AudioRecognition.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4o2mHJO6skA5"
   },
   "source": [
    "# <font color = 'red'> Interact with Google Colab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "k2uD1ctlYaqg"
   },
   "outputs": [],
   "source": [
    "# !ln -sf /opt/bin/nvidia-smi /usr/bin/nvidia-smi\n",
    "# !pip install gputil -q\n",
    "# !pip install psutil -q\n",
    "# !pip install humanize -q\n",
    "# import psutil\n",
    "# import humanize\n",
    "# import os\n",
    "# import GPUtil as GPU\n",
    "# GPUs = GPU.getGPUs()\n",
    "# # XXX: only one GPU on Colab and isn’t guaranteed\n",
    "# gpu = GPUs[0]\n",
    "# def printm():\n",
    "#  process = psutil.Process(os.getpid())\n",
    "#  print(\"Gen RAM Free: \" + humanize.naturalsize( psutil.virtual_memory().available ), \" | Proc size: \" + humanize.naturalsize( process.memory_info().rss))\n",
    "#  print(\"GPU RAM Free: {0:.0f}MB | Used: {1:.0f}MB | Util {2:3.0f}% | Total {3:.0f}MB\".format(gpu.memoryFree, gpu.memoryUsed, gpu.memoryUtil*100, gpu.memoryTotal))\n",
    "# printm()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 35680,
     "status": "ok",
     "timestamp": 1549508802645,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "kvjd_AWDsZ-v",
    "outputId": "da9032bb-450a-4668-f286-eb379faab7b7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[?25l\r",
      "\u001b[K    1% |▎                               | 10kB 27.3MB/s eta 0:00:01\r",
      "\u001b[K    2% |▋                               | 20kB 2.2MB/s eta 0:00:01\r",
      "\u001b[K    3% |█                               | 30kB 3.2MB/s eta 0:00:01\r",
      "\u001b[K    4% |█▎                              | 40kB 2.1MB/s eta 0:00:01\r",
      "\u001b[K    5% |█▋                              | 51kB 2.6MB/s eta 0:00:01\r",
      "\u001b[K    6% |██                              | 61kB 3.1MB/s eta 0:00:01\r",
      "\u001b[K    7% |██▎                             | 71kB 3.5MB/s eta 0:00:01\r",
      "\u001b[K    8% |██▋                             | 81kB 4.0MB/s eta 0:00:01\r",
      "\u001b[K    9% |███                             | 92kB 4.5MB/s eta 0:00:01\r",
      "\u001b[K    10% |███▎                            | 102kB 3.5MB/s eta 0:00:01\r",
      "\u001b[K    11% |███▋                            | 112kB 3.5MB/s eta 0:00:01\r",
      "\u001b[K    12% |████                            | 122kB 4.9MB/s eta 0:00:01\r",
      "\u001b[K    13% |████▎                           | 133kB 4.9MB/s eta 0:00:01\r",
      "\u001b[K    14% |████▋                           | 143kB 9.4MB/s eta 0:00:01\r",
      "\u001b[K    15% |█████                           | 153kB 9.4MB/s eta 0:00:01\r",
      "\u001b[K    16% |█████▎                          | 163kB 9.4MB/s eta 0:00:01\r",
      "\u001b[K    17% |█████▋                          | 174kB 9.3MB/s eta 0:00:01\r",
      "\u001b[K    18% |██████                          | 184kB 9.3MB/s eta 0:00:01\r",
      "\u001b[K    19% |██████▎                         | 194kB 9.3MB/s eta 0:00:01\r",
      "\u001b[K    20% |██████▋                         | 204kB 44.0MB/s eta 0:00:01\r",
      "\u001b[K    21% |███████                         | 215kB 10.3MB/s eta 0:00:01\r",
      "\u001b[K    22% |███████▎                        | 225kB 10.2MB/s eta 0:00:01\r",
      "\u001b[K    23% |███████▋                        | 235kB 10.2MB/s eta 0:00:01\r",
      "\u001b[K    24% |████████                        | 245kB 10.2MB/s eta 0:00:01\r",
      "\u001b[K    25% |████████▎                       | 256kB 10.1MB/s eta 0:00:01\r",
      "\u001b[K    26% |████████▋                       | 266kB 9.9MB/s eta 0:00:01\r",
      "\u001b[K    27% |█████████                       | 276kB 10.1MB/s eta 0:00:01\r",
      "\u001b[K    29% |█████████▎                      | 286kB 10.1MB/s eta 0:00:01\r",
      "\u001b[K    30% |█████████▋                      | 296kB 10.0MB/s eta 0:00:01\r",
      "\u001b[K    31% |██████████                      | 307kB 10.2MB/s eta 0:00:01\r",
      "\u001b[K    32% |██████████▎                     | 317kB 43.2MB/s eta 0:00:01\r",
      "\u001b[K    33% |██████████▋                     | 327kB 43.9MB/s eta 0:00:01\r",
      "\u001b[K    34% |███████████                     | 337kB 45.0MB/s eta 0:00:01\r",
      "\u001b[K    35% |███████████▎                    | 348kB 40.8MB/s eta 0:00:01\r",
      "\u001b[K    36% |███████████▋                    | 358kB 41.9MB/s eta 0:00:01\r",
      "\u001b[K    37% |████████████                    | 368kB 48.0MB/s eta 0:00:01\r",
      "\u001b[K    38% |████████████▎                   | 378kB 47.8MB/s eta 0:00:01\r",
      "\u001b[K    39% |████████████▋                   | 389kB 46.9MB/s eta 0:00:01\r",
      "\u001b[K    40% |█████████████                   | 399kB 12.3MB/s eta 0:00:01\r",
      "\u001b[K    41% |█████████████▎                  | 409kB 12.1MB/s eta 0:00:01\r",
      "\u001b[K    42% |█████████████▋                  | 419kB 12.2MB/s eta 0:00:01\r",
      "\u001b[K    43% |██████████████                  | 430kB 12.3MB/s eta 0:00:01\r",
      "\u001b[K    44% |██████████████▎                 | 440kB 12.3MB/s eta 0:00:01\r",
      "\u001b[K    45% |██████████████▋                 | 450kB 12.6MB/s eta 0:00:01\r",
      "\u001b[K    46% |███████████████                 | 460kB 12.5MB/s eta 0:00:01\r",
      "\u001b[K    47% |███████████████▎                | 471kB 12.5MB/s eta 0:00:01\r",
      "\u001b[K    48% |███████████████▋                | 481kB 12.4MB/s eta 0:00:01\r",
      "\u001b[K    49% |████████████████                | 491kB 12.4MB/s eta 0:00:01\r",
      "\u001b[K    50% |████████████████▎               | 501kB 48.7MB/s eta 0:00:01\r",
      "\u001b[K    51% |████████████████▋               | 512kB 49.3MB/s eta 0:00:01\r",
      "\u001b[K    52% |█████████████████               | 522kB 49.6MB/s eta 0:00:01\r",
      "\u001b[K    53% |█████████████████▎              | 532kB 49.9MB/s eta 0:00:01\r",
      "\u001b[K    54% |█████████████████▋              | 542kB 48.4MB/s eta 0:00:01\r",
      "\u001b[K    55% |██████████████████              | 552kB 49.5MB/s eta 0:00:01\r",
      "\u001b[K    57% |██████████████████▎             | 563kB 51.5MB/s eta 0:00:01\r",
      "\u001b[K    58% |██████████████████▋             | 573kB 51.0MB/s eta 0:00:01\r",
      "\u001b[K    59% |███████████████████             | 583kB 52.4MB/s eta 0:00:01\r",
      "\u001b[K    60% |███████████████████▎            | 593kB 52.9MB/s eta 0:00:01\r",
      "\u001b[K    61% |███████████████████▋            | 604kB 52.6MB/s eta 0:00:01\r",
      "\u001b[K    62% |████████████████████            | 614kB 54.4MB/s eta 0:00:01\r",
      "\u001b[K    63% |████████████████████▎           | 624kB 53.2MB/s eta 0:00:01\r",
      "\u001b[K    64% |████████████████████▋           | 634kB 53.6MB/s eta 0:00:01\r",
      "\u001b[K    65% |█████████████████████           | 645kB 54.7MB/s eta 0:00:01\r",
      "\u001b[K    66% |█████████████████████▎          | 655kB 55.0MB/s eta 0:00:01\r",
      "\u001b[K    67% |█████████████████████▋          | 665kB 43.1MB/s eta 0:00:01\r",
      "\u001b[K    68% |██████████████████████          | 675kB 15.2MB/s eta 0:00:01\r",
      "\u001b[K    69% |██████████████████████▎         | 686kB 14.9MB/s eta 0:00:01\r",
      "\u001b[K    70% |██████████████████████▋         | 696kB 14.9MB/s eta 0:00:01\r",
      "\u001b[K    71% |███████████████████████         | 706kB 14.8MB/s eta 0:00:01\r",
      "\u001b[K    72% |███████████████████████▎        | 716kB 14.9MB/s eta 0:00:01\r",
      "\u001b[K    73% |███████████████████████▋        | 727kB 14.8MB/s eta 0:00:01\r",
      "\u001b[K    74% |████████████████████████        | 737kB 14.7MB/s eta 0:00:01\r",
      "\u001b[K    75% |████████████████████████▎       | 747kB 14.7MB/s eta 0:00:01\r",
      "\u001b[K    76% |████████████████████████▋       | 757kB 14.8MB/s eta 0:00:01\r",
      "\u001b[K    77% |████████████████████████▉       | 768kB 16.0MB/s eta 0:00:01\r",
      "\u001b[K    78% |█████████████████████████▏      | 778kB 48.9MB/s eta 0:00:01\r",
