main_keras_realization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**File name**: main_keras_realization.ipynb<br>\n",
    "**Authors**: Anton Lukashchuk &lt;[academik116@gmail.com](mailto:academik116@gmail.com)&gt;, Anton Karazeev &lt;[anton.karazeev@gmail.com](mailto:anton.karazeev@gmail.com)&gt;, Maxim Karpov &lt;[maxim.karpov@epfl.ch](mailto:maxim.karpov@epfl.ch)&gt; <br>\n",
    "This file is part of REDE project (https://github.com/akarazeev/REDE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "%config InlineBackend.figure_format='retina'\n",
    "%matplotlib inline\n",
    "\n",
    "# Silence warnings\n",
    "import warnings\n",
    "warnings.simplefilter(action=\"ignore\", category=FutureWarning)\n",
    "warnings.simplefilter(action=\"ignore\", category=UserWarning)\n",
    "warnings.simplefilter(action=\"ignore\", category=RuntimeWarning)\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import interpolate\n",
    "import os\n",
    "import tqdm\n",
    "import pickle\n",
    "import scipy.io as spio\n",
    "\n",
    "from IPython.display import Image\n",
    "\n",
    "np.random.seed(123)\n",
    "import matplotlib.pyplot as plt\n",
    "#plt.rcParams[\"figure.figsize\"] = (8, 8)\n",
    "plt.rcParams[\"font.size\"] = 14"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Model\n",
    "from keras.layers import Conv2D\n",
    "from keras.layers import Input, Dense\n",
    "from keras.layers import Activation, Flatten\n",
    "from keras.layers import MaxPool2D, Dropout, BatchNormalization\n",
    "from keras.utils import plot_model\n",
    "from keras import utils\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error as mse\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 718,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('data/rede/raw/5070-62-111-images.pkl', 'rb') as f:\n",
    "    a = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 719,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5070, 62, 111)"
      ]
     },
     "execution_count": 719,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 599,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "\n",
    "with open('data/rede/raw/5070-62-111-images.pkl', 'rb') as f:\n",
    "    img = pickle.load(f)\n",
    "    \n",
    "with open('data/rede/raw/5070-frequencies_modes.pkl', 'rb') as f:\n",
    "    Disp = pickle.load(f)    \n",
    "    \n",
    "with open('data/rede/raw/5070-5-parameters.pkl', 'rb') as f:\n",
    "    parameters_list = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 600,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5070, 62, 111)"
      ]
     },
     "execution_count": 600,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 601,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.50e-07, 1.55e-06, 8.00e-07, 2.00e-05, 8.80e-07])"
      ]
     },
     "execution_count": 601,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters_list[3010]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 603,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x18d0235fd30>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 226,
       "width": 375
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# import matplotlib.cm as cm\n",
    "# import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "plt.imshow(img[5000])#, cmap=cm.Greys_r)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 604,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5070, 2, 50)"
      ]
     },
     "execution_count": 604,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Disp.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 605,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.50e+14, 1.55e+14, 1.60e+14, 1.65e+14, 1.70e+14, 1.75e+14,\n",
       "       1.80e+14, 1.85e+14, 1.90e+14, 1.95e+14, 2.00e+14, 2.05e+14,\n",
       "       2.10e+14, 2.15e+14, 2.20e+14, 2.25e+14, 2.30e+14, 2.35e+14,\n",
       "       2.40e+14, 2.45e+14, 2.50e+14, 2.55e+14, 2.60e+14, 2.65e+14,\n",
       "       2.70e+14, 2.75e+14, 2.80e+14, 2.85e+14, 2.90e+14, 2.95e+14,\n",
       "       3.00e+14, 3.05e+14, 3.10e+14, 3.15e+14, 3.20e+14, 3.25e+14,\n",
       "       3.30e+14, 3.35e+14, 3.40e+14, 3.45e+14, 3.50e+14, 3.55e+14,\n",
       "       3.60e+14, 3.65e+14, 3.70e+14, 3.75e+14, 3.80e+14, 3.85e+14,\n",
       "       3.90e+14, 3.95e+14])"
      ]
     },
     "execution_count": 605,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "freq_grid = Disp[0,0,:]\n",
    "freq_grid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 606,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5070, 50)"
      ]
     },
     "execution_count": 606,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Dint = Disp[:,1,:]\n",
