notebooks/check_tfrecords.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import IPython.display as ipd\n",
    "import os\n",
    "import librosa\n",
    "import glob\n",
    "import time\n",
    "import logging\n",
    "import os\n",
    "import glob\n",
    "import dcase_util\n",
    "import itertools\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from scipy import signal\n",
    "from os.path import normpath, basename\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"] = ''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def count_examples(record_list):\n",
    "    return sum([1 for record in record_list for _ in tf.python_io.tf_record_iterator(record)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Standard parsing function with all fields."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SCENE: beach\n",
      "Number of training files for scene beach: 34768\n",
      "Number of val files for scene beach: 35220\n",
      "Number of testing files for scene beach: 37352\n",
      "Number of testing negative files for scene beach: 36581\n",
      "Number of testing positive files for scene beach: 771\n",
      "Percentage negative: 2.11%\n",
      "SCENE: bus\n",
      "Number of training files for scene bus: 26482\n",
      "Number of val files for scene bus: 28076\n",
      "Number of testing files for scene bus: 33810\n",
      "Number of testing negative files for scene bus: 33060\n",
      "Number of testing positive files for scene bus: 750\n",
      "Percentage negative: 2.27%\n",
      "SCENE: cafe_restaurant\n",
      "Number of training files for scene cafe_restaurant: 36097\n",
      "Number of val files for scene cafe_restaurant: 32150\n",
      "Number of testing files for scene cafe_restaurant: 30590\n",
      "Number of testing negative files for scene cafe_restaurant: 29906\n",
      "Number of testing positive files for scene cafe_restaurant: 684\n",
      "Percentage negative: 2.29%\n",
      "SCENE: car\n",
      "Number of training files for scene car: 30303\n",
      "Number of val files for scene car: 33744\n",
      "Number of testing files for scene car: 36386\n",
      "Number of testing negative files for scene car: 35777\n",
      "Number of testing positive files for scene car: 609\n",
      "Percentage negative: 1.70%\n",
      "SCENE: city_center\n",
      "Number of training files for scene city_center: 34343\n",
      "Number of val files for scene city_center: 31168\n",
      "Number of testing files for scene city_center: 33488\n",
      "Number of testing negative files for scene city_center: 32781\n",
      "Number of testing positive files for scene city_center: 707\n",
      "Percentage negative: 2.16%\n",
      "SCENE: forest_path\n",
      "Number of training files for scene forest_path: 27132\n",
      "Number of val files for scene forest_path: 25594\n",
      "Number of testing files for scene forest_path: 30590\n",
      "Number of testing negative files for scene forest_path: 29994\n",
      "Number of testing positive files for scene forest_path: 596\n",
      "Percentage negative: 1.99%\n",
      "SCENE: grocery_store\n",
      "Number of training files for scene grocery_store: 35336\n",
      "Number of val files for scene grocery_store: 27944\n",
      "Number of testing files for scene grocery_store: 28658\n",
      "Number of testing negative files for scene grocery_store: 28153\n",
      "Number of testing positive files for scene grocery_store: 505\n",
      "Percentage negative: 1.79%\n",
      "SCENE: home\n",
      "Number of training files for scene home: 29774\n",
      "Number of val files for scene home: 36692\n",
      "Number of testing files for scene home: 30590\n",
      "Number of testing negative files for scene home: 29857\n",
      "Number of testing positive files for scene home: 733\n",
      "Percentage negative: 2.46%\n",
      "SCENE: library\n",
      "Number of training files for scene library: 34054\n",
      "Number of val files for scene library: 35334\n",
      "Number of testing files for scene library: 32200\n",
      "Number of testing negative files for scene library: 31546\n",
      "Number of testing positive files for scene library: 654\n",
      "Percentage negative: 2.07%\n",
      "SCENE: metro_station\n",
      "Number of training files for scene metro_station: 33153\n",
      "Number of val files for scene metro_station: 26453\n",
      "Number of testing files for scene metro_station: 28980\n",
      "Number of testing negative files for scene metro_station: 28363\n",
      "Number of testing positive files for scene metro_station: 617\n",
      "Percentage negative: 2.18%\n",
      "SCENE: office\n",
      "Number of training files for scene office: 32985\n",
      "Number of val files for scene office: 28841\n",
      "Number of testing files for scene office: 34454\n",
      "Number of testing negative files for scene office: 33488\n",
      "Number of testing positive files for scene office: 966\n",
      "Percentage negative: 2.88%\n",
      "SCENE: park\n",
      "Number of training files for scene park: 28245\n",
      "Number of val files for scene park: 36928\n",
      "Number of testing files for scene park: 30590\n",
      "Number of testing negative files for scene park: 29906\n",
      "Number of testing positive files for scene park: 684\n",
      "Percentage negative: 2.29%\n",
      "SCENE: residential_area\n",
      "Number of training files for scene residential_area: 29524\n",
      "Number of val files for scene residential_area: 32673\n",
      "Number of testing files for scene residential_area: 30268\n",
      "Number of testing negative files for scene residential_area: 29711\n",
      "Number of testing positive files for scene residential_area: 557\n",
      "Percentage negative: 1.87%\n",
      "SCENE: train\n",
      "Number of training files for scene train: 28107\n",
      "Number of val files for scene train: 31594\n",
      "Number of testing files for scene train: 35742\n",
      "Number of testing negative files for scene train: 35109\n",
      "Number of testing positive files for scene train: 633\n",
      "Percentage negative: 1.80%\n",
      "SCENE: tram\n",
      "Number of training files for scene tram: 24894\n",
      "Number of val files for scene tram: 28753\n",
      "Number of testing files for scene tram: 29302\n",
      "Number of testing negative files for scene tram: 28674\n",
      "Number of testing positive files for scene tram: 628\n",
      "Percentage negative: 2.19%\n"
     ]
    }
   ],
   "source": [
    "scenes=[\n",
    "    \"beach\",\n",
    "    \"bus\",\n",
    "    \"cafe_restaurant\",\n",
    "    \"car\",\n",
    "    \"city_center\",\n",
    "    \"forest_path\",\n",
    "    \"grocery_store\",\n",
    "    \"home\",\n",
    "    \"library\",\n",
    "    \"metro_station\",\n",
    "    \"office\",\n",
    "    \"park\",\n",
    "    \"residential_area\",\n",
    "    \"train\",\n",
    "    \"tram\"\n",
    "]\n",
    "data_dir= \"/home/ellen/RED/\"\n",
    "for scene in scenes:\n",
    "    train_filenames = [f for f in glob.glob(data_dir+'data_binaries/train/*/{}/*.tfrecord'.format(scene)) if 'neg' in f]\n",
    "    test_filenames = glob.glob(data_dir+'data_binaries/test/*/{}/*.tfrecord'.format(scene))\n",
    "    val_neg_filenames = [f for f in glob.glob(data_dir+'data_binaries/val/*/{}/*.tfrecord'.format(scene)) if 'neg' in f]\n",
    "    test_neg_filenames = [f for f in glob.glob(data_dir+'data_binaries/test/*/{}/*.tfrecord'.format(scene)) if 'neg' in f]\n",
    "    test_pos_filenames = [f for f in glob.glob(data_dir+'data_binaries/test/*/{}/*.tfrecord'.format(scene)) if 'pos' in f]\n",
    "    print('SCENE: {}'.format(scene))\n",