      "\u001b[K    79% |█████████████████████████▌      | 788kB 51.8MB/s eta 0:00:01\r",
      "\u001b[K    80% |█████████████████████████▉      | 798kB 51.7MB/s eta 0:00:01\r",
      "\u001b[K    81% |██████████████████████████▏     | 808kB 51.1MB/s eta 0:00:01\r",
      "\u001b[K    82% |██████████████████████████▌     | 819kB 50.9MB/s eta 0:00:01\r",
      "\u001b[K    83% |██████████████████████████▉     | 829kB 52.1MB/s eta 0:00:01\r",
      "\u001b[K    85% |███████████████████████████▏    | 839kB 51.8MB/s eta 0:00:01\r",
      "\u001b[K    86% |███████████████████████████▌    | 849kB 50.9MB/s eta 0:00:01\r",
      "\u001b[K    87% |███████████████████████████▉    | 860kB 46.9MB/s eta 0:00:01\r",
      "\u001b[K    88% |████████████████████████████▏   | 870kB 46.4MB/s eta 0:00:01\r",
      "\u001b[K    89% |████████████████████████████▌   | 880kB 47.4MB/s eta 0:00:01\r",
      "\u001b[K    90% |████████████████████████████▉   | 890kB 48.1MB/s eta 0:00:01\r",
      "\u001b[K    91% |█████████████████████████████▏  | 901kB 48.5MB/s eta 0:00:01\r",
      "\u001b[K    92% |█████████████████████████████▌  | 911kB 49.6MB/s eta 0:00:01\r",
      "\u001b[K    93% |█████████████████████████████▉  | 921kB 13.9MB/s eta 0:00:01\r",
      "\u001b[K    94% |██████████████████████████████▏ | 931kB 13.8MB/s eta 0:00:01\r",
      "\u001b[K    95% |██████████████████████████████▌ | 942kB 13.8MB/s eta 0:00:01\r",
      "\u001b[K    96% |██████████████████████████████▉ | 952kB 13.9MB/s eta 0:00:01\r",
      "\u001b[K    97% |███████████████████████████████▏| 962kB 14.2MB/s eta 0:00:01\r",
      "\u001b[K    98% |███████████████████████████████▌| 972kB 14.3MB/s eta 0:00:01\r",
      "\u001b[K    99% |███████████████████████████████▉| 983kB 14.1MB/s eta 0:00:01\r",
      "\u001b[K    100% |████████████████████████████████| 993kB 10.1MB/s \n",
      "\u001b[?25h  Building wheel for PyDrive (setup.py) ... \u001b[?25ldone\n",
      "\u001b[?25h"
     ]
    }
   ],
   "source": [
    "# Install the PyDrive wrapper & import libraries.\n",
    "# This only needs to be done once per notebook.\n",
    "\n",
    "# PyDrive simplifies many common Google Drive API tasks.\n",
    "!pip install -U -q PyDrive\n",
    "# -q means quiet\n",
    "\n",
    "# provide the ability to authenticate to Google APIs using various methods.\n",
    "from pydrive.auth import GoogleAuth\n",
    "from pydrive.drive import GoogleDrive\n",
    "from google.colab import auth\n",
    "from oauth2client.client import GoogleCredentials\n",
    "\n",
    "# Authenticate and create the PyDrive client.\n",
    "# This only needs to be done once per notebook.\n",
    "auth.authenticate_user()\n",
    "gauth = GoogleAuth()\n",
    "gauth.credentials = GoogleCredentials.get_application_default()\n",
    "drive = GoogleDrive(gauth)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 122
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 18768,
     "status": "ok",
     "timestamp": 1549508823990,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "2KCznvL0moU6",
    "outputId": "bb96e9bc-6a1c-4046-c827-73f257e7cbb4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3Aietf%3Awg%3Aoauth%3A2.0%3Aoob&scope=email%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdocs.test%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive.photos.readonly%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fpeopleapi.readonly&response_type=code\n",
      "\n",
      "Enter your authorization code:\n",
      "··········\n",
      "Mounted at /content/drive\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Be able to work within google folder\n",
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 170
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1281,
     "status": "ok",
     "timestamp": 1549508826274,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "VrrO1zP5Shbd",
    "outputId": "c2b9b2d7-d235-4594-dba1-e550a69723cd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filesystem     1G-blocks  Used Available Use% Mounted on\n",
      "overlay             359G   22G      319G   7% /\n",
      "tmpfs                 7G    0G        7G   0% /dev\n",
      "tmpfs                 7G    0G        7G   0% /sys/fs/cgroup\n",
      "tmpfs                 7G    1G        7G   1% /var/colab\n",
      "/dev/sda1           365G   26G      340G   7% /opt/bin\n",
      "shm                   6G    0G        6G   0% /dev/shm\n",
      "tmpfs                 7G    0G        7G   0% /sys/firmware\n",
      "drive               100G   44G       57G  44% /content/drive\n"
     ]
    }
   ],
   "source": [
    "# tells you how much data you have on their servers. look at /dev/sda1\n",
    "!df -BG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1480,
     "status": "ok",
     "timestamp": 1549508827248,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "YR0YVSHmqB5N",
    "outputId": "f5aea667-a093-404c-c7f7-539a4944ca2e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/content\n"
     ]
    }
   ],
   "source": [
    "!pwd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1788,
     "status": "ok",
     "timestamp": 1549508827950,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "TmWNLpbrTvm5",
    "outputId": "b0f0cde3-cd03-44db-a3d4-acaa9d98f797"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "audio_df2.pkl  audio_df.pkl\t       model.json\t train_audio\n",
      "audio_df3.pkl  AudioRecognition.ipynb  model_weights.h5  vox1_meta.csv\n",
      "audio_df4.pkl  full_df.pkl\t       test_audio\n"
     ]
    }
   ],
   "source": [
    "# Change directory to the one this notebook is in\n",
    "import os\n",
    "os.chdir(\"drive/My Drive/Colab Notebooks/\")\n",
    "!ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 2081,
     "status": "ok",
     "timestamp": 1549508828862,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "iR3uPcYNKt5H",
    "outputId": "2505859d-915d-4890-a736-522e11266dca"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/content/drive/My Drive/Colab Notebooks\n"
     ]
    }
   ],
   "source": [
    "!pwd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "m8bK1dEAkHfR"
   },
   "source": [
    "# <font color = 'red' >Prep data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5H8v-kjAmEuH"
   },
   "source": [
    "data came from: \n",
    "http://www.robots.ox.ac.uk/~vgg/data/voxceleb/vox1.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "iwXC1c9zW1AD"
   },
   "outputs": [],
   "source": [
    "# unzip the test set\n",
    "!unzip vox1_test_aac.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "A3tbRrmbW2bE"
   },
   "outputs": [],
   "source": [
    "# concat the 4 parts of training set\n",
    "cat vox1_dev_wav_* > vox1_dev_wav.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "61aOb5wVW3c8"
   },
   "outputs": [],
   "source": [
    "# unzip the concat file\n",
    "!unzip vox1_dev_wav.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1247,
     "status": "ok",
     "timestamp": 1549508844495,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "Qer3SJJrvZPE",
    "outputId": "234df5b9-59c1-4290-ade8-42c48d459596"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# import necessary libraries\n",
    "import re\n",
    "import wave \n",
    "import itertools\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from os import listdir\n",
    "from os.path import isfile, join\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "import librosa\n",
    "import librosa.feature \n",
    "import librosa.display\n",
    "# import talos as ta\n",
    "# from talos import live\n",
    "# from talos.model import hidden_layers\n",
    "from keras import layers, optimizers\n",
    "from keras.models import model_from_json\n",
    "from keras.models import Sequential, Model\n",
    "from keras.optimizers import Adam, sgd, Adamax\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.layers.advanced_activations import LeakyReLU, PReLU\n",
    "from keras.layers import Dense, LSTM, Dropout, Flatten, Input, Conv1D, MaxPooling1D\n",
