CHI.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np # linear algebra\n",
    "import matplotlib.pyplot as plt\n",
    "import time\n",
    "%matplotlib inline\n",
    "import tensorflow\n",
    "#from sklearn.model_selection import train_test_split\n",
    "#from sklearn.metrics import confusion_matrix\n",
    "from keras.utils.np_utils import to_categorical # convert to one-hot-encoding\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D, BatchNormalization\n",
    "from keras.layers.advanced_activations import LeakyReLU, PReLU\n",
    "from keras.optimizers import Adam\n",
    "#from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.callbacks import LearningRateScheduler\n",
    "#import pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "523.8856117725372\n"
     ]
    }
   ],
   "source": [
    "t=time.time()\n",
    "train_data = np.array(list(np.loadtxt(\"train.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "train_labels = np.array(list(np.loadtxt(\"train_label.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "eval_data = np.array(list(np.loadtxt(\"eval.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "eval_labels = np.array(list(np.loadtxt(\"eval_label.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "\n",
    "#eval_data = pandas.read_csv(\"eval.csv\", delimiter=\",\",header=None,dtype='float32')\n",
    "#eval_labels = pandas.read_csv(\"eval_label.csv\", delimiter=\",\",header=None,dtype='float32')\n",
    "print(time.time()-t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(400, 1728) (400, 4)\n",
      "1.3678913116455078\n"
     ]
    }
   ],
   "source": [
    "t=time.time()\n",
    "test_data = np.array(list(np.loadtxt(\"H4_CG.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "test_labels = np.array(list(np.loadtxt(\"H4CG_label.csv\", delimiter=\",\"))).astype(\"float32\")\n",
    "print(np.shape(test_data),np.shape(test_labels))\n",
    "print(time.time()-t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = train_data.reshape(-1, 24,24,3)\n",
    "x_val = eval_data.reshape(-1, 24,24,3)\n",
    "x_test = test_data.reshape(-1, 24,24,3)\n",
    "\n",
    "y_train = train_labels.reshape(-1, 22,22,1)\n",
    "y_val = eval_labels.reshape(-1, 22,22,1)\n",
    "#y_test = test_label.reshape(-1, 22,22,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Conv2D(filters = 1, kernel_size = (3, 3), activation='tanh', input_shape = (24, 24, 3)))\n",
    "#model.add(BatchNormalization())\n",
    "#model.add(Conv2D(filters = 1, kernel_size = (1, 1), activation='tanh', padding='same'))\n",
    "#model.add(BatchNormalization())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='mean_squared_error', optimizer = Adam(lr=1e-4))#, metrics=[\"accuracy\"]) #binary_crossentropy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "annealer = LearningRateScheduler(lambda x: 1e-3 * 0.9 ** x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 64000 samples, validate on 16000 samples\n",
      "Epoch 1/10\n",
      " - 8s - loss: 0.0878 - val_loss: 0.0492\n",
      "Epoch 2/10\n",
      " - 7s - loss: 0.0324 - val_loss: 0.0199\n",
      "Epoch 3/10\n",
      " - 7s - loss: 0.0142 - val_loss: 0.0101\n",
      "Epoch 4/10\n",
      " - 8s - loss: 0.0080 - val_loss: 0.0064\n",