    "    print('Number of training files for scene {}: {}'.format(scene, count_examples(train_filenames)))\n",
    "    print('Number of val files for scene {}: {}'.format(scene, count_examples(val_neg_filenames)))\n",
    "\n",
    "    print('Number of testing files for scene {}: {}'.format(scene, count_examples(test_filenames)))\n",
    "    print('Number of testing negative files for scene {}: {}'.format(scene, count_examples(test_neg_filenames))) \n",
    "    print('Number of testing positive files for scene {}: {}'.format(scene, count_examples(test_pos_filenames)))\n",
    "    print('Percentage negative: {0:.2f}%'.format((count_examples(test_pos_filenames)/count_examples(test_neg_filenames)*100)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _parse_function(file):\n",
    "    features = {\n",
    "        'label': tf.FixedLenFeature([], tf.int64),\n",
    "        'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "        'scene': tf.FixedLenFeature([], tf.string),\n",
    "        'source_file': tf.FixedLenFeature([], tf.string)\n",
    "    }\n",
    "    parsed_features = tf.parse_single_example(file, features=features)\n",
    "    label = tf.cast(parsed_features['label'], tf.int64)\n",
    "    # Quanitised values in the range 0 to 255\n",
    "    audio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "    # The data is first scaled (between 0 and 2) by dividing by\n",
    "    # 128, -1 then adjusts the range to values between -1 and 1.\n",
    "    # tensorflow division: '[...] if one of x or y is a float, \n",
    "    # then the result will be a float.'\n",
    "    # See: https://www.tensorflow.org/api_docs/python/tf/div\n",
    "    audio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "    scene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "    source_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "    return (audio_inputs, audio_quant), label\n",
    "\n",
    "repo_dir = \"\"\n",
    "data_binaries_log = os.path.join(repo_dir,'data_binaries/train/babycry/bus/')\n",
    "filenames_neg = ['{}babycry_bus_neg.tfrecord'.format(data_binaries_log)]\n",
    "filenames_pos = ['{}babycry_bus_pos.tfrecord'.format(data_binaries_log)]\n",
    "\n",
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_parse_function)\n",
    "# dataset = dataset.repeat()\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_pos})\n",
    "next_data, next_label= iterator.get_next()\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c2efa0748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.squeeze(data[0]))\n",
    "plt.savefig('pos_example')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs = 44100\n",
    "ipd.Audio(np.squeeze(data[0]), rate=44100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "librosa.output.write_wav('pos_wave', np.squeeze(data[0]), fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_parse_function)\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_data, next_label= iterator.get_next()\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])\n",
    "data, label = sess.run([next_data, next_label])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c2f269978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.squeeze(data[0]))\n",
    "plt.savefig('neg_example')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs = 44100\n",
    "ipd.Audio(np.squeeze(data[0]), rate=44100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "librosa.output.write_wav('neg_wave', np.squeeze(data[0]), fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Negative example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c2573e9b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c2581af98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_0\n",
      "label 1/0 (positive/negative): [0]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(targets[0][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0][0]\n",
    "ipd.Audio(data[0][0], rate=fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c25735c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, t, Sxx = signal.spectrogram(data[0][0], fs)\n",
    "plt.pcolormesh(t, f, Sxx)\n",
    "plt.ylabel('Frequency [Hz]')\n",
    "plt.xlabel('Time [sec]')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Positive example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_pos})\n",
    "next_batch_label, next_batch_data, next_batch_targets, scene, source_file = iterator.get_next()\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n",
    "label, data, targets, s, s_f = sess.run([next_batch_label, next_batch_data, next_batch_targets, scene, source_file])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c25735b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c25a69080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): [1]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(targets[0][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0][0]\n",
    "ipd.Audio(data[0][0], rate=fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c25a305f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, t, Sxx = signal.spectrogram(data[0][0], fs)\n",
    "plt.pcolormesh(t, f, Sxx)\n",
    "plt.ylabel('Frequency [Hz]')\n",
    "plt.xlabel('Time [sec]')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Check that this is what we expect (i.e. files have been written properly)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f='data/TUT-rare-sound-events-2017-development/data/mixture_data/devtrain/20b255387a2d0cddc0a3dff5014875e7/audio/mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2.wav' \n",
    "# ipd.Audio(f)\n",
    "audio = dcase_util.containers.AudioContainer().load( filename=f, mono=True)\n",
    "\n",
    "segment_length = 0.1\n",
    "fs = 44100\n",
    "# Float, needs to be converted to an int. \n",
    "duration_in_secs = int(audio.duration_sec)\n",
    "\n",
    "# segment_meta is a MetaDataContainer\n",
    "# Parameter exists to skip segments: https://dcase-repo.github.io/dcase_util/tutorial_audio.html\n",
    "data, segment_meta = audio.segments(segment_length_seconds=segment_length)\n",
    "# ipd.Audio(data[14], rate=fs)\n",
    "\n",
    "data_quantised = linear2mu(data[5])\n",
    "ipd.Audio(data_quantised, rate=fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### CAE parsing function\n",
    "Should just be inputs between -1 and 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _cae_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = [] # No labels being passed in training, just a placeholder.\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Range between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn audio_inputs, label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_cae_parse_function)\n",
    "# dataset = dataset.repeat()\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c309d6320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): []\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _cae_test_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = tf.cast(parsed_features['label'], tf.int64)\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Range between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn audio_inputs, label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_cae_test_parse_function)\n",