    "from keras.activations import relu, elu, softmax, selu\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import f1_score, precision_score, recall_score, accuracy_score\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import roc_curve, auc\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QpThAmiBzuNL"
   },
   "source": [
    "## get labels into dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 728,
     "status": "ok",
     "timestamp": 1549508847874,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "b2f6CUOEztY3",
    "outputId": "926abda4-fbbc-4079-cc70-acc65a8b778b"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>VoxCeleb1 ID</th>\n",
       "      <th>VGGFace1 ID</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Nationality</th>\n",
       "      <th>Set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id10001</td>\n",
       "      <td>A.J._Buckley</td>\n",
       "      <td>m</td>\n",
       "      <td>Ireland</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>id10002</td>\n",
       "      <td>A.R._Rahman</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>id10003</td>\n",
       "      <td>Aamir_Khan</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>id10004</td>\n",
       "      <td>Aaron_Tveit</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id10005</td>\n",
       "      <td>Aaron_Yoo</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  VoxCeleb1 ID   VGGFace1 ID Gender Nationality  Set\n",
       "0      id10001  A.J._Buckley      m     Ireland  dev\n",
       "1      id10002   A.R._Rahman      m       India  dev\n",
       "2      id10003    Aamir_Khan      m       India  dev\n",
       "3      id10004   Aaron_Tveit      m         USA  dev\n",
       "4      id10005     Aaron_Yoo      m         USA  dev"
      ]
     },
     "execution_count": 9,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read the meta csv file\n",
    "meta_df = pd.read_csv('vox1_meta.csv', delimiter = '\\t' )\n",
    "meta_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 822,
     "status": "ok",
     "timestamp": 1549506163211,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "j5dBXkM50RHw",
    "outputId": "7dcbd849-a800-48cc-d59c-97728b95e905"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>name</th>\n",
       "      <th>gender</th>\n",
       "      <th>nationality</th>\n",
       "      <th>set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id10001</td>\n",
       "      <td>A.J._Buckley</td>\n",
       "      <td>m</td>\n",
       "      <td>Ireland</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>id10002</td>\n",
       "      <td>A.R._Rahman</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>id10003</td>\n",
       "      <td>Aamir_Khan</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>id10004</td>\n",
       "      <td>Aaron_Tveit</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id10005</td>\n",
       "      <td>Aaron_Yoo</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id          name gender nationality  set\n",
       "0  id10001  A.J._Buckley      m     Ireland  dev\n",
       "1  id10002   A.R._Rahman      m       India  dev\n",
       "2  id10003    Aamir_Khan      m       India  dev\n",
       "3  id10004   Aaron_Tveit      m         USA  dev\n",
       "4  id10005     Aaron_Yoo      m         USA  dev"
      ]
     },
     "execution_count": 14,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# change the column names to be more intuitive and lowercased\n",
    "meta_df.columns = ['id', 'name', 'gender', 'nationality', 'set'] # rename columns\n",
    "meta_df.head() # shape of df: (1251, 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nBYEDTv0wm49"
   },
   "source": [
    "## get training data\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "vi0HuVi3wtHZ"
   },
   "source": [
    "- **mfcc**: Mel-frequency cepstral coefficients\n",
    "- **melspectrogram**: Compute a Mel-scaled power spectrogram\n",
    "- **chorma-stft**: Compute a chromagram from a waveform or power spectrogram\n",
    "    - in music, the term chroma feature or chromagram closely relates to the twelve different pitch classes. Chroma-based features, which are also referred to as \"pitch class profiles\", are a powerful tool for analyzing music whose pitches can be meaningfully categorized and whose tuning approximates to the equal-tempered scale. \n",
    "- **spectral_contrast**: Compute spectral contrast\n",
    "    - Spectral contrast is defined as the level difference between peaks and valleys in the spectrum\n",
    "- **tonnetz**: Computes the tonal centroid features (tonnetz), following the method of [2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "pRp6ejq4BYZd"
   },
   "outputs": [],
   "source": [
    "def extract_feat(filename, filepath):\n",
    "  \n",
    "  # load in audio file\n",
    "  y, sr = librosa.load(filepath + filename) # y = audio file, sr = sample rate\n",
    "\n",
    "  # extract the various features of the audio\n",
    "  mfcc = np.mean(librosa.feature.mfcc(y = y, sr = sr, n_mfcc=40).T, axis = 0)  \n",
    "  mel = np.mean(librosa.feature.melspectrogram(y = y, sr = sr).T, axis = 0)\n",
    "  stft = np.abs(librosa.stft(y))\n",
    "  chroma = np.mean(librosa.feature.chroma_stft(S = stft, y = y, sr = sr).T, axis = 0)\n",
    "  contrast = np.mean(librosa.feature.spectral_contrast(S = stft, y = y, sr = sr).T, axis = 0)\n",
    "  tonnetz =  np.mean(librosa.feature.tonnetz(y = librosa.effects.harmonic(y), sr = sr).T, axis = 0)\n",
    "  \n",
    "  return mfcc,chroma,mel,contrast,tonnetz # shape: (40,), (12,), (128,), (7,), (6,)\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 98
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 373,
     "status": "ok",
     "timestamp": 1549338602561,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "NuRZVU5T1mNP",
    "outputId": "bd56f0ae-2da2-4eb5-93d7-5561cd849e76"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>...</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>0 rows × 194 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [id, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, mfcc, chroma, chroma, chroma, chroma, chroma, chroma, chroma, chroma, chroma, chroma, chroma, chroma, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, mel, ...]\n",
       "Index: []\n",
       "\n",
       "[0 rows x 194 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Instatiate a dataframe the train audio features will be in \n",
    "columns = ['id'] + ['mfcc']*40 + ['chroma']*12 + ['mel']*128 + ['contrast']*7 + ['tonnetz']*6\n",
    "audio_df = pd.DataFrame(columns = columns)\n",
    "audio_df # 0 rows × 194 columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "DDQXYtGgLK9B"
   },
   "outputs": [],
   "source": [
    "# Get data, extract features, append features into dataframe\n",
    "ids = [f for f in listdir('train_audio/')] # get the filenames (ids) in the train_audio/ directory\n",
    "for idd in ids[344:]:\n",
    "  subfiles = [f for f in listdir('train_audio/{}'.format(idd))] # go into each celeb file\n",
    "  \n",
    "  for subfile in subfiles:\n",
    "    audiofiles = [f for f in listdir('train_audio/{}/{}'.format(idd, subfile))] # extract the audio files in each subfile\n",
    "    \n",
    "    for audiofile in audiofiles:\n",
    "      pathname = 'train_audio/{}/{}/'.format(idd, subfile)\n",
    "      mfcc,chroma,mel,contrast,tonnetz = extract_feat(audiofile, pathname)\n",
    "      features = np.hstack([mfcc,chroma,mel,contrast,tonnetz])\n",
    "     \n",
    "      # get the id in integer form\n",
    "      idd_int = int(idd[2:])\n",
    "\n",
    "      # add id in the front of the features array\n",
    "      labelled = np.insert(features, 0, idd_int, axis = 0)\n",
    "      fill = np.empty((0,194))\n",
    "      row = np.vstack([fill,labelled]) # shape (1,193)\n",
    "      \n",
    "      # put row in a dataframe\n",
    "      row_df = pd.DataFrame(row, columns = columns)\n",
    "\n",
    "      # append row_df into the dataframe\n",
    "      audio_df = audio_df.append(row_df, ignore_index = True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "8KD0EH5hLYUs"