    "Dint.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 607,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(np.array(Dint), np.array(parameters_list)*10**7*[1,1,1,1e-2,1],\n",
    "                                                  test_size=0.2,\n",
    "                                                  random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 608,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4056, 50)"
      ]
     },
     "execution_count": 608,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 609,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lukashch\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\sklearn\\preprocessing\\data.py:675: DeprecationWarning: The parameter y on transform() is deprecated since 0.19 and will be removed in 0.21\n",
      "  DeprecationWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[-0.90867512, -0.86860248, -0.82651408, ...,  0.46336936,\n",
       "         0.48335698,  0.5020384 ],\n",
       "       [ 0.55700102,  0.61819261,  0.67060301, ...,  1.28108283,\n",
       "         1.21810288,  1.16333774],\n",
       "       [ 0.40467559,  0.44131716,  0.47404122, ...,  0.89987021,\n",
       "         0.88401011,  0.86797456],\n",
       "       ...,\n",
       "       [-0.17308039, -0.19390666, -0.21365399, ..., -0.70282562,\n",
       "        -0.68739114, -0.6719052 ],\n",
       "       [ 1.44915211,  1.39588821,  1.34358009, ..., -0.26849714,\n",
       "        -0.30052332, -0.33099853],\n",
       "       [-0.39775457, -0.43636651, -0.47323738, ...,  0.30557833,\n",
       "         0.34187546,  0.37661035]])"
      ]
     },
     "execution_count": 609,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sc = StandardScaler()\n",
    "sc.fit_transform(X_train,y_train)\n",
    "sc.transform(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 612,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = Input(shape=(50,))\n",
    "x = BatchNormalization()(inputs)\n",
    "x = Dense(1000,activation='relu')(x)\n",
    "# x = Dense(50,activation='relu')(x)\n",
    "# x = Dropout(0.05, noise_shape=None, seed=None)(x)\n",
    "# x = Dense(22,activation='relu')(inputs)\n",
    "\n",
    "x = BatchNormalization()(x)\n",
    "x = Dense(20,activation='relu')(x)\n",
    "# x = Dropout(0.05, noise_shape=None, seed=None)(x)\n",
    "\n",
    "x = BatchNormalization()(x)\n",
    "out = Dense(5,activation='linear')(x)\n",
    "\n",
    "model = Model(inputs=inputs, outputs=out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 613,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='Nadam',\n",
    "                loss='mean_squared_error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 614,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3244 samples, validate on 812 samples\n",
      "Epoch 1/80\n",
      " - 8s - loss: 66.9275 - val_loss: 50.2269\n",
      "Epoch 2/80\n",
      " - 4s - loss: 29.3780 - val_loss: 13.6793\n",
      "Epoch 3/80\n",
      " - 2s - loss: 5.8865 - val_loss: 2.2129\n",
      "Epoch 4/80\n",
      " - 2s - loss: 1.0133 - val_loss: 0.5081\n",
      "Epoch 5/80\n",
      " - 2s - loss: 0.3464 - val_loss: 0.3396\n",
      "Epoch 6/80\n",
      " - 2s - loss: 0.2843 - val_loss: 0.2748\n",
      "Epoch 7/80\n",
      " - 2s - loss: 0.2723 - val_loss: 0.2665\n",
      "Epoch 8/80\n",
      " - 2s - loss: 0.2728 - val_loss: 0.2845\n",
      "Epoch 9/80\n",
      " - 3s - loss: 0.2737 - val_loss: 0.2789\n",
      "Epoch 10/80\n",
      " - 2s - loss: 0.2688 - val_loss: 0.2694\n",
      "Epoch 11/80\n",
      " - 2s - loss: 0.2668 - val_loss: 0.2657\n",
      "Epoch 12/80\n",
      " - 2s - loss: 0.2687 - val_loss: 0.2789\n",
      "Epoch 13/80\n",
      " - 2s - loss: 0.2661 - val_loss: 0.2704\n",
      "Epoch 14/80\n",
      " - 4s - loss: 0.2631 - val_loss: 0.2806\n",
      "Epoch 15/80\n",
      " - 3s - loss: 0.2638 - val_loss: 0.2725\n",
      "Epoch 16/80\n",
      " - 4s - loss: 0.2625 - val_loss: 0.2809\n",
      "Epoch 17/80\n",
      " - 3s - loss: 0.2633 - val_loss: 0.2755\n",
      "Epoch 18/80\n",
      " - 2s - loss: 0.2619 - val_loss: 0.2604\n",
      "Epoch 19/80\n",
      " - 2s - loss: 0.2625 - val_loss: 0.2790\n",
      "Epoch 20/80\n",
      " - 2s - loss: 0.2607 - val_loss: 0.2751\n",
      "Epoch 21/80\n",
      " - 2s - loss: 0.2597 - val_loss: 0.2675\n",
      "Epoch 22/80\n",
      " - 2s - loss: 0.2603 - val_loss: 0.2631\n",
      "Epoch 23/80\n",
      " - 2s - loss: 0.2600 - val_loss: 0.2659\n",
      "Epoch 24/80\n",
      " - 2s - loss: 0.2600 - val_loss: 0.2617\n",
      "Epoch 25/80\n",
      " - 2s - loss: 0.2584 - val_loss: 0.2668\n",
      "Epoch 26/80\n",
      " - 2s - loss: 0.2576 - val_loss: 0.2606\n",
      "Epoch 27/80\n",
      " - 2s - loss: 0.2574 - val_loss: 0.2644\n",
      "Epoch 28/80\n",
      " - 2s - loss: 0.2569 - val_loss: 0.2735\n",
