{
  "cells": [
    {
      "metadata": {
        "_cell_guid": "21830c4b-3f47-4a7f-9697-95c05419871d",
        "_uuid": "6d170584ce2dcc8ddc7d555c52e00c637810bd42",
        "trusted": true
      },
      "cell_type": "code",
      "source": "import numpy as np\nimport matplotlib.pyplot as plt",
      "execution_count": 59,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "0aa7d1cd-9c1e-494d-a5a1-56e7d8f6ffca",
        "_uuid": "dbfd497e8ca31e01cb32a8e395be7568f2c8b64b",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# Ensure results are reproducable\nfrom numpy.random import seed\nseed(1)\nfrom tensorflow import set_random_seed\nset_random_seed(2)",
      "execution_count": 60,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "27f99be4-d278-4b40-a6cd-6eac99f29198",
        "_uuid": "a197bd7e470a98cf0542005a10db6369affb9e8c",
        "trusted": true
      },
      "cell_type": "code",
      "source": "def mnist_load_data(path='mnist.npz'):\n    with np.load(path) as f:\n        x_train, y_train = f['x_train'], f['y_train']\n        x_test, y_test = f['x_test'], f['y_test']\n    return (x_train, y_train), (x_test, y_test)\n        \n\n(X_train, y_train), (X_test, y_test) = mnist_load_data(path='../input/mnist.npz')",
      "execution_count": 61,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "65c1a879-f719-4b08-be79-4272be794c20",
        "_uuid": "3fe506f3b81e8d5aeb556f4ca78c7c9d5eb7e01c",
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_train = X_train.astype('float32') / 255.\nX_test = X_test.astype('float32') / 255.",
      "execution_count": 62,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "cd701e9e-952b-4e94-a460-566a01c40903",
        "_uuid": "5609dd8f1788af46098d9c786b1047dcfcc95bdb",
        "trusted": true
      },
      "cell_type": "code",
      "source": "plt.imshow(X_test[0])",
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 63,
          "data": {
            "text/plain": "<matplotlib.image.AxesImage at 0x7f524345c588>"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<Figure size 432x288 with 1 Axes>",
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAADaVJREFUeJzt3X+MXOV1xvHnib1e4jU0GILrGgcnhKA6NDjVxiSCVo4IKZAgEyWhWKrlSpRFLUhQRW2Rq6iWWqUUhSC3SSM5wY1BBGgCCCtx01CrrYVKHS/I2IBpTajT2DVewLQ2AfwDn/6x19EGdt5d5ted9fl+pNXO3HPv3KPrfXzvzDszryNCAPJ5R90NAKgH4QeSIvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kNT0bu5shvvjJA10c5dAKq/rZzochzyZdVsKv+1LJa2WNE3SNyPiltL6J2lAF/jiVnYJoGBzbJz0uk1f9tueJulrki6TtFDSMtsLm308AN3VynP+xZKejYjnIuKwpHslLW1PWwA6rZXwz5P00zH3d1fLfoHtIdvDtoeP6FALuwPQTh1/tT8i1kTEYEQM9qm/07sDMEmthH+PpPlj7p9ZLQMwBbQS/i2SzrH9XtszJF0taX172gLQaU0P9UXEUds3SPpHjQ71rY2Ip9rWGYCOammcPyI2SNrQpl4AdBFv7wWSIvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU4QeSIvxAUoQfSIrwA0kRfiCplmbptb1L0kFJb0g6GhGD7WgKQOe1FP7KxyPixTY8DoAu4rIfSKrV8IekH9p+zPZQOxoC0B2tXvZfFBF7bJ8h6WHbz0TEprErVP8pDEnSSZrZ4u4AtEtLZ/6I2FP9HpH0oKTF46yzJiIGI2KwT/2t7A5AGzUdftsDtk8+flvSJyU92a7GAHRWK5f9cyQ9aPv443w7In7Qlq4AdFzT4Y+I5ySd38ZeAHQRQ31AUoQfSIrwA0kRfiApwg8kRfiBpNrxqb4UXrr2Yw1r71n+bHHbZ0bmFOuHD/UV6/PuKddn7n6lYe3Y1qeL2yIvzvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBTj/JP0x3/07Ya1zw68XN747BZ3vqRc3nX01Ya11S98vMWdT10/GjmrYW3gtl8qbjt942PtbqfncOYHkiL8QFKEH0iK8ANJEX4gKcIPJEX4gaQcEV3b2SmeHRf44q7tr51+9rkLGtZe/FD5/9BTd5SP8cu/6mJ9xof+t1i/9bwHGtYueedrxW2//+qsYv1TMxt/V0CrXovDxfrmQwPF+pKTjjS97/d//7pi/QNDW5p+7Dptjo06EPvLf1AVzvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kNSEn+e3vVbSpyWNRMR51bLZku6TtEDSLklXRcQEH2qf2ga+u7lQa+2xT2ltc/3NLy9pWPuLCxeU9/2v5TkHbl3y/iY6mpzprx0r1ge27S3WT9t0f7H+azMaz3cwc1d5LoQMJnPm/5akS9+07GZJGyPiHEkbq/sAppAJwx8RmyTtf9PipZLWVbfXSbqyzX0B6LBmn/PPiYjj12TPSyrPRwWg57T8gl+Mfjig4ZvXbQ/ZHrY9fESHWt0dgDZpNvz7bM+VpOr3SKMVI2JNRAxGxGCf+pvcHYB2azb86yWtqG6vkPRQe9oB0C0Tht/2PZIelXSu7d22r5F0i6RLbO+U9InqPoApZMJx/ohY1qA0NT+YfwI6+vy+hrWB+xvXJOmNCR574LsvNdFRe+z7vY8V6x+cUf7z/fL+cxvWFvzdc8VtjxarJwbe4QckRfiBpAg/kBThB5Ii/EBShB9Iiim6UZvpZ80v1r+68qvFep+nFevfWf2JhrXT9j5a3DYDzvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBTj/KjNM384r1j/SH95pumnDpenH5/99Ktvu6dMOPMDSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKM86OjDn3qIw1rj3/u9gm2Ls/w9Ps33lisv/PffjTB4+fGmR9IivADSRF+ICnCDyRF+IGkCD+QFOEHkppwnN/2WkmfljQSEedVy1ZJulbSC9VqKyNiQ6eaxNT135c1Pr/Mcnkcf9l/XVKsz/zBE8V6FKuYzJn/W5IuHWf57RGxqPoh+MAUM2H4I2KTpP1d6AVAF7XynP8G29tsr7V9ats6AtAVzYb/65LOlrRI0l5JtzVa0faQ7WHbw0d0qMndAWi3psIfEfsi4o2IOCbpG5IWF9ZdExGDETHYN8EHNQB0T1Phtz13zN3PSHqyPe0A6JbJDPXdI2mJpNNt75b0Z5KW2F6k0dGUXZKu62CPADpgwvBHxLJxFt/RgV4wBb3j5JOL9eW/8UjD2oFjrxe3HfnS+4r1/kNbinWU8Q4/ICnCDyRF+IGkCD+QFOEHkiL8QFJ8dTdasnPVB4v1753+tw1rS3d+trht/waG8jqJMz+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJMU4P4r+73c+Wqxv++2/LtZ/fPRIw9orf3Vmcdt+7S3W0RrO/EBShB9IivADSRF+ICnCDyRF+IGkCD+QFOP8yU2f9yvF+k1fvK9Y73f5T+jqJ5Y3rL37H/i8fp048wNJEX4gKcIPJEX4gaQIP5AU4QeSIvxAUhOO89ueL+lOSXMkhaQ1EbHa9mxJ90laIGmXpKsi4uXOtYpmeHr5n/j87+0u1j8/66Vi/e6DZxTrc77Y+PxyrLglOm0yZ/6jkr4QEQslfVTS9bYXSrpZ0saIOEfSxuo+gCliwvBHxN6IeLy6fVDSDknzJC2VtK5abZ2kKzvVJID2e1vP+W0vkPRhSZslzYmI49+z9LxGnxYAmCImHX7bsyTdL+mmiDgwthYRodHXA8bbbsj2sO3hIzrUUrMA2mdS4bfdp9Hg3x0RD1SL99meW9XnShoZb9uIWBMRgxEx2Kf+dvQMoA0mDL9tS7pD0o6I+MqY0npJK6rbKyQ91P72AHTKZD7Se6Gk5ZK2295aLVsp6RZJf2/7Gkk/kXRVZ1pES84/t1j+8zPuaunhv/alzxfr73ri0ZYeH50zYfgj4hFJblC+uL3tAOgW3uEHJEX4gaQIP5AU4QeSIvxAUoQfSIqv7j4BTFv4gYa1oXtbe+/VwrXXF+sL7vr3lh4f9eHMDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJMc5/AnjmD05tWLti5oGGtck4818Ol1eIcb+9DVMAZ34gKcIPJEX4gaQIP5AU4QeSIvxAUoQfSIpx/ing9SsWF+sbr7itUJ3Z3mZwwuDMDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJTTjOb3u+pDslzZEUktZExGrbqyRdK+mFatWVEbGhU41m9j8XTivW3zO9+bH8uw+eUaz3HSh/np9P809dk3mTz1FJX4iIx22fLOkx2w9Xtdsj4sudaw9Ap0wY/ojYK2lvdfug7R2S5nW6MQCd9bae89teIOnDkjZXi26wvc32WtvjfpeU7SHbw7aHj+hQS80CaJ9Jh9/2LEn3S7opIg5I+rqksyUt0uiVwbhvMI+INRExGBGDfepvQ8sA2mFS4bfdp9Hg3x0RD0hSROyLiDci4pikb0gqf/oEQE+ZMPy2LekOSTsi4itjls8ds9pnJD3Z/vYAdMpkXu2/UNJySdttb62WrZS0zPYijY727JJ0XUc6REv+8qWFxfqjv7WgWI+929vYDXrJZF7tf0SSxykxpg9MYbzDD0iK8ANJEX4gKcIPJEX4gaQIP5CUo4tTLJ/i2XGBL+7a/oBsNsdGHYj94w3NvwVnfiApwg8kRfiBpAg/kBThB5Ii/EBShB9Iqqvj/LZfkPSTMYtOl/Ri1xp4e3q1t17tS6K3ZrWzt7Mi4t2TWbGr4X/Lzu3hiBisrYGCXu2tV/uS6K1ZdfXGZT+QFOEHkqo7/Gtq3n9Jr/bWq31J9NasWnqr9Tk/gPrUfeYHUJNawm/7Utv/YftZ2zfX0UMjtnfZ3m57q+3hmntZa3vE9pNjls22/bDtndXvcadJq6m3Vbb3VMduq+3La+ptvu1/tv207ads31gtr/XYFfqq5bh1/bLf9jRJ/ynpEkm7JW2RtCwinu5qIw3Y3iVpMCJqHxO2/ZuSXpF0Z0ScVy27VdL+iLil+o/z1Ij4kx7pbZWkV+qeubmaUGbu2JmlJV0p6XdV47Er9HWVajhudZz5F0t6NiKei4jDku6VtLSGPnpeRGyStP9Ni5dKWlfdXqfRP56ua9BbT4iIvRHxeHX7oKTjM0vXeuwKfdWijvDPk/TTMfd3q7em/A5JP7T9mO2hupsZx5xq2nRJel7SnDqbGceEMzd305tmlu6ZY9fMjNftxgt+b3VRRPy6pMskXV9d3vakGH3O1kvDNZOaublbxplZ+ufqPHbNznjdbnWEf4+k+WPun1kt6wkRsaf6PSLpQfXe7MP7jk+SWv0eqbmfn+ulmZvHm1laPXDsemnG6zrCv0XSObbfa3uGpKslra+hj7ewPVC9ECPbA5I+qd6bfXi9pBXV7RWSHqqxl1/QKzM3N5pZWjUfu56b8Toiuv4j6XKNvuL/Y0l/WkcPDfp6n6Qnqp+n6u5N0j0avQw8otHXRq6RdJqkjZJ2SvonSbN7qLe7JG2XtE2jQZtbU28XafSSfpukrdXP5XUfu0JftRw33uEHJMULfkBShB9IivADSRF+ICnCDyRF+IGkCD+QFOEHkvp/uK0ZUt56JeQAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "9dfd1c24-c1fd-40b2-ad68-67a6bf031d7c",
        "_uuid": "2340ec54c0c740174baf46db68b090a4915ae023",
        "trusted": true
      },
      "cell_type": "code",
      "source": "from keras.layers import Input, Dense\nfrom keras.models import Model\n\n# this is the size of our encoded representations\nencoding_dim = 32  # 32 floats -> compression of factor 24.5, assuming the input is 784 floats\n\n# this is our input placeholder\ninput_img = Input(shape=(784,))\n# \"encoded\" is the encoded representation of the input\nencoded = Dense(encoding_dim, activation='relu')(input_img)\n# \"decoded\" is the lossy reconstruction of the input\ndecoded = Dense(784, activation='sigmoid')(encoded)\n\n# this model maps an input to its reconstruction\nautoencoder = Model(input_img, decoded)",
      "execution_count": 64,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "802f400e8bdf0aba00bc3a5a97ccf32b0b23a8b8"
      },
      "cell_type": "code",
      "source": "from keras.utils import plot_model\nplot_model(autoencoder, to_file='model.png', show_shapes=True)\nplt.figure(figsize=(10,10))\nplt.imshow(plt.imread('model.png'))",
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 65,
          "data": {
            "text/plain": "<matplotlib.image.AxesImage at 0x7f5243396ef0>"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<Figure size 720x720 with 1 Axes>",
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "28b28eb7-ab0a-47d0-b802-8d21e31e464f",
        "_uuid": "bea5dced59e6a74ec2ec205bb5f10510ec1e1472",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# this model maps an input to its encoded representation\nencoder = Model(input_img, encoded)",
      "execution_count": 66,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "7ec50d66-e7cb-437f-b6bb-c727113ba2a7",
        "_uuid": "b011eba420a8ebef932f3dfce3e95b58cd37788d",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# create a placeholder for an encoded (32-dimensional) input\nencoded_input = Input(shape=(encoding_dim,))\n# retrieve the last layer of the autoencoder model\ndecoder_layer = autoencoder.layers[-1]\n# create the decoder model\ndecoder = Model(encoded_input, decoder_layer(encoded_input))",
      "execution_count": 67,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "72018a5e-ba7c-4204-a0da-4e74816635d5",
        "_uuid": "4ddd442c4abd7504aeb7d3b18479342751458309",
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_train_flat = X_train.reshape((len(X_train), np.prod(X_train.shape[1:])))\nX_test_flat = X_test.reshape((len(X_test), np.prod(X_test.shape[1:])))\nprint(X_train_flat.shape)\nprint(X_test_flat.shape)",
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "stream",
          "text": "(60000, 784)\n(10000, 784)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "d204cb0a-270f-4703-8d0e-580164e4b015",
        "_uuid": "aabe441f9ded49bb70a32c9cdf2d3682e20370d3",
        "trusted": true
      },
      "cell_type": "code",
      "source": "autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')",
      "execution_count": 69,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "26b92141-2632-4221-850a-5ae2055a1928",
        "_uuid": "23f7c371afb9c9ac0db39fd41a383c56cb37da2d",
        "trusted": true
      },
      "cell_type": "code",
      "source": "autoencoder.fit(X_train_flat, X_train_flat,\n                epochs=50,\n                batch_size=256,\n                shuffle=True,\n                validation_data=(X_test_flat, X_test_flat))",
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "stream",