   },
   "source": [
    "<font color = 'blue'> **because it takes hours to collect ~100 ids, had to collect in iteration. Saved these iterations in audio_df(#).pkl files. There are four pickle files, all together collected ~500 IDs. **"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XNkPK857E7wA"
   },
   "outputs": [],
   "source": [
    "# Save dataframe in a pickle file\n",
    "audio_df.to_pickle('audio_df4.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 277,
     "status": "ok",
     "timestamp": 1549371670337,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "NCoo85Prvur4",
    "outputId": "b96530a8-6a5a-400d-9d39-b66aa1e0537c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(194,)"
      ]
     },
     "execution_count": 19,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df4 = pd.read_pickle('audio_df4.pkl')\n",
    "df4.id.unique().shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "fCH3a9jfgctc"
   },
   "source": [
    "## Test/Train/Split final dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "mVq0TMkLA0gX"
   },
   "outputs": [],
   "source": [
    "# Read in all the pickle files and concat/join/merge\n",
    "df1 = pd.read_pickle('audio_df.pkl')\n",
    "df2 = pd.read_pickle('audio_df2.pkl')\n",
    "df3 = pd.read_pickle('audio_df3.pkl')\n",
    "df4 = pd.read_pickle('audio_df4.pkl')\n",
    "df = pd.concat([df1,df2,df3,df4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zb7S7D48NbvS"
   },
   "outputs": [],
   "source": [
    "# Set display options to 200 columns to be able to see all 193 columns\n",
    "pd.set_option('display.max_columns', 200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 309
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 5071,
     "status": "ok",
     "timestamp": 1549506178398,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "jR3_uuBuLeaN",
    "outputId": "6280f3cb-b72a-44fd-897f-8c07f5b16bbf"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>chroma</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>mel</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>contrast</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24778</th>\n",
       "      <td>11100.0</td>\n",
       "      <td>-503.800568</td>\n",
       "      <td>159.991782</td>\n",
       "      <td>-16.537968</td>\n",
       "      <td>37.070034</td>\n",
       "      <td>-4.459876</td>\n",
       "      <td>1.584697</td>\n",
       "      <td>-3.264234</td>\n",
       "      <td>6.204571</td>\n",
       "      <td>-6.923838</td>\n",
       "      <td>-0.757389</td>\n",
       "      <td>2.842366</td>\n",
       "      <td>-3.018444</td>\n",
       "      <td>5.399441</td>\n",
       "      <td>2.729970</td>\n",
       "      <td>4.665108</td>\n",
       "      <td>-9.940535</td>\n",
       "      <td>-2.158047</td>\n",
       "      <td>-1.898706</td>\n",
       "      <td>-0.797725</td>\n",
       "      <td>-0.538610</td>\n",
       "      <td>-3.721104</td>\n",
       "      <td>-6.202170</td>\n",
       "      <td>-9.058631</td>\n",
       "      <td>0.302317</td>\n",
       "      <td>-6.527288</td>\n",
       "      <td>-4.754056</td>\n",
       "      <td>-3.112750</td>\n",
       "      <td>0.613957</td>\n",
       "      <td>0.508048</td>\n",
       "      <td>-2.096162</td>\n",
       "      <td>1.506327</td>\n",
       "      <td>1.187206</td>\n",
       "      <td>1.999975</td>\n",
       "      <td>1.129699</td>\n",
       "      <td>6.913423</td>\n",
       "      <td>2.863409</td>\n",
       "      <td>-0.923728</td>\n",
       "      <td>-0.756724</td>\n",
       "      <td>-0.185686</td>\n",
       "      <td>-0.967237</td>\n",
       "      <td>0.497434</td>\n",
       "      <td>0.479296</td>\n",
       "      <td>0.526856</td>\n",
       "      <td>0.480500</td>\n",
       "      <td>0.492818</td>\n",
       "      <td>0.517734</td>\n",
       "      <td>0.575788</td>\n",
       "      <td>0.639266</td>\n",
       "      <td>0.614953</td>\n",
       "      <td>0.520258</td>\n",
       "      <td>0.534425</td>\n",
       "      <td>0.521624</td>\n",
       "      <td>0.000793</td>\n",
       "      <td>0.005434</td>\n",
       "      <td>0.030386</td>\n",
       "      <td>0.006876</td>\n",
       "      <td>0.006212</td>\n",
       "      <td>0.009032</td>\n",
       "      <td>0.059345</td>\n",
       "      <td>0.094872</td>\n",
       "      <td>0.042270</td>\n",
       "      <td>0.020747</td>\n",
       "      <td>0.055137</td>\n",
       "      <td>0.036292</td>\n",
       "      <td>0.003616</td>\n",
       "      <td>0.011298</td>\n",
       "      <td>0.057143</td>\n",
       "      <td>0.090732</td>\n",
       "      <td>0.046198</td>\n",
       "      <td>0.043124</td>\n",
       "      <td>0.039419</td>\n",
       "      <td>0.044442</td>\n",
       "      <td>0.026989</td>\n",
       "      <td>0.021998</td>\n",
       "      <td>0.088756</td>\n",
       "      <td>0.081188</td>\n",
       "      <td>0.095127</td>\n",
       "      <td>0.057110</td>\n",
       "      <td>0.063391</td>\n",
       "      <td>0.052137</td>\n",
       "      <td>0.061667</td>\n",
       "      <td>0.052075</td>\n",
       "      <td>0.037905</td>\n",
       "      <td>0.012853</td>\n",
       "      <td>0.021025</td>\n",
       "      <td>0.008507</td>\n",
       "      <td>0.006129</td>\n",
       "      <td>0.012864</td>\n",
       "      <td>0.035194</td>\n",
       "      <td>0.050898</td>\n",
       "      <td>0.052126</td>\n",
       "      <td>0.017207</td>\n",
       "      <td>0.024111</td>\n",
       "      <td>0.013910</td>\n",
       "      <td>0.011910</td>\n",
       "      <td>0.016019</td>\n",
       "      <td>0.018280</td>\n",
       "      <td>0.013507</td>\n",
       "      <td>0.003757</td>\n",
       "      <td>0.002855</td>\n",
       "      <td>0.003046</td>\n",
       "      <td>0.001857</td>\n",
       "      <td>0.002147</td>\n",
       "      <td>0.002103</td>\n",
       "      <td>0.003201</td>\n",
       "      <td>0.008896</td>\n",
       "      <td>0.005241</td>\n",
       "      <td>0.003217</td>\n",
       "      <td>0.005578</td>\n",
       "      <td>0.004855</td>\n",
       "      <td>0.004800</td>\n",
       "      <td>0.001446</td>\n",
       "      <td>0.000582</td>\n",
       "      <td>0.001034</td>\n",
       "      <td>0.001469</td>\n",
       "      <td>0.000948</td>\n",
       "      <td>0.000737</td>\n",
       "      <td>0.001306</td>\n",
       "      <td>0.002862</td>\n",
       "      <td>0.001120</td>\n",
       "      <td>0.000777</td>\n",
       "      <td>0.000475</td>\n",
       "      <td>0.000245</td>\n",
       "      <td>0.000148</td>\n",
       "      <td>0.000372</td>\n",
       "      <td>0.001103</td>\n",
       "      <td>0.001979</td>\n",
       "      <td>0.001484</td>\n",
       "      <td>0.001756</td>\n",
       "      <td>0.000432</td>\n",
       "      <td>0.000250</td>\n",
       "      <td>0.000339</td>\n",
       "      <td>0.000319</td>\n",
       "      <td>0.000238</td>\n",
       "      <td>0.000163</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000072</td>\n",
       "      <td>0.000047</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000105</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>0.000019</td>\n",
       "      <td>0.000019</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>7.056818e-06</td>\n",
       "      <td>9.596689e-06</td>\n",
       "      <td>6.661124e-06</td>\n",
       "      <td>5.487163e-06</td>\n",
       "      <td>3.610811e-06</td>\n",
       "      <td>1.066116e-06</td>\n",
       "      <td>3.940568e-08</td>\n",
       "      <td>4.386587e-09</td>\n",
       "      <td>3.934327e-09</td>\n",
       "      <td>3.616842e-09</td>\n",
       "      <td>3.364994e-09</td>\n",
       "      <td>3.155534e-09</td>\n",
       "      <td>2.986337e-09</td>\n",
       "      <td>2.844588e-09</td>\n",
       "      <td>2.727968e-09</td>\n",
       "      <td>2.635663e-09</td>\n",
       "      <td>2.563692e-09</td>\n",
       "      <td>2.513333e-09</td>\n",
       "      <td>2.481457e-09</td>\n",