      "Epoch 5/10\n",
      " - 7s - loss: 0.0056 - val_loss: 0.0050\n",
      "Epoch 6/10\n",
      " - 8s - loss: 0.0046 - val_loss: 0.0043\n",
      "Epoch 7/10\n",
      " - 7s - loss: 0.0041 - val_loss: 0.0040\n",
      "Epoch 8/10\n",
      " - 7s - loss: 0.0039 - val_loss: 0.0038\n",
      "Epoch 9/10\n",
      " - 7s - loss: 0.0038 - val_loss: 0.0038\n",
      "Epoch 10/10\n",
      " - 7s - loss: 0.0037 - val_loss: 0.0037\n"
     ]
    }
   ],
   "source": [
    "hist = model.fit(x=x_train, y=y_train, batch_size=512,\n",
    "                           epochs=10, #Increase this when not on Kaggle kernel\n",
    "                           verbose=2,  #1 for ETA, 0 for silent\n",
    "                           validation_data=(x_val, y_val), #For speed\n",
    "                           callbacks=[annealer])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "plt.plot(hist.history['loss'], color='b')\n",
    "plt.plot(hist.history['val_loss'], color='r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_hat = model.predict(x_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16000, 22, 22, 1) (16000, 22, 22, 1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.3428540345129893"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(np.shape(y_val),np.shape(y_hat))\n",
    "(np.sum((y_val-y_hat)**2)/(20*20*40))**0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "cannot reshape array of size 7744000 into shape (20,20,40)",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-39-9c7e1ec10eac>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0my_val\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_val\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m40\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0my_hat\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_hat\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m40\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mimshow\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_val\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\numpy\\core\\fromnumeric.py\u001b[0m in \u001b[0;36mreshape\u001b[1;34m(a, newshape, order)\u001b[0m\n\u001b[0;32m    277\u001b[0m            [5, 6]])\n\u001b[0;32m    278\u001b[0m     \"\"\"\n\u001b[1;32m--> 279\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0m_wrapfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'reshape'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnewshape\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0morder\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0morder\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    280\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    281\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\tensorflow\\lib\\site-packages\\numpy\\core\\fromnumeric.py\u001b[0m in \u001b[0;36m_wrapfunc\u001b[1;34m(obj, method, *args, **kwds)\u001b[0m\n\u001b[0;32m     49\u001b[0m \u001b[1;32mdef\u001b[0m \u001b[0m_wrapfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     50\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 51\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     52\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     53\u001b[0m     \u001b[1;31m# An AttributeError occurs if the object does not have\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mValueError\u001b[0m: cannot reshape array of size 7744000 into shape (20,20,40)"