    "# dataset = dataset.repeat()\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJztnXeYHNWV9t/beaImSxqNpFECSSgzEiCBwIANGC9ibYxxwLDGizGLvf6cPmF7cd7FsOuAjY1ZmzWYBQxYnyWTZEAiGCQhoZw1yiONNEGa0aTO9/uj6nZXd1dXV+5Q9/c8etRTXdN1p6r61LnnnvMeQikFh8PhcJyFK98D4HA4HI79cOPP4XA4DoQbfw6Hw3Eg3PhzOByOA+HGn8PhcBwIN/4cDofjQLjx53A4HAfCjT+Hw+E4EG78ORwOx4F48j2AbDQ0NNDW1tZ8D4PD4XCKivfff7+HUtqYa7+CNf6tra3YtGlTvofB4XA4RQUh5Kia/XjYh8PhcBwIN/4cDofjQLjx53A4HAfCjT+Hw+E4EG78ORwOx4Fw48/hcDgOhBt/DofDcSDc+DuQY73D+NwfNmJP57l8D4XD4eQJbvwdyI9e3I01e7tw3S/ezvdQOBxOnuDG34GcPhfM9xA4HE6eKVh5B451bOvoT7z+r7/tg8/twq2XTERNuS+Po+JwOHbCjb/D6BsOp/z8q7XtoBRoqvbjEwsn5GlUHA7HbrjxdxgDwSgA4MGb5uDjbePRPxLB3O//LbGdw+E4Ax7zdxCUUvzmzYMAgAq/8Nyv8LkBAD96cQ++85cdeRsbh8OxF278HURnfxBPbTgGAJjSWAkA8LiTt8CT64+ha4AvBnM4ToAbfwcxGBJCO7/61HycP6Yqsf3mtpbkPjz8w+E4Am78HUR71yAAoMKXutRDQBKv1x3qtXVMHA4nP3Dj7yD2nhoAALQ2VKRsXzavOfH61d2nbR0Th8PJD9z4O4hQJAa/x4VJacZ/8dQGHLn/eixqrUMwEsvT6Dgcjp1w418A/HXbSazadtLy42w51ocyMbtHjnK/G+sPnUE0Frd8LE7mwOkBPPDKXuyQFNtxOHbDjX8B8KWnt+DLT2+x/DgjkRiGw9k9+0ox/XPnSS74ZiWPvXMEv37jIB5562C+h8JxMNz455l4nMq+toKRSAxXz2jK+v5nL2kFwDN+rGYgGAEAjCg8iDkcq+HGP4+8faAb//32ocTPVmbaUErR3jWYkekjpcIvhIRe3X3KsnE4mf2nB3Djw+9g5wkh3LNmbxceXtuOCA+zcfIAl3fII7f+/r2Unx/7+2EsmdpgybEO9QwBADxuknWflppyAMD6Q2csGYPTufm369A3HIFLcgkeXL0PF02qQ1trXf4GxnEk3PMvEAJeF4bC1oVbmHbP1TNGZ91nVLkXN85rxgjP+LGEvmEh3PODZbNStg+EeJiNYz/c+OeBHR39+I+X96Rsa6oKYP2hMxmqm2bx8s5OAElNn2xU+D3oGQzhiXVHLBmHU0nPoCIS7/9Y77DNo+FwuPHPC4+8dRC/fTMZ6x9THUBbay0A4I193ZYcc82eLgBAa32F4n5zW2oQjVHct3IXzg5Z8yByIqzADgAWTarD8munw+8Rvn5Wzvg4nGzwmH8ekMon/+KWeVg2bxxO9QexYvMJxVRMI4Rjcdw4rxljRgUU97t54XiAAN98fjsGQ1HUVvAGL2bAruuTd1yE80ZX4bzRVfjnyyZj8rdeQpBn/XDyAPf8baZnMIS39ie9e5ZbX+YVMm2sirePhGOKBV5S2Ji2HO+zZCxOhF3XMl/yK+dyEfg9Ljy0pp2rqXJshxt/m1m9K5lGWVPuTUgr+73CpbBKXiEYicHvUWf8pzUJY1p3sMeSsTgRdl3Tr8GscaMAAKt3cU0ljr2YYvwJIdcSQvYRQtoJIctl3v8qIWQ3IWQ7IeR1QshEM45bjLCQz+4fXIOt930oIbLm97hAiJXGP67a8582ugrjasoQivL8c7MIJjz/1GvwxOcWAQBGeNyfYzOGjT8hxA3gYQDXAZgJ4JOEkJlpu20B0EYpnQPgeQAPGD1uMUIpxe/Eoi4W5mEQQuB1ubDhsPk59sd6hxGOxRFQ6fkL4wFWbD6B7oGQ6eNxImwhP5B23dnPI2H+oOXYixme/yIA7ZTSQ5TSMIBnACyT7kApXUspZfls6wG0wIEc6R1Gz2AY5T43CMkstgrH4ugXc8HNZMWWDgDAzOZq1b/TcXYEAPDbN7n+jBmwbJ/6tAV0t4vA53YhGOWLvhx7McP4jwNwXPJzh7gtG3cAeFnuDULInYSQTYSQTd3d1qQ85hOm6fLQLfNl319mUYHVUCiKMq8bH5yZvcArG2d4uqcpBCMxfGTO2AzPHxAK/LjOD8duzDD+cnoBsgplhJDPAGgD8KDc+5TSRymlbZTStsbGRhOGVlic7BO86WyFVhV+D46dGTY97v/ani6Uq4z3p+N2ZZeD4KijdzCEwz1DiSyqdHweN7byzCqOzZhh/DsAjJf83AIgQ5yeEHI1gG8DuIFS6shA8jlxsbcqIG8ERpV5AQCbj5019bid/SOaF2/v+cBUAEBVwGvqWJwI646W3kGNMRSKcnE3ju2YYfw3AphGCJlECPEBuAXAKukOhJD5AH4LwfB3mXDMoiQsGuCmKr/s+9fPHgvAfEnlWJzi1ku0JVh9/Zrzs3qqHG0Mito9n7poguz7V85o4npKHNsxbPwppVEA9wBYDWAPgGcppbsIIT8ghNwg7vYggEoAzxFCthJCVmX5uJKGed9+mbgvkAwH3f/yXgyZJPYVjcURidGM7CI1DIWjeOydw4mHFkcfLKSTTU670ufBoe4hO4fEUcGqbSfxyJsH8eDqvTgXND8RI9+Y4tpRSl8C8FLatvskr6824zjFTrLQR/6ZO6Y6AI+L4FDPEDYc7sWV07Uv0GYcUzTcAa/25zwVV27W7uvCNReMMTwWp8IWc7Otn7BMn2gsDo+b110WCtLuerOaR+E6cWZeKvA7zUYSnn8W41/mc+OVrywFkKr/YwRmePR4/umfwdHHcDiGhaJwnxwXiCm4QT7DKlhKUXabG38bCUViYiVv9gwaFmc/2DVoyjEfXtsOIHuoSQ1MDpqjj10n+xWltNmD2arqbo520q/FN5/fjr2nSqu3NTf+NhKKxrN6/Yw6sQgoHDPez5dSij+8ewQAUKVj8fZbH54OAHjPgqpjJ0EIwXAou2H3J6p8ufEvFKQS3Iw/rjuah5FYBzf+NhKMxGSLfKT4PC5UBzymaL1I0ztnjFVf3cu4c+kU3HHpJK7xY5BgJIb5E2qyvs89/8KDJVw8c+fFiW2DJRb64cbfRt492AtfDs8fELJ+dncan2Ie7E6GjtQcN9tYhsMxnvGjk46zwwhF44phH+YQmHHNOebwpii7Lk13Xrn1pCXyK/mCG38bGQpFVU3th0JRxXUBtUgbsTdUytcW5MIjZqjsP505DebkhhV4TW7M3kFtqiihvaOj35YxcXKz84RwLcbXluPmtqQU2TslJHPOjb+NBCMxLJunJHsksGBirSnxXxY62v+j63R7/nPHC+GKEBce0wULmV05vSnrPpMaKlBX4eOFXgXEcDiGy6Y1YFS5Fw/cNBdvf/MDie2lAjf+NrH1eB+GwjFU+nNn3VT4Pdhxoh+UGlv03XNqAC6iP+QDAAEPazLDwz56CInnzZcjf7/c58b/bjiGfTILjRx7eWt/N7Ye70sJ+VjdbCkfcONvEzf95l0ASalkJVio5WS/sdZ+BEDcYNIQy0Thnr8+wrEY3C6Ss3hr8ZR6AMBTG0oro6QYeej1AwCAtta6xLZSXJTnxt8moqIVHl9XnnPfq2cIlb1GJR5C0biuLB8prDI4xD1/XYQiudN7AeCBm+ZiTHWAh34KgMFQFB+aORp3XDopsS3AjT/HDth0s9Og5y+klhq7xKznLG82oo9wTJ3xB4QHLQ+v5ZfBUBR7Tw1kiBp6xZnbthJalOfG32b+cX7uBV8m+dw3bKyRSjASMyTrACSlKLjnr49QJK56zSXgdXPPP89sPCJkyDVkUd4tpZRnrtlrE01Vflw5vSmrpruU5poyAMYrPoORuGE9/lKc7tpJKBpLzJ5yEfC6+XnOM+w799EFmU7aJZPrS6oKm3v+NqGmupfBpH+Xr9hhKONnx4l+0zz/Qz1cclgP4Zh6z7/M68bbB3p43+Q8wh6+AZkHdoXfjfeOnCmZ5Adu/G0iGImrNv7STl/SQi0tsOmpUWkGow8Pp6N2wRdAotXmf7y8t6QqSYsJFnYrk2l7WlMu6G6xArBihxt/G4jFKcKxuOrFV5dE912vl8EyhS6dWq/r96VjaR4VKKniFjvRsuD72cWtideDJmg7cbTDFtzlPP9PLhI6sZklt55vuPG3gRNibr8eL/qszkXftfuEbpnlJrRiLPd7EgthSjy5/ihe2tGJP208ZviYpYKWBd8KibfZbpKkN0cbb4jfm4Av85pViAWa6w722jomq+ALvjaw/rBws4wVF3LV0DaxFpuOnkVEp7TzSzsEDf6ZBvP8AWEWka0LFaPrXBDf+cvOxM+fWCjfr9ZphKKxRLggF5MkyQAHTg/g8vMarRoWJwvdAyEA8hXZLbVCjc7hEln/4p6/DbD4+yWT1Ydgfv3pBQCASExfzD4YiWPBhBrMGjdK1+9LWTqtEdEcD6H0d41KU5QKoah6z7++0o8DP74OQGlpyBQTw+EYbpzXLCusWOn3YP6EmpJJx+XG3waSjdvVn26WHqg3v15LdlHOsXhdOdce0o3VEDdeAISmIGpj/oBQTOR2ETz9njWhM0opHl7bjrdEyWJOksffPYKugaBiqLTC59HU3GjVtpO484lNZgzPdHjYxwaY4dRiBNiDIqzT8x+JxDCqzFiOf2IsHlfOrKGewVDKz519I5g2usqU4xczPrcL5zQuEPrcLnT2B019gDN6BsN4cPU+eN0EB378YVM/u5ihlOK7q3YBABZJNH3SCUVjmjLoWBP4SCyeqBIuFAprNCWKWmVHKWzfgvD8Pe6cN3y6589lCgTilGJWs7Z1l69fcz4Aa6qqWRaY3rWkUoXd3//32um4UaEKf8nUBgBCBp8WCjGMx42/DbDevVoatLhcBF430ZXqGY9THOweMtH4uxCLU0QVZiGviU1LGD98YXdJlcLrIRaniMapZkltlhJshZ7SPklTnvtW7sT9L+9NNJxxMonirhyh2YBOlds/vHMErctfxONiT+1CgBt/GxBK/LWfap/bpcuAMtloE5qBAUiGoJS8/31pnb7eO3IGexzelpBdO7XyDowyCxu6S1t7PrHuKB558yD+uUBj0naSKO7K4TDpvTY/e20/AOC7q3YVTDIEN/42IGR8aPfC/d7c4Rb54wk3plmpgokQlMJYRsIxXDm9CS986dLENqOS1MUOM/7aPX/rlFRLSZvGTBLFXTmMf3JWpu57WR3IXFY1WnVvFtz428BTG45lLIiqQVho1f5lTWQXGejglTKOHFPdzv4R7DjRj3KfO6Usfq/Du1KFYtoX+oGkd7n/tPmFXis2nzD9M0sB9v3MbfyF988Mqiu+lDP0OwpEHoIbf4sZNOD9+lRk2ciRTC01L+YPZF+AZPH+mc3VGDsqgOZRAQDGJamLncRCv0bjz4q9jppcTEQpxYk+ISTYmCZZrLeepFQ4NyJoKfk8yrFSlrGjpvI+Hqey399CSbPlxt9ijIQ+/B59MX/moWvJLlIeB/P85cfC4qWfvaQV5T4P3r33KlXpoaUOS9PV6vm3NlTA6yYYNrmYiF2Pb157PjZ+++qU9woxG8VO2LlpzlGFP16s8lVzb7N9mFDjJxdNQKXfg6FQYZxrbvwthpWL68G452+W8Rc+p+PscMr2M0NhfHflTvz7S3sBJJu9szH89q1D2HvKuYu+zPPXE34r87qx66S5545dPyYZLuWd9h5Tj1VsMIdJTtBNSkBDI/feIeG731CZnGV53ASPvXMYR3vzLxFhinUghFxLCNlHCGknhCyXeX8pIWQzISRKCLnJjGMWC91iLHFyY+4mLukI+fXavYTEQqNJnn9thaBNk/4ge/tANx5fl2w4Ltek/Il1zm1Izjx/rWEfQFiAdJuUrcXYLrYgHF0tGKMPzRydeO/u/91s7sGKDLbgm8thYjF/NRIPR3qEh+1Fk+owsb4cH5zZhD5Rqvv+l/caGa4pGLYOhBA3gIcBXAdgJoBPEkJmpu12DMDtAJ4yerxiY1ic4jGtHi3oD/uwzAVzjP/EemGqmx4XVrOeoVQbUOroTfUEgPkTakyXyGCft2BCLQDg0c+2YcXdi009RrESUmjiIiWR56/C+LPZwacumoA3v/EBXDl9NBoqBUfqpMH+3GZghnVYBKCdUnqIUhoG8AyAZdIdKKVHKKXbATjKEpw+F8S/rRSULuWm2rnweVzYrWPqb8ToyMHCFulVu2v2dOX83Wc3deCP646YMo5iI7H2osPzr/Rr05BRw7bjfQCACol2TXqjcqcSVBkqZdls21U0cn9m43EA8hlEuwog48cM4z8OwHHJzx3iNsezdm8XzgyF0VJbhqZq+YbQSgwEo7qqdI0YHTmyVTWeHhC8l8mNFRk9T//5skmJ1/+2cpcp4yg2zAi/mdnRi4UqyiXpuOM0yIyXMsn1GeXvW7n4XVCjufXaHiELTq5wzKzvphHMGIFcZFJXCRsh5E5CyCZCyKbu7sJIhzICC4u8+OXLdHnhs8eN0iXsZmShUQ6Pi8BFMj3/4VAMH5kzFmu+dgV+evO8lPe+ff3MjIwSp2Fk4f3KGU3iZ5gX+hkORTFrXHWKzEiF34NFk+pw0aTsYmZOIBiNwesmOftWuFwE05oqNaXGyl1/l1nl9wYwwzp0ABgv+bkFwEk9H0QpfZRS2kYpbWtsLP5GFns6hSKnCpl+oGoIeN2axb1O9Qfxoxd3AzAv7EMIEcaSZogO9Qwphg2cHlIw4vmza/f9v+42LQd/x4lzsuHHMq9bV3ixlHh192l4XOquk9/ryilcKJVwkHr+5eL5z7/pN8f4bwQwjRAyiRDiA3ALgFUmfG7Rwx7uclkwaijzuhGOxTUpCP7i9QNgu5s5tfR7Um94ZpCUsh7MWnAuVsIGiu3YrO3FHZ3YZ1KltN/jku0/OxyOwlsAYYh80j8SUf2QVZOFd1YSrpM+cL93g5ALUykj+2A3hq84pTQK4B4AqwHsAfAspXQXIeQHhJAbAIAQspAQ0gHg4wB+SwhxRBB4OBzFFB0pngwtOcUMaTpmrimstrGk3vBsTLOas3cKI4Tg85dOSokxOwkjxXbSkJ1ZGkmRWBxzWjKv17zxNY7X/InG4okG7bkIqPD8mVP0k4/NhkvyPbxy+mjcsnC8ZkloKzDlcU8pfYlSeh6ldAql9MfitvsopavE1xsppS2U0gpKaT2l9AIzjms2wUgM31u1C2v2Gpe4HQhG8NKOU4ZCHyyz4O0Dyusf+08P4BevHQCl1DQlz3T8Hhd2S1Q6R1RK4Jb53BgOxzAcdp7I28FuoZBHzwxMOlt4YXunKeMJx+RbSlb4PRiJxHAuaN7istW8tvs03j1oXmFaWJRdV4Maz589TOUSNjxugq6BEI6fGc54z06cPddLY8eJfvzh3SP4+nPbDX/WlmNCWt24Wv3ZFHNaagAAL+04pbjfN57fjp+9th9nhyMw0dlP4VwwmhK7DKlUQWRphduO5z+1zW7YzEuPAyA1RNtNSgsMR+OysxDW8W2reM8WA59/YhM+9d8bTPu8UDSuemFejeef7A+Q+f1ggn0Prt6ncZTmwo2/hEExHto/YtwDYhf/7ium6v6MeeNrcP7oqpxexsEu4WYaicQSWQRmPwTmtoxKkZpI6J/nCOksFFvimZm1UiyEojHUVfh0hd+kxl9NQZEawlmaybdNrEu870SisTiicao6QUKN5x9U0R/gVJ4LvfK/6lBA/PtLewAIqY1GMavKNuBzY0Rlxs/Hfv0uTp0TbiizM23KfG6MnE3e8F3nRAncHF+YhCKoAw3LSDieszlINqSGSMuaTzaUuoqpadZTymgV4At4XTh+ZgT9wxGMKpfvk73p6Flx38zrz44Ty3NTF+75SzggetDVJjQ+D5lUZRvwuFR/+ZnhB4Af/eNsQ8fNHIc7pbkIq2FQ2/zCiV5lMBrTLa4nNdJm9ENWaiyTfEAX3+zMjPtKa13MmGohlPvekewV2IfF9R45Ta8vXjEFQFI2JV9w4y+DXm9NCvsiGS20KvO5dXl+N8xtNnTcdAI+N0bCyS8a+/vG1gQUfy+XHHQpEwzHdN9L0tmnGhGxXCjVHBTbNZLm0JuRCaW1/8X1c8YAUJ6RDYWjmNxQkaLoyVg8pQH1Fb68Z8HxsI8M5noTRj3/3NK+RhrGaBlHz2AI8TiFy0US2Qy5jFsxe5VGiMUpXt/bhQsn1ur6fWlAoH8kImZy6Q9HJrqKKYQhzFpbMMJb+7ux99Q5TBtdBVDgA9ObMvaRPqQ2HO7FtbPGaj7O+0fP4s393ZjaVIkvP70FgHpHTc3D8oXtnZg9LnsatF+nbpeZcOMvIi3w0COpkI5ZmvqDoajiTWmH4QeSBWsn+kYwvq5cMZtBSuKLYkLoophgaXx6Pf9xNWWJRfa9pwZwdjiCOlFaWw+JgjM5z7+AYv63/897kKbAH7n/+ox9pB73XU9ult0nF196anOGsqbqBd8c9TcsrVlp6VD4XufX8+dhHxHphTTDAzKrm9YF46oVC0LYtPe+j6SraJvL/AlC2ik7T2wROqfnX0CGxU7Y36u2cCgdn8eFlfdcitsXtwIwvuirFPNn92ghXCM1tU9mhMHkJJXV1mPk7Gwnzoo/dmFL1s+YP6E277Uv3PiLsKbKVQGPKV+CkJhT7TKYOeR3C928aJbMAJZ1Y8QrVENZWhOLA6cFyYFcU2VmWH7yyl4c681vUYseegdD+N6qXYkG32oxc80HAO5buQtxA1WhSo1lPG4XPC5iaWjuqQ3HcOlP1uBQ9yDWH+rFC9t1yX8BEEI2VuBV2T0nV+X9iIpZcaXfg20d/XntncyNv8j6g70AhLz0aJwaLr8ORdRXDCrhzyEhyxpJB7wuXD9nLGrLvbhbzCYwk0QHI9Gr8YhflFwPN+n7v1xzwPRxWc2KzSfwh3eP4NlNx3PvLMGssB/zMl/bcxqd5/TnhecSmfN7XJaG5r71/3ag4+wIvvfX3bjl0fW456ktsvupSbPeIilGm9ygXz4lnfNGV6naLzFTymL8gyoKIKvLhIj7/tPm6Dbpgcf8RYbCMZT73FjYWoc1e7sQjsZzFjApETKQ5ieF3WhC+XnmeFjYZ0JdBR7+lPZuYWphNzJrehGMxNGqMVWtmOQDGOyhOygjiKb4e2al+kruISOhH6WwDyA4GXaEfdL7QKczbXQV9nQqL4SGojHUV/jQWOXH+Dpz0iV3fv8a1bUxhBDhYZnlfKkp8Prw7LF4+r3jeW3mzj1/kZ0n+lHucye8daMZP3tPDcBrQg/dXDFzFq6yWj6ZGaFjvUPYdrwPq7adxJmhsKbP0Ktumk9Yde6v3zioaXHdrLCPNMPHiPhaKIfx97hITqOrF+l9cqg7e+Pyd9t7MsbQlTbb+duuU3hy/TH4PS4EvG68uvs0Vu/KlD95p70Hr+xUr4kU0HidfG4X/rhevj91p7ieoFTgyWRPjuSxkXvxfRstwu91YzAUTXw5jMY/A16XZm9Rdlw5HkbDokEYPUp7pzAt1JYLawrROMVj7xwGIOj9qOETbUK7h0aZnOdCx6MzbBVS2RA8Fxc0Vydem+H5Z3sYDYWihma6Sqw/1Ktqvzse35Sx7e/tqeJtd/7xfQBMZVb4m74gbpPy6d9twF1PqmtKf/HkOs2OSUAULByQmc2yhVylPP56cY0un+m13PiLBCMxzG2pMU2OYCgUwzwxQ8YIvhzjCUVjaKzyW542xsS/IrE4tFal/+SmOajyeyxTHLUS6XnXMhs0q8K7odKP5+66BICxSt9cnv/8CbWmSUeno9YJksviyfY3+zwuLJ3WoHtMAa8Ldy6djCP3X49n7rxE8+9/+cqpWcfH7pOmquwFkKz4ayiPUtrc+IsIqn7uxALrXU++j75hbWENBqUUW4/3mRKKSaaVZcksMFBFqoXEQygS16V95Pe6ik7i4VR/MEV5UUsYz6ywD5DUT2I9YfWQS7+mwu/GZotUPbMtakZVZLq8096Drz67FXtPpYaDAl53SiP6l3YIIZ51B3vxjee2KX4mpRShaFxzqCf9+ID8bCyUY5YFJGcFv3njYNZMPqvhxl8kFIkh4HHBKxq2XSfP4b3D2bU7lDgtpl+aEvPPEfYJRuK2dMxifXxD0XjCkHzvH9TXFghKiMVl/N9K66PQVKU+bKXGAKhlYoOwqHmoR398OJntI+8oRGNU/N++a5TtfhhXU4bzxcybF3d0YsXmE3hhW2r83u9xYZGk7/Dd/yuEeH7z5kE8936H4nHD4uxVT4c1hjrjn/3zCSGoDnjQPxJJ6fplJ9z4izDPf4Ikg2VIZxEGWxi8akZmabpWcoV9RiL2eP5ChoPQVjIUjWP6mCrcvmSS6t9Xyo4oVNK/2FqkmZMxf+PXpjrgxaJJdYbiw7myfRaKhtSM6vZ0gllmrdnuh3kTavDKVy5LCROmXwufx4WLJ9fj85em3oNq1kXMeDCn172kfj6T0lD+/O+IhZlWhdtywY0/gGO9wzjcMwS/x5USqnltT5euz+vsHwEA2WbZWmHeg5znv+tkP97c322KgVGDz+NCKBLT1PUo/XeLifTsGi1hq1fFEI0Znj8gGBtjC75ixXm2VE9JWM9spIKAUp5+7xj+fkCmGxcVnA1peDH9QcEiJW5JYdZwOJpztv7Gvi7c+YSwsGyG57/pSGbBWa6aCgazNX/bbbxzoB648Qfw+Sc2AgD6hiMYMyqANlGMq1djVSeDLXCV+40bZaXsI9be70oZ8SsrYN57tqYgir9rUx65mbDFvI8tEMr0tYyf3QNm9IYA1HWPUiLXgi/bbqXnn16Q9eDqffjM7zO7cVFR1i4SS8bC0x9818wSlDWXTEku+u7oyOx4lh5P/90rpWN1AAAgAElEQVTbh7H+kPCAMHJpzhtdCUA+VVNtdT/rf/3+UX3hZaNw4w8kCi1cRPC0n//iYlw2rUH3AiW72ceO0t/CkaHkkQUjMVT5PbjrcvMreuXweYRF22y9YJUQHhxF5vlHYvC5Xfivm+dqlj+IxSmuuWC0ISVOKWVetyFNm1ypnkozTKMEwzFc0FyNv9yzRNPvSR8W7OFV5nXjzqWTcevFEwEAS89rxM1tLRhTHZAN0yo9zKY0Vmoaj5Sm6gBGV/tlv5dqq/sn1Jdj1rhqU/o16IEbfyQ1PaQ+gs/t0u0FsWmuGbF4dhOt2HIi4z22TmEXKZ6/xsVsv8eVU5q60HhlZ2fi3ojGKR5ee1D170bjcXhc5n29Al43jp0Z1r0gyxoVZZuJmFXfIsd7h88g4HWrDoGx8ybN5lm17STe3N8tVM6nfQ5bizrYlemFHz8zgp+9uj8xA4jGk+fPaDZetgeylur+Mq/bUPGeEbjxh1BSDgCfEb0JQMjU0e35J4SdjJ/e5hph9nBYJtMjGMn8IlgJy9jRE/Y5NxIxZQ3ETgZDUVUqk3LE4jShf2QG7Hwf0SmOx+6TbDMRK5U9CRHWT3xuFz44czQ+NHN01v0A4AuXTwaAjPN322OC3HOm8RfWk6LixfqHuc2J8/Wlp7fgF68fwNbjQhor+/sWTKgx3EkrkGUdJpRFiiXrZ+RpRsyNP4DR1X7UV/hw+XmNiW0sxKEHNap+agl43fjo/HGyHlkoak+aJ8Mnhm4E70vb3zazeVTiy1kshKJxfGLh+JRtaj3vaJzqatyejcVibFuvZx6MxjFJQQTNSuntSIxiydR6EELw359tw3euT6YIs9g5AJR73bjj0km4QIyFu7M8qNLvPb9XmJEOhaJwEeChW+bhh8suAIBErQ5bPwhH47h6RhNW3L0EVQFj7VoDWT1/9QkRAe75548zQ2E8uf5YxhfViPFnqoNmeeV+r1s2YyIU0W6EDY3DyIKvx5VX+Vo9BCOxDMmDoMp7Ihqjpi32ArnrPXIxEo4pOiN+t7HPz8buk+cwEomlyCdIHZZQNI6RcAz3rtiBoXAs5b7KLkKXGfaJxilWbjuBgNedSEsGkjo7g6EIvrtyJ46fGdZ872Yj4HXh7QM9GYkhu072a+gN4MLeUwM4JdNfwGocb/x/9MJuAEDXQOoF9Hn0x/wZZi72yaVJ5sPz17vg63WToqrwjcbiiMRoorqW/b1q0y0Fz9+8a2NUdiQUjSneK8ygmn2N/m3lTgDAibMjiW3S3hPhaBy7O8/h6feOCeOQ3Ff/8oGpWDChBjelNUVJn8Gw3+kZSFbkp9+fL+04hcfXHcW5oHkdtNiC8ca0dM8qvwd9Kgu3WG+Crz231ZQxacHxxj/bl8nn1l+UFI7FMXe8cV0fRpnPlUX3xG7PX4j5h3Qs+HrdxeX5Mw+/zCf8nT+6cRYA9cqasXhcdXMQNRg1zrlkQFjlr9lhHybhHJekXEpnAaFoPMWxkRrtJVMbsOLuJfjPj8/F9bOTfXoXtiYre4Gk8R+JxHCzKCKYPus+N5I0xka76zH+aUkrgMyMolA0jtkt2fv3SmHrjfmQdna88c+Wi+s3EPYJR2OyvVL1EhCntekSyrtOnrPV8/d7XdjTeU53kVc0Tg11o7ITZixYqIT9zyS0c2F2zN+ocd509Kxi2MdnMKyUDSYbkS0EForEUv6mbM6M9H7LWPCV/F3sIZn+OdJCKjP6bADJa5J+zrTMjCt81qXY5sLxxj+bc+YzEKMWUjDNO7V1lcI0eVtHqvBWuc9tWwN3QChcYnF/7WEf64qIrKB3MPVBy6SVD5weVPX7sbjJMf/EgqwRTf/sv5sMK5nrgV4oFkzeuTS1FuUrV09LzK6DWTz/lPFJDHx6OFWar8+MvpI8tXlrcfLnTEsXP1ajY4YCsFYcb/yzef4+twtxqk/oSk8evBJsmjucNjUMReOYNU7d9NIMZjZX69ZFYfsXS+iHfaFb64X48pTGSriI+vFHY9bE/PV4iKwl6aLW+qz7WOX5l/ncmFhfjpmSvgQA8JWrz8OXrpyKaJymyBpnmzGzv3/G2OqM92ZLvgNsPyURPrMWfLNdk3BMvfGfO74GPrcrIZluJ6acBULItYSQfYSQdkLIcpn3/YSQP4nvbyCEtJpxXKPE4xQrNmcWTwHGyt31ZMMoweRfX5dI+nadC2IgGDXca1gL0htat+dfJIu+cg85tbPB/acHEI7FDckHpJNL4E+JXKJuQPLvXL5ihyENofTjrtx6Ekez1CYwz/msJJyZbcbMQlYVMh69TyYkpGR8zVDblR43/Zpodf6MZBYawfBZIIS4ATwM4DoAMwF8khCSrvV7B4CzlNKpAH4G4CdGj2sGpweS6VU/FBf0GEY8IT3ZMEqMrhaaQnRLUsp+/3ehm9aavfrE5/QgjaPqWfAFUvVaCpmgTK2GV2XV93dX7gIAnOgbybGneozIL4RyiLqlv2dWU/H2LuUQGav7kPb1zWaYF0+px/QxVbjmgjEZ70nXVlh4SKnyXW0mTi6SjdxljL+G73++pE/MsFCLALRTSg9RSsMAngGwLG2fZQAeF18/D+AqYlYepAGkK+y3Sqp7AWPGX0vMTw1etwsXT65LucnYg8DOPhCpXrC2LCNf0YV9Mtsw+lRWfTNVVzMnZUbkF3Lp+kg/X7q/UXJpEU1uEGL10nag2apul57XiFe+shT/vHSy4mfKef7/9fG5KftIJR6M4HG74HERhGPp4VhtWXhChXIRev4AxgE4Lvm5Q9wmuw+lNAqgH0D2AKRNnFTwzHKVu//6jXa0Ln8R//DLv+NY2rTWbM8fYD1LkzcZ86DNPo4SKYZQR54/YE0FqRUkPH/pbEdjEoCZmU1GJJdzKXoCqTM5s4x/Lglv9jdJM6iM6mHJGf/0v9tMh8mXZrijsTjiVNv3w4yaIj2YYTnkPPj006tmHxBC7iSEbCKEbOru7pb5FXNRaqKQK+b/wCtCe78dJ/rxyFupgl9CzM/c/Hu/J1XSd6qY4fCfaV6NlfjTDKEWfO7i9/yFWoXcloNNas38W1knNV1rUDlaOALCmCfUCV53yKRxM8//owvSfUGkjIeJrl1+XiPG1RpTwmX3qDSwcOHEWixsrUXzKCF8+umLJsr+rr7jpdYDsXOtLezjzovnb4bSVgcAqQBKC4CTWfbpIIR4AIwCkCFiTSl9FMCjANDW1mZ5QENpWurTsECZ3qDa7AVfgBVYJcfLimbmm1hMlnsM2XOtc1FsYZ9snr+a+4GZHTO1jAghoraSgQXfHOs0v/70Anzkl383zfNnzko2yXH2YO0fiWB8XRke/9wiw8eUWzBurinDc3ctNvzZcqTfE2rPdfpnFGvMfyOAaYSQSYQQH4BbAKxK22cVgNvE1zcBWEPz1bVYgpKOtpaY/1FJQ4d4nGpK9VJLwOtC72AY1/zsLXz9uW14/+hZuF0kZ8MIM5FWrJZ6ts9+MZ8/3fPX4nmbWeQFCFLHb+7TPiNWk+0DGNcPSmfnSSGcky2Uw7z0nsEwjp8xZ3HcTpVbQOjX/adNxxNZd3IzxlwMhaJYu6/b1AQBNRg+U2IM/x4AqwHsAfAspXQXIeQHhJAbxN1+D6CeENIO4KsAMtJB8wHz/G9ua8l4T0uqp3Sqqmfapwa/x42BUBT7Tg/gebFBtZ1pngJJY1auMTZrZacoK2APOqnhUuv5M2mPL185zdQxhaPxRJcrLahpKA6Yn+vPPqepWj7n3gpDLZUN/9ySSfiBqO5pNUy7SI/nf0iUa39y/VHzB6aAKQLrlNKXALyUtu0+yesggI+bcSwzYVP7HyyblfFerrCPtK2eNHlATXxVD+kyDgPBCC6b1pBlb6tIGh7WZ0Atxeb5D4ViGFdTlhI79rmJqrCV3+NCU5Vftb6LWq6dNQbb06q81aDW8zf7AT0SiaGh0q8g2WC+LpX0QXPfP6RnnFsHC9uoWVzPht5GPXpxdIXv3lNCPrOcoVbygiilKSEjaTMGtV80raR/UbZ19Oe1OYrWLkjsYXr7/2zMw4xFO9s7+lCR1oP5UPcQ3j3Yq/h7lFI8s/F4hkqsGejVm2KpiDmNv8kP6GBEnZKomQRsFDqUwuwBm5Xrcf48JqoCqMHRxj8gXiC5kgOlikrmGVWJBlCq9KhX/iAXCyfVZWzTEwIwwiViQ5HW+nLN5eheT/IcbzikbEALgTKfO0NpsVesRFXy0E5aqMtu9YKv2WGfYERZSVT6HUkvstSLkqaPFbDZN/P8H3lTyPzTMqthjWeMtpXUiqONf0ihu5FfYQrMnvL/evU0LD2vMaXBh1We/1KZEM/Fk+0tlRhV5sWR+6/HG9/4gOaF5pTl/byX9+VmJBzD3PHyYRulsEjEwrAWk9TWitpQhNlhn2AkrtxARmIg04ss9WL3gu+XrxLWddKTR7R8/z8lpp7anQnncOOfvQduNrlWICn36/e6UeZ1YdvxvkSuctLLMtcDkZudVNjsKRhBemOvyxE6KQSC0VjWEIJSTjZrzmEFWjRgzgUjWP7n7TjVH1Q9GzWrj+8L20/ilZ2nsGZvl3IPAQsMtd3CAeycpushadEPcruIUCls83pY8VgPC1Dqtak0BZZqg7B0vjNDYdRX+i3z/AHg9sWtaKktw2t7TiMUjWPBhFrTj2EV08ck1Rh/uaYdX/vQ+XkcTW5GwnEE0kIIF02qw4bDZxSN48FuIUVULoPMKH6xEjQepzlnXi/v6MQzG4+jptyHFjEbLZfxJ4SolrBQ4p6ntiRexxQyut0ughvnNWN2i/FalZ9/Yh42HLbfqWCzl/R7Qmsme3qxmB042/hHshdjJY2/XPtEJvdbDq+rES/tOJW4cOw9K6af37tBiA1+/jJlfZNCxOdxYdU9S3DDr97J91BUEYpkev6fWDheNP7ZC3JGIjFU+T144CbzK68T3bxicQRcyjPLYXEdKhiJaXJIzFaYvGpGk+L7P79lvinHuXH+ONw4X76S2ErYgnYwEkuZ3WrNafB73bZ7/jzsk2VqrxT/lOZNs5jm9/+6C197dhveO3Im5fc5SYzqttjFYCiKgVA00cKRwe6V//OnrVllj4ORuKKipBG0hGVW7zoFAHARoqn2RNCZMa/aNF/ZN3bB7omvPbct5Z7Q6vkLjW3srfJ1tucfjaO2PFvMP3vYR+rdMw9/9S5Ba786IJzSqoCjT60s4+vkFRsLjd0nzwEAqgKpGU3sWm8+1od1h3rxgfMzvdpgJJbx0DAL9lAR7j/lbCvW8rO23Kup8MiMsI8UK9I5C4mGymQzeqlczIKJ2kKyfi8P+9iKUrtFVuEpa/wjSU8qPZuBydMyDX5OkoDXjcumNWAgaF/rST2wh3tb2hdYjSELyoSLzEKL/ALTFYpRimiMwkXU5ZGbHfYpdc9fek6DYeG8PXjTHMUsJzmM9AzXS2k/lnOgFPZhWQMPrWnPeO/BvwmKnlLPPx27c3aLhXzc5FphaXvpX2DpveKSySoJRmJ4eecpzV98tST77OY+f4e6BcmAn792AIOhqOowpNnywqXu+Uvp6BOk3fXUGuit4TCCc66MDHqarkRj8cQXy+91Z43v2p1vXCykq5MWIix2m34NpT/LPcBY5yqrZn1qPf/0PgJbjp1VXXRkdtin1D1/ILmW0i1Wdev5m/0evuBrK3rUN6VPZyXPvwAalRUk+WpcoQV2jTM8f4kXK7fgy7bdttg8vXgpavv4BtMerpEY1eT5G/FA0xc6nZD48F1RQ6hnUFhnmZClG5kS+VjwLf0rk4UzQ2H0DUdUZWZItWjSjb+SdgknE7M9SyvI7vkn7xU5A5ktXGQW7PhPvHtEcb+1e1Nln3d3nlOtMqk35j8YimLhj1/DpHtT9B0TfSdKGZbF9sMXdgPQV3zp97qw8chZW3WvHGu5tonqiE1V8nKzgPw0W/p09nvcaKktx7SmSotGWXqYvaBoBUlN9lQj3lwTwOxxo8R9Mr00lu1hVUoryyBbseWE4n7/8tTmjG3nguqalvt1zsxe33M6EfaQMs/GZkP5In0hXcmmZIOtESq1ljUbxxp/lrFzyZTs+jjLr5sOIHWKLy3t97oJAl43nrnzYotGWXrko5JRK0qe/5N3XCTuI+f5i92/LJoNjtG4lnC1pMBqrsoqWr0zMzmPtTrgQU25T2bv0mJ8WutJLdIOjA/PHgtAmEHZhXONfyJXP7fwFDNWP311Px5acyDxPovr260kWMwUg+f/P+8cASC/aM/i/q/s7Mx4j0k7WBX2kYYT+obDOfeXjkPtA0nv9ZFb4qrWqPxarJhxvdm17TjLPX/LYTe40oIv+8KwB8VDrx/Ais3ClHtyY1INVDrN/8f54/DQJ80pWS9FfB4XonGakZFSKFBK0TMYQsDrkl20Z/eLXP9nt7h/Q6X2ab8ayiVOxoOr92Xdb/qYKgCpjcrVGii9C/IkTap1WlMlfmSSTHOhI/3+f+FyfdIrLKQ3HLbP83dsMroapUPm+Qcj8Ywshu/fkGwPJzUSP/vEPDOHWXJIZTNy6dPkg0hMuM73fGCq7PuEEFw3a0zCy5cSisbhcRHLPH/pfaYk/1tT7sWi1roUuXLVxl9n2EfarP7I/ddr/v1iRnpu771uhq7PYOsESoqxZsM9f4WwD/P8712xPdGhh2FFCzonkAil2XiTa4GlSSrr0MuvWyhJhJsFWxjMNnGilGL9oTMI+NwpoR61i9A+j0uXrnyh125YCTu3ehZ6Gex+S0/TtRLu+SvEQlmxzuZjfdh8LLV3avqX/JvXno9g2LlfALUkctVjufVp8gG7hrmakMg9vAS5EGudgvs+MhPf/PN2TMiik8QWol0EqJZoE2UToktHb54/O65dDdMLiaqAB8vmNePjF47X/RnsATJiow1xrOfPPBWl/GclIbL0h8bdV0zFVwtco74Q8Bd4I3c1ufqCCJeM1LeOinGtXDd7DIDsnjwb1+XnNcLlIvjYAqGvwIyx1bL7p6N3wZcd9+Y2/QawWHG5CH5xy3xcKtNtTy0Jz5+HfawnHI3D6yaKTTEqFLJ4eNhHH2b3iTWbPacERc9cvWfzFfZh992hnkF8/vFNuPX3G/AXSd5/KEs4U23M3+8W/jatxUabxQ5mXNZEH24XASHA2n1dth3T0WGfXAZcSQWxuYardupBizhZPjh+RhDnkmZzpZOtl66ae8ooXjeBiwBPv3c8sW3Lsb5EIxMWjko3wh6VPZdZ562ugSDGjirLsXcS1tGOy5roh1J7K6Id+5gOR7N38ZLyoZmjZbdzz18fhe75sw5YUxWqtn0eF2JximjawmhI5T1lBEJIxr0nLQxK1K/oLDQ7b7SQJjoU0hZ7DkXjmNMi3/Ceo47rZo1RvTZjBo40/qf6g/jj+qOJhhdKuFV6TBx1MOPY2R/M80jkeX1vF9wuolilybzqrz+3LSUF2I6wD5DsNSGldfmLuOZnbyUeXuwBkXgYqbyNWTbRyq3KEhLp2LHeUeoEvG7Z+hGrcGTY530xPqmG9HLrT180QdEr5CjDKhkLVdkzFs+s6UiHGbm/bD2J798wC6PKhayaYCRui9BftgfTvtMDONI7lDLGzy1pxelzQVxxfqOqz54teu+sm5lawrF40bTpLFQCXjdf8LUaLT1Km9Pinj/+x9n4pyWTzB6SY2gUq1/tnN5qIRSJ49pZYxT3kaZzhiT3klJzIDO56cKWrO8xZ4V5/NNGV+Gx2xeiqUrdGlVTVQAXTqzVnG8eisYcId9sJWVet63p4o68WloKjGIOkKS1E6aDVIjG/8XtnTjQNZhT/lga3pDeSztPnLMl9KEkGfzImwcBGMu6KfO6Neebh6M87GOUgNeFgVDUtoI5R4Z9WMjh7ium5Nz3m9eej2gsDpeLOEKe1mrKvIVr/JkUcq6p94WS3r7S8FW5z40hGzw3pWWo42cEYTAjM5CA14UzQ9rCDyFu/A3DrtnR3uHEwruVONL4M2/tiyqMf1NVAD+/hQu1mUUgUclYmDF/IFWnRo6J9RV45DMLcNeTm1M8/0gsjgua1RVTGWFqU6phuHrGaLy253TKNiO9c4XYs8awT8T6NNdSZ854Yb3FLllnQ49qQkgdIeRVQsgB8f/aLPu9QgjpI4S8YOR4ZqFGzpljDSx76q/bT+Z5JNk5q0IuOSn3LdxL0VgckRi1ZdGzMi3sU1OeKZOhtnOXHAGvG4d6hjT9TjhmfZprqcOu65v7unPsaQ5Gr9ZyAK9TSqcBeF38WY4HAdxq8FimwXLM5VLmONbjdZOCFgJTU4OQXq8QTPT9td4AXtBcjaumN2H5ddPxhcsny85gx4zSX4TI/iYtstuhiD1prqXM+aIUd2e/PZr+Rq/WMgCPi68fB3Cj3E6U0tcBDBg8lmmw+CSvRswPH1vQUrBFXkDusA+QWanMFkjt8PxrK3z4/e0LcdflU3DvdTMwpbESL3zp0pR99HSTYswaJ4SuhjWEfgRRO278jVAd8KK1vty2dE+jV2s0pbQTAMT/m3Lsrwgh5E5CyCZCyKbubuumPnxxKr8UYitHaW5/rjx/IBn2YQ+x7WJPaKtVPbMh1eJhjUH0Uu4Tfv/J9UdVHzsap/C5eRjVKHYWeuW0gISQ1wghO2X+LTN7MJTSRymlbZTStsZGdUUpehDK8PmNmi/0LChaTd9wssH5L1V0YvOlef77TgsT23xJHFzQXJ2I8//m0xca+qyLJwt9rd9QKTKW6I3BPX/D2PndyOkiUEqvzvYeIeQ0IWQspbSTEDIWgH2SdAawqwyfIw/z/CmlBRN6GxLb5z1w0xxMU5Fmlwz7CF/U4VAMLgKcb0OKnhwetwv7f3ydKZ81takSV5zfqEr+BFDXEpWjjjIbjb/Rq7UKwG3i69sArDT4ebbAC1Lyi9/rBqWFJfFwStQaqvCpC5mkL/iuO9SLgNddMA8zo1T4PNje0a+qUXz/iDBr4tk+xgl4Xdh4RL38jBGMXq37AXyQEHIAwAfFn0EIaSOE/I7tRAh5G8BzAK4ihHQQQq4xeFxD2KG+yMkOe/DaqWOSC2bAyv0qde/Twj5uQlQtFBcLE+qFRka7O3Nr/PQOhQCol43mZIfdQWrWnYxiyAJSSnsppVdRSqeJ/58Rt2+ilH5est9llNJGSmkZpbSFUrra6MCNELah3R4nO+zcF1K6JyusGV+bvXubFPY3JFM9Y1g8pd6aweUBJmWuRgqFKYlOauCCh0a5aJJwD9nhGDnS/d15oj/RTpBjPwHRa77/pb15HonAwe5B/OszWwEAFSo9f7a4+tYBISttJBwrKVXL9CI2JQ51DwIQ5C04xmD33/7T1mfGO9IClvvdiQU+jv2wkNuKLdo0461i3cFeAMClUxtUq1+yAkGWYhmMxlS3SiwGWLGaGg80EhPOQWOV39IxOYGWWkFFmElzW4kjjf9IOIa5XKQtb0ysz94iMR+w7IqHP71AdfMeQggundqQ+N2RcLzEjL96z5/JP48qy5SZ4GiDCbppUR7Wi+OMfzxO0TMYVmzOzrGWQjv3zIBrDdsIBTlxDIWi6BkM2SLtYBdaei0HE93DSufvzxdawm1GcdzVYlkdpZSZUWyMlujOaNGPsYpgJA4X0a71FPC6EIrEcKBLiHnXlvusGF5eCGiQ3g5GhQ5mpZLmmk9YoZwdFfCOM/7spE5ryk8xDgeokqhSFkKu/0hEWKzVarzKxFL8YTFTaNGkOiuGlxcSnr+K8EOpLXbnEy0zLqM40PjzKWq+kRpZO2Kbudh1sl9X3UfA60ZnfxA7TvQDUF8gVgx4xGymPady5/mvP9TL62ZMgmWRceNvASGuQ1IQ1FUIIRKtvWKtoNzn0dWBi4UOmQzC5MbCWsg2AzWL2D6Pq6BVWosJQogof8Jj/qaT1CHh09R88q0PzwBQGJ5/MBLDnHHaBdlmjhVChz2DYfg8LsXeusXI1KZKVdcnGIlh8ZQGG0bkDPweF8/2sQIe9ikMWGbMm/vzrwU4EoklGstrgRn7P2/uKEnPt0ylvHAwUlpprvnG73XzsI8VsCcqN/75hTVBbxczZfJJUGf/2VJa4JUj4HWpy/aJxEoqzTXf8LCPRSRj/txTySdjR5VhdLW/IMTdgjo9/xaVOkDFitrGIixbimMOdjU7cpzxv3fFDgDc8y8E1IYVrOL59zvwmd9twOGeoYTeECdJwOvG9o5+xX1icYrhcGlJW+Qbn8eNF7d3Wn6c0lqhUsGpc4JuOzf++SffHb2+/ty2xOvzdDZhue2SiXh83VF8ctEEs4ZVMISi8ZwefdeA8H2iyH+xXqnwyUXjcdCGcKijjL80jsbDPvnHzn6lubhqhr72099fNgvfXzbL5NEUBrPHVeOd9h7FjmtDYoGb3ocnJ5PPXtJqy3EcZfxP9gUTr31c0jnvBLwuvNPeY+sxd3T044cv7sZ7h8+kbK8ssTRNM6jwexCLU3ScHcH4Ovn1jf2nBQ+Vn7/iw1EW8PiZYQCC7nh9RenosBQrlCYrSe3i7fbuDMN/c1sLGiq5HHE642oEeeG9p7Jry7PG92NHldkyJo55OMr4synqn7+4GC7eci7vXDy5HuFo3FZxt6FQFNJLP6dlFB64aS6/H2SYLRa+se+NHCxs11yjrg8Cp3BwlPHnU9TCgl2HznPBHHuaww2/+jseXnswJTPFy8N/WWHX53t/3ZW1p+zDa9sBqJOB4BQWjrrzWUbC6GrupRQCo8qF5h/HeoctP1YoGkukLd7cNh7f+vB0AMBXrp5m+bGLlYZKPyp8bvQNR7JqHzFdI549V3w4ygUeCkVR5nVzBcICYfoYIUNEKaxgFkOhpPH69vUz4HW7cOfSKZYft5hxuQju/fAMfOcvOzEUiirOmLmWf/HhKCu4vaOfN5kuIJg2zpbjZ1O2r93XhdblL2Lhj1/D2r3maP+wfHSAh3q0wAz+qf7M0FwhNOLh6MdR38+GyhcAAAw1SURBVIIyX+HklXOACWL6YLoo2h1/2AgA6B4I4Z/E10Y5OyRkpbTWl7Ykg9kwZ4l1wJMit41TPDjK+I+EY4kMBk7+8bpdaKj0YzCU+kC2wqFkoaWHPjnf/A8vYZh+kVxoblDc9sBNc2wdE8ccHBXzD0bjqCnz5nsYHAkVfje2He9L/ByTsfyUUnzlT1txpGcIt17SipsubMn5uZuPncVHf/0uFrbW4oa5zfi3lbvE4znqljcMC/vsOz2A62aPTXlv3aFeAKXVwcxJOOqqBcMxlPFMn4IiFImjoTK5DtNxNjPzZygcw8qtJwEATbtOqTL+P3xhNwBg45Gz2HgkuaYwIUulKkeepmqh+M0ts6DLiiYXttbaOiaOOTgr7MN1xwuOOS2jUkIKbkmx1ccWtKB5VCDlfbVCcP3D8vFovtirjYDXDbeLyLbbHAnHUOFzo4k7VEWJY74J24734diZYV6MUmBU+j3Ye2oAD7yyFwBwz1NbJO+5cbI/iJ0nkrLCbx/owbmg8kJjMBLDoZ4hawbsQAJpbQXjcYpvPLcNG4+e5d+nIsYxxv/dg0J88oMzR+d5JBwpN84fB0BohQgAW8X4/5eunIpaUX9pQ5oWzz4FrRkA6Dg7Irv93uumGxqrU/F73Sme/+mBIJ57vwPbjvdx41/EGDL+hJA6QsirhJAD4v8ZwT9CyDxCyDpCyC5CyHZCyCeMHFMvTNPlyun6pHs51rD0vEbcvrgVw2kVpDe3jU+0euwZDKW8lyv0w8JEv7+tDf/92TYAQlz6C5fzoi49pDcUj0STi/I8jFq8GL1yywG8TimdBuB18ed0hgF8llJ6AYBrAfycEFJj8Liaeeydw/B73LwSsQCp8LsxEIzi2U3HJds8iUYiKzafSNn/+fc7FD9vd+e5xGdU+IXP4HpO+gl43QiKtRg//ds+LH1wbcp7nOLEqPFfBuBx8fXjAG5M34FSup9SekB8fRJAF4BGg8fVBKVCq7n6Si7jXIjMHy94+L998yAA4ILmatSWe1MMy+2LW/H45xYBAHoHw4qfNyCuCUxuqMDscaPw0fnj8NUPnm/F0B2B4PkLs62H1rSnvGe3JDfHPIy6Q6MppZ0AQCntJIQoxlQIIYsA+AAcNHhcTbBmyKXYaq8UWNhaBwDoFUXCVv7LEhBCEsZ/THUA37vhAgDAkqn1Oau0WdFYQ6UfLhfBTz8xz6qhOwK/xPNP5yoeRi1achp/QshrAMbIvPVtLQcihIwF8EcAt1FKZe8kQsidAO4EgAkTzDHUz206jn9/aQ8APvUvVPxi3Jg1BmHeJFOKHFebbBQS8LgT+2VjR0cf/B4X1+g3Cb/Hhbf2d6N1+Yuy73GKk5zWkFJ6dbb3CCGnCSFjRa9/LISQjtx+1QBeBPAdSul6hWM9CuBRAGhrazOlyH/1rtM4KxoLruZZmPg9LhAidPaS0lxThlsvnohbFo1PbAuo0Gfye9yIctEx05g9blRG9zMA+MicsbiaZ88VLUat4SoAt4mvbwOwMn0HQogPwP8D8ASl9DmDx9OMtECotb7C7sNzVEAIQcCTuXDodhH88MZZuKA5qccU8LhTMk/kGApHMaeFaziZhVx69OIp9fjVpxZgSmNlHkbEMQOjxv9+AB8khBwA8EHxZxBC2gghvxP3uRnAUgC3E0K2iv8sD8JSSvGlp7ck8sYBHvYpZNTOysp8LpzoG0lIC8jx9oEefq1NJC7TxauO98Auegx9QyilvQCuktm+CcDnxddPAnjSyHH0EIrG8ddtJzGtqRIXNFfD7XJh2mjupRQqX7xiCu5/eS++fJVyZy02C9je0Y/xWXR6XETeYHH0cdGkelw6tQF/b+/BxPpyLGqtw90fmJrvYXEMUrLuEQv3fPqiCbh9yaQ8j4aTi7sun4K7VBRhXTq1AUD27l+UUsQpcOHEOlPH52TcLoInP39RvofBMZmSXQH9zl92AuASvqUGC+ekSz4wWFovrzzlcJQp2W/IO+09AIAloqfIKQ1qxKbvFPJhHbYY7JdZQOZwOElK1vgHI3F88YopaK4py70zp2gghGD6mKqsYZ+QKEDGPX8OR5mS+4b0DYcx+d4XEY7FUcGbtZck5T43Vu86LavxE+SeP4ejipIz/i4XSfSAndlcnd/BcCzhtsWtAIDX95zOeI97/hyOOkruG1IdSPbondtiu3goxwaWzRuHeeNrEg3EpXDPn8NRR0mmwowdFUBnf5Bn+pQwFX433j7Qg3tXbMcb+7rR2R/ExZPrEIkJ0z6uOcPhKFOS1vFXn5qPNXu7uNZ4CVPhE27dp99L9gBYf+hMogfA3PF81sfhKFGS7tGFE+vwjWt4y75S5rLz5FtCjERi+NKVUzGqzCv7PofDEShJ488pfZQyucp9JTmh5XBMhRt/TlEiLd7zulN1+7miJ4eTG278OUXJ6OpA4vWar12B1V9Zmvh51jhu/DmcXPD5Mado8bldQjFfWlYXL+7jcHLDjT+naPntrRdi3aFe1JZ7UVPmxe2LW1Fd5uVNxTkcFRBaoLrnbW1tdNOmTfkeBofD4RQVhJD3KaVtufbjLhKHw+E4EG78ORwOx4Fw48/hcDgOhBt/DofDcSDc+HM4HI4D4cafw+FwHAg3/hwOh+NAuPHncDgcB1KwRV6EkG4ARw18RAOAHpOGU8rw86QOfp7Uw8+VOqw6TxMppfKa5xIK1vgbhRCySU2Vm9Ph50kd/Dyph58rdeT7PPGwD4fD4TgQbvw5HA7HgZSy8X803wMoEvh5Ugc/T+rh50odeT1PJRvz53A4HE52Stnz53A4HE4WSs74E0KuJYTsI4S0E0KW53s8+YAQ8hghpIsQslOyrY4Q8ioh5ID4f624nRBCHhLP13ZCyALJ79wm7n+AEHJbPv4WKyGEjCeErCWE7CGE7CKE/Ku4nZ8rCYSQACHkPULINvE8fV/cPokQskH8m/9ECPGJ2/3iz+3i+62Sz7pX3L6PEHJNfv4iayGEuAkhWwghL4g/F+Z5opSWzD8AbgAHAUwG4AOwDcDMfI8rD+dhKYAFAHZKtj0AYLn4ejmAn4ivPwzgZQAEwMUANojb6wAcEv+vFV/X5vtvM/k8jQWwQHxdBWA/gJn8XGWcJwKgUnztBbBB/PufBXCLuP0RAF8UX98N4BHx9S0A/iS+nil+J/0AJonfVXe+/z4LztdXATwF4AXx54I8T6Xm+S8C0E4pPUQpDQN4BsCyPI/JdiilbwE4k7Z5GYDHxdePA7hRsv0JKrAeQA0hZCyAawC8Sik9Qyk9C+BVANdaP3r7oJR2Uko3i68HAOwBMA78XKUg/r2D4o9e8R8FcCWA58Xt6eeJnb/nAVxFCCHi9mcopSFK6WEA7RC+syUDIaQFwPUAfif+TFCg56nUjP84AMclP3eI2zjAaEppJyAYPQBN4vZs58xR51Kccs+H4NXyc5WGGMrYCqALwsPtIIA+SmlU3EX6NyfOh/h+P4B6OOA8Afg5gG8CiIs/16NAz1OpGX8is42nMymT7Zw55lwSQioB/BnAVyil55R2ldnmiHNFKY1RSucBaIHghc6Q203835HniRDyEQBdlNL3pZtldi2I81Rqxr8DwHjJzy0ATuZpLIXGaTFEAfH/LnF7tnPmiHNJCPFCMPz/SyldIW7m5yoLlNI+AG9AiPnXEEI84lvSvzlxPsT3R0EIQ5b6eVoC4AZCyBEIIecrIcwECvI8lZrx3whgmri67oOwiLIqz2MqFFYBYFkotwFYKdn+WTGT5WIA/WKoYzWADxFCasVslw+J20oGMb76ewB7KKU/lbzFz5UEQkgjIaRGfF0G4GoI6yNrAdwk7pZ+ntj5uwnAGiqsZK4CcIuY5TIJwDQA79nzV1gPpfReSmkLpbQVgu1ZQyn9NAr1POV7ZdzsfxAyMvZDiEl+O9/jydM5eBpAJ4AIBC/iDgixxNcBHBD/rxP3JQAeFs/XDgBtks/5HITFpnYA/5Tvv8uC83QphOn0dgBbxX8f5ucq4zzNAbBFPE87Adwnbp8MwSi1A3gOgF/cHhB/bhffnyz5rG+L528fgOvy/bdZeM6uQDLbpyDPE6/w5XA4HAdSamEfDofD4aiAG38Oh8NxINz4czgcjgPhxp/D4XAcCDf+HA6H40C48edwOBwHwo0/h8PhOBBu/DkcDseB/H/bThcbwH/hvQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c31de05c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LABEL [[[-0.109375 -0.09375  -0.125    ...  0.375     0.34375   0.328125]]]\n",
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): 0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0])\n",
    "plt.show()\n",
    "print('LABEL', targets)\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### WaveNet parsing function\n",
    "- Input should be between -1 and 1\n",
    "- Output should be between 0 and 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _wavenet_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t    'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t    'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t    'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t    'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = []\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn (audio_inputs, audio_quant), label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_wavenet_parse_function)\n",
    "# dataset = dataset.repeat()\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c31e9e940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c31f50fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): []\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(data[1][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parsing function for conditioned WaveNet\n",
    "- Input 1 should be input of conditioning section of network between -1 and 1. \n",
    "\n",
    "- Input 2 should be input of current example between -1 and 1. \n",
    "\n",
    "- Output should be current example between 0 and 255. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _wavenet_test_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t    'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t    'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t    'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t    'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = tf.cast(parsed_features['label'], tf.int64)\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn (audio_inputs, audio_quant), label\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_wavenet_test_parse_function)\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c31f8ae80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c31f5a4a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): 0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(data[1][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _wavenetcond_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = [] # No labels being passed in training, just a placeholder.\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\t# There can only be one input so they need to be in a list\n",
    "\treturn (audio_inputs, audio_inputs, audio_quant), label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_wavenetcond_parse_function)\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c3214f8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c321f7c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c3220f208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): []\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,UklGRiQgAABXQVZFZm10IBAAAAABAAEARKwAAIhYAQACABAAZGF0YQAgAADUktSS1JLUktSS1JIQjxCPEI/UktSS1JLUktSS1JKYlpiWmJaYlpiWmJaYlpiWmJaYlpiWmJZcmlyaXJpcmlyaXJpcmlyaXJpcmlyaXJpcmlyaXJpcmlyaXJofnh+eH54fnh+eH54fnh+eH54fnh+eH54fnlyaXJpcmlyaXJpcmlyaH54fnh+e46HjoeOh46HjoeOh46Gnpaelp6Wnpaelp6Wnpaelp6VrqWupLq0urfKw8rDysLa0trS2tLa0erh6uHq4erh6uHq4trS2tLa0erg9vD28AcDFw8XDiMeIx4jHTMsQz9TS1NKX1lvaW9pb2lvaW9qX1pfWl9aX1lvaW9of3uPhpuVq6Wrpaulq6WrpLu3y8LX0efh5+Hn4Pfw9/AAAAADDA8MDwwPDAwAAPfw9/D38AACHB0sLSwuHB4cHwwPDA8MDwwPDA4cHSwsOD9IS0hIODw4PDg8ODw4PDg/SEpYWWhpaGloaHR4dHh0eHR4dHh0eHR7hIeEhpSWlJaUlpSXhIaUlpSWlJWkpaSlpKWkpaSlpKWkpaSmlJeEhHR4dHloalhaWFtIS0hLSEtIS0hLSEpYWlhaWFpYWlhaWFpYWlhbSEtIS0hLSEtIS0hIOD0sLhwfDAwAAefi19PLw8vAu7S7tauku7S7tLu0u7Wrpaumm5ePh4+Hj4R/eH94f3h/eH94f3h/eW9pb2lval9Zb2lvaH94f3uPh4+Hj4R/eH94f3h/eH94f3lvaW9pb2pfWl9aX1pfWl9Zb2lvaH94f3h/eH94f3h/e4+Hj4ablpuWm5ablpuXj4ePh4+Hj4ePh4+Ef3lvaW9pb2lvaW9of3h/eH94f3h/e4+Gm5WrpLu0u7Wrpaulq6WrppuWm5Wrpauku7fLwtfS19Hn4Pfw9/AAAAAAAAAAAPfw9/AAAAAAAAAAAAAAAAD38Pfx5+Hn4efg9/D38PfwAAAAAAAAAAD38Pfw9/Hn4efi19LX0tfS19PLw8vAu7Wrpaulq6Wrpaulq6S7t8vDy8C7tLu0u7S7taulq6ablpuXj4ePh4+Hj4ablauku7S7t8vDy8PLwLu0u7S7tLu0u7S7tLu0u7S7tLu0u7S7taumm5ablpuWm5Wrpauku7S7tLu3y8PLw8vC19Hn4PfwAAMMDhweHB0sLSwtLCw4PDg/SEtISlhaWFpYWlhaWFloaWhpaGh0e4SGlJaUlaSlpKWkpaSlpKWkpaSlpKSwtLC0sLfAw8DDwMPAw8DDwMPAw8DDwMPAw8DDwMPAw8DDwMLQ0tDS0NHg4eDh4OLQ0tDS0NHg4eDg7PP8/w0PDQ8NDw0PDQ4ZHhkdKS0pLDk8OTw5PDk8OT0pLSksOTw5PDk/SUtJS0lLSUtJSlVaVVllaHV4dXuFh4WHhYeFh4WHhYeFh4WHhYeFhpGWkZaRlpGWkZaRlpGWkZaRlpGVoaWhpaGloaWhpaGloaWhpaGloaWhpaGloaaRlpGWkZaRlpGWkZaRlpGVoaWhpaGloaSxtLG0sbSxtLG0sbWhpaGloaWhpaGloaWhpaGloaWhpaGloaWhpaGloaSxtLG0sbSxt8HDwcPBw8HDwcPBw8HAsbSxtLG0sbSxtLG0sbSxtLG0sbWhpaGloaWhpaGloaWhpaGloaaRlpGXhYeFh4WHhYR1eHV4dXh1eHV4dXllaWVpZWllaWVpZWllaWVpZWllaWVpZWllaWVqVVpVWlVbSUtJS0lIOTw5PDk8OTw5PSktKS0pLhkeGR4ZHw0PDQ8NDw0PDQ8NDhkeGR8NDw0PDQ/8/Ozx4OLQ0tDTwMPAwLC0sLWkpaSlpKWkpaSlpKWkpaSmlJaUl4SHhIR0eWhqWFtIS0hIODw4PDg8ODw4PDg8OD0sLSwuHB4cHhweHB8MDwwPDA8MDwwPDAwAAPfx5+Hn4efh5+Hn4efh5+Hn4tfS19LX0efh5+Hn4efh5+Hn4Pfw9/D38Pfw9/Hn4efh5+Hn4Pfw9/Hn4efh5+LX0tfS19LX0tfS19LX0tfS19LX0tfS19LX0tfS19LX0efh5+Hn4efg9/D38Pfw9/AAAwwPDA8MDAAAAAMMDwwOHB4cHSwsODw4P0hLSEtIS0hLSEtIS0hLSEtIS0hLSEpYWlhZaGloaWhodHh0eHR7hIeEh4SHhIeEh4SGlJaUlaSlpKaUl4SHhIeEh4SHhIeEhpSWlJaUlpSWlJaUl4SHhIeEh4SHhIR0eHR4dHuEh4SGlJaUlpSWlJaUlpSXhIeEhpSXhIeEh4SEdHuEhpSWlJWkpLC0sLWkpaSmlJaUlpSVpKWkpaSlpKWkpLC3wMLQ0eDh4ODs8Ozw7PDs8/z//P8NDw0PDQ8NDw0PDQ8NDw0OGR4ZHhkeGR0pLSksOTw5P0lLSUtJS0lLSUtJSlVaVVpVWlVaVVpVWlVbSUtJSlVaVVpVWlVZZWllaHV4dXh1eHV4dXuFh4WHhYeFh4WHhYeFh4WGkZaRlaGloaWhpaGloaWhpaGloaSxtLG0sbSxtLG3wcPBw8HDwcLN0s3SzdLN0s3R3eHd4d3h3eHd4d3h3eHd4d3h3eHd4d3h3eHd4O3w7fDt8O3w7fDt8O3z/f/9//3//f/9//387fDt8O3z/f/9/O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3w7fDt8O3x3eHd4d3h3eHd4d3h3eHd4s3SzdLN0s3R3eHd4d3h3eHd4s3SzdHd4d3h3eHd4s3SzdLN0s3SzdLN0s3SzdPBw8HDwcPBw8HDwcPBwLG0sbSxtLG0sbSxtLG0sbWhpaGloaWhpaGloaWhpaGloaWhpaGloaWhpaGloaWhpaGmkZaRlpGWkZaRlpGWkZeFhHV4dXh1eHV4dXh1e4WHhYeFh4WEdXh1eWVpZWllalVaVVpVWWVpZWllalVaVVpVWlVaVVpVWlVaVVpVW0lLSUtJS0lIOTw5PSkuGR4ZHw0PDQ8NDw0P/P/8/Ozx4OLQ0tDS0NPAw8DAsLWkpaSmlJaUl4SHhIeEhHR4dHloaWhqWFtISDg8OD0sLhweHB8MDAAA9/Hn4efh5+D38AADDA8MDAAB5+Hn4efh5+D38Pfw9/AAAAADDA0sLDg/SEpYWlhZaGloaWhpaGloaWhpaGpYWlhaWFpYWWhpaGh0eHR7hIeEhpSWlJWkpLC0sLSwt8DDwMPAw8DDwMPAwtDS0NLQ0tDS0NLQ0tDS0NLQ0tDS0NLQ0tDS0NLQ08DDwMPAw8DDwMPAw8DDwMPAw8DC0NLQ0eDh4OHg4eDh4OHg4eDh4ODs8Ozw7PDs8Ozw7PDs8/z//P/8/w0P/P/8//z//Pzs8Ozw7PDs8Ozw7PDs8/z//P/8//z//P/8/w0PDQ8ND/z//P/8/Ozw7PHg4eDh4OHg4eDh4OHg4Ozw7PDs8Ozw7PHg4eDh4OHg4tDS0NPAw8DDwMPAwtDS0NLQ0tDS0NLQ0tDS0NPAw8DDwMPAw8DDwMPAw8DDwMPAw8DAsLSwt8DDwMPAw8DAsLWkpaSlpKWkpaSmlJaUl4SEdHloaWhpaGloaWhpaGloaWhodHh0eHR4dHloalhbSEtISDg8OD0sLhwfDAwAAefi19LX0efh5+LX0tfS19LX0tfS19Hn4efi19PLw8vDy8PLw8vC19PLwLu1q6ePh4+Ef3h/eW9pb2pfWl9bU0tTS1NLU0tTSEM9My4jHiMeIx8XDxcPFwwHAAcABwMXDxcOIx8XDxcMBwD28Pbw9vD28Pbw9vHq4eri2tLa0trTysPKw8rDysC6tLq0urS6ta6lrqWupp6Wnpaelp6Wnpaelp6Wnpaelp6Wnpaelp6Wnpael46HjoeOh46HjoeOh46Efnh+eXJpcmlyaXJpcmlyaXJpcmlyaXJpcmlyaXJofnh+eXJofnh+eH57joeOh46HjoeOh46Hjoaelp6Wnpaelp6XjoeOh46Gnpaelp6WnpWupp6WnpWupa6lrqS6tLq0urWupa6lrqWupLq0urS6tLq0urWupa6lrqWupa6lrqWupa6lrqWupa6kurWupa6lrqaelp6Wnpaelp6XjoeOhH54fnuOh46HjoeOh46HjoeOh46HjoeOh46HjoeOh46HjoeOh46HjoeOh46Hjoaelp6Wnpaela6lrqWupa6lrqWupa6lrqWupa6lrqWupa6lrqWupa6lrqWupp6WnpeOhp6Wnpaelp6XjoeOh46HjoeOh46HjoR+eH55cmlyaXJpcmlyaXJqYlpiWmJaYltSS1JLUktSS1JLUktSS1JLUktSS1JLUktSS1JIQjxCPEI8QjxCPTYtNi02LTYtNi02LTYtNi02LiYeJh4mHiYeJh02LTYtNi02LTYsQjxCPEI9Ni02LTYtNi02LTYtNi02LTYtNi02LTYsQjxCPEI8QjxCPEI8QjxCPEI8Qj02LTYsQjxCPEI8QjxCPEI8Qj02LTYtNi02LiYeJh4mHiYeJh4mHiYeJh4mHiYdNi02LTYtNi02LTYsQjxCPEI8QjxCPEI8QjxCPEI8QjxCP1JLUkpiWXJpcmlyaH54fnh+e46HjoeOh46Gnpaelp6Wnpaelp6Wnpaelp6VrqWupa6mnpaelp6Wnpaela6mnpaelp6Wnpaelp6Wnpaelp6Wnpaelp6Wnpaela6mnpaelp6Wnpaelp6XjoeOh46HjoeOhp6Wnpaelp6Wnpaelp6Wnpaela6lrqWupa6lrqaelp6Wnpaela6lrqWupa6kurS6tLq3ysPKw8rAurS6tLq0urS6t8rDysPKw8rDysPKw8rDysPKw8rDysPKw8rC2tLa0trS2tLa0trS2tLa0trS2tLa0trS2tLa0trS2tLa0trS2tLa0trS2tLa08rDysPKwtrS2tLa08rDysPKw8rDysPKw8rC2tLa0erh6uHq4eri2tLa0trS2tLa0trS2tLa0trS2tLa0trS2tLa0erh6uHq4erh6uHq4erg9vD28AcABwMXDxcOIx4jHiMeIx4jHiMeIx4jHiMeIx4jHiMeIx4jHTMtMy0zLEM8QzxDPEM/U0pfWl9aX1pfWl9bU0tTS1NIQzxDP1NLU0pfWW9pb2lvaW9of3h/eH94f3h/e4+Hj4ablpuWm5ablpuWm5ablaulq6S7t8vDy8PLw8vC19LX0efh5+D38Pfw9/AAAAADDA4cHhwdLC0sLDg8ODw4PDg8OD9IS0hKWFpYWlhaWFpYWlhaWFpYWlhbSEtISlhZaGloaHR4dHh0eWhpaGloaWhpaGloaWhqWFpYWlhaWFloaWhqWFtISDg8OD0sLDg8ODw4PDg9LC0sLhweHB4cHhweHB8MDPfw9/Hn4tfR5+Hn4tfS19PLwLu1q6Wrpauku7S7tLu0u7S7tLu3y8PLw8vAu7Wrpaumm5ablpuWm5abl4+Ef3lval9aX1pfWl9aX1pfWl9aX1pfWW9pb2lvaW9pb2pfW1NIQz0zLTMuIx4jHiMeIx8XDxcMBwAHAPbw9vD28Pbw9vD28Pbx6uHq4erh6uHq4Pbx6uHq4erh6uHq4erg9vD28erh6uHq4Pbw9vAHAAcABwMXDxcPFw8XDxcOIx4jHTMtMy0zLEM8QzxDPEM8QzxDP1NLU0pfWl9Zb2h/e4+Gm5ablpuVq6WrpLu0u7S7tLu0u7S7tLu3y8PLwtfS19LX0tfS19Hn4efi19LX0tfS19LX0PfwAAMMDwwMAAD38Pfw9/MMDSwsOD9IS0hLSEtIS0hKWFloaHR7hIeEh4SHhIeEh4SHhIaUlpSVpKWkpLC3wMLQ0eDg7PP8//z//P8NDhkdKSw5P0lLSUtJS0lLSUtJS0lKVVpVWlVZZWllaWVpZWh1eHV4dXh1eHV4dXh1eHV7hYeFh4WHhYeFh4WHhYaRlpGVoaWhpaGloaWhpaGloaSxtLG0sbWhpaGloaWhpLG0sbSxtLG0sbSxtLG0sbSxtLG0sbSxtLG0sbSxtLG0sbSxtLG0sbfBw8HDwcPBw8HDwcPBw8HDwcPBw8HDwcCxtLG1oaWhpaGloaSxtLG0sbSxtLG0sbSxtLG0sbSxtLG0sbSxtLG0sbWhpaGloaaRlpGWkZaRlpGWkZaRlpGWkZaRlpGWkZaRl4WHhYR1eHV4dXllaWVpZWpVWlVbSUtJS0lLSUtJSDk8OTw5PDk8OTw5PDk8OTw5PDk8OTw5PDk8OT0pLSkuGR4ZHhkeGR4ZHhkfDQ8NDw0PDQ8NDw0PDQ8NDw0OGR4ZHhkfDQ8NDw0PDQ8ND/z//P/8//z//P/8//z//Pzs8Ozx4OHg4eDh4OLQ0tDS0NPAw8DDwMPAw8DAsLSwtaSmlJeEhpSWlJWkpaSlpKWkpLC0sLSwtLC1pKWkpaSlpKWkpaSlpKWkppSWlJeEh4SEdHloaWhpaGloaHR4dHh0eHR4dHh0eHR4dHh0eWhqWFpYWlhaWFpYWlhaWFtIS0hIODw4PDg8OD0sLhwfDAwAAPfw9/D38Pfw9/Hn4efi19PLwLu1q6WrppuWm5ablpuVq6ablpuXj4ePh4+Hj4ePhpuXj4ePhH95b2lvaW9pb2lval9aX1pfW1NLU0tTS1NLU0pfWl9aX1lvaW9pb2lvaH94f3h/eW9pb2lvaW9pb2lvaW9qX1tTS1NKX1pfWW9of3h/eH94f3h/eH94f3lvaW9qX1pfWl9aX1pfWl9aX1pfWl9Zb2lvaW9pb2lval9aX1tTS1NKX1lvaW9pb2pfW1NLU0tTSl9aX1lvaW9qX1tTS1NIQz9TS1NLU0tTSEM8QzxDPEM8QzxDP1NIQzxDPEM8Qz9TSl9aX1pfWl9bU0tTS1NKX1lvaH97j4ePh4+Hj4ePhpuWm5Wrpaulq6ablpuWm5Wrpauku7S7t8vDy8PLw8vDy8PLw8vC19LX0tfTy8PLwLu0u7Wrpaulq6WrpLu0u7S7taumm5abl4+Hj4ePh4+Hj4ePh4+Ef3lvaW9pb2lvaH94f3uPh4+Gm5ablaulq6S7t8vC19Hn4PfwAAMMDwwOHB4cHhwdLC0sLDg8OD9ISlhZaGh0e4SGlJWkpaSksLSwt8DDwMPAw8DC0NLQ0tDS0NLQ0eDh4OHg4eDh4OHg4eDg7PDs8Ozw7PDs8/z//P/8//z//P/8/Ozw7PHg4eDh4OHg4eDh4OHg4eDh4OHg4Ozw7PDs8Ozw7PDs8Ozw7PDs8Ozw7PHg4eDh4OLQ0tDR4OHg4eDi0NLQ0tDS0NHg4eDg7PDs8Ozw7PHg4eDh4OLQ0tDS0NPAw8DDwMPAwtDTwMPAwLC0sLWkpaSlpKWkpaSmlJeEhHR5aGloaWhqWFpYWlhaWFtISDg8OD0sLhwfDAwAAAAAAAAAAwwPDA8MDwwPDA8MDAAA9/LX08vAu7Wrpaumm5abl4+Ef3h/eW9pb2lvaW9pb2lvaW9pb2pfWl9bU0tTSEM8QzxDPEM8QzxDPEM8QzxDPEM9My0zLTMtMy0zLTMtMy4jHiMfFwwHAAcA9vD28erh6uLa0trS2tLa0trTysPKw8rDysPKw8rDysPKw8rDysC6tLq0urWupa6lrqael46HjoR+eH54fnh+eH54fnh+eH55cmh+eH54fnh+eH54fnh+eXJpcmlyaXJpcmlyaXJpcmlyaH54fnh+eXJpcmlyaXJpcmh+eH54fnh+eH57joeOh46HjoeOh46Efnh+eH55cmlyaXJofnh+eH54fnh+e46HjoeOh46HjoeOh46Gnpael46HjoeOh46HjoeOhH54fnh+eH54fnh+eH57joeOh46HjoeOhp6VrqWupa6lrqWupa6kurS6tLq3ysPKwLq0urS6t8rDysLa0trS2tHq4erh6uD28AcABwAHAAcDFw8XDiMeIx0zLTMtMyxDPEM/U0pfWW9of3h/eH97j4ablpuVq6WrppuWm5ePh4+Gm5ablaulq6Wrpaulq6Wrpaulq6Wrpaulq6ablpuWm5Wrpaulq6WrppuWm5ablauku7bX0tfS19C7taulq6Wrpaulq6WrppuWm5ablpuVq6S7tLu0u7S7taulq6Wrpauku7fLwtfS19LX0tfS19Hn4Pfw9/D38PfwAAAAAwwPDA8MDwwMAAAAAwwOHB4cHSwtLC0sLhweHB4cHSwtLCw4PDg8OD9IS0hKWFpYWWhpaGloaWhpaGloaHR4dHh0eHR5aGpYW0hLSEpYWWhodHh0eHR5aGloalhaWFpYWlhZaGloaWhpaGpYW0hLSEg4PDg8ODw4PDg/SEtIS0hLSEpYWlhZaGloaWhpaGloaWhodHh0eWhpaGpYW0hLSEg4PDg9LC0sLSwuHB4cHSwtLCw4PDg8ODw4PDg8ODw4PDg8OD9IS0hIOD0sLSwuHB8MDwwMAAD38Pfx5+LX0tfS19LX0efh5+D38Pfw9/D38Pfw9/D38Pfx5+Hn4efg9/D38AAAAAMMDhweHB0sLSwtLC0sLSwuHB4cHwwPDAwAAAAAAAMMDwwPDA4cHhwdLCw4P0hKWFloaWhpaGpYW0hIODw4PDg/SEpYWWhodHh0e4SHhIeEhHR4dHh0eHR4dHh0eHR4dHh0eHR4dHh0eWhpaGpYWlhaWFpYWWhpaGh0eHR7hIeEhHR4dHloaWhpaGloaWhpaGpYW0hLSEtIS0hLSEg4PDg9LC0sLSwtLCw4P0hIODw4PDg8OD9IS0hKWFpYWlhaWFpYWlhaWFpYWlhbSEg4PSwtLC0sLDg8ODw4PSwtLC0sLSwtLC0sLDg8OD0sLhwfDA8MDwwPDA8MDwwPDA8MDwwPDA8MDwwOHB4cHhwdLC0sLSwtLC0sLSwsOD9ISlhaWFloaWhqWFpYW0hLSEtIS0hLSEtIS0hLSEtISlhaWFloaHR4dHh0eHR4dHh0e4SGlJaUlpSVpKaUlpSWlJaUl4SHhIeEhHR4dHuEh4SEdHh0eWhpaGloalhaWFpYWlhaWFpYWlhaWFpYWlhZaGloalhaWFtIS0hLSEtIS0hLSEtISlhaWFpYWlhbSEtISSwtLC4cHhweHB4cHhwfDAwAAPfx5+LX0tfS19Hn4efg9/D38AADDA4cHSwtLC0sLSwtLC0sLDg8ODw4P0hLSEtIS0hLSEtIS0hLSEpYWlhaWFpYWlhaWFpYWlhaWFtIS0hIODw4PDg8OD9IS0hLSEtIS0hLSEtISlhaWFpYWlhaWFloalhaWFtISDg9LC4cHhwfDAwAAAAA9/D38AAAAAAAAwwPDA8MDhweHB0sLSwtLC4cHhwfDA8MDwwOHB0sLDg9LC0sLhwfDA8MDwwPDA8MDhweHB8MDwwPDAwAAAAAAAAAAAADDA4cHhweHB4cHhweHB4cHhwdLC0sLSwuHB8MDAAAAAAAAwwPDA8MDwwPDAwAAAAAAAAAAwwNLCw4P0hLSEtIS0hLSEtIS0hLSEtIS0hLSEg4PDg8ODw4PDg8ODw4PDg8ODw4P0hLSEtISlhZaGloaHR4dHuEh4SHhIeEhHR4dHloaWhodHh0eHR4dHh0eHR5aGloaWhpaGh0eHR4dHh0eHR4dHh0e4SHhIR0eHR5aGh0eHR7hIeEh4SHhIeEh4SEdHh0e4SHhIR0eHR4dHh0eHR7hIaUlaSlpKSwtLC0sLSwtLC3wMLQ0tDR4OHg4Ozz/P/8//z//P/8//z//P/8//z//P/8//z//P/8//z//P8NDw0PDQ8NDw0PDQ4ZHhkdKS0pLSktKS0pLSksOTw5PDk8OTw5PDk8OTw5PDk8OTw5PDk8OTw5P0lLSUtJS0lLSUtJS0lKVVpVWlVaVVllaWVpZWpVWlVaVVpVWlVaVVpVWlVaVVpVWlVaVVpVWWVpZWllaWVpZWllaWVpZWllalVaVVpVW0lLSUtJS0lKVVpVWlVaVVpVWlVaVVpVWlVaVVtJS0lLSUtJS0lLSUtJS0lLSUtJSDk8OT0pLSktKS4ZHhkeGR4ZHhkeGR4ZHw0PDQ8NDw0PDQ8ND/z//P/8//z87PDs8eDh4OLQ0tDTwMPAw8DAsLSwtaSmlJaUlpSWlJaUlpSXhIR0eHR5aGloaWhpaGloaWhpaGpYWlhbSEtIS0hLSEtISDg8OD0sLSwtLC0sLSwtLC0sLSwuHB4cHwwMAAD38efh5+D38Pfw9/D38efh5+LX08vDy8C7tLu0u7fLw8vDy8C7tLu1q6WrppuXj4ePh4+Gm5ablpuWm5ablpuWm5Wrpaulq6Wrpaumm5ablpuWm5ablpuVq6Wrpauku7S7tLu3y8PLwtfS19Hn4efh5+Hn4tfS19PLwtfS19Hn4AADDA4cHSwtLC0sLDg8OD9ISlhZaGloaHR4dHuEh4SHhIeEh4SGlJaUlpSWlJaUlpSVpKWkpLC0sLfAw8DDwMPAwLC0sLSwt8DC0NLQ0tDS0NHg4eDg7PDs8/z//Pzs8/z87PDs8Ozw7PDs8Ozx4OHg4eDh4OHg4eDh4OLQ0tDS0NLQ0tDS0NLQ0eDh4OHg4Ozw7PDs8Ozw7PHg4Ozw7PDs8Ozw7PDs8Ozw7PDs8Ozw7PHg4eDh4OHg4eDh4OHg4tDS0NLQ0tDTwMPAwLC0sLfAw8DC0NLQ0tDTwMPAw8DDwMCwtaSlpKaUlpSWlJWkpaSlpKaUlpSXhIeEhHR4dHh0eHR4dHh0eWhpaGloalhaWFpYWlhaWFpYWlhaWFpYW0hLSEtIS0hLSEtIS0hLSEg4PDg9LC0sLhweHB4cHhwdLC0sLhweHBwAAPfw9/Hn4efh5+Hn4Pfw9/D38Pfx5+Hn4tfTy8C7taulq6Wrpaulq6ablpuWm5ePh4+Ef3h/eH94f3lvaW9qX1pfWl9bU0tTS1NLU0tTS1NIQzxDPTMtMy0zLiMeIx4jHiMeIx8XDxcMBwD28erh6uA==\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(data[1][0])\n",