          "text": "Train on 60000 samples, validate on 10000 samples\nEpoch 1/50\n60000/60000 [==============================] - 3s 46us/step - loss: 0.3558 - val_loss: 0.2710\nEpoch 2/50\n60000/60000 [==============================] - 2s 35us/step - loss: 0.2642 - val_loss: 0.2538\nEpoch 3/50\n60000/60000 [==============================] - 2s 33us/step - loss: 0.2441 - val_loss: 0.2320\nEpoch 4/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.2239 - val_loss: 0.2134\nEpoch 5/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.2075 - val_loss: 0.1992\nEpoch 6/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.1955 - val_loss: 0.1893\nEpoch 7/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.1868 - val_loss: 0.1819\nEpoch 8/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.1798 - val_loss: 0.1754\nEpoch 9/50\n60000/60000 [==============================] - 2s 25us/step - loss: 0.1739 - val_loss: 0.1701\nEpoch 10/50\n60000/60000 [==============================] - 1s 24us/step - loss: 0.1689 - val_loss: 0.1655\nEpoch 11/50\n60000/60000 [==============================] - 1s 22us/step - loss: 0.1646 - val_loss: 0.1613\nEpoch 12/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1606 - val_loss: 0.1575\nEpoch 13/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1571 - val_loss: 0.1541\nEpoch 14/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1538 - val_loss: 0.1509\nEpoch 15/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1507 - val_loss: 0.1479\nEpoch 16/50\n60000/60000 [==============================] - 1s 23us/step - loss: 0.1478 - val_loss: 0.1450\nEpoch 17/50\n60000/60000 [==============================] - 1s 25us/step - loss: 0.1450 - val_loss: 0.1422\nEpoch 18/50\n60000/60000 [==============================] - 1s 23us/step - loss: 0.1423 - val_loss: 0.1397\nEpoch 19/50\n60000/60000 [==============================] - 1s 22us/step - loss: 0.1398 - val_loss: 0.1372\nEpoch 20/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1374 - val_loss: 0.1348\nEpoch 21/50\n60000/60000 [==============================] - 1s 20us/step - loss: 0.1351 - val_loss: 0.1326\nEpoch 22/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1330 - val_loss: 0.1304\nEpoch 23/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1309 - val_loss: 0.1284\nEpoch 24/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1290 - val_loss: 0.1266\nEpoch 25/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1271 - val_loss: 0.1247\nEpoch 26/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1254 - val_loss: 0.1231\nEpoch 27/50\n60000/60000 [==============================] - 2s 26us/step - loss: 0.1238 - val_loss: 0.1215\nEpoch 28/50\n60000/60000 [==============================] - 2s 27us/step - loss: 0.1223 - val_loss: 0.1200\nEpoch 29/50\n60000/60000 [==============================] - 2s 28us/step - loss: 0.1208 - val_loss: 0.1185\nEpoch 30/50\n60000/60000 [==============================] - 1s 25us/step - loss: 0.1195 - val_loss: 0.1172\nEpoch 31/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1182 - val_loss: 0.1160\nEpoch 32/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1170 - val_loss: 0.1149\nEpoch 33/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1159 - val_loss: 0.1138\nEpoch 34/50\n60000/60000 [==============================] - 1s 22us/step - loss: 0.1149 - val_loss: 0.1128\nEpoch 35/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1140 - val_loss: 0.1119\nEpoch 36/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1131 - val_loss: 0.1110\nEpoch 37/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1123 - val_loss: 0.1103\nEpoch 38/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1116 - val_loss: 0.1096\nEpoch 39/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1108 - val_loss: 0.1089\nEpoch 40/50\n60000/60000 [==============================] - 1s 23us/step - loss: 0.1102 - val_loss: 0.1082\nEpoch 41/50\n60000/60000 [==============================] - 1s 24us/step - loss: 0.1096 - val_loss: 0.1076\nEpoch 42/50\n60000/60000 [==============================] - 1s 24us/step - loss: 0.1090 - val_loss: 0.1071\nEpoch 43/50\n60000/60000 [==============================] - 1s 23us/step - loss: 0.1085 - val_loss: 0.1066\nEpoch 44/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1080 - val_loss: 0.1061\nEpoch 45/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1075 - val_loss: 0.1057\nEpoch 46/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1071 - val_loss: 0.1052\nEpoch 47/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1066 - val_loss: 0.1048\nEpoch 48/50\n60000/60000 [==============================] - 1s 20us/step - loss: 0.1062 - val_loss: 0.1044\nEpoch 49/50\n60000/60000 [==============================] - 1s 21us/step - loss: 0.1059 - val_loss: 0.1041\nEpoch 50/50\n60000/60000 [==============================] - 1s 20us/step - loss: 0.1055 - val_loss: 0.1037\n",
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "execution_count": 70,
          "data": {
            "text/plain": "<keras.callbacks.History at 0x7f520169f0b8>"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "266da819-0d1e-47ef-b638-8c31de72d0f0",
        "_uuid": "fb7787cf702f90fa6ce428dc6ed78d147bb6d1e3",
        "trusted": true
      },
      "cell_type": "code",
      "source": "original = np.expand_dims(X_test_flat[0],0)\nseven = autoencoder.predict(original)",