     ]
    }
   ],
   "source": [
    "y_val=np.reshape(y_val,(20,20,40))\n",
    "y_hat=np.reshape(y_hat,(20,20,40))\n",
    "\n",
    "plt.imshow(y_val[:,:,20])\n",
    "plt.show()\n",
    "plt.imshow(y_hat[:,:,20])\n",
    "plt.show()\n",
    "\n",
    "print(np.shape(y_hat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_hat = model.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Error Analysis\n",
    "n=len(y_test)\n",
    "print(np.mean(y_hat-y_test))\n",
    "\n",
    "print((np.sum((y_hat-y_test)**2)/(n))**0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "(np.sum(np.sum((y_test-y_hat)**2,axis=0),axis=0)/(20*20))**0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test=np.reshape(y_test,(20,20))\n",
    "y_hat=np.reshape(y_hat,(20,20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "x=np.arange(0,4,4/20)\n",
    "plt.plot(x,y_test[:,19],'r')\n",
    "plt.scatter(x,y_hat[:,19])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.contour.QuadContourSet at 0x16cb405e1d0>"
      ]
     },
     "execution_count": 310,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.contour(y_hat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "W=model.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'name': 'conv2d_18', 'trainable': True, 'batch_input_shape': (None, 24, 24, 3), 'dtype': 'float32', 'filters': 1, 'kernel_size': (3, 3), 'strides': (1, 1), 'padding': 'valid', 'data_format': 'channels_last', 'dilation_rate': (1, 1), 'activation': 'tanh', 'use_bias': True, 'kernel_initializer': {'class_name': 'VarianceScaling', 'config': {'scale': 1.0, 'mode': 'fan_avg', 'distribution': 'uniform', 'seed': None}}, 'bias_initializer': {'class_name': 'Zeros', 'config': {}}, 'kernel_regularizer': None, 'bias_regularizer': None, 'activity_regularizer': None, 'kernel_constraint': None, 'bias_constraint': None}\n",
      "[array([[[[-8.1317371e-04],\n",
      "         [ 5.1146219e-03],\n",
      "         [ 6.5396344e-03]],\n",
      "\n",
      "        [[ 2.2287475e-02],\n",
      "         [-9.8511798e-04],\n",
      "         [-1.8057708e-02]],\n",
      "\n",
      "        [[-1.0302169e-03],\n",
      "         [-4.6328022e-03],\n",
      "         [ 7.9389019e-03]]],\n",
      "\n",
      "\n",
      "       [[[-6.7465684e-05],\n",
      "         [ 2.2198537e-02],\n",
      "         [-1.0959053e-03]],\n",
      "\n",
      "        [[-5.9630307e-03],\n",
      "         [ 5.6676039e-05],\n",
      "         [-2.7690774e-03]],\n",
      "\n",
      "        [[ 4.2593097e-03],\n",
      "         [-2.2524429e-02],\n",
      "         [-3.0477650e-03]]],\n",
      "\n",
      "\n",
      "       [[[ 2.6667150e-04],\n",
      "         [-4.2880583e-04],\n",
      "         [ 1.2110445e-03]],\n",
      "\n",
      "        [[-2.1619616e-02],\n",
      "         [-6.7683571e-04],\n",
      "         [-1.6188229e-03]],\n",
      "\n",
      "        [[-9.3402911e-04],\n",
      "         [ 3.7595505e-04],\n",
      "         [ 3.5609982e-03]]]], dtype=float32), array([-0.00958441], dtype=float32)]\n"