    "plt.show()\n",
    "plt.plot(data[2][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _wavenetcond_parse_test_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = tf.cast(parsed_features['label'], tf.int64)\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\t# There can only be one input so they need to be in a list\n",
    "\treturn (audio_inputs, audio_inputs, audio_quant), label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_wavenetcond_parse_test_function)\n",
    "# dataset = dataset.repeat()\n",
    "dataset = dataset.batch(1)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "next_batch_data, next_batch_targets= iterator.get_next()\n",
    "data, label = sess.run([next_batch_data, next_batch_targets])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c32338208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c3236e0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c324100f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scene: bus\n",
      "File name, segment:  mixture_devtrain_babycry_023_05c4cff75186054af7fbb4c6623d19d2_6\n",
      "label 1/0 (positive/negative): 0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.plot(data[0][0])\n",
    "plt.show()\n",
    "plt.plot(data[1][0])\n",
    "plt.show()\n",
    "plt.plot(data[2][0])\n",
    "plt.show()\n",
    "print('Scene:', bytes(s[0]).decode('utf8'))\n",
    "print('File name, segment: ', bytes(s_f[0]).decode('utf8'))\n",
    "print('label 1/0 (positive/negative):', label[0])\n",
    "fs=44100\n",
    "data_input = data[0]\n",
    "ipd.Audio(data[0], rate=fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### How uneven batches are handled\n",
    "#### The last batch is just uneven, see below :) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8338\n",
      "Number of batches per epoch: 260.5625\n",
      "Number of batches processed: 261\n",
      "Shape of last batch:  18\n"
     ]
    }
   ],
   "source": [
    "def _cae_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = [] # No labels being passed in training, just a placeholder.\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Range between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn audio_inputs,  label\n",
    "\n",
    "data_binaries_log = 'data_binaries/train/babycry/bus/'\n",
    "filenames_neg = ['{}babycry_bus_neg.tfrecord'.format(data_binaries_log)]\n",
    "filenames_pos = ['{}babycry_bus_pos.tfrecord'.format(data_binaries_log)]\n",
    "\n",
    "batch_size = 32\n",
    "print(num_examples)\n",
    "print(\"Number of batches per epoch:\", num_examples/batch_size)\n",
    "\n",
    "filenames_pl = tf.placeholder(tf.string, shape=[None])\n",
    "dataset = tf.data.TFRecordDataset(filenames_pl)\n",
    "dataset = dataset.map(_cae_parse_function)\n",
    "dataset = dataset.batch(batch_size)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "next_batch = iterator.get_next()\n",
    "iter_init = iterator.initializer\n",
    "next_batch_data, next_batch_label = iterator.get_next()\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init, feed_dict={filenames_pl: filenames_neg})\n",
    "c = 0\n",
    "while True:\n",
    "    try:\n",
    "        label, data= sess.run([next_batch_label, next_batch_data])\n",
    "        c+=1\n",
    "    except tf.errors.OutOfRangeError:\n",
    "        break\n",
    "print('Number of batches processed:', c)\n",
    "print('Shape of last batch: ', label.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check shuffle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of batches per epoch: 362.25\n",
      "[array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
      "       1, 1, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
      "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), array([0, 0, 0, 0, 0, 0, 0, 0])]\n"
     ]
    }
   ],
   "source": [
    "def _cae_parse_function(file):\n",
    "\tfeatures = {\n",
    "\t\t'label': tf.FixedLenFeature([], tf.int64),\n",
    "\t\t'audio_inputs': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'scene': tf.FixedLenFeature([], tf.string),\n",
    "\t\t'source_file': tf.FixedLenFeature([], tf.string)\n",
    "\t}\n",
    "\tparsed_features = tf.parse_single_example(file, features=features)\n",
    "\tlabel = tf.cast(parsed_features['label'], tf.int64) # No labels being passed in training, just a placeholder.\n",
    "\t# Quanitised value\n",
    "\taudio_quant = tf.decode_raw(parsed_features['audio_inputs'], tf.int64)\n",
    "\t# Range between -1 and 1\n",
    "\taudio_inputs = tf.cast(audio_quant, dtype=tf.float32)/128 - 1\n",
    "\tscene = tf.decode_raw(parsed_features['scene'], tf.uint8)\n",
    "\tsource_file = tf.decode_raw(parsed_features['source_file'], tf.uint8)\n",
    "\treturn audio_inputs,  label\n",
    "\n",
    "data_binaries_log = 'data_binaries/train/babycry/bus/'\n",
    "filenames_neg = ['{}babycry_bus_neg.tfrecord'.format(data_binaries_log),\n",
    "                '{}babycry_bus_pos.tfrecord'.format(data_binaries_log)]\n",
    "filenames_pos = glob.glob('data_binaries/val/babycry/beach/*.tfrecord')\n",
    "batch_size = 32\n",
    "num_examples = count_examples(filenames_pos)\n",
    "print(\"Number of batches per epoch:\", num_examples/batch_size)\n",
    "dataset = tf.data.TFRecordDataset(filenames_pos)\n",
    "dataset = dataset.map(_cae_parse_function)\n",
    "dataset = dataset.batch(batch_size)\n",
    "iterator = dataset.make_initializable_iterator()\n",
    "iter_init = iterator.initializer\n",
    "next_batch_data, next_batch_label = iterator.get_next()\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(iter_init)\n",
    "c = 0\n",
    "collect_labels = []\n",
    "while True:\n",
    "    try:\n",
    "        data, label = sess.run([next_batch_data, next_batch_label])\n",
    "        c+=1\n",
    "        collect_labels.append(label)\n",
    "    except tf.errors.OutOfRangeError:\n",
    "        break\n",
    "print(collect_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of examples in each set. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Ref: fhttps://stackoverflow.com/questions/40472139/obtaining-total-number-of-records-from-tfrecords-file-in-tensorflow\n",
    "num_neg = sum([1 for record in filenames_neg for _ in tf.python_io.tf_record_iterator(record)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_pos =sum([1 for record in filenames_pos for _ in tf.python_io.tf_record_iterator(record)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of negative examples: 8338\n",
      "Number of positive examples: 356\n",
      "percentage if positive examples: 4.27%\n"
     ]
    }
   ],
   "source": [
    "print('Number of negative examples: {}'.format(num_neg))\n",
    "print('Number of positive examples: {}'.format(num_pos))\n",
    "print('percentage of positive examples: {:.2f}%'.format((num_pos/num_neg)*100))"
   ]
  }
 ],
 "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}