      "execution_count": 71,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "8a9c98f7-dc26-4b5b-8560-b69f8d1367e3",
        "_uuid": "c49e70475499522d1b92cbb410116a86b1941733",
        "trusted": true
      },
      "cell_type": "code",
      "source": "seven = seven.reshape(1,28,28)",
      "execution_count": 72,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "a7c617b0-182d-4b4a-8866-ba29d9a77282",
        "_uuid": "7191a54197d25eef483bf5b923be84c562098e00",
        "trusted": true
      },
      "cell_type": "code",
      "source": "original = original.reshape(1,28,28)",
      "execution_count": 73,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "c8a9a80e-3ad0-47fc-a48c-f73e259dcfb8",
        "_uuid": "511f6f026a4c32d15a7dbf03af70b862a21e0018",
        "trusted": true
      },
      "cell_type": "code",
      "source": "fig = plt.figure(figsize=(7, 10))\na=fig.add_subplot(1,2,1)\na.set_title('Original')\nimgplot = plt.imshow(original[0,:,:])\n\nb=fig.add_subplot(1,2,2)\nb.set_title('Autoencoder')\nimgplot = plt.imshow(seven[0,:,:])",
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<Figure size 504x720 with 2 Axes>",
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAawAAADgCAYAAABfJVIPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAGmZJREFUeJzt3XucXGV9x/Hvb69JdpMACYQQwFVIkEhLbCPgrYTiBVTES42hitgqsV76El/UlqZVaWupUhWpeGnUCFVAaIBKLaIQq8irEE2AgFwsATcYyI2QkA1JNtndX/84J+2QPs/MzuzszD57Pu/XK6/d+Z05z3nOZp75zTnzO88xdxcAAGNdS7M7AADAcJCwAABJIGEBAJJAwgIAJIGEBQBIAgkLAJAEElZCzGyJmX2j3s8dRltuZsfWoy0A/4exVR3jOqzmMbP3SrpQ0jGSdki6SdJfuvv2ZvbrQGbmkma7+9pm9wXFYGY/kXSipMPdvX+Y6yT3Ok2xz83EEVaTmNmFkj4r6eOSpko6RdILJN1mZh2B57c1todAc5hZj6RXS3JJb25qZ8aoor4fkLCawMymSPobSX/q7re6+z5375W0UFKPpHeb2cVmttzMvmNmOyS9N499p6Sd95jZOjPbamafMLNeM3tNvux/n2tmPfmph/PM7Akze9rM/qqknZPM7C4z225mG8zsilDSBBrkPZLulnSlpPP2B83sJ2b2/pLH7zWzO/Pf78jDa8xsp5m9M4+fb2ZrzewZM7vZzI4oWf/FZnZbvuxXZrawZNmVZvZlM/sPM+szs5VmdkzJ8peUrLvJzJbk8U4z+6KZPZX/+6KZdZas9/F8jD1lZn9cutP5up/Lx+gmM/uamU3Mly0ws/Vm9hdmtlHSt0b+Z04PCas5XiFpgqQbS4PuvlPSLZJem4fOlrRc0kGSri59rpnNlfQVSe+SNFPZUdqsCtt9laTjJJ0u6ZNmdnweH5T0MUnTJb08X/6hGvYLqIf3KHu9Xy3p9WY2o9IK7v57+a8nunu3u19nZr8v6R+UfRCcKWmdpO9Kkpl1SbpN0jWSDpO0SNJX8nG13yJlHywPlrRW0t/n606WdLukWyUdIelYSSvydf5K2dmSecpOaZ4k6a/z9c6Q9GfKxvdsSa85YDc+I2lOvu6xysbzJ0uWHy7pEGVnYhZX+puMRySs5pgu6Wl3Hwgs25Avl6S73P3f3H3I3Xcf8Lw/kPTv7n6nu+9V9sKu9IXk37j7bndfI2mNsgEld1/t7ne7+0B+pPfPkk6tbdeA2pnZq5S9IV/v7qslPSbpD2ts7l2Slrn7Pfn3YH8p6eX5Kcc3Sep192/lr/t7Jd0g6R0l69/k7j/Px+nVyhKJ8nU3uvvn3X2Pu/e5+8qSbf6tu2929y3KEt65+bKFkr7l7r909+ckXVyy36YsCX3M3Z9x9z5JlyhLmvsNSfqUu/cH3g8KoZDnQceApyVNN7O2QNKamS+XpN+UaeOI0uXuvsvMtlbY7saS33dJ6pYkM5sj6QuS5kuapOx1sbrSTgCj4DxJP3L3/WPgmjx2WQ1tHSHpnv0P3H1nPkZmKUuKJ5tZaYFTm6RvlzwOjhdJRylLpLFtrit5vC6P7V+2+oBl+x2qbOytznKXJMkktZY8Z4u774lstxA4wmqOuyT1S3pbadDMuiWdqf87vVDuiGmDpCNL1p0oaVqN/fmqpEeUVStNkbRE2WABGiZ/DS+UdKqZbcy/q/mYpBPN7ERJzyl7U9/v8ApNPqUsMe1vv0vZGHlS2Ye9n7r7QSX/ut39g8Po6m8kvWg425R0dB6TsjF71AHL9nta0m5JLynpz1R37y55TuFLuklYTeDuzyo7VfAlMzvDzNrz0xTXS1qv53/Ki1ku6Swze0VeIHGxak8yk5WV1e80sxdLGs6gBertLcq+T52r7PTbPEnHS/qZsu+17pP0NjOblF+79L4D1t+k5yeSayX9kZnNywsfLpG0Mj/t/X1Jc8zs3Hz8tZvZy0q+1y3n+5JmmtkFeaHEZDM7uWSbf21mh5rZdGWn6vcXSl2vrHhqrplNkvSp/Q26+5Ckr0u6zMwOkyQzm2Vmrx9GfwqDhNUk7n6psiOZzylLFiuVfXI7fTjXnbj7g5L+VNmXyBsk7ZS0WdmRW7X+TNn3BH3KBs11NbQBjNR5yr7jecLdN+7/J+kKZd8NXSZpr7LEdJUOKERS9qHtqrzadaG73y7pE8q+m9qg7HrHRZKUf0f0uvzxU8pO/31WUqcqyNd9raSz8vUelXRavvjTklZJul/SA8pOSX46X+8Hkr4o6cfKijh+fEDTf5HH784rg29XViSFHBcOjxP56cTtyk7r/brZ/QGAeuMIK2FmdlZ+eqRL2ZHaA5J6m9srABgdJKy0na3sdMZTyq7rWOQcMgMYpzglCABIAkdYAIAkjOjC4XyqkcuVXdz2DXf/TLnnd1inT1DXSDYJNNUePae93j9q16gxplBEfdr2tLsfWul5NScsM2uV9GVl5Z3rJf3CzG5294di60xQl06202vdJNB0K31F5SfViDGForrdl6+r/KyRnRI8SdJad388n8vuu8qKAADUhjEFlDGShDVLz5/rbr0Cs4Wb2WIzW2Vmq/bVdE0rUBiMKaCMUS+6cPel7j7f3ee3V76IHEAFjCkU1UgS1pN6/kSOR+YxALVhTAFljCRh/ULSbDN7YT756iJJN9enW0AhMaaAMmquEnT3ATP7iKQfKivBXZZPyAqgBowpoLwRXYfl7rcou6U7gDpgTAFxzHQBAEgCCQsAkAQSFgAgCSQsAEASSFgAgCSQsAAASSBhAQCSQMICACSBhAUASAIJCwCQBBIWACAJJCwAQBJIWACAJJCwAABJIGEBAJJAwgIAJIGEBQBIAgkLAJAEEhYAIAkkLABAEkhYAIAkkLAAAEkgYQEAktA2kpXNrFdSn6RBSQPuPr8enQKKijGF5JnFl7mPqOkRJazcae7+dB3aAZBhTAEBnBIEACRhpAnLJf3IzFab2eLQE8xssZmtMrNV+9Q/ws0B4x5jCogY6SnBV7n7k2Z2mKTbzOwRd7+j9AnuvlTSUkmaYoeM7AQmMP4xpoCIER1hufuT+c/Nkm6SdFI9OgUUFWMKiKv5CMvMuiS1uHtf/vvrJP1t3XoGFEySYypSEdbS2Vl9Wy3Vf372wcHwgkjch8ockPpQ1duvG4vvu7WE/8bRfR9hJd6IjeL2R3JKcIakmyx7wbZJusbdb61Lr4BiYkwBZdScsNz9cUkn1rEvQKExpoDyKGsHACSBhAUASAIJCwCQhHpMzZS8ree/PBg/+ty10XUe2TwjGN/b3x5dZ9a14WWT1u8MxofueyjaFlB3La3BcOvUKdFVfFZkHMzoCsYHJ8Q/I/dPCW+/fXe8eq9lb7girWPb3mC87dnd0baGJobH576pE4Lxwc74vnhruLJvcEI4vrc73tZQ5C1l2prw+0brI+uibQ3ufC68oFyFZKzqr9ycgdW2NUwcYQEAkkDCAgAkgYQFAEgCCQsAkAQSFgAgCSQsAEASKGuX9OcfvyYYf3vXtvhKx9SwoQXhcO/ArmD88i2n1bCRsennm18QjHd9fmp0nbYVq0erO4Vm7R3BeMuU7mB872/1RNvaNSPc1rPHhD8L75kbLytvbQvf26utLV5y3b87XPM9tC8cb58UL6s+9YWPBeMvm/LrYHyChUvnJemn218cjG/bOzEYf930+CUsE1r2BeOX3HdGMP6ifzwq2pat+VUw7pF5dGtSZiLfkW6IIywAQBJIWACAJJCwAABJIGEBAJJAwgIAJIEqQUn/tGRRMP7J347n84MfDlcbbTs+PiFkx29vD8YvPeHGYPyymSujbf3HrnBF1xsnhSfErMVuD1dBrewPT2wqSQsmhCuaFNmXY9/5gWhbc1bE+4YRiN1yfdZhwfjAxPCktJK0oydSDXhYuLJv6pRwRawkdbSFK8i29U2KrmObO4Px9v7wPu47PNqUtuwJj6ld3eFKyDuenRNt6+7enmD8yOnh94CejqejbZ3QsTUYX/vie4PxO2ecEm1rQrSCr0z1XrWT3A7Vs+Tw+TjCAgAkgYQFAEgCCQsAkAQSFgAgCSQsAEASSFgAgCRULGs3s2WS3iRps7ufkMcOkXSdpB5JvZIWunuZmWLHtq7l4ZLrruXVtzWlhu1/6fAFwfinX9kT385P1wbjly44toYehLXtDpcmd92/IbrOtDtuCMZ/qyM8Gemk3nB8vGvquBoKX5LRsi18SYQdMTnaVNdT4bZa94Q/C/dvmhZtq3VTuK0je+OTzHZsfTYYt13hiXSHDopfkrF+3ouC8Wv2huPT7guXqEvScdvDpejbT54VjK+4aG68rWk/C8Z3DoZL+lv2xicLjl3SUAtrDV/u4AMDddvGgYZzhHWlpAOnBb5I0gp3ny1pRf4YwPBdKcYVUJWKCcvd75D0zAHhsyVdlf9+laS31LlfwLjGuAKqV+tMFzPcff95oY2SZsSeaGaLJS2WpAmKX7EOYHjjijGFohpx0YW7u6ToXdHcfam7z3f3+e0Kn3MF8HzlxhVjCkVVa8LaZGYzJSn/ubl+XQIKi3EFlFHrKcGbJZ0n6TP5z+/VrUcFNLBxUzDedUM4LsWnquxaHq5OqqdN7395dNlLOsIvqc89c1ww3vOtx6NtjV6t0ZjVmHEVm/x2b7gab+IT4Uo8SercErlFfeR10NoXrt6TpJat4aq7oR190XV8b3iy5dgpH2uNf0Y/9KH4JL/BbfTH9yU2PqeuCR8Rb+qPV2K2Rgr7Ht4ensl3whPx6sWhwRompvXwX3M0qwFjKh5hmdm1ku6SdJyZrTez9ykbUK81s0clvSZ/DGCYGFdA9SoeYbn7OZFFp9e5L0BhMK6A6jHTBQAgCSQsAEASSFgAgCTUWiWIAmh7wVHB+BVLroiu027hSqt/vfw1wfi0DXdV3zGMSKy6zfsicwmWqQZrbQ9XCcbq7XzPnmhbg8/tDq9TrrItdjv2yG3dvVxhW+z28V5mbr5YU5F59nb3HBSMn3rQndG2+obCbT12T3h8znnmvyv0rk4if+NYVWE9cIQFAEgCCQsAkAQSFgAgCSQsAEASSFgAgCSQsAAASaCsHVGPfCx8O++XdcZvs/3g3nBp8iEP7apLnzB6hvaEy92tTFm5tUXeQmITpkYm2JXKlK/XUiZd0zpVTgwbK+uWZBMnBuO9bw+v87sT1kXb+smuOcH4zLvC5fa+KzwGpTHwNx4hjrAAAEkgYQEAkkDCAgAkgYQFAEgCCQsAkASqBKH+N74sGL/nDy6LrBG+zbckffCjHw3GJ/7Xz6vtFhotNsnrULwaLHaL+lhbDbutegMmZo1NcCtJ++bPDsY/9IofB+PtFp9g9zvrTgrGD35kWzA+VKYSsxmVffXEERYAIAkkLABAEkhYAIAkkLAAAEkgYQEAklCxStDMlkl6k6TN7n5CHrtY0vmStuRPW+Lut4xWJzG6njgz/Lml28LVgOf8+rXRtibduiYYT7s2qb7G7JiKzf83EKkErKGtmpSZs6+et7Wvdvut06dFV3n0j8P7/+pJ4dvX39d/ZLStXbfOCMYPfurBYDw6X+A4MJwjrCslnRGIX+bu8/J/JCtg+K4UYwqoWsWE5e53SHqmAX0BCoExBdRmJN9hfcTM7jezZWZ2cN16BBQXYwooo9aE9VVJx0iaJ2mDpM/Hnmhmi81slZmt2qfw/XYAMKaASmpKWO6+yd0H3X1I0tclhecOyZ671N3nu/v89jJT+gBFxpgCKqspYZnZzJKHb5X0y/p0BygmxhRQ2XDK2q+VtEDSdDNbL+lTkhaY2Txl1cq9kj4win1EHbRMnhxddu6r7wzGdwztCcY3X/KiaFud/b+ormMFNK7GVCMmU42VrqvMBLQeLkWvpeS7pbs7GN98Znwc/MlLfxSMP9g/Kxi/dPlbo20de8O6YHxgx87wColPcFtOxYTl7ucEwt8chb4AhcCYAmrDTBcAgCSQsAAASSBhAQCSQMICACShYtEFxodHL35JdNn3p38lGD/70bcH4523UAlYKGO56iwyya0PVd9na2sPxgd+59hgfN6f3B9t67jODcH4R1e8Oxg//orHom0NbN4SXjCW/19GCUdYAIAkkLAAAEkgYQEAkkDCAgAkgYQFAEgCVYLjzLPvPiUYv/+d/xRd57HILdB3fjZ82+5OhSuggFFR5nb3HpsasIYKupaJE4LxJy8I38Lla4f/MNrWjX0nBuPHX74tGB/ctLlC7yBxhAUASAQJCwCQBBIWACAJJCwAQBJIWACAJJCwAABJoKw9UW2zjgjGL/jEdcF4p8X/qxetOTcYP/QHTHKLMaCek7yaRRftOWl2MH7D74QvCZnZ2hFt6yt3nxaMH/fofWU6h0o4wgIAJIGEBQBIAgkLAJAEEhYAIAkkLABAEipWCZrZUZL+RdIMSS5pqbtfbmaHSLpOUo+kXkkL3T08syNqYm3x/54Tv78+GH9H99Zg/Oq+w6JtzfhE+HNLfMpRjARjqnlaJ0+OLpt6cW8wPqe9Kxh/bN/OaFvHfW1PMO4DA/HOoaLhHGENSLrQ3edKOkXSh81srqSLJK1w99mSVuSPAVTGmAJqUDFhufsGd78n/71P0sOSZkk6W9JV+dOukvSW0eokMJ4wpoDaVHXhsJn1SHqppJWSZrj7/hsjbVR2eiO0zmJJiyVpgibV2k9gXGJMAcM37KILM+uWdIOkC9x9R+kyd3dl5+L/H3df6u7z3X1+uzpH1FlgPGFMAdUZVsIys3ZlA+tqd78xD28ys5n58pmSuGUmMEyMKaB6w6kSNEnflPSwu3+hZNHNks6T9Jn85/dGpYdFduJx0UV/d9i3q2rqy5e8I7rsoDV3VdUWRmbMjqnYPHv1nMuv3iJ9bukMH3k+8cETok3ddvSlwfigh0+7vu3e86NtHfHQ48H4GP5LJmE432G9UtK5kh4ws/0zNy5RNqiuN7P3SVonaeHodBEYdxhTQA0qJix3v1NSbIrj0+vbHWD8Y0wBtWGmCwBAEkhYAIAkkLAAAEkgYQEAklDVTBcYHa1z5wTji79bfVXz3GUfDsZ7vn131W2hYGLl62VuK9+Qkvcy24+Vr+85NVy+/uZFd0bbmtTSGoz/bE/4bXLGP8Yv2h7avTu6DLXjCAsAkAQSFgAgCSQsAEASSFgAgCSQsAAASaBKcAx45EMHB+NnTdoRjJdz5E/2hheM5QlMka5yFYRVtxX+/Gzt8bcpO3pWML7uDeF1PtLVG23r3v6uYPz9Ny0Oxuc88t/RtgYZb6OCIywAQBJIWACAJJCwAABJIGEBAJJAwgIAJIGEBQBIAmXtDbTnrJOC8RVnfT6yxqTR6wxQB9bRUbfnW0d7MD54TLh0XZJ+febkYPz0U9YE4602FG3rG5t+LxifsjZSuj9UpnQ9Vu7f5MmCoxIpw+cICwCQBBIWACAJJCwAQBJIWACAJJCwAABJqFglaGZHSfoXSTMkuaSl7n65mV0s6XxJW/KnLnH3W0aro+PBU68M34L76LbqqwGv7jssGG/fEZ78No0aoGJIbkyVqyAbHAyGLXLremst8xn58EOD4SdPC1cCStILF/QG42cdcm8wPq11Z7StmROeDcbvPThSdXdEeAxKkvX1BeM+MBBdp2qxasDIJMKSpKHw/1cqhlPWPiDpQne/x8wmS1ptZrflyy5z98+NXveAcYkxBdSgYsJy9w2SNuS/95nZw5LiF0YAKIsxBdSmqu+wzKxH0kslrcxDHzGz+81smZkFb+pkZovNbJWZrdqn/hF1FhhvGFPA8A07YZlZt6QbJF3g7jskfVXSMZLmKfu0GJyuwd2Xuvt8d5/frvB5baCIGFNAdYaVsMysXdnAutrdb5Qkd9/k7oPuPiTp65LC8w4B+H8YU0D1hlMlaJK+Kelhd/9CSXxmfi5ekt4q6Zej08Xi+oetc6PL7np9TzDuGx4Ypd6gXsbTmIpVvVlb5K2lO3wbeknytvDn54lb4lWKW57rDsbX9h8ejD9eZi7BH/SGx9v0ByL7uHV7tC2PzTNYyzx/0Y1EtuFpVwKWM5wqwVdKOlfSA2Z2Xx5bIukcM5unrCy3V9IHRqWHwPjDmAJqMJwqwTslhT4WNP/6ECBBjCmgNsx0AQBIAgkLAJAEEhYAIAkkLABAEswbeGvkKXaIn2ynN2x7QL2t9BXa4c/UsTZ5ZMbVmGoJTw5ddpWO9nD8oKnhFcqVlUfeC2PvkUNbn4k3FZsUuDW8j3WdFDdBt/vy1e4+v9LzOMICACSBhAUASAIJCwCQBBIWACAJJCwAQBIaWiVoZlskrcsfTpf0dMM2PvYUef9T3vcXuHv4Xu5NwJh6niLvf+r7Pqxx1dCE9bwNm60aThnjeFXk/S/yvo+mov9di7z/Rdl3TgkCAJJAwgIAJKGZCWtpE7c9FhR5/4u876Op6H/XIu9/Ifa9ad9hAQBQDU4JAgCSQMICACShKQnLzM4ws1+Z2Vozu6gZfWgUM1tmZpvN7JclsUPM7DYzezT/eXAz+zhazOwoM/tPM3vIzB40s4/m8ULsfyMxporxmir6mGp4wjKzVklflnSmpLmSzjGzuY3uRwNdKemMA2IXSVrh7rMlrcgfj0cDki5097mSTpH04fz/uij73xCMKUnFeU0Vekw14wjrJElr3f1xd98r6buSzm5CPxrC3e+QdOCNc86WdFX++1WS3tLQTjWIu29w93vy3/skPSxplgqy/w3EmCrIa6roY6oZCWuWpN+UPF6fx4pkhrtvyH/fKGlGMzvTCGbWI+mlklaqgPs/yhhTBXxNFXFMUXTRZJ5dVzCury0ws25JN0i6wN13lC4rwv6jsYrwmirqmGpGwnpS0lElj4/MY0WyycxmSlL+c3OT+zNqzKxd2cC62t1vzMOF2f8GYUwV6DVV5DHVjIT1C0mzzeyFZtYhaZGkm5vQj2a6WdJ5+e/nSfpeE/syaszMJH1T0sPu/oWSRYXY/wZiTBXkNVX0MdWUmS7M7A2SviipVdIyd//7hneiQczsWkkLlE3/v0nSpyT9m6TrJR2t7NYQC939wC+Rk2dmr5L0M0kPSBrKw0uUnXMf9/vfSIwpxpSKsP9MzQQASAFFFwCAJJCwAABJIGEBAJJAwgIAJIGEBQBIAgkLAJAEEhYAIAn/A2H1Rs7YUWLuAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "e6da03db-00cf-4203-b5a5-bf52f93db0db",
        "_uuid": "b28a0d516d4f851f4b3751dc59b8d35ae265ca5f"
      },
      "cell_type": "markdown",
      "source": "# VAE"
    },
    {