     ]
    }
   ],
   "source": [
    "for layer in model.layers:\n",
    "    g=layer.get_config()\n",
    "    h=layer.get_weights()\n",
    "    print(g)\n",
    "    print(h)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 3, 3, 1)\n",
      "(1,)\n"
     ]
    }
   ],
   "source": [
    "for i in range(2):\n",
    "    print(np.shape(W[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([-0.00958441], dtype=float32)]\n"
     ]
    }
   ],
   "source": [
    "plt.subplot(131)\n",
    "plt.imshow(W[0][:,:,0,0])\n",
    "plt.subplot(132)\n",
    "plt.imshow(W[0][:,:,1,0])\n",
    "plt.subplot(133)\n",
    "plt.imshow(W[0][:,:,2,0])\n",
    "plt.show()\n",
    "print(W[1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savetxt('Weight.csv',np.reshape(W[0],(1,-1)),delimiter=',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "\n",
    "inp = model.input                                           # input placeholder\n",
    "outputs = [layer.output for layer in model.layers]          # all layer outputs\n",
    "functors = [K.function([inp, K.learning_phase()], [out]) for out in outputs]    # evaluation functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"batch_normalization_1/keras_learning_phase:0\", shape=(), dtype=bool)\n",
      "Tensor(\"conv2d_52/Relu:0\", shape=(?, 19, 19, 16), dtype=float32)\n",
      "Tensor(\"batch_normalization_49/cond/Merge:0\", shape=(?, 19, 19, 16), dtype=float32)\n",
      "Tensor(\"conv2d_53/Relu:0\", shape=(?, 17, 17, 16), dtype=float32)\n",
      "Tensor(\"batch_normalization_50/cond/Merge:0\", shape=(?, 17, 17, 16), dtype=float32)\n",
      "Tensor(\"max_pooling2d_34/MaxPool:0\", shape=(?, 8, 8, 16), dtype=float32)\n",
      "Tensor(\"dropout_71/cond/Merge:0\", shape=(?, 8, 8, 16), dtype=float32)\n",
      "Tensor(\"conv2d_54/Relu:0\", shape=(?, 6, 6, 32), dtype=float32)\n",
      "Tensor(\"batch_normalization_51/cond/Merge:0\", shape=(?, 6, 6, 32), dtype=float32)\n",
      "Tensor(\"conv2d_55/Relu:0\", shape=(?, 4, 4, 32), dtype=float32)\n",
      "Tensor(\"batch_normalization_52/cond/Merge:0\", shape=(?, 4, 4, 32), dtype=float32)\n",
      "Tensor(\"max_pooling2d_35/MaxPool:0\", shape=(?, 2, 2, 32), dtype=float32)\n",
      "Tensor(\"dropout_72/cond/Merge:0\", shape=(?, 2, 2, 32), dtype=float32)\n",
      "Tensor(\"flatten_19/Reshape:0\", shape=(?, ?), dtype=float32)\n",
      "Tensor(\"dense_56/BiasAdd:0\", shape=(?, 2048), dtype=float32)\n",
      "Tensor(\"leaky_re_lu_38/LeakyRelu:0\", shape=(?, 2048), dtype=float32)\n",
      "Tensor(\"dropout_73/cond/Merge:0\", shape=(?, 2048), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "print(K.learning_phase())\n",
    "for i in range(16):\n",
    "    print(outputs[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Testing\n",
    "#test = x_test[10,:,:,:][np.newaxis,...]\n",
    "test=np.random.rand(1,24,24,3)\n",
    "layer_outs = [func([test, 0]) for func in functors]\n",
    "#print(layer_outs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 19, 19, 16)\n",
      "(1, 1, 19, 19, 16)\n",
      "(1, 1, 17, 17, 16)\n",
      "(1, 1, 17, 17, 16)\n",
      "(1, 1, 8, 8, 16)\n",
      "(1, 1, 8, 8, 16)\n",
      "(1, 1, 6, 6, 32)\n",
      "(1, 1, 6, 6, 32)\n",
      "(1, 1, 4, 4, 32)\n",