       "      <td>18.123065</td>\n",
       "      <td>16.864310</td>\n",
       "      <td>20.108288</td>\n",
       "      <td>18.542917</td>\n",
       "      <td>19.385362</td>\n",
       "      <td>19.506113</td>\n",
       "      <td>60.697950</td>\n",
       "      <td>0.007471</td>\n",
       "      <td>0.020963</td>\n",
       "      <td>0.050210</td>\n",
       "      <td>-0.102916</td>\n",
       "      <td>-0.026035</td>\n",
       "      <td>0.005766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24779</th>\n",
       "      <td>11100.0</td>\n",
       "      <td>-442.607326</td>\n",
       "      <td>160.302527</td>\n",
       "      <td>-31.399481</td>\n",
       "      <td>21.130893</td>\n",
       "      <td>-10.974108</td>\n",
       "      <td>4.796752</td>\n",
       "      <td>-5.693317</td>\n",
       "      <td>3.232060</td>\n",
       "      <td>0.999348</td>\n",
       "      <td>-0.711150</td>\n",
       "      <td>2.550994</td>\n",
       "      <td>-4.104016</td>\n",
       "      <td>3.072689</td>\n",
       "      <td>3.371486</td>\n",
       "      <td>3.884172</td>\n",
       "      <td>-9.338989</td>\n",
       "      <td>-4.311083</td>\n",
       "      <td>-2.956424</td>\n",
       "      <td>-4.952378</td>\n",
       "      <td>-3.204237</td>\n",
       "      <td>-1.798397</td>\n",
       "      <td>-3.337122</td>\n",
       "      <td>-2.487623</td>\n",
       "      <td>3.405535</td>\n",
       "      <td>-3.631716</td>\n",
       "      <td>-1.138690</td>\n",
       "      <td>-3.985064</td>\n",
       "      <td>0.040738</td>\n",
       "      <td>4.584732</td>\n",
       "      <td>-1.081874</td>\n",
       "      <td>-0.537442</td>\n",
       "      <td>0.576348</td>\n",
       "      <td>0.446506</td>\n",
       "      <td>-1.680391</td>\n",
       "      <td>-0.161449</td>\n",
       "      <td>-2.273001</td>\n",
       "      <td>-3.235756</td>\n",
       "      <td>-5.072984</td>\n",
       "      <td>-3.912861</td>\n",
       "      <td>-1.930381</td>\n",
       "      <td>0.632974</td>\n",
       "      <td>0.534964</td>\n",
       "      <td>0.563326</td>\n",
       "      <td>0.562226</td>\n",
       "      <td>0.545896</td>\n",
       "      <td>0.483770</td>\n",
       "      <td>0.442797</td>\n",
       "      <td>0.494035</td>\n",
       "      <td>0.542529</td>\n",
       "      <td>0.576757</td>\n",
       "      <td>0.589167</td>\n",
       "      <td>0.653003</td>\n",
       "      <td>0.001122</td>\n",
       "      <td>0.006539</td>\n",
       "      <td>0.036587</td>\n",
       "      <td>0.013001</td>\n",
       "      <td>0.013548</td>\n",
       "      <td>0.013358</td>\n",
       "      <td>0.015413</td>\n",
       "      <td>0.063964</td>\n",
       "      <td>0.060538</td>\n",
       "      <td>0.070826</td>\n",
       "      <td>0.159765</td>\n",
       "      <td>0.199303</td>\n",
       "      <td>0.054368</td>\n",
       "      <td>0.009223</td>\n",
       "      <td>0.063817</td>\n",
       "      <td>0.053219</td>\n",
       "      <td>0.120010</td>\n",
       "      <td>0.245420</td>\n",
       "      <td>0.271008</td>\n",
       "      <td>0.344488</td>\n",
       "      <td>0.079667</td>\n",
       "      <td>0.072464</td>\n",
       "      <td>0.166110</td>\n",
       "      <td>0.132764</td>\n",
       "      <td>0.255801</td>\n",
       "      <td>0.269304</td>\n",
       "      <td>0.199609</td>\n",
       "      <td>0.077808</td>\n",
       "      <td>0.054156</td>\n",
       "      <td>0.045447</td>\n",
       "      <td>0.033343</td>\n",
       "      <td>0.043260</td>\n",
       "      <td>0.080957</td>\n",
       "      <td>0.161356</td>\n",
       "      <td>0.161980</td>\n",
       "      <td>0.165373</td>\n",
       "      <td>0.261900</td>\n",
       "      <td>0.237476</td>\n",
       "      <td>0.177378</td>\n",
       "      <td>0.128297</td>\n",
       "      <td>0.091435</td>\n",
       "      <td>0.041402</td>\n",
       "      <td>0.071076</td>\n",
       "      <td>0.108694</td>\n",
       "      <td>0.132639</td>\n",
       "      <td>0.028384</td>\n",
       "      <td>0.018485</td>\n",
       "      <td>0.010136</td>\n",
       "      <td>0.008793</td>\n",
       "      <td>0.010192</td>\n",
       "      <td>0.014159</td>\n",
       "      <td>0.006357</td>\n",
       "      <td>0.041276</td>\n",
       "      <td>0.035302</td>\n",
       "      <td>0.019856</td>\n",
       "      <td>0.015073</td>\n",
       "      <td>0.012102</td>\n",
       "      <td>0.008323</td>\n",
       "      <td>0.010512</td>\n",
       "      <td>0.018413</td>\n",
       "      <td>0.007697</td>\n",
       "      <td>0.002625</td>\n",
       "      <td>0.002576</td>\n",
       "      <td>0.002031</td>\n",
       "      <td>0.004848</td>\n",
       "      <td>0.004092</td>\n",
       "      <td>0.004719</td>\n",
       "      <td>0.003144</td>\n",
       "      <td>0.001334</td>\n",
       "      <td>0.000752</td>\n",
       "      <td>0.000601</td>\n",
       "      <td>0.000957</td>\n",
       "      <td>0.002569</td>\n",
       "      <td>0.003178</td>\n",
       "      <td>0.002321</td>\n",
       "      <td>0.001084</td>\n",
       "      <td>0.000905</td>\n",
       "      <td>0.000736</td>\n",
       "      <td>0.000702</td>\n",
       "      <td>0.000594</td>\n",
       "      <td>0.000371</td>\n",
       "      <td>0.000377</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000216</td>\n",
       "      <td>0.000227</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000093</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.000171</td>\n",
       "      <td>0.000140</td>\n",
       "      <td>0.000114</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>0.000022</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000034</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0.000090</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.000018</td>\n",
       "      <td>0.000023</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>1.272312e-05</td>\n",
       "      <td>8.955307e-06</td>\n",
       "      <td>1.367225e-05</td>\n",
       "      <td>7.171851e-06</td>\n",
       "      <td>4.043730e-06</td>\n",
       "      <td>1.242139e-06</td>\n",
       "      <td>8.010924e-08</td>\n",
       "      <td>1.141345e-08</td>\n",
       "      <td>9.478378e-09</td>\n",
       "      <td>8.492636e-09</td>\n",
       "      <td>7.798572e-09</td>\n",
       "      <td>7.266692e-09</td>\n",
       "      <td>6.860868e-09</td>\n",
       "      <td>6.536293e-09</td>\n",
       "      <td>6.277775e-09</td>\n",
       "      <td>6.078889e-09</td>\n",
       "      <td>5.926311e-09</td>\n",
       "      <td>5.821557e-09</td>\n",
       "      <td>5.755594e-09</td>\n",
       "      <td>16.209617</td>\n",
       "      <td>15.989443</td>\n",
       "      <td>18.808640</td>\n",
       "      <td>18.342743</td>\n",
       "      <td>19.651828</td>\n",
       "      <td>19.176303</td>\n",
       "      <td>58.457264</td>\n",
       "      <td>0.010165</td>\n",
       "      <td>0.010087</td>\n",
       "      <td>0.037996</td>\n",
       "      <td>-0.079704</td>\n",
       "      <td>-0.029291</td>\n",
       "      <td>-0.001756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24780</th>\n",
       "      <td>11100.0</td>\n",
       "      <td>-485.288794</td>\n",
       "      <td>183.993909</td>\n",
       "      <td>-26.067565</td>\n",
       "      <td>11.609672</td>\n",
       "      <td>-8.942526</td>\n",
       "      <td>-4.337396</td>\n",
       "      <td>1.094897</td>\n",
       "      <td>0.014207</td>\n",
       "      <td>-5.998587</td>\n",
       "      <td>6.376794</td>\n",
       "      <td>5.000262</td>\n",
       "      <td>-11.034626</td>\n",
       "      <td>-0.015949</td>\n",
       "      <td>4.268690</td>\n",
       "      <td>-3.595981</td>\n",
       "      <td>-8.744786</td>\n",
       "      <td>-1.217848</td>\n",
       "      <td>-5.092902</td>\n",
       "      <td>-2.665413</td>\n",
       "      <td>-1.058037</td>\n",
       "      <td>-4.575139</td>\n",