      "metadata": {
        "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
        "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a",
        "trusted": true
      },
      "cell_type": "code",
      "source": "import numpy as np\nimport matplotlib.pyplot as plt\n#from scipy.stats import norm\n\nfrom keras.layers import Input, Dense, Lambda\nfrom keras.models import Model\nfrom keras import backend as K\nfrom keras import metrics",
      "execution_count": 75,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "2b9d51b1-f9a6-4a4e-8802-3f18a4b867b4",
        "_uuid": "4d25a45182ad846782f201521e304dd3941a0922",
        "trusted": true
      },
      "cell_type": "code",
      "source": "batch_size = 100\noriginal_dim = 784\nlatent_dim = 32\nintermediate_dim = 256\nepochs = 50\nepsilon_std = 1.0",
      "execution_count": 76,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "73c11906-ac31-458b-b51f-f129869a8884",
        "_uuid": "4a26674b0946a0a84241b19c023b277929bdc3ce",
        "trusted": true
      },
      "cell_type": "code",
      "source": "x = Input(shape=(original_dim,))\nh = Dense(intermediate_dim, activation='relu')(x)\nz_mean = Dense(latent_dim)(h)\nz_log_var = Dense(latent_dim)(h)",
      "execution_count": 77,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "e8f2325c-90bf-429a-937e-52eaf755146a",
        "_uuid": "acbf98a32f2da0c14cf19bf74e26e9e4677e6586",
        "trusted": true
      },
      "cell_type": "code",
      "source": "def sampling(args):\n    z_mean, z_log_var = args\n    epsilon = K.random_normal(shape=(K.shape(z_mean)[0], latent_dim), mean=0.,\n                              stddev=epsilon_std)\n    return z_mean + K.exp(z_log_var / 2) * epsilon",
      "execution_count": 78,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "75b19cd0-e9bb-486a-826a-f553a655f472",
        "_uuid": "a54430ccc010ea7deb7fe1a08fc156ea93bde41d",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# note that \"output_shape\" isn't necessary with the TensorFlow backend\nz = Lambda(sampling)([z_mean, z_log_var])",
      "execution_count": 79,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "e3bc9deb-50c4-48bc-8cab-6fcdd2bc1cdf",
        "_uuid": "4f40029827608bc289328ec5a94a3e913a6e5e45",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# we instantiate these layers separately so as to reuse them later\nh_decoded = Dense(intermediate_dim, activation='relu')(z)\n\nx_decoded = Dense(original_dim, activation='sigmoid')(h_decoded)",
      "execution_count": 80,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "4f40c339-bbe7-4225-96d3-ea4986e2b663",
        "_uuid": "53ce840a87f49bbcb3906a6a4597754f23e14a7d",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# instantiate VAE model\nvae = Model(x, x_decoded)",
      "execution_count": 81,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "6ca1311f-bd97-41ef-b11f-6c177f5bb01d",
        "_uuid": "a5959f2c4a4c1e99243dc9c04de640e78a818320",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# Compute VAE loss\nreconstruction_loss = original_dim * metrics.binary_crossentropy(x, x_decoded)\nkl_loss = - 0.5 * K.sum(1 + z_log_var - K.square(z_mean) - K.exp(z_log_var), axis=-1)\nvae_loss = K.mean(reconstruction_loss + kl_loss)",
      "execution_count": 82,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "c9c24ff9-ebd8-4bcc-9b66-4721e0b3ca6f",
        "_uuid": "9838fd084a881c1e3bbc4627ba0e107d2731e03d",
        "trusted": true
      },
      "cell_type": "code",
      "source": "vae.add_loss(vae_loss)\nvae.compile(optimizer='rmsprop')\nvae.summary()",
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "text": "__________________________________________________________________________________________________\nLayer (type)                    Output Shape         Param #     Connected to                     \n==================================================================================================\ninput_8 (InputLayer)            (None, 784)          0                                            \n__________________________________________________________________________________________________\ndense_12 (Dense)                (None, 256)          200960      input_8[0][0]                    \n__________________________________________________________________________________________________\ndense_13 (Dense)                (None, 32)           8224        dense_12[0][0]                   \n__________________________________________________________________________________________________\ndense_14 (Dense)                (None, 32)           8224        dense_12[0][0]                   \n__________________________________________________________________________________________________\nlambda_2 (Lambda)               (None, 32)           0           dense_13[0][0]                   \n                                                                 dense_14[0][0]                   \n__________________________________________________________________________________________________\ndense_15 (Dense)                (None, 256)          8448        lambda_2[0][0]                   \n__________________________________________________________________________________________________\ndense_16 (Dense)                (None, 784)          201488      dense_15[0][0]                   \n==================================================================================================\nTotal params: 427,344\nTrainable params: 427,344\nNon-trainable params: 0\n__________________________________________________________________________________________________\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "fdc3fea6-6af7-4736-a345-d686a159158e",