      "(1, 1, 4, 4, 32)\n",
      "(1, 1, 2, 2, 32)\n",
      "(1, 1, 2, 2, 32)\n",
      "(1, 1, 128)\n",
      "(1, 1, 2048)\n",
      "(1, 1, 2048)\n",
      "(1, 1, 2048)\n",
      "(1, 1, 512)\n",
      "(1, 1, 512)\n",
      "(1, 1, 512)\n",
      "(1, 1, 1024)\n"
     ]
    }
   ],
   "source": [
    "for i in range(20):\n",
    "    print(np.shape(layer_outs[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 24, 24, 3)\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 19, 19, 16) 7.5127034\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 19, 19, 16) -191.24802\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 17, 17, 16) 69.89182\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 17, 17, 16) -86.77472\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 8, 8, 16) 25.455803\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 8, 8, 16) 25.455803\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 6, 6, 32) 10.93575\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 6, 6, 32) -1.3638282\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 4, 4, 32) 7.066024\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 4, 4, 32) 2.4494896\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 2, 2, 32) 4.9899244\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 2, 2, 32) 4.9899244\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"
    }
   ],
   "source": [
    "print(np.shape(test))\n",
    "plt.imshow(test[0,:,:,2])\n",
    "plt.show()\n",
    "for i in range(12):\n",
    "    print(np.shape(layer_outs[i]), np.sum(np.array(layer_outs[i])[0,0,:,:,0]))\n",
    "    plt.imshow(np.array(layer_outs[i])[0,0,:,:,0])\n",
    "    plt.show()\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "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"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAHlxJREFUeJzt3Xl8VPW9//HXJ/tKICTsSwBZ1SIacV+oG9Kr2P5sRa9rvdLe6vV2ufVnl2ut7e3Pq78uau1V3LrcXrFqvXIrimvdKgoUWQKEfQkBAmRPZjJLvvePTDCmAQaYzJmZvJ+PxzzmzJkvOZ98HzNvvvmezZxziIhIaknzugAREYk9hbuISApSuIuIpCCFu4hIClK4i4ikIIW7iEgKUriLiKQghbuISApSuIuIpKAMrzZcUlLiysrKvNq8iEhSWrZs2T7nXOnh2nkW7mVlZSxdutSrzYuIJCUz2xZNO03LiIikIIW7iEgKUriLiKQghbuISApSuIuIpCCFu4hIClK4i4ikIIW7iEictAZC/PTVSlbsqO/1bSncRUTipMkf4qE3N1JR3djr21K4i4jESTDcDkBmuvX6thTuIiJxEgw7ALIyej96Fe4iInHSOXLPSFO4i4ikjEBI0zIiIinnwJy7pmVERFJH55x7ZiJMy5jZk2ZWY2arD/K+mdmDZrbRzFaa2cmxL1NEJPn5g2EAcjITINyBXwMzD/H+pcD4yGMu8B/HXpaISOr5JNzTe31bhw1359w7QO0hmswGfus6LAb6m9nQWBUoIpIqfIkU7lEYDuzo8roqsk5ERLroHLnnZiVHuPd0TI/rsaHZXDNbamZL9+7dG4NNi4gkD3+w42iZnCQ5WqYKGNnl9QiguqeGzrl5zrly51x5aelhb94tIpJSfEk2cl8AXB85auZ0oME5tysGP1dEJKUc2KGa0fvhnnG4Bmb2NHA+UGJmVcAPgEwA59wjwEJgFrARaAVu6q1iRUSSmS8YJisjjbS03j9D9bDh7py7+jDvO+DWmFUkIpKi/IEwuXE4UgZ0hqqISNz4g+0KdxGRVOMLhuNydioo3EVE4sYfDMflBCZQuIuIxI0vGI7LYZCgcBcRiRt/MByXwyBB4S4iEjf+YLtG7iIiqUY7VEVEUpB2qIqIpCB/UCcxiYikHF9AI3cRkZTinMMf0hmqIiIpJRh2hNudjpYREUkl/lDH5X6z43CjDlC4i4jEhT8Qvxt1gMJdRCQufHG8UQco3EVE4qLz/qkauYuIpJAD90/V0TIiIqmj8/6p2br8gIhI6tjd4AegX05mXLancBcRiYPqBh8A40oL4rI9hbuISBy0tIVITzNdFVJEJJU0+0PkZ6VjZnHZnsJdRCQOmtvCFGRnxG17CncRkThoaQuRr3AXEUktjf4gBTkJFu5mNtPMKs1so5nd2cP7o8zsLTNbbmYrzWxW7EsVEUlO4XbH6p0NjB8UnyNlIIpwN7N04GHgUmAKcLWZTenW7PvAH5xz04A5wK9iXaiISLLyBcM0+kMcl0jhDkwHNjrnNjvnAsB8YHa3Ng7oF1kuAqpjV