       "      <td>-8.019194</td>\n",
       "      <td>-7.963319</td>\n",
       "      <td>-1.761272</td>\n",
       "      <td>-10.511799</td>\n",
       "      <td>-9.388931</td>\n",
       "      <td>-6.599813</td>\n",
       "      <td>-3.779780</td>\n",
       "      <td>-1.163284</td>\n",
       "      <td>-2.449122</td>\n",
       "      <td>-1.393838</td>\n",
       "      <td>-0.807064</td>\n",
       "      <td>-0.655574</td>\n",
       "      <td>-4.885306</td>\n",
       "      <td>0.219818</td>\n",
       "      <td>4.619997</td>\n",
       "      <td>8.157972</td>\n",
       "      <td>11.245900</td>\n",
       "      <td>11.076964</td>\n",
       "      <td>6.106433</td>\n",
       "      <td>0.472416</td>\n",
       "      <td>0.370513</td>\n",
       "      <td>0.329820</td>\n",
       "      <td>0.346764</td>\n",
       "      <td>0.456383</td>\n",
       "      <td>0.665538</td>\n",
       "      <td>0.642137</td>\n",
       "      <td>0.534971</td>\n",
       "      <td>0.478117</td>\n",
       "      <td>0.462401</td>\n",
       "      <td>0.381104</td>\n",
       "      <td>0.360215</td>\n",
       "      <td>0.000522</td>\n",
       "      <td>0.001570</td>\n",
       "      <td>0.005693</td>\n",
       "      <td>0.006369</td>\n",
       "      <td>0.009270</td>\n",
       "      <td>0.033299</td>\n",
       "      <td>0.194070</td>\n",
       "      <td>0.129483</td>\n",
       "      <td>0.045789</td>\n",
       "      <td>0.003905</td>\n",
       "      <td>0.001021</td>\n",
       "      <td>0.002371</td>\n",
       "      <td>0.014030</td>\n",
       "      <td>0.046403</td>\n",
       "      <td>0.106909</td>\n",
       "      <td>0.164091</td>\n",
       "      <td>0.090521</td>\n",
       "      <td>0.086963</td>\n",
       "      <td>0.045958</td>\n",
       "      <td>0.074889</td>\n",
       "      <td>0.110980</td>\n",
       "      <td>0.049130</td>\n",
       "      <td>0.049725</td>\n",
       "      <td>0.061341</td>\n",
       "      <td>0.051585</td>\n",
       "      <td>0.066949</td>\n",
       "      <td>0.188229</td>\n",
       "      <td>0.188903</td>\n",
       "      <td>0.157806</td>\n",
       "      <td>0.096955</td>\n",
       "      <td>0.022311</td>\n",
       "      <td>0.068481</td>\n",
       "      <td>0.096177</td>\n",
       "      <td>0.057816</td>\n",
       "      <td>0.095490</td>\n",
       "      <td>0.054344</td>\n",
       "      <td>0.041118</td>\n",
       "      <td>0.021336</td>\n",
       "      <td>0.008077</td>\n",
       "      <td>0.032519</td>\n",
       "      <td>0.128685</td>\n",
       "      <td>0.061266</td>\n",
       "      <td>0.018746</td>\n",
       "      <td>0.029230</td>\n",
       "      <td>0.013624</td>\n",
       "      <td>0.015305</td>\n",
       "      <td>0.034468</td>\n",
       "      <td>0.014040</td>\n",
       "      <td>0.009790</td>\n",
       "      <td>0.006121</td>\n",
       "      <td>0.004840</td>\n",
       "      <td>0.015512</td>\n",
       "      <td>0.008388</td>\n",
       "      <td>0.017746</td>\n",
       "      <td>0.013110</td>\n",
       "      <td>0.005659</td>\n",
       "      <td>0.008049</td>\n",
       "      <td>0.007609</td>\n",
       "      <td>0.011967</td>\n",
       "      <td>0.004412</td>\n",
       "      <td>0.001054</td>\n",
       "      <td>0.001223</td>\n",
       "      <td>0.001630</td>\n",
       "      <td>0.001185</td>\n",
       "      <td>0.000486</td>\n",
       "      <td>0.000696</td>\n",
       "      <td>0.000943</td>\n",
       "      <td>0.000466</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000140</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000067</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0.000344</td>\n",
       "      <td>0.000324</td>\n",
       "      <td>0.000386</td>\n",
       "      <td>0.000257</td>\n",
       "      <td>0.000242</td>\n",
       "      <td>0.000322</td>\n",
       "      <td>0.000309</td>\n",
       "      <td>0.000192</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000057</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000012</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>2.407622e-06</td>\n",
       "      <td>1.514996e-06</td>\n",
       "      <td>1.679450e-06</td>\n",
       "      <td>1.593449e-06</td>\n",
       "      <td>1.080508e-06</td>\n",
       "      <td>2.516526e-07</td>\n",
       "      <td>7.536460e-08</td>\n",
       "      <td>4.531382e-08</td>\n",
       "      <td>3.777022e-08</td>\n",
       "      <td>3.345398e-08</td>\n",
       "      <td>3.045017e-08</td>\n",
       "      <td>2.817581e-08</td>\n",
       "      <td>2.645287e-08</td>\n",
       "      <td>2.508716e-08</td>\n",
       "      <td>2.400698e-08</td>\n",
       "      <td>2.317983e-08</td>\n",
       "      <td>2.254846e-08</td>\n",
       "      <td>2.211556e-08</td>\n",
       "      <td>2.184430e-08</td>\n",
       "      <td>19.568966</td>\n",
       "      <td>17.647668</td>\n",
       "      <td>21.506072</td>\n",
       "      <td>21.893552</td>\n",
       "      <td>20.946511</td>\n",
       "      <td>20.155151</td>\n",
       "      <td>57.492762</td>\n",
       "      <td>-0.009299</td>\n",
       "      <td>0.001061</td>\n",
       "      <td>-0.017164</td>\n",
       "      <td>-0.085846</td>\n",
       "      <td>-0.004636</td>\n",
       "      <td>0.017903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24781</th>\n",
       "      <td>11100.0</td>\n",
       "      <td>-245.065071</td>\n",
       "      <td>151.064511</td>\n",
       "      <td>-38.958297</td>\n",
       "      <td>64.883404</td>\n",
       "      <td>-13.837212</td>\n",
       "      <td>17.243977</td>\n",
       "      <td>27.705282</td>\n",
       "      <td>1.842497</td>\n",
       "      <td>-5.586764</td>\n",
       "      <td>-33.966514</td>\n",
       "      <td>8.911245</td>\n",
       "      <td>-10.995370</td>\n",
       "      <td>-23.489653</td>\n",
       "      <td>-1.413575</td>\n",
       "      <td>1.952128</td>\n",
       "      <td>12.300586</td>\n",
       "      <td>-15.156877</td>\n",
       "      <td>4.555376</td>\n",
       "      <td>6.508424</td>\n",
       "      <td>-22.563778</td>\n",
       "      <td>-0.481283</td>\n",
       "      <td>-14.654238</td>\n",
       "      <td>-10.456867</td>\n",
       "      <td>12.072084</td>\n",
       "      <td>-9.021553</td>\n",
       "      <td>-0.989671</td>\n",
       "      <td>14.412788</td>\n",
       "      <td>3.237931</td>\n",
       "      <td>1.414593</td>\n",
       "      <td>1.183669</td>\n",
       "      <td>0.336710</td>\n",
       "      <td>3.661523</td>\n",
       "      <td>-3.547981</td>\n",
       "      <td>-1.043097</td>\n",
       "      <td>2.833205</td>\n",
       "      <td>-2.267681</td>\n",
       "      <td>1.285617</td>\n",
       "      <td>-4.658701</td>\n",
       "      <td>-1.119360</td>\n",
       "      <td>5.155084</td>\n",
       "      <td>0.536894</td>\n",
       "      <td>0.523960</td>\n",
       "      <td>0.559828</td>\n",
       "      <td>0.618278</td>\n",
       "      <td>0.647445</td>\n",
       "      <td>0.541000</td>\n",
       "      <td>0.509117</td>\n",
       "      <td>0.456515</td>\n",
       "      <td>0.526898</td>\n",
       "      <td>0.657898</td>\n",
       "      <td>0.624820</td>\n",
       "      <td>0.595695</td>\n",
       "      <td>0.108796</td>\n",
       "      <td>0.541510</td>\n",
       "      <td>1.494819</td>\n",
       "      <td>0.864747</td>\n",
       "      <td>1.438759</td>\n",
       "      <td>2.012480</td>\n",
       "      <td>3.114951</td>\n",
       "      <td>21.496714</td>\n",
       "      <td>31.904784</td>\n",
       "      <td>25.889306</td>\n",
       "      <td>36.432177</td>\n",
       "      <td>13.280785</td>\n",
       "      <td>1.619520</td>\n",
       "      <td>0.770733</td>\n",
       "      <td>3.219379</td>\n",
       "      <td>7.050670</td>\n",
       "      <td>6.812521</td>\n",
       "      <td>4.209394</td>\n",
       "      <td>7.169079</td>\n",
       "      <td>5.664094</td>\n",