        "_uuid": "9c9d4173646a9c3936a3f72902c1081c5a9e96df",
        "trusted": true
      },
      "cell_type": "code",
      "source": "vae.fit(X_train_flat,\n        shuffle=True,\n        epochs=epochs,\n        batch_size=batch_size,\n        validation_data=(X_test_flat, None))",
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "stream",
          "text": "Train on 60000 samples, validate on 10000 samples\nEpoch 1/50\n60000/60000 [==============================] - 4s 64us/step - loss: 159.8752 - val_loss: 130.9053\nEpoch 2/50\n60000/60000 [==============================] - 4s 70us/step - loss: 126.9431 - val_loss: 120.5206\nEpoch 3/50\n60000/60000 [==============================] - 5s 80us/step - loss: 119.3553 - val_loss: 116.8157\nEpoch 4/50\n60000/60000 [==============================] - 4s 73us/step - loss: 115.5697 - val_loss: 113.6902\nEpoch 5/50\n60000/60000 [==============================] - 4s 68us/step - loss: 113.2996 - val_loss: 111.5416\nEpoch 6/50\n60000/60000 [==============================] - 4s 66us/step - loss: 111.8469 - val_loss: 110.2709\nEpoch 7/50\n60000/60000 [==============================] - 4s 63us/step - loss: 110.7690 - val_loss: 110.2429\nEpoch 8/50\n60000/60000 [==============================] - 4s 63us/step - loss: 109.9677 - val_loss: 109.4120\nEpoch 9/50\n60000/60000 [==============================] - 3s 56us/step - loss: 109.3722 - val_loss: 109.0644\nEpoch 10/50\n60000/60000 [==============================] - 3s 55us/step - loss: 108.8697 - val_loss: 107.7731\nEpoch 11/50\n60000/60000 [==============================] - 3s 57us/step - loss: 108.4473 - val_loss: 108.0265\nEpoch 12/50\n60000/60000 [==============================] - 4s 59us/step - loss: 108.0508 - val_loss: 107.5796\nEpoch 13/50\n60000/60000 [==============================] - 4s 64us/step - loss: 107.7435 - val_loss: 107.2199\nEpoch 14/50\n60000/60000 [==============================] - 4s 62us/step - loss: 107.4772 - val_loss: 107.0171\nEpoch 15/50\n60000/60000 [==============================] - 3s 57us/step - loss: 107.1876 - val_loss: 106.3389\nEpoch 16/50\n60000/60000 [==============================] - 3s 56us/step - loss: 107.0171 - val_loss: 106.2242\nEpoch 17/50\n60000/60000 [==============================] - 3s 56us/step - loss: 106.8215 - val_loss: 106.2699\nEpoch 18/50\n60000/60000 [==============================] - 4s 59us/step - loss: 106.6459 - val_loss: 106.4430\nEpoch 19/50\n60000/60000 [==============================] - 4s 63us/step - loss: 106.4755 - val_loss: 105.5310\nEpoch 20/50\n60000/60000 [==============================] - 3s 56us/step - loss: 106.4311 - val_loss: 106.0317\nEpoch 21/50\n60000/60000 [==============================] - 3s 57us/step - loss: 106.2307 - val_loss: 105.7900\nEpoch 22/50\n60000/60000 [==============================] - 4s 68us/step - loss: 106.1153 - val_loss: 105.7806\nEpoch 23/50\n60000/60000 [==============================] - 5s 76us/step - loss: 106.0152 - val_loss: 105.5900\nEpoch 24/50\n60000/60000 [==============================] - 5s 79us/step - loss: 105.9018 - val_loss: 105.9176\nEpoch 25/50\n60000/60000 [==============================] - 4s 74us/step - loss: 105.7904 - val_loss: 105.5335\nEpoch 26/50\n60000/60000 [==============================] - 4s 67us/step - loss: 105.6816 - val_loss: 105.3009\nEpoch 27/50\n60000/60000 [==============================] - 3s 57us/step - loss: 105.6579 - val_loss: 105.5394\nEpoch 28/50\n60000/60000 [==============================] - 3s 57us/step - loss: 105.5281 - val_loss: 104.9429\nEpoch 29/50\n60000/60000 [==============================] - 4s 61us/step - loss: 105.4698 - val_loss: 105.1891\nEpoch 30/50\n60000/60000 [==============================] - 4s 62us/step - loss: 105.3761 - val_loss: 104.7125\nEpoch 31/50\n60000/60000 [==============================] - 3s 56us/step - loss: 105.3125 - val_loss: 105.3194\nEpoch 32/50\n60000/60000 [==============================] - 3s 57us/step - loss: 105.2293 - val_loss: 105.7832\nEpoch 33/50\n60000/60000 [==============================] - 3s 56us/step - loss: 105.2023 - val_loss: 105.1130\nEpoch 34/50\n60000/60000 [==============================] - 3s 57us/step - loss: 105.0885 - val_loss: 105.2185\nEpoch 35/50\n60000/60000 [==============================] - 4s 62us/step - loss: 105.0555 - val_loss: 104.5444\nEpoch 36/50\n60000/60000 [==============================] - 3s 56us/step - loss: 104.9912 - val_loss: 104.8461\nEpoch 37/50\n60000/60000 [==============================] - 3s 56us/step - loss: 104.9423 - val_loss: 104.7661\nEpoch 38/50\n60000/60000 [==============================] - 3s 56us/step - loss: 104.8586 - val_loss: 104.2338\nEpoch 39/50\n60000/60000 [==============================] - 3s 57us/step - loss: 104.7935 - val_loss: 104.5815\nEpoch 40/50\n60000/60000 [==============================] - 4s 61us/step - loss: 104.7809 - val_loss: 104.7738\nEpoch 41/50\n60000/60000 [==============================] - 3s 57us/step - loss: 104.7650 - val_loss: 104.8135\nEpoch 42/50\n60000/60000 [==============================] - 4s 68us/step - loss: 104.7377 - val_loss: 104.5791\nEpoch 43/50\n60000/60000 [==============================] - 4s 70us/step - loss: 104.6725 - val_loss: 104.7893\nEpoch 44/50\n60000/60000 [==============================] - 4s 69us/step - loss: 104.6280 - val_loss: 104.3656\nEpoch 45/50\n60000/60000 [==============================] - 5s 75us/step - loss: 104.5809 - val_loss: 104.5995\nEpoch 46/50\n60000/60000 [==============================] - 4s 66us/step - loss: 104.5388 - val_loss: 104.5873\nEpoch 47/50\n60000/60000 [==============================] - 3s 56us/step - loss: 104.4693 - val_loss: 104.3842\nEpoch 48/50\n60000/60000 [==============================] - 3s 58us/step - loss: 104.4929 - val_loss: 104.1723\nEpoch 49/50\n60000/60000 [==============================] - 3s 56us/step - loss: 104.4392 - val_loss: 105.1005\nEpoch 50/50\n60000/60000 [==============================] - 3s 55us/step - loss: 104.4380 - val_loss: 104.2996\n",
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "execution_count": 84,
          "data": {
            "text/plain": "<keras.callbacks.History at 0x7f5242f3ccf8>"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "7e794937-526c-4bd6-a2e7-4ae25fc470aa",
        "_uuid": "dcce46bb299521cbc13929527ff480fded47340e",
        "trusted": true
      },
      "cell_type": "code",
      "source": "one_seven = X_test_flat[0]",
      "execution_count": 85,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "205a9a27-0369-4e6f-b804-426125e1dca5",
        "_uuid": "2cde45e88b576f56b6cb51c0e9ef4ae45ae9e2c9",
        "trusted": true
      },
      "cell_type": "code",
      "source": "one_seven = np.expand_dims(one_seven,0)",
      "execution_count": 86,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "31cd32dc-bf0f-4052-b209-c80df787b273",
        "_uuid": "7c3e1d3b3c304037dd7a6ba1ed010c9a0c75bfd1",
        "trusted": true
      },
      "cell_type": "code",
      "source": "one_seven.shape",
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 87,
          "data": {
            "text/plain": "(1, 784)"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "39aeeef0-a9fd-4ed2-a4a4-aafc1d074779",
        "_uuid": "76d07aca4c96e14b7745ae8d098c5024731021c5",
        "trusted": true
      },
      "cell_type": "code",
      "source": "one_seven = one_seven.repeat(4,axis=0)",
      "execution_count": 88,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "f87e1bde-2ee1-494a-b967-60f78a25ea3b",
        "_uuid": "1fd22752d1bbc89e6e0c6513b297f5715f15f742",
        "trusted": true
      },
      "cell_type": "code",
      "source": "s = vae.predict(one_seven)",
      "execution_count": 89,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "f843288c-332b-40ac-a286-b1a0e5a630af",
        "_uuid": "2a001c55c3e044d44c93a9f6f898016f49ef7f8e",
        "trusted": true
      },
      "cell_type": "code",
      "source": "s.shape",
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 90,
          "data": {
            "text/plain": "(4, 784)"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "6f00c1ae-95ba-4d3d-be42-64b6211f3acf",
        "_uuid": "aa57b61e44a4f400723d291e54a6741f9fb47505",
        "trusted": true
      },
      "cell_type": "code",
      "source": "s= s.reshape(4,28,28)",
      "execution_count": 91,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "3feb2334-72c3-4214-876c-7aaa1df6c94b",
        "_uuid": "a6d5a95acb0fffc428abd318cf401c55f6cc59fb",
        "trusted": true
      },
      "cell_type": "code",
      "source": "fig=plt.figure(figsize=(8, 8))\ncolumns = 2\nrows = 2\nfor i in range(1, columns*rows +1):\n    img = s[i-1]\n    fig.add_subplot(rows, columns, i)\n    plt.imshow(img)\nplt.show()",
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<Figure size 576x576 with 4 Axes>",
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "86afa688-b66e-4ad4-94ec-1bf79c355862",
        "_uuid": "ef0f7e51cae005dc0e010330c3b9db7fcff556a7",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# build a model to project inputs on the latent space\nencoder = Model(x, z_mean)",
      "execution_count": 93,
      "outputs": []
    },
    {
      "metadata": {
        "_cell_guid": "bc72f7f2-2f03-448b-b461-c53a8942957d",
        "_uuid": "5765635730b54ab1a9ff8fc9b0581667a7c3ac39",
        "trusted": true
      },
      "cell_type": "code",
      "source": "# display a 2D plot of the digit classes in the latent space\nx_test_encoded = encoder.predict(X_test_flat, batch_size=batch_size)\nplt.figure(figsize=(6, 6))\nplt.scatter(x_test_encoded[:, 0], x_test_encoded[:, 1], c=y_test)\nplt.colorbar()\nplt.show()",
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<Figure size 432x432 with 2 Axes>",
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "metadata": {
        "_cell_guid": "d0836515-016a-4f9c-b9cd-62abd0d8ca8f",
        "_uuid": "4d9f9686cf6c9f4b97ee3049bfffdc442cd26e71",
        "trusted": true
      },
      "cell_type": "code",
      "source": "",
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.6.6",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 1
}