      "Epoch 29/80\n",
      " - 2s - loss: 0.2570 - val_loss: 0.2698\n",
      "Epoch 30/80\n",
      " - 2s - loss: 0.2568 - val_loss: 0.2829\n",
      "Epoch 31/80\n",
      " - 2s - loss: 0.2568 - val_loss: 0.2610\n",
      "Epoch 32/80\n",
      " - 2s - loss: 0.2569 - val_loss: 0.2669\n",
      "Epoch 33/80\n",
      " - 2s - loss: 0.2590 - val_loss: 0.2598\n",
      "Epoch 34/80\n",
      " - 2s - loss: 0.2558 - val_loss: 0.2589\n",
      "Epoch 35/80\n",
      " - 2s - loss: 0.2549 - val_loss: 0.2577\n",
      "Epoch 36/80\n",
      " - 2s - loss: 0.2569 - val_loss: 0.2612\n",
      "Epoch 37/80\n",
      " - 2s - loss: 0.2552 - val_loss: 0.2582\n",
      "Epoch 38/80\n",
      " - 2s - loss: 0.2564 - val_loss: 0.2630\n",
      "Epoch 39/80\n",
      " - 2s - loss: 0.2554 - val_loss: 0.2608\n",
      "Epoch 40/80\n",
      " - 2s - loss: 0.2543 - val_loss: 0.2660\n",
      "Epoch 41/80\n",
      " - 2s - loss: 0.2548 - val_loss: 0.2645\n",
      "Epoch 42/80\n",
      " - 2s - loss: 0.2548 - val_loss: 0.2551\n",
      "Epoch 43/80\n",
      " - 2s - loss: 0.2544 - val_loss: 0.2707\n",
      "Epoch 44/80\n",
      " - 2s - loss: 0.2546 - val_loss: 0.2615\n",
      "Epoch 45/80\n",
      " - 2s - loss: 0.2540 - val_loss: 0.2686\n",
      "Epoch 46/80\n",
      " - 2s - loss: 0.2543 - val_loss: 0.2697\n",
      "Epoch 47/80\n",
      " - 2s - loss: 0.2539 - val_loss: 0.2579\n",
      "Epoch 48/80\n",
      " - 2s - loss: 0.2550 - val_loss: 0.2637\n",
      "Epoch 49/80\n",
      " - 2s - loss: 0.2535 - val_loss: 0.2578\n",
      "Epoch 50/80\n",
      " - 2s - loss: 0.2532 - val_loss: 0.2626\n",
      "Epoch 51/80\n",
      " - 2s - loss: 0.2545 - val_loss: 0.2622\n",
      "Epoch 52/80\n",
      " - 2s - loss: 0.2532 - val_loss: 0.2595\n",
      "Epoch 53/80\n",
      " - 2s - loss: 0.2530 - val_loss: 0.2569\n",
      "Epoch 54/80\n",
      " - 2s - loss: 0.2528 - val_loss: 0.2609\n",
      "Epoch 55/80\n",
      " - 2s - loss: 0.2538 - val_loss: 0.2577\n",
      "Epoch 56/80\n",
      " - 2s - loss: 0.2530 - val_loss: 0.2607\n",
      "Epoch 57/80\n",
      " - 2s - loss: 0.2525 - val_loss: 0.2566\n",
      "Epoch 58/80\n",
      " - 2s - loss: 0.2538 - val_loss: 0.2600\n",
      "Epoch 59/80\n",
      " - 2s - loss: 0.2522 - val_loss: 0.2592\n",
      "Epoch 60/80\n",
      " - 2s - loss: 0.2528 - val_loss: 0.2581\n",
      "Epoch 61/80\n",
      " - 2s - loss: 0.2534 - val_loss: 0.2782\n",
      "Epoch 62/80\n",
      " - 2s - loss: 0.2525 - val_loss: 0.2615\n",
      "Epoch 63/80\n",
      " - 2s - loss: 0.2527 - val_loss: 0.2613\n",
      "Epoch 64/80\n",
      " - 2s - loss: 0.2522 - val_loss: 0.2603\n",
      "Epoch 65/80\n",
      " - 2s - loss: 0.2518 - val_loss: 0.2719\n",
      "Epoch 66/80\n",
      " - 2s - loss: 0.2524 - val_loss: 0.2605\n",
      "Epoch 67/80\n",
      " - 2s - loss: 0.2536 - val_loss: 0.2559\n",
      "Epoch 68/80\n",
      " - 2s - loss: 0.2517 - val_loss: 0.2624\n",
      "Epoch 69/80\n",
      " - 2s - loss: 0.2520 - val_loss: 0.2574\n",
      "Epoch 70/80\n",
      " - 3s - loss: 0.2502 - val_loss: 0.2631\n",
      "Epoch 71/80\n",
      " - 3s - loss: 0.2516 - val_loss: 0.2624\n",
      "Epoch 72/80\n",
      " - 2s - loss: 0.2521 - val_loss: 0.2643\n",
      "Epoch 73/80\n",
      " - 2s - loss: 0.2518 - val_loss: 0.2635\n",
      "Epoch 74/80\n",
      " - 2s - loss: 0.2511 - val_loss: 0.2545\n",
      "Epoch 75/80\n",
      " - 2s - loss: 0.2516 - val_loss: 0.2526\n",
      "Epoch 76/80\n",
      " - 2s - loss: 0.2511 - val_loss: 0.2621\n",
      "Epoch 77/80\n",
      " - 2s - loss: 0.2510 - val_loss: 0.2551\n",
      "Epoch 78/80\n",
      " - 2s - loss: 0.2518 - val_loss: 0.2571\n",
      "Epoch 79/80\n",
      " - 2s - loss: 0.2502 - val_loss: 0.2620\n",
      "Epoch 80/80\n",
      " - 2s - loss: 0.2497 - val_loss: 0.2550\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train,\n",
    "                               batch_size=32,\n",
    "                               epochs=80, verbose=2,\n",
    "                               validation_split=.20,\n",
    "                            )#validation_data=(X_test,y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 621,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Adam-opt & MSE')"
      ]
     },
     "execution_count": 621,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x18d0cf8f048>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 271,
       "width": 389
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['val_loss'], label='val_loss')\n",
    "plt.plot(history.history['loss'], label='loss')\n",
    "# plt.plot(history.history['val_acc'], label='val_acc')\n",
    "# plt.plot(history.history['acc'], label='acc')\n",
    "plt.ylim([0.22, 0.35])\n",