       "      <td>1.260896</td>\n",
       "      <td>0.852129</td>\n",
       "      <td>0.549142</td>\n",
       "      <td>0.194683</td>\n",
       "      <td>0.094934</td>\n",
       "      <td>0.271491</td>\n",
       "      <td>0.580219</td>\n",
       "      <td>5.189417</td>\n",
       "      <td>8.306912</td>\n",
       "      <td>4.077939</td>\n",
       "      <td>2.129786</td>\n",
       "      <td>2.072622</td>\n",
       "      <td>5.182110</td>\n",
       "      <td>6.749469</td>\n",
       "      <td>3.834243</td>\n",
       "      <td>3.135310</td>\n",
       "      <td>2.685067</td>\n",
       "      <td>4.049170</td>\n",
       "      <td>1.950059</td>\n",
       "      <td>1.223609</td>\n",
       "      <td>2.247702</td>\n",
       "      <td>3.884009</td>\n",
       "      <td>1.774382</td>\n",
       "      <td>1.170372</td>\n",
       "      <td>1.111732</td>\n",
       "      <td>1.442600</td>\n",
       "      <td>1.909349</td>\n",
       "      <td>2.164764</td>\n",
       "      <td>1.883826</td>\n",
       "      <td>0.609570</td>\n",
       "      <td>0.370517</td>\n",
       "      <td>0.463228</td>\n",
       "      <td>0.441120</td>\n",
       "      <td>0.462427</td>\n",
       "      <td>0.293852</td>\n",
       "      <td>0.194079</td>\n",
       "      <td>0.171825</td>\n",
       "      <td>0.133636</td>\n",
       "      <td>1.031969</td>\n",
       "      <td>0.540905</td>\n",
       "      <td>0.091394</td>\n",
       "      <td>0.020530</td>\n",
       "      <td>0.002251</td>\n",
       "      <td>0.012173</td>\n",
       "      <td>0.062106</td>\n",
       "      <td>0.123772</td>\n",
       "      <td>0.368276</td>\n",
       "      <td>0.797135</td>\n",
       "      <td>0.357009</td>\n",
       "      <td>0.314608</td>\n",
       "      <td>0.526035</td>\n",
       "      <td>0.464763</td>\n",
       "      <td>0.494539</td>\n",
       "      <td>0.858060</td>\n",
       "      <td>1.988180</td>\n",
       "      <td>1.387661</td>\n",
       "      <td>0.792763</td>\n",
       "      <td>0.673456</td>\n",
       "      <td>1.040225</td>\n",
       "      <td>0.848740</td>\n",
       "      <td>0.204700</td>\n",
       "      <td>0.034160</td>\n",
       "      <td>0.019386</td>\n",
       "      <td>0.024073</td>\n",
       "      <td>0.043195</td>\n",
       "      <td>0.051860</td>\n",
       "      <td>0.063158</td>\n",
       "      <td>0.088456</td>\n",
       "      <td>0.097171</td>\n",
       "      <td>0.160235</td>\n",
       "      <td>0.138793</td>\n",
       "      <td>0.066637</td>\n",
       "      <td>0.031959</td>\n",
       "      <td>0.007413</td>\n",
       "      <td>0.002262</td>\n",
       "      <td>0.002185</td>\n",
       "      <td>0.009627</td>\n",
       "      <td>0.038512</td>\n",
       "      <td>0.076637</td>\n",
       "      <td>0.118891</td>\n",
       "      <td>0.109286</td>\n",
       "      <td>0.052654</td>\n",
       "      <td>0.018820</td>\n",
       "      <td>0.005708</td>\n",
       "      <td>0.004651</td>\n",
       "      <td>0.005241</td>\n",
       "      <td>0.001435</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>4.148847e-07</td>\n",
       "      <td>3.101894e-07</td>\n",
       "      <td>3.152449e-07</td>\n",
       "      <td>2.569581e-07</td>\n",
       "      <td>1.733312e-07</td>\n",
       "      <td>8.675330e-08</td>\n",
       "      <td>4.705479e-08</td>\n",
       "      <td>3.916420e-08</td>\n",
       "      <td>3.369339e-08</td>\n",
       "      <td>2.982289e-08</td>\n",
       "      <td>2.693547e-08</td>\n",
       "      <td>2.465684e-08</td>\n",
       "      <td>2.287610e-08</td>\n",
       "      <td>2.143892e-08</td>\n",
       "      <td>2.028743e-08</td>\n",
       "      <td>1.939513e-08</td>\n",
       "      <td>1.871170e-08</td>\n",
       "      <td>1.823823e-08</td>\n",
       "      <td>1.794401e-08</td>\n",
       "      <td>14.076762</td>\n",
       "      <td>17.334168</td>\n",
       "      <td>23.895092</td>\n",
       "      <td>16.602850</td>\n",
       "      <td>24.084574</td>\n",
       "      <td>30.465436</td>\n",
       "      <td>46.417020</td>\n",
       "      <td>0.000386</td>\n",
       "      <td>0.020902</td>\n",
       "      <td>0.035148</td>\n",
       "      <td>-0.054016</td>\n",
       "      <td>-0.017736</td>\n",
       "      <td>0.006993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24782</th>\n",
       "      <td>11100.0</td>\n",
       "      <td>-248.627770</td>\n",
       "      <td>155.198986</td>\n",
       "      <td>-34.390508</td>\n",
       "      <td>46.996799</td>\n",
       "      <td>-19.865917</td>\n",
       "      <td>8.591392</td>\n",
       "      <td>24.312880</td>\n",
       "      <td>14.931386</td>\n",
       "      <td>15.091263</td>\n",
       "      <td>-24.988573</td>\n",
       "      <td>-4.616273</td>\n",
       "      <td>-8.151826</td>\n",
       "      <td>-5.866359</td>\n",
       "      <td>-10.660226</td>\n",
       "      <td>-26.923120</td>\n",
       "      <td>8.691167</td>\n",
       "      <td>3.961607</td>\n",
       "      <td>-6.163567</td>\n",
       "      <td>-0.486175</td>\n",
       "      <td>0.843321</td>\n",
       "      <td>-0.647939</td>\n",
       "      <td>-7.275344</td>\n",
       "      <td>0.497309</td>\n",
       "      <td>-14.372342</td>\n",
       "      <td>-5.901338</td>\n",
       "      <td>5.480820</td>\n",
       "      <td>-12.958535</td>\n",
       "      <td>-0.361551</td>\n",
       "      <td>2.497690</td>\n",
       "      <td>3.464214</td>\n",
       "      <td>7.193841</td>\n",
       "      <td>-1.106161</td>\n",
       "      <td>7.100684</td>\n",
       "      <td>3.597750</td>\n",
       "      <td>0.144246</td>\n",
       "      <td>1.816157</td>\n",
       "      <td>-4.186132</td>\n",
       "      <td>-3.398857</td>\n",
       "      <td>-1.743872</td>\n",
       "      <td>-0.519501</td>\n",
       "      <td>0.586439</td>\n",
       "      <td>0.582491</td>\n",
       "      <td>0.615522</td>\n",
       "      <td>0.557301</td>\n",
       "      <td>0.505101</td>\n",
       "      <td>0.517623</td>\n",
       "      <td>0.517072</td>\n",
       "      <td>0.577805</td>\n",
       "      <td>0.631334</td>\n",
       "      <td>0.622539</td>\n",
       "      <td>0.611907</td>\n",
       "      <td>0.599925</td>\n",
       "      <td>0.105920</td>\n",
       "      <td>0.467849</td>\n",
       "      <td>1.076475</td>\n",
       "      <td>0.793000</td>\n",
       "      <td>1.439990</td>\n",
       "      <td>2.131597</td>\n",
       "      <td>9.019579</td>\n",
       "      <td>32.419324</td>\n",
       "      <td>35.022377</td>\n",
       "      <td>20.609777</td>\n",
       "      <td>13.506990</td>\n",
       "      <td>2.347880</td>\n",
       "      <td>0.966765</td>\n",
       "      <td>1.660041</td>\n",
       "      <td>3.588389</td>\n",
       "      <td>4.513102</td>\n",
       "      <td>3.573969</td>\n",
       "      <td>2.361245</td>\n",
       "      <td>1.280359</td>\n",
       "      <td>0.467326</td>\n",
       "      <td>0.134435</td>\n",
       "      <td>0.144517</td>\n",
       "      <td>0.577176</td>\n",
       "      <td>1.394501</td>\n",
       "      <td>1.915011</td>\n",
       "      <td>8.929031</td>\n",
       "      <td>9.435276</td>\n",
       "      <td>2.215534</td>\n",
       "      <td>4.382540</td>\n",
       "      <td>5.678197</td>\n",
       "      <td>7.023208</td>\n",
       "      <td>12.068122</td>\n",
       "      <td>7.764633</td>\n",
       "      <td>5.108208</td>\n",
       "      <td>6.621411</td>\n",
       "      <td>8.702986</td>\n",
       "      <td>7.613229</td>\n",
       "      <td>9.046283</td>\n",
       "      <td>10.816843</td>\n",
       "      <td>9.685508</td>\n",
       "      <td>5.519378</td>\n",
       "      <td>3.727723</td>\n",
       "      <td>4.177566</td>\n",
       "      <td>6.447048</td>\n",
       "      <td>5.834067</td>\n",
       "      <td>3.144169</td>\n",
       "      <td>2.718003</td>\n",
       "      <td>1.080829</td>\n",
