_columns', 500)\n",
    "#pd.set_option('display.width', 1000)\n",
    "import sys\n",
    "np.set_printoptions(threshold=sys.maxsize)\n",
    "\n",
    "sim.run_steps(n_steps, data={inp: test_data[inp][:minibatch_size]})\n",
    "\n",
    "for i in range(1):\n",
    "    plt.figure()\n",
    "    plt.subplot(1, 2, 1)\n",
    "    plt.imshow(np.reshape(test_data[inp][i, 0], (28, 28)),\n",
    "               cmap=\"gray\")\n",
    "    plt.axis('off')\n",
    "\n",
    "    plt.subplot(1, 2, 2)\n",
    "    plt.plot(sim.trange(), sim.data[out_p_filt][i])\n",
    "    plt.legend([str(i) for i in range(10)], loc=\"upper left\")\n",
    "    plt.xlabel(\"time\")\n",
    "    \n",
    "    plt.subplot(2, 1, 1)\n",
    "    plt.plot(sim.trange(), sim.data[A_spikes][i])\n",
    "    plt.legend([str(i) for i in range(10)], loc=\"upper left\")\n",
    "    plt.xlabel(\"time\")\n",
    "    #plt.subplot(2, 1, 1)\n",
    "    #plt.plot(sim.trange(), sim.data[temp][i])\n",
    "    #plt.legend([str(i) for i in range(10)], loc=\"best\")\n",
    "    #plt.xlabel(\"time\")\n",
    "    \n",
    "    #print(sim.data[inp]) #temp gives input current given to the 784 neurons for all input images\n",
    "    #print(sim.data[inp][i,0].shape) #i denotes the index of image in the minibatch\n",
    "    #print(sim.data[A_spikes][i]) #w\n",
    "    \n",
    "    print(\"oooo\")\n",
    "    \n",
    "    print(sim.data[temp].shape) #temp gives input current given to the 784 neurons for all input images\n",
    "    print(sim.data[temp][i,0].shape) #i denotes the index of image in the minibatch\n",
    "    print(sim.data[temp][i,0]) #we are giving the same input for 30 steps so just i,o would give the input without repetition\n",
    "    \n",
    "    print(sim.data[out_p_filt][i].shape) #this is for output current\n",
    "    print(sim.data[out_p_filt][i])\n",
    "    \n",
    "    print(\"we are here\")\n",
    "    #print(sim.data[conn1].weights)  # connection weights\n",
    "    #print(sim.data[conn2].weights)\n",
    "    #print(sim.data[conn3].weights)\n",
    "    print(sim.data[conn5].weights)\n",
    "    \n",
    "    #print(sim.data[ens].bias)  # bias values\n",
    "    #print(sim.data[ens].encoders)  # encoder values\n",
    "    #print(sim.data[ens])  # to see all the parameters for an object\n",
    "\n",
    "    \n",
    "    #with open('out.txt', 'w') as f:\n",
    "    #print >> f, 'Filename:', filename     # Python 2.x\n",
    "        #print(sim.data[temp][i], file=f) \n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Sess = tf.compat.v1.Session()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#trial = tf.trainable_variables(scope=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Sess. run(tf.trainable_weights())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "te = np.arange(18)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "te.reshape(3,6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "aligned(8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "time  = np.arange(1, 201, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "14"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot([1,2,3],[1,2,3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "pot = [1,0,0,9,7]\n",
    "tada = max(pot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tada"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8288545086937564"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import random\n",
    "random.uniform(0.5,1)\n"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}