    "plt.legend()\n",
    "plt.title('Adam-opt & MSE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 622,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1014/1014 [==============================] - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - 0s 345us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.25523141443846964"
      ]
     },
     "execution_count": 622,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 623,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mse score  =  0.25523139403093553\n"
     ]
    }
   ],
   "source": [
    "print('mse score  = ', mse(model.predict(X_test),y_test) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 466,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([154,  54,  14,  32, 277, 383, 123, 272, 508, 551])"
      ]
     },
     "execution_count": 466,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a= np.random.randint(1,len(y_test),10)\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 468,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 3.5 , 14.5 ,  7.5 ,  2.4 ,  8.  ],\n",
       "       [ 3.  , 15.5 ,  7.6 ,  1.65,  9.6 ],\n",
       "       [ 2.5 , 15.5 ,  6.4 ,  1.55,  9.6 ],\n",
       "       [ 3.5 , 15.5 ,  8.  ,  2.2 ,  8.  ],\n",
       "       [ 3.  , 15.5 ,  7.2 ,  1.55,  8.4 ],\n",
       "       [ 3.5 , 15.5 ,  6.8 ,  1.55,  9.6 ],\n",
       "       [ 3.  , 15.5 ,  8.5 ,  1.8 , 10.  ],\n",
       "       [ 4.  , 15.  ,  6.  ,  1.75,  7.6 ],\n",
       "       [ 2.5 , 14.5 ,  7.  ,  2.4 ,  8.4 ],\n",
       "       [ 4.  , 14.5 ,  8.5 ,  1.9 , 10.  ]])"
      ]
     },
     "execution_count": 468,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[a]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 467,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 3.3143342, 14.9358635,  7.576552 ,  1.9447037,  8.548236 ],\n",
       "       [ 2.9594233, 14.9880295,  7.8288393,  1.9011168,  9.482163 ],\n",
       "       [ 3.1460183, 15.228081 ,  7.058155 ,  1.9260627,  9.692702 ],\n",
       "       [ 3.4182646, 15.498358 ,  7.0852118,  1.9343007,  8.853269 ],\n",
       "       [ 3.3239105, 15.008889 ,  7.32713  ,  1.9090928,  8.689564 ],\n",
       "       [ 3.0878284, 15.083664 ,  7.5174565,  1.9432529,  9.540515 ],\n",
       "       [ 3.3618717, 15.127085 ,  8.252455 ,  1.8655368,  9.490334 ],\n",
       "       [ 3.4425998, 15.472049 ,  7.1377964,  1.9095007,  8.77194  ],\n",
       "       [ 3.243337 , 15.221086 ,  7.312777 ,  2.0029273,  8.727796 ],\n",
       "       [ 2.924205 , 14.94679  ,  8.190481 ,  1.8111858,  9.131056 ]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 467,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_test)[a]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 412,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.22842622, -0.24478626, -0.17896032, -5.30073547, -0.02171631])"
      ]
     },
     "execution_count": 412,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_test)[0]-y_test[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 728,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 728,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(model, to_file='img/basic-net.png')\n",
    "\n",
    "Image(filename=\"img/basic-net.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 615,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_im, X_test_im, y_train_im, y_test_im = train_test_split(img, \n",
    "                                                np.array(parameters_list)*10**7*[1,1,1,1e-2,1],\n",
    "                                                test_size=0.2,\n",
    "                                                random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 616,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_im = X_train_im.reshape(X_train.shape[0], 62, 111, 1).astype(np.float64)\n",
    "X_test_im = X_test_im.reshape(X_test.shape[0], 62, 111, 1).astype(np.float64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 617,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4056, 62, 111, 1)"
      ]
     },
     "execution_count": 617,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_im.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 720,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_image = Input(shape=(62, 111, 1))\n",
    "x = Conv2D(16, 5, strides=1, activation='relu')(input_image)\n",
    "x = MaxPool2D(2, strides=2)(x)\n",
    "x = Conv2D(32, 4, strides=2, activation='relu')(input_image)\n",
    "x = MaxPool2D(2, strides=2)(x)\n",
    "x = Conv2D(64, 3, strides=2, activation='relu')(x)\n",
    "x = MaxPool2D(2, strides=2)(x)\n",
    "\n",
    "x = Flatten()(x)\n",
    "\n",
    "x = Dense(100, activation='relu')(x)\n",
    "x = Dense(5, activation='linear')(x)\n",