6KISHILhDquK5OVHr+Z8GgmgIYDO7q8rgJO69bmbuBVM/snIB+4MCbViYikgAPhHqcrQkJ0I/eeDsp03V5fDfzaOTcCmAX8zsz+5meb2VwzW2pmS/fu3Xvk1YqIJKHOcI/XjTogunCvAkZ2eT2Cv512uRn4A4Bz7gMgByjp/oOcc/Occ+XOufLS0tKjq1hEJMkEwh0XDctKsHBfAow3szFmlkXHDtMF3dpsBy4AMLPJdIS7huYiIkCDLwhAYSIdCumcCwG3AYuAtXQcFVNhZveY2eWRZt8CbjGzFcDTwI3Oue5TNyIifVJNYxsAgwpz4rbNqP4bcc4tBBZ2W3dXl+U1wFmxLU1EJDVs2tsMwOB+2XHbps5QFRHpZa+vraFsYB7F+Vlx26bCXUSkFznn2F7byuljB8btipCgcBcR6VXVDX5qWwJMGdbv8I1jSOEuItKL1u9uAmDyUIW7iEjKqNzTEe4TBhXGdbsKdxGRXlS5u4mhRTkU5cXnxtidFO4iIr1o/Z4mJgyO76gdFO4iIr3GOcfWfS2MKcmP+7YV7iIiveSlVbtoCYQZPTAv7ttWuIuI9IKd9T6++8dVDO+fy+VTh8V9+wp3EZEY8wfDfPV3ywi1Ox67vpyBBfG77ECn+F2iTESkj1hUsZtVOxt45NqT437yUieN3EVEYmzT3hbSDGZMGuRZDQp3EZEYW7xpP8P655Idx9vqdadwFxGJob1NbXy0tZZRxfE/QqYrhbuISAwtXLULgK+eN87TOhTuIiIxtGlvM9kZaZwz/m9uIx1XCncRkRhaurWOwf1y4nrt9p4o3EVEYmTZtlrW7GpkSpwv79sThbuISIwsXLUbgHuuON7jShTuIiIxUdPoZ/5H2zl/YimDCnO8LkfhLiISC798ayMtgTD/9NnjvC4FULiLiMREdb2fSUMKOWV0sdelAAp3EZFj1t7uqKhuoLQw/hcIOxiFu4jIMXpm6Q52Nfj53IlDvS7lgKjC3cxmmlmlmW00szsP0uZLZrbGzCrM7L9iW6aISOJat6uRnMw0rjp1pNelHHDYS/6aWTrwMHARUAUsMbMFzrk1XdqMB74DnOWcqzMz7y6FJiISR6FwO2+v38u40gLPT1zqKpqR+3Rgo3Nus3MuAMwHZndrcwvwsHOuDsA5VxPbMkVEEtNvP9jG1v2tnl9Lprtown04sKPL66rIuq4mABPM7H0zW2xmM2NVoIhIogqG23nozQ1MHtqPWQk03w7R3Ympp78zXA8/ZzxwPjACeNfMTnDO1X/qB5nNBeYCjBo16oiLFRFJJD9/bT11rUHuv3Iq6WmJMyUD0Y3cq4CuewlGANU9tHnRORd0zm0BKukI+09xzs1zzpU758pLS0uPtmYREc9t3dfCvHc2c8GkQVw4ZbDX5fyNaMJ9CTDezMaYWRYwB1jQrc1/AzMAzKyEjmmazbEsVEQkkSzZWkuo3XHHzElel9Kjw4a7cy4E3AYsAtYCf3DOVZjZPWZ2eaTZImC/ma0B3gK+7Zzb31tFi4h47Z0N+wAYMSDX40p6Fs2cO865hcDCbuvu6rLsgG9GHiIiKa+63sfI4lzys6OK0bjTGaoiIkfo/kXrWLatjtlTux84mDgU7iIiR+CjLbU8+d5Wpo4o4rYEuQJkTxTuIiJH4P+9vJbi/CwevHoaOZnpXpdzUAp3EZEoBULtVFQ3cukJQxg9MN/rcg5J4S4iEqX1e5oIhNqZOrK/16UclsJdRCRKH+/oOOl+6giFu4hISvAHwzy7rIqSgixGFifmse1dKdxFRKJw94IKVuyo5//OnJRQl/Y9GIW7iMhhPPX+FuYv2cGtM8bxxfLEuSHHoSTmqVUiIgni3pfX8cjbmzh3QinfvGii1+VETSN3EZGDeOK9LTzy9iYuOX4wj11/SsJd1vdQFO4iIj3YWe/j3pfXMqwohwfmTCM7I3FPWOqJwl1EpJvqeh/XPfEh6WnGH756RkKfiXowmnMXEeli895mrnzkAxp8QZ64oZwRA/K8LumoaOQuIhKxv7mN65/8iJa2EI9eewrnTxzkdUlHTSN3ERGgoTXIZQ+9R3WDn998eTrnTUjuW4Fq5C4ifV57u+P7L65mT1Mbj153StIHOyjcRaSPC7c77nh+Jf+zopqvXzCeS44f4nVJMaFpGRHps3bW+/jq75axamcDt804LqFvvnGkFO4i0iftrPdx7eMfsmVfCzeeWca/XJI8Z59GQ+EuIn1OayDEHc+tYMu+lpTYedoTzbmLSJ8SCrdz238t54NN+/nOpZNSMthBI3cR6UOcc/z4pbW8ua6GH15+PDecWeZ1Sb1G4S4ifUIw3M7PX1vPr/+ylQsnD07pYAdNy4hIHzHvnc386s8dV3h8YM5JXpfT66IKdzObaWaVZrbRzO48RLsrzcyZWXnsShQROTYNviCPv7uZ8tEDePS6cvKzU3/S4rDhbmbpwMPApcAU4Gozm9JDu0LgduDDWBcpInK0/MEw33zmY+pag9x+wXivy4mbaEbu04GNzrnNzrkAMB+Y3UO7HwH3Af4Y1icictT8wTC3P72cN9bVcO3pozhnfInXJcVNNOE+HNjR5XVVZN0BZjYNGOmc+1MMaxMROWrOOW56agmvrtnD/zl5BD++4sSkuLF1rEQz8dRTb7gDb5qlAT8HbjzsDzKbC8wFGDVqVHQViogchXtfWccHm/dz64xxfPuSSV6XE3fRjNyrgK63+x4BVHd5XQicAPzZzLYCpwMLetqp6pyb55wrd86Vl5am5okDIuK9j3fU8+jbmzlheD++ceEEr8vxRDThvgQYb2ZjzCwLmAMs6HzTOdfgnCtxzpU558qAxcDlzrmlvVKxiMghtLc7bn96Of1yMnjqxulkpPfNI74P+1s750LAbcAiYC3wB+dchZndY2aX93aBIiLRCoXbmffuZrbXtvLD2cdTWpjtdUmeiepgT+fcQmBht3V3HaTt+cdelojIkdnV4OOO51by7oZ9nDi8iMs+M8zrkjyV+kfyi0hKC7c77nhuJc//tQoz+JeLJzD33HF9djqmk8JdRJLag29s4Pm/VnHjmWXcdFYZowfme11SQlC4i0hS8gfD/PP85Syq2MMZYwfyg8um9Knj2A+nb//dIiJJ63svrD4Q7E/cWK5g70YjdxFJKtv3t/L7j7bx/F+ruO700fzoihO8LikhKdxFJGm0BkJ86dEP2N3o55zxJSl339NYUriLSFJwzvG9F1azu9HPEzeUc8HkwV6XlNAU7iKS8Bp8Qa5/4kNWVDXwD2ePUbBHQeEuIgmtyR/khwsqWFHVwHdnTeIfzh7rdUlJQeEuIgmrriXAlY/8hU17W7h1xjjmnjvO65KShsJdRBLS0x9t5ycvraUlEGLedadw8fFDvC4pqSjcRSSh7G7w89T7W3jivS2cPHoAd1wykfKyYq/LSjoKdxFJGM8u3cGP/rSG5rYQl00dxj2zT6AoN9PrspKSwl1EPFfT6OeO51fy58q9TBnaj1/MOYkJgwu9LiupKdxFxFMvrdzFt59bQVuonStPGcFPPn8iWRm6MsqxUriLiCecczy7tIrvv7iayUP78dMvfobjBmm0HisKdxGJK+ccO2p9PPHeZn7zwTbOHDeQB6+eRklB371rUm9QuItI3ATD7fzjfy7j9bU1AHzh5OHcf+VU0tN0RcdYU7iLSFws21bLd/+4mso9TZw7oZQfzz6BUQPzvC4rZSncRaRXNfiC/NtLa3hh+U6KcjN5+JqTmXXiEF1/vZcp3EWkVzjnWLuriX99cTXLttVx+dRh3DP7ePrnZXldWp+gcBeRmKpvDfD2+r08/u4WVu1sIDsjjYeunsZlU4d5XVqfonAXkZjwBcI8u2wH971SSXNbiOL8LL7/ucl87jNDGVqU63V5fY7CXUSO2Wtr9vCDF1dT3eDnxOFF3DrjOD47aZBORvKQwl1Ejsl/L9/J15/5mHGl+Tx106mcP6FUO0sTQFT/rZrZTDOrNLONZnZnD+9/08zWmNlKM3vDzEbHvlQRSTT7mtu4f1ElA/Iyeen2c5gxcZCCPUEcNtzNLB14GLgUmAJcbWZTujVbDpQ75z4DPAfcF+tCRSSx1DT5uempJeys9/Grvz+FnMx0r0uSLqKZlpkObHTObQYws/nAbGBNZwPn3Ftd2i8Gro1lkSKSONbvaeKRP2/ixRXVOOf42Zemcsa4gV6XJd1EE+7DgR1dXlcBpx2i/c3Ayz29YWZzgbkAo0aNirJEEUkENU1+HntnM4+/t4XM9DSuP2M0f3/aaI4bVOB1adKDaMK9pwk012NDs2uBcuC8nt53zs0D5gGUl5f3+DNEJLFU1/u4f1El/7OimlC749SyAfzsSycxsliXDkhk0YR7FTCyy+sRQHX3RmZ2IfA94DznXFtsyhMRr2ysaeIXr29g4apdtDu48cwyrjtjNONKNVJPBtGE+xJgvJmNAXYCc4BrujYws2nAo8BM51xNzKsUkbhp9Ad5dmkVP321EgNuOXcsV5WPZKxCPakcNtydcyEzuw1YBKQDTzrnKszsHmCpc24BcD9QADwbOQxqu3Pu8l6sW0RibF9zGw+8voHn/1pFayBM+eiO6RdduTE5RXUSk3NuIbCw27q7uixfGOO6RCQO2tsdb1XWsHx7Pb/5YCu+QJgrpg3nxjPLOGF4kdflyTHQGaoifdSybXX8ZOFalm2rwwxOLSvmrr+bolBPEQp3kT6mwRfkx39aw7PLqijKzeT2C8Zz89ljKMrN9Lo0iSGFu0gf4AuEmb9kO0u21vLmuhr8wXYmD+3HM185nX45CvVUpHAXSXG+QJhrHl/M8u31jBiQyyXHD+Ga6aM4bazOKk1lCneRFFZd7+O+V9axfHs9N55Zxt2XH+91SRInCneRFOOcY/HmWl5fu4dnluygNRDiy2eN4a7Lul/vT1KZwl0khaysqufBNzbw+toastLTOHt8CXf93RTKSvK9Lk3iTOEukgLeXr+Xp97fwp8r91KUm8m3L5nIl88aQ26WLsPbVyncRZLY9v2t3LdoHX9auYthRTl89bxxfG3GOB0BIwp3kWTT0hbi7fV7eWX1bhZV7KYt1M4t54zhWxdP1A0z5ACFu0gSCIXb+cum/bywfCcLV+2iLdROSUEW54wv4bbPjuekkf29LlESjMJdJIEFQu089OYGXli+k6o6H0W5mXzh5BFcPnUYp5YNICM9qtsgSx+kcBdJUBXVDXzjmY9Zv6eZ6WXF3HnpJC6cPFhTLxIVhbtIgmnwBbl/0Tr+c/F28rPSeejqaVw2dZjXZUmSUbiLJAh/MMwDb2zg94u30egPcc74Ev7/F6cyuF+O16VJElK4i3jMOceb62q49+V1bKhp5pLjB3PrjOM4cXgRkZvfiBwxhbuIh2oa/Xznj6t4Y10NQ4ty+PVNp3L+xEFelyUpQOEu4oENe5pYsKKa336wjSZ/kK+dP45vXDSBTB39IjGicBeJk9ZAiOeXVfFKxW7+smk/BkwfU8z3P6e7H0nsKdxFetm2/S3MX7KD+R9tp641yITBBfzjeeP48tljKCnI9ro8SVEKd5Fe9PPX1vPAGxsAuGDSIL42YxynjC72uCrpCxTuIjG2bX8Lr1bsYVHFbpZuq+PCyYO5Z/bxDOuf63Vp0oco3EVi5KMttTzwxnre37gfgElDCvnWRRP4ynnjyMrQjlKJL4W7yDGq3N3Eo29v4o/LdzKoMJs7Zk7kss8MY2RxntelSR+mcBc5SqFwO0+8t4V7X1lHuhk3nDGaO2ZOIj9bXyvxXlSfQjObCTwApAOPO+fu7fZ+NvBb4BRgP3CVc25rbEsVSQzOOSr3NHHfK5W8ua6GCyYN4t8+fyJDinSZAEkchw13M0sHHgYuAqqAJWa2wDm3pkuzm4E659xxZjYH+Hfgqt4oWMQr4XZHRXUD//piBSt21JNm8N1Zk7jlnLG6TIAknGhG7tOBjc65zQBmNh+YDXQN99nA3ZHl54Bfmpk551wMaxXpdf5gmAZf8JNHa8fz5n3N/P7D7dS3BinIzuCms8q45ZyxOgJGElY04T4c2NHldRVw2sHaOOdCZtYADAT2dW1kZnOBuQCjRo06ypJFPhEKt9PSFqapLUhLW5iWQIjWzudAiJa28IHnlrYQLd2WWwNhmv0hGv0hGv1BAqH2g27rwsmDmXXiEGZMHMSA/Kw4/pYiRy6acO/p783uI/Jo2uCcmwfMAygvL9eovg8KhdtpCUTCtS1Ec1voQCh/sq4jkJvbOteFaW4LRdZ9+t+2HSKMu8vLSic/O4OC7IyO5awMBuRlMbI4j345GfTLyaRfbiZFPTwG5GVRlKebTkvyiCbcq4CRXV6PAKoP0qbKzDKAIqA2JhWKpzpHxs2BEK1dwrgzbDvDuDNsO0fJzV0CuDUQPuIwTjPIz8ogPzuD/Ox0CrI7lof3z6Igu2tIZ1CQk/GpdfldwjsvO528rAzyMtNJS9O8uPQd0YT7EmC8mY0BdgJzgGu6tVkA3AB8