    "convnet = Model(inputs=input_image, outputs=x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 721,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_90 (InputLayer)        (None, 62, 111, 1)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_66 (Conv2D)           (None, 30, 54, 32)        544       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_56 (MaxPooling (None, 15, 27, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_67 (Conv2D)           (None, 7, 13, 64)         18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_57 (MaxPooling (None, 3, 6, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_19 (Flatten)         (None, 1152)              0         \n",
      "_________________________________________________________________\n",
      "dense_243 (Dense)            (None, 100)               115300    \n",
      "_________________________________________________________________\n",
      "dense_244 (Dense)            (None, 5)                 505       \n",
      "=================================================================\n",
      "Total params: 134,845\n",
      "Trainable params: 134,845\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "convnet.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 693,
   "metadata": {},
   "outputs": [],
   "source": [
    "convnet.compile(optimizer='adam',\n",
    "                loss='mean_squared_error'\n",
    "                )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 694,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3244 samples, validate on 812 samples\n",
      "Epoch 1/100\n",
      " - 12s - loss: 11.3847 - val_loss: 1.2981\n",
      "Epoch 2/100\n",
      " - 8s - loss: 0.9760 - val_loss: 0.8620\n",
      "Epoch 3/100\n",
      " - 8s - loss: 0.6764 - val_loss: 0.6610\n",
      "Epoch 4/100\n",
      " - 8s - loss: 0.5476 - val_loss: 0.5631\n",
      "Epoch 5/100\n",
      " - 8s - loss: 0.4679 - val_loss: 0.5158\n",
      "Epoch 6/100\n",
      " - 8s - loss: 0.4233 - val_loss: 0.4626\n",
      "Epoch 7/100\n",
      " - 8s - loss: 0.3811 - val_loss: 0.4651\n",
      "Epoch 8/100\n",
      " - 8s - loss: 0.3600 - val_loss: 0.4291\n",
      "Epoch 9/100\n",
      " - 8s - loss: 0.3387 - val_loss: 0.4065\n",
      "Epoch 10/100\n",
      " - 8s - loss: 0.3193 - val_loss: 0.4005\n",
      "Epoch 11/100\n",
      " - 8s - loss: 0.3066 - val_loss: 0.3754\n",
      "Epoch 12/100\n",
      " - 8s - loss: 0.2943 - val_loss: 0.3865\n",
      "Epoch 13/100\n",
      " - 8s - loss: 0.2869 - val_loss: 0.3661\n",
      "Epoch 14/100\n",
      " - 8s - loss: 0.2815 - val_loss: 0.3739\n",
      "Epoch 15/100\n",
      " - 8s - loss: 0.2782 - val_loss: 0.3846\n",
      "Epoch 16/100\n",
      " - 8s - loss: 0.2680 - val_loss: 0.3603\n",
      "Epoch 17/100\n",
      " - 8s - loss: 0.2673 - val_loss: 0.3556\n",
      "Epoch 18/100\n",
      " - 8s - loss: 0.2654 - val_loss: 0.3600\n",
      "Epoch 19/100\n",
      " - 8s - loss: 0.2502 - val_loss: 0.3452\n",
      "Epoch 20/100\n",
      " - 8s - loss: 0.2508 - val_loss: 0.3405\n",
      "Epoch 21/100\n",
      " - 8s - loss: 0.2450 - val_loss: 0.3492\n",
      "Epoch 22/100\n",
      " - 8s - loss: 0.2485 - val_loss: 0.3488\n",
      "Epoch 23/100\n",
      " - 8s - loss: 0.2365 - val_loss: 0.3356\n",
      "Epoch 24/100\n",
      " - 8s - loss: 0.2379 - val_loss: 0.3382\n",
      "Epoch 25/100\n",
      " - 8s - loss: 0.2352 - val_loss: 0.3306\n",
      "Epoch 26/100\n",
      " - 8s - loss: 0.2395 - val_loss: 0.3453\n",
      "Epoch 27/100\n",
      " - 8s - loss: 0.2272 - val_loss: 0.3346\n",
      "Epoch 28/100\n",
      " - 8s - loss: 0.2266 - val_loss: 0.3475\n",
      "Epoch 29/100\n",
      " - 8s - loss: 0.2262 - val_loss: 0.3381\n",
      "Epoch 30/100\n",
      " - 8s - loss: 0.2187 - val_loss: 0.3341\n",
      "Epoch 31/100\n",
      " - 8s - loss: 0.2270 - val_loss: 0.3303\n",
      "Epoch 32/100\n",
      " - 8s - loss: 0.2250 - val_loss: 0.3384\n",
      "Epoch 33/100\n",
      " - 8s - loss: 0.2242 - val_loss: 0.3491\n",
      "Epoch 34/100\n",
      " - 8s - loss: 0.2131 - val_loss: 0.3589\n",
      "Epoch 35/100\n",
      " - 8s - loss: 0.2110 - val_loss: 0.3339\n",
      "Epoch 36/100\n",
      " - 8s - loss: 0.2160 - val_loss: 0.3328\n",
      "Epoch 37/100\n",
      " - 8s - loss: 0.2097 - val_loss: 0.3698\n",
      "Epoch 38/100\n",
      " - 8s - loss: 0.2144 - val_loss: 0.3698\n",
      "Epoch 39/100\n",
      " - 8s - loss: 0.2133 - val_loss: 0.3273\n",
      "Epoch 40/100\n",
      " - 8s - loss: 0.2082 - val_loss: 0.3283\n",
      "Epoch 41/100\n",
      " - 8s - loss: 0.2287 - val_loss: 0.3874\n",
      "Epoch 42/100\n",
      " - 8s - loss: 0.2039 - val_loss: 0.3333\n",
      "Epoch 43/100\n",
      " - 8s - loss: 0.2087 - val_loss: 0.4095\n",
      "Epoch 44/100\n",
      " - 8s - loss: 0.2014 - val_loss: 0.3301\n",
      "Epoch 45/100\n",
      " - 8s - loss: 0.1943 - val_loss: 0.3357\n",
      "Epoch 46/100\n",
      " - 8s - loss: 0.2153 - val_loss: 0.3390\n",
      "Epoch 47/100\n",
      " - 8s - loss: 0.2006 - val_loss: 0.3321\n",
      "Epoch 48/100\n",
      " - 8s - loss: 0.1968 - val_loss: 0.3677\n",
      "Epoch 49/100\n",
      " - 8s - loss: 0.1909 - val_loss: 0.3564\n",
      "Epoch 50/100\n",
      " - 8s - loss: 0.1907 - val_loss: 0.3507\n",
      "Epoch 51/100\n",
      " - 8s - loss: 0.1976 - val_loss: 0.3497\n",
      "Epoch 52/100\n",
      " - 8s - loss: 0.1888 - val_loss: 0.3345\n",
      "Epoch 53/100\n",
      " - 8s - loss: 0.2067 - val_loss: 0.3413\n",
      "Epoch 54/100\n",
      " - 8s - loss: 0.1923 - val_loss: 0.3484\n",
      "Epoch 55/100\n",
      " - 8s - loss: 0.1935 - val_loss: 0.3275\n",
      "Epoch 56/100\n",
      " - 8s - loss: 0.1874 - val_loss: 0.3820\n",
      "Epoch 57/100\n",
      " - 8s - loss: 0.1904 - val_loss: 0.3361\n",
      "Epoch 58/100\n",
      " - 8s - loss: 0.1816 - val_loss: 0.3437\n",
      "Epoch 59/100\n",
      " - 8s - loss: 0.1807 - val_loss: 0.3601\n",
      "Epoch 60/100\n",
      " - 8s - loss: 0.1918 - val_loss: 0.3563\n",
      "Epoch 61/100\n",
      " - 8s - loss: 0.1841 - val_loss: 0.3500\n",
      "Epoch 62/100\n",
      " - 8s - loss: 0.1810 - val_loss: 0.3627\n",
      "Epoch 63/100\n",
      " - 8s - loss: 0.1785 - val_loss: 0.3461\n",
      "Epoch 64/100\n",
      " - 8s - loss: 0.1872 - val_loss: 0.3280\n",
      "Epoch 65/100\n",
      " - 8s - loss: 0.1773 - val_loss: 0.3456\n",
      "Epoch 66/100\n",
      " - 8s - loss: 0.1760 - val_loss: 0.3469\n",
      "Epoch 67/100\n",
      " - 8s - loss: 0.1777 - val_loss: 0.3392\n",
      "Epoch 68/100\n",
      " - 8s - loss: 0.1795 - val_loss: 0.3517\n",
      "Epoch 69/100\n",
      " - 8s - loss: 0.1740 - val_loss: 0.4189\n",
      "Epoch 70/100\n",
      " - 8s - loss: 0.1866 - val_loss: 0.3502\n",
      "Epoch 71/100\n",
      " - 8s - loss: 0.1665 - val_loss: 0.3505\n",
      "Epoch 72/100\n",
      " - 8s - loss: 0.1735 - val_loss: 0.3538\n",
      "Epoch 73/100\n",
      " - 8s - loss: 0.1692 - val_loss: 0.3435\n",
      "Epoch 74/100\n",
      " - 8s - loss: 0.1667 - val_loss: 0.3432\n",
      "Epoch 75/100\n",
      " - 8s - loss: 0.1610 - val_loss: 0.3781\n",
      "Epoch 76/100\n",
      " - 8s - loss: 0.1684 - val_loss: 0.3641\n",
      "Epoch 77/100\n",
      " - 8s - loss: 0.1600 - val_loss: 0.3735\n",
      "Epoch 78/100\n",
      " - 8s - loss: 0.1635 - val_loss: 0.3708\n",
      "Epoch 79/100\n",
      " - 8s - loss: 0.1649 - val_loss: 0.3452\n",
      "Epoch 80/100\n",
      " - 8s - loss: 0.1637 - val_loss: 0.3744\n",
      "Epoch 81/100\n",
      " - 8s - loss: 0.1617 - val_loss: 0.3566\n",
      "Epoch 82/100\n",
      " - 8s - loss: 0.1536 - val_loss: 0.3468\n",
      "Epoch 83/100\n",
      " - 8s - loss: 0.1546 - val_loss: 0.3560\n",
      "Epoch 84/100\n",
      " - 8s - loss: 0.1502 - val_loss: 0.3682\n",
      "Epoch 85/100\n",
      " - 8s - loss: 0.1532 - val_loss: 0.3622\n",
      "Epoch 86/100\n",
      " - 8s - loss: 0.1482 - val_loss: 0.3745\n",
      "Epoch 87/100\n",
      " - 8s - loss: 0.1539 - val_loss: 0.3614\n",
      "Epoch 88/100\n",
      " - 8s - loss: 0.1484 - val_loss: 0.3631\n",
      "Epoch 89/100\n",
      " - 8s - loss: 0.1482 - val_loss: 0.3786\n",
      "Epoch 90/100\n",
      " - 8s - loss: 0.1425 - val_loss: 0.3638\n",
      "Epoch 91/100\n",
      " - 8s - loss: 0.1388 - val_loss: 0.3720\n",
      "Epoch 92/100\n",
      " - 8s - loss: 0.1435 - val_loss: 0.3615\n",
      "Epoch 93/100\n",
      " - 8s - loss: 0.1333 - val_loss: 0.3596\n",
      "Epoch 94/100\n",
      " - 8s - loss: 0.1407 - val_loss: 0.3820\n",
      "Epoch 95/100\n",
      " - 8s - loss: 0.1360 - val_loss: 0.4044\n",
      "Epoch 96/100\n",
      " - 8s - loss: 0.1386 - val_loss: 0.3914\n",
      "Epoch 97/100\n",
      " - 8s - loss: 0.1495 - val_loss: 0.3772\n",
      "Epoch 98/100\n",
      " - 8s - loss: 0.1377 - val_loss: 0.3753\n",
      "Epoch 99/100\n",
      " - 8s - loss: 0.1410 - val_loss: 0.3823\n",
      "Epoch 100/100\n",
      " - 8s - loss: 0.1403 - val_loss: 0.3778\n"
     ]
    }
   ],
   "source": [
    "\n",
    "history_callback = convnet.fit(X_train_im, y_train_im,\n",
    "                               batch_size=32,\n",
    "                               epochs=50, verbose=2,\n",
    "                               validation_data=(X_test_im, y_test_im))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 696,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Adam-opt & MSE')"
      ]
     },
     "execution_count": 696,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x18cf550ad68>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 271,