AFwJvKn5du+E2x1N/iBN/hB1rQEafV1GwV2mJTrXtQY+HcadUxuxCOMReXk9Bm/n8sHW5WSmaSelyDE4bLhH5tBvAxbRcSjkk865CjO7B1jqnFsAPAH8zsw20jFin9ObRfcFgVA7+5rb2NfcRoMvSLM/RJM/RFNbKLIcpK41SH1rgHpfkEZfR5g3+YO0BMKH/flpRkcQZ306jEfm55Gf9ekw7mjXfV36gTBXGIsknqiOc3fOLQQWdlt3V5dlP/DF2JaWepxz1LUGqW1po7blk+f9zW2s3d3IvuYA+5vb2NccoMEXPOTPystKp39uJv3zsuifl8m40gIKczIozMmkX27Hc2FOxoE2CmORvkWn0sVAuN2xp9FPdb2PnfU+qus7lqvrfQdG3g2+II3+EOH2nmerRhbnMqwol4lDCjmrIJvSgmxKCrMpKcimKDczEtwZFGZnkp+dToZu6iAih6BwP4hwu2N/Sxt7mz551LYEIlMlAWpbAgde1zS1/U1oF+VmMqx/LqWF2YwamE9RbgZFuZmUFGQzsCCb4rwsBuRnUpyfxYC8LHIy0z36TUUkFfXJcG9vd2zd38LOeh+76v3savCzu9HHrgY/exo7g7yNngbZ2RlplBZmU5yfRXF+FhOHFDKoMJvhA3IZ3r/jMbR/LgW6voiIeKhPJNDepjZeWlnNml2NVO5ppnJ3I/7gp48CKSnIZmhRDsP753DSyP6UFmZ3PAo+eS4pzCI3M11z1SKS8FIy3H2BMMu21fHGuj28u2EfG2uaASgtzGZcaT7XTB/NpCGFlJXkM7Qoh0H9ssnO0LSIiKSOlAn35dvreOzdzVRUN7K9thXnICs9jdPHDeTz04ZzweRBTBrSz+syRUTiIqnDva4lwGPvbubjHfUs3ryf4vwsThszkC9MG8HY0nxmTBqkuW8R6ZOSNvmq6lq56tHF7Kz3ceLwIm4+ewy3fXY8Rbm6/oeISFKGe11LgMt/+T61LQEeufYUZp4wxOuSREQSSlKeCfPU+1uobQnwi6tOUrCLiPQg6cLdFwjz1PtbmV5WzBXThntdjohIQkq6cH91zW6a2kJ85byxXpciIpKwki7c87MyuGjKYGZMHOR1KSIiCSvpdqheOGUwF04Z7HUZIiIJLelG7iIicngKdxGRFKRwFxFJQQp3EZEUpHAXEUlBCncRkRSkcBcRSUEKdxGRFGTO9XCj0Hhs2GwvsO0o/3kJsC+G5aQy9VV01E/RUT9Fr7f6arRzrvRwjTwL92NhZkudc+Ve15EM1FfRUT9FR/0UPa/7StMyIiIpSOEuIpKCkjXc53ldQBJRX0VH/RQd9VP0PO2rpJxzFxGRQ0vWkbuIiBxC0oW7mc00s0oz22hmd3pdj9fMbKuZrTKzj81saWRdsZm9ZmYbIs8DIuvNzB6M9N1KMzvZ2+p7l5k9aWY1Zra6y7oj7hszuyHSfoOZ3eDF79KbDtJPd5vZzsjn6mMzm9Xlve9E+qnSzC7psj6lv5tmNtLM3jKztWZWYWb/HFmfmJ8p51zSPIB0YBMwFsgCVgBTvK7L4z7ZCpR0W3cfcGdk+U7g3yPLs4CXAQNOBz70uv5e7ptzgZOB1UfbN0AxsDnyPCCyPMDr3y0O/XQ38C89tJ0S+d5lA2Mi38f0vvDdBIYCJ0eWC4H1kf5IyM9Uso3cpwMbnXObnXMBYD4w2+OaEtFs4DeR5d8AV3RZ/1vXYTHQ38yGelFgPDjn3gFqu60+0r65BHjNOVfrnKsDXgNm9n718XOQfjqY2cB851ybc24LsJGO72XKfzedc7ucc3+NLDcBa4HhJOhnKtnCfTiwo8vrqsi6vswBr5rZMjObG1k32Dm3Czo+kEDnDWfVf0feN325z26LTCc82TnVgPoJADMrA6YBH5Kgn6lkC3frYV1fP9znLOfcycClwK1mdu4h2qr/Du5gfdNX++w/gHHAScAu4KeR9X2+n8ysAHge+LpzrvFQTXtYF7e+SrZwrwJGdnk9Aqj2qJaE4JyrjjzXAC/Q8efxns7plshzTaS5+u/I+6ZP9plzbo9zLuycawceo+NzBX28n8wsk45g/71z7o+R1Qn5mUq2cF8CjDezMWaWBcwBFnhck2fMLN/MCjuXgYuB1XT0Sece+BuAFyPLC4DrI3vxTwcaOv+c7EOOtG8WAReb2YDI1MTFkXUprdu+mM/T8bmCjn6aY2bZZjYGGA98RB/4bpqZAU8Aa51zP+vyVmJ+przeA30Ue6xn0bGXehPwPa/r8bgvxtJxVMIKoKKzP4CBwBvAhshzcWS9AQ9H+m4VUO7179DL/fM0HVMKQTpGSzcfTd8AX6Zjx+FG4Cavf6849dPvIv2wko6QGtql/fci/VQJXNplfUp/N4Gz6Zg+WQl8HHnMStTPlM5QFRFJQck2LSMiIlFQuIuIpCCFu4hIClK4i4ikIIW7iEgKUriLiKQghbuISApSuIuIpKD/BU6ntJAh916iAAAAAElFTkSuQmCC\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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-1.4002874 -2.1922913  1.9350648 ... -2.3408444 -1.5874015 -1.0232062]\n"
     ]
    }
   ],
   "source": [
    "for i in range(12,19):\n",
    "    plt.plot(np.sort(np.array(layer_outs[i])[0,0,:]))\n",
    "    plt.show()\n",
    "print(np.array(layer_outs[19])[0,0,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"batch_normalization_1/keras_learning_phase:0\", shape=(), dtype=bool)\n"
     ]
    }
   ],
   "source": [
    "print(K.learning_phase())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(400, 24, 24, 3)"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.shape(x_test)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}