       "width": 380
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history_callback.history['val_loss'], label='val_loss')\n",
    "plt.plot(history_callback.history['loss'], label='loss')\n",
    "# plt.plot(history.history['val_acc'], label='val_acc')\n",
    "# plt.plot(history.history['acc'], label='acc')\n",
    "plt.ylim([0.22, 0.79])\n",
    "plt.legend()\n",
    "plt.title('Adam-opt & MSE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 697,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1014/1014 [==============================] - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - 1s 1ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.37259184003346535"
      ]
     },
     "execution_count": 697,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "convnet.evaluate(X_test_im,y_test_im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 626,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mse score  =  0.30836219374642665\n"
     ]
    }
   ],
   "source": [
    "print('mse score  = ', mse(convnet.predict(X_test_im),y_test_im) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 727,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 727,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(convnet, to_file='img/conv-net.png')\n",
    "\n",
    "Image(filename=\"img/conv-net.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's try to combine both types of NN. One with vector input and convolutional with image input."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 634,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers.merge import concatenate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 707,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs1 = Input(shape=(50,))\n",
    "x = BatchNormalization()(inputs1)\n",
    "x = Dense(500,activation='relu')(x)\n",
    "# x = Dense(50,activation='relu')(x)\n",
    "# x = Dropout(0.05, noise_shape=None, seed=None)(x)\n",
    "# x = Dense(22,activation='relu')(inputs)\n",
    "\n",
    "x = BatchNormalization()(x)\n",
    "x = Dense(20,activation='relu')(x)\n",
    "# x = Dropout(0.05, noise_shape=None, seed=None)(x)\n",
    "\n",
    "# x = BatchNormalization()(x)\n",
    "out1 = Dense(5,activation='linear')(x)\n",
    "\n",
    "inputs2 = Input(shape=(62, 111, 1))\n",
    "x = Conv2D(32, 4, strides=2, activation='relu')(inputs2)\n",
    "x = MaxPool2D(2, strides=2)(x)\n",
    "x = Conv2D(64, 2, strides=1, activation='relu')(x)\n",
    "x = MaxPool2D(2, strides=2)(x)\n",
    "\n",
    "out2 = Flatten()(x)\n",
    "\n",
    "# merge input models\n",
    "merge = concatenate([out1, out2])\n",
    "# interpretation model\n",
    "x = Dense(10, activation='relu')(merge)\n",
    "x = Dense(10, activation='relu')(x)\n",
    "output = Dense(5, activation='linear')(x)\n",
    "combined = Model(inputs=[inputs1, inputs2], outputs=output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 708,
   "metadata": {},
   "outputs": [],
   "source": [
    "combined.compile(optimizer='adam',\n",
    "                loss='mean_squared_error'\n",
    "                )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 709,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4056 samples, validate on 1014 samples\n",
      "Epoch 1/60\n",
      " - 16s - loss: 19.4097 - val_loss: 1.9594\n",
      "Epoch 2/60\n",
      " - 11s - loss: 0.8243 - val_loss: 0.4763\n",
      "Epoch 3/60\n",
      " - 15s - loss: 0.3806 - val_loss: 0.3617\n",
      "Epoch 4/60\n",
      " - 16s - loss: 0.3697 - val_loss: 0.3568\n",
      "Epoch 5/60\n",
      " - 12s - loss: 0.3563 - val_loss: 0.3640\n",
      "Epoch 6/60\n",
      " - 13s - loss: 0.3238 - val_loss: 0.3360\n",
      "Epoch 7/60\n",
      " - 12s - loss: 0.3182 - val_loss: 0.3650\n",
      "Epoch 8/60\n",
      " - 13s - loss: 0.3281 - val_loss: 0.3359\n",
      "Epoch 9/60\n",
      " - 14s - loss: 0.3305 - val_loss: 0.3421\n",
      "Epoch 10/60\n",
      " - 12s - loss: 0.3222 - val_loss: 0.3647\n",
      "Epoch 11/60\n",
      " - 14s - loss: 0.3258 - val_loss: 0.3336\n",
      "Epoch 12/60\n",
      " - 16s - loss: 0.3108 - val_loss: 0.3320\n",
      "Epoch 13/60\n",
      " - 15s - loss: 0.3062 - val_loss: 0.3223\n",
      "Epoch 14/60\n",
      " - 17s - loss: 0.3100 - val_loss: 0.3299\n",
      "Epoch 15/60\n",
      " - 16s - loss: 0.3013 - val_loss: 0.3174\n",
      "Epoch 16/60\n",
      " - 12s - loss: 0.3091 - val_loss: 0.3394\n",
      "Epoch 17/60\n",
      " - 12s - loss: 0.3103 - val_loss: 0.4122\n",
      "Epoch 18/60\n",
      " - 12s - loss: 0.2998 - val_loss: 0.3503\n",
      "Epoch 19/60\n",
      " - 12s - loss: 0.2963 - val_loss: 0.3186\n",
      "Epoch 20/60\n",
      " - 12s - loss: 0.3037 - val_loss: 0.3147\n",
      "Epoch 21/60\n",
      " - 12s - loss: 0.2952 - val_loss: 0.3327\n",
      "Epoch 22/60\n",
      " - 12s - loss: 0.3090 - val_loss: 0.3276\n",
      "Epoch 23/60\n",
      " - 12s - loss: 0.3025 - val_loss: 0.3163\n",
      "Epoch 24/60\n",
      " - 12s - loss: 0.3145 - val_loss: 0.3302\n",
      "Epoch 25/60\n",
      " - 12s - loss: 0.2999 - val_loss: 0.3379\n",
      "Epoch 26/60\n",
      " - 12s - loss: 0.2945 - val_loss: 0.3215\n",
      "Epoch 27/60\n",
      " - 12s - loss: 0.2967 - val_loss: 0.3119\n",
      "Epoch 28/60\n",
      " - 12s - loss: 0.2950 - val_loss: 0.3147\n",
      "Epoch 29/60\n",
      " - 12s - loss: 0.2985 - val_loss: 0.3098\n",
      "Epoch 30/60\n",
      " - 12s - loss: 0.2953 - val_loss: 0.3173\n",
      "Epoch 31/60\n",
      " - 12s - loss: 0.3050 - val_loss: 0.3572\n",
      "Epoch 32/60\n",
      " - 12s - loss: 0.3028 - val_loss: 0.3474\n",
      "Epoch 33/60\n",
      " - 12s - loss: 0.2966 - val_loss: 0.3195\n",
      "Epoch 34/60\n",
      " - 12s - loss: 0.2958 - val_loss: 0.3517\n",
      "Epoch 35/60\n",
      " - 12s - loss: 0.2956 - val_loss: 0.3917\n",
      "Epoch 36/60\n",
      " - 12s - loss: 0.2999 - val_loss: 0.3391\n",
      "Epoch 37/60\n",
      " - 12s - loss: 0.2924 - val_loss: 0.3310\n",
      "Epoch 38/60\n",
      " - 12s - loss: 0.3032 - val_loss: 0.3247\n",
      "Epoch 39/60\n",
      " - 12s - loss: 0.2938 - val_loss: 0.3249\n",
      "Epoch 40/60\n",
      " - 12s - loss: 0.2945 - val_loss: 0.3189\n",
      "Epoch 41/60\n",
      " - 12s - loss: 0.3114 - val_loss: 0.3223\n",
      "Epoch 42/60\n",
      " - 12s - loss: 0.2931 - val_loss: 0.3180\n",
      "Epoch 43/60\n",
      " - 12s - loss: 0.2958 - val_loss: 0.3139\n",
      "Epoch 44/60\n",
      " - 12s - loss: 0.2962 - val_loss: 0.3722\n",
      "Epoch 45/60\n",
      " - 12s - loss: 0.2959 - val_loss: 0.3181\n",
      "Epoch 46/60\n",
      " - 12s - loss: 0.2932 - val_loss: 0.3144\n",
      "Epoch 47/60\n",
      " - 12s - loss: 0.2947 - val_loss: 0.3355\n",
      "Epoch 48/60\n",
      " - 12s - loss: 0.2904 - val_loss: 0.3300\n",
      "Epoch 49/60\n",
      " - 12s - loss: 0.2984 - val_loss: 0.3426\n",
      "Epoch 50/60\n",
      " - 12s - loss: 0.2883 - val_loss: 0.3239\n",
      "Epoch 51/60\n",
      " - 12s - loss: 0.3031 - val_loss: 0.3947\n",
      "Epoch 52/60\n",
      " - 12s - loss: 0.2960 - val_loss: 0.3155\n",
      "Epoch 53/60\n",
      " - 12s - loss: 0.2868 - val_loss: 0.3319\n",
      "Epoch 54/60\n",
      " - 12s - loss: 0.2893 - val_loss: 0.3770\n",
      "Epoch 55/60\n",
      " - 12s - loss: 0.3027 - val_loss: 0.3154\n",
      "Epoch 56/60\n",
      " - 12s - loss: 0.2926 - val_loss: 0.3144\n",
      "Epoch 57/60\n",
      " - 12s - loss: 0.2857 - val_loss: 0.3220\n",
      "Epoch 58/60\n",
      " - 12s - loss: 0.2948 - val_loss: 0.3284\n",
      "Epoch 59/60\n",
      " - 12s - loss: 0.2878 - val_loss: 0.3241\n",
      "Epoch 60/60\n",
      " - 12s - loss: 0.2996 - val_loss: 0.3562\n"
     ]
    }
   ],
   "source": [
    "history_combined = combined.fit( [X_train, X_train_im], y_train_im,\n",
    "                               batch_size=32,\n",
    "                               epochs=60, verbose=2,\n",
    "                               validation_data=([X_test, X_test_im], y_test_im))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 710,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Adam-opt & MSE')"
      ]
     },
     "execution_count": 710,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x18d143192e8>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 271,
       "width": 380
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history_combined.history['val_loss'], label='val_loss')\n",
    "plt.plot(history_combined.history['loss'], label='loss')\n",
    "# plt.plot(history.history['val_acc'], label='val_acc')\n",
    "# plt.plot(history.history['acc'], label='acc')\n",
    "plt.ylim([0.22, 0.79])\n",
    "plt.legend()\n",
    "plt.title('Adam-opt & MSE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 712,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1014/1014 [==============================] - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - ETA:  - 1s 1ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.35622153091712816"
      ]
     },
     "execution_count": 712,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined.evaluate([X_test,X_test_im],y_test_im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 726,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 726,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(combined, to_file='img/double-net.png')\n",
    "\n",
    "Image(filename=\"img/double-net.png\")"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}