       "      <td>0.877817</td>\n",
       "      <td>0.646774</td>\n",
       "      <td>0.726231</td>\n",
       "      <td>0.339065</td>\n",
       "      <td>0.121204</td>\n",
       "      <td>0.057631</td>\n",
       "      <td>0.046279</td>\n",
       "      <td>0.045617</td>\n",
       "      <td>0.040016</td>\n",
       "      <td>0.020995</td>\n",
       "      <td>0.007697</td>\n",
       "      <td>0.031670</td>\n",
       "      <td>0.093844</td>\n",
       "      <td>0.167326</td>\n",
       "      <td>0.213018</td>\n",
       "      <td>0.240931</td>\n",
       "      <td>0.270418</td>\n",
       "      <td>0.334137</td>\n",
       "      <td>0.375846</td>\n",
       "      <td>0.398335</td>\n",
       "      <td>0.218672</td>\n",
       "      <td>0.513310</td>\n",
       "      <td>0.353635</td>\n",
       "      <td>0.145692</td>\n",
       "      <td>0.174510</td>\n",
       "      <td>0.272515</td>\n",
       "      <td>0.142709</td>\n",
       "      <td>0.051990</td>\n",
       "      <td>0.039068</td>\n",
       "      <td>0.091555</td>\n",
       "      <td>0.136489</td>\n",
       "      <td>0.160709</td>\n",
       "      <td>0.172265</td>\n",
       "      <td>0.203239</td>\n",
       "      <td>0.118917</td>\n",
       "      <td>0.078777</td>\n",
       "      <td>0.130462</td>\n",
       "      <td>0.046902</td>\n",
       "      <td>0.027368</td>\n",
       "      <td>0.012811</td>\n",
       "      <td>0.012970</td>\n",
       "      <td>0.006990</td>\n",
       "      <td>0.009978</td>\n",
       "      <td>0.027250</td>\n",
       "      <td>0.025636</td>\n",
       "      <td>0.025410</td>\n",
       "      <td>0.027141</td>\n",
       "      <td>0.020168</td>\n",
       "      <td>0.021944</td>\n",
       "      <td>0.011948</td>\n",
       "      <td>0.008220</td>\n",
       "      <td>0.026534</td>\n",
       "      <td>0.040798</td>\n",
       "      <td>0.040358</td>\n",
       "      <td>0.039717</td>\n",
       "      <td>0.028370</td>\n",
       "      <td>0.018373</td>\n",
       "      <td>0.005395</td>\n",
       "      <td>0.000444</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>5.600791e-07</td>\n",
       "      <td>4.436708e-07</td>\n",
       "      <td>3.612102e-07</td>\n",
       "      <td>3.039560e-07</td>\n",
       "      <td>1.974624e-07</td>\n",
       "      <td>7.464940e-08</td>\n",
       "      <td>3.639231e-08</td>\n",
       "      <td>3.299891e-08</td>\n",
       "      <td>3.035463e-08</td>\n",
       "      <td>2.815176e-08</td>\n",
       "      <td>2.636237e-08</td>\n",
       "      <td>2.484636e-08</td>\n",
       "      <td>2.361095e-08</td>\n",
       "      <td>2.257019e-08</td>\n",
       "      <td>2.170813e-08</td>\n",
       "      <td>2.102621e-08</td>\n",
       "      <td>2.049209e-08</td>\n",
       "      <td>2.011938e-08</td>\n",
       "      <td>1.988239e-08</td>\n",
       "      <td>14.235088</td>\n",
       "      <td>16.730982</td>\n",
       "      <td>22.263361</td>\n",
       "      <td>20.216165</td>\n",
       "      <td>21.885407</td>\n",
       "      <td>29.649005</td>\n",
       "      <td>47.240011</td>\n",
       "      <td>0.014806</td>\n",
       "      <td>-0.003009</td>\n",
       "      <td>-0.001559</td>\n",
       "      <td>-0.052127</td>\n",
       "      <td>-0.002941</td>\n",
       "      <td>-0.016477</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            id        mfcc        mfcc       mfcc       mfcc       mfcc  \\\n",
       "24778  11100.0 -503.800568  159.991782 -16.537968  37.070034  -4.459876   \n",
       "24779  11100.0 -442.607326  160.302527 -31.399481  21.130893 -10.974108   \n",
       "24780  11100.0 -485.288794  183.993909 -26.067565  11.609672  -8.942526   \n",
       "24781  11100.0 -245.065071  151.064511 -38.958297  64.883404 -13.837212   \n",
       "24782  11100.0 -248.627770  155.198986 -34.390508  46.996799 -19.865917   \n",
       "\n",
       "            mfcc       mfcc       mfcc       mfcc       mfcc      mfcc  \\\n",
       "24778   1.584697  -3.264234   6.204571  -6.923838  -0.757389  2.842366   \n",
       "24779   4.796752  -5.693317   3.232060   0.999348  -0.711150  2.550994   \n",
       "24780  -4.337396   1.094897   0.014207  -5.998587   6.376794  5.000262   \n",
       "24781  17.243977  27.705282   1.842497  -5.586764 -33.966514  8.911245   \n",
       "24782   8.591392  24.312880  14.931386  15.091263 -24.988573 -4.616273   \n",
       "\n",
       "            mfcc       mfcc       mfcc       mfcc       mfcc       mfcc  \\\n",
       "24778  -3.018444   5.399441   2.729970   4.665108  -9.940535  -2.158047   \n",
       "24779  -4.104016   3.072689   3.371486   3.884172  -9.338989  -4.311083   \n",
       "24780 -11.034626  -0.015949   4.268690  -3.595981  -8.744786  -1.217848   \n",
       "24781 -10.995370 -23.489653  -1.413575   1.952128  12.300586 -15.156877   \n",
       "24782  -8.151826  -5.866359 -10.660226 -26.923120   8.691167   3.961607   \n",
       "\n",
       "           mfcc      mfcc       mfcc      mfcc       mfcc       mfcc  \\\n",
       "24778 -1.898706 -0.797725  -0.538610 -3.721104  -6.202170  -9.058631   \n",
       "24779 -2.956424 -4.952378  -3.204237 -1.798397  -3.337122  -2.487623   \n",
       "24780 -5.092902 -2.665413  -1.058037 -4.575139  -8.019194  -7.963319   \n",
       "24781  4.555376  6.508424 -22.563778 -0.481283 -14.654238 -10.456867   \n",
       "24782 -6.163567 -0.486175   0.843321 -0.647939  -7.275344   0.497309   \n",
       "\n",
       "            mfcc       mfcc      mfcc       mfcc      mfcc      mfcc  \\\n",
       "24778   0.302317  -6.527288 -4.754056  -3.112750  0.613957  0.508048   \n",
       "24779   3.405535  -3.631716 -1.138690  -3.985064  0.040738  4.584732   \n",
       "24780  -1.761272 -10.511799 -9.388931  -6.599813 -3.779780 -1.163284   \n",
       "24781  12.072084  -9.021553 -0.989671  14.412788  3.237931  1.414593   \n",
       "24782 -14.372342  -5.901338  5.480820 -12.958535 -0.361551  2.497690   \n",
       "\n",
       "           mfcc      mfcc      mfcc      mfcc      mfcc      mfcc      mfcc  \\\n",
       "24778 -2.096162  1.506327  1.187206  1.999975  1.129699  6.913423  2.863409   \n",
       "24779 -1.081874 -0.537442  0.576348  0.446506 -1.680391 -0.161449 -2.273001   \n",
       "24780 -2.449122 -1.393838 -0.807064 -0.655574 -4.885306  0.219818  4.619997   \n",
       "24781  1.183669  0.336710  3.661523 -3.547981 -1.043097  2.833205 -2.267681   \n",
       "24782  3.464214  7.193841 -1.106161  7.100684  3.597750  0.144246  1.816157   \n",
       "\n",
       "           mfcc       mfcc       mfcc      mfcc    chroma    chroma    chroma  \\\n",
       "24778 -0.923728  -0.756724  -0.185686 -0.967237  0.497434  0.479296  0.526856   \n",
       "24779 -3.235756  -5.072984  -3.912861 -1.930381  0.632974  0.534964  0.563326   \n",
       "24780  8.157972  11.245900  11.076964  6.106433  0.472416  0.370513  0.329820   \n",
       "24781  1.285617  -4.658701  -1.119360  5.155084  0.536894  0.523960  0.559828   \n",
       "24782 -4.186132  -3.398857  -1.743872 -0.519501  0.586439  0.582491  0.615522   \n",
       "\n",
       "         chroma    chroma    chroma    chroma    chroma    chroma    chroma  \\\n",
       "24778  0.480500  0.492818  0.517734  0.575788  0.639266  0.614953  0.520258   \n",
       "24779  0.562226  0.545896  0.483770  0.442797  0.494035  0.542529  0.576757   \n",
       "24780  0.346764  0.456383  0.665538  0.642137  0.534971  0.478117  0.462401   \n",
       "24781  0.618278  0.647445  0.541000  0.509117  0.456515  0.526898  0.657898   \n",
       "24782  0.557301  0.505101  0.517623  0.517072  0.577805  0.631334  0.622539   \n",
       "\n",
       "         chroma    chroma       mel       mel       mel       mel       mel  \\\n",
       "24778  0.534425  0.521624  0.000793  0.005434  0.030386  0.006876  0.006212   \n",
       "24779  0.589167  0.653003  0.001122  0.006539  0.036587  0.013001  0.013548   \n",
       "24780  0.381104  0.360215  0.000522  0.001570  0.005693  0.006369  0.009270   \n",
       "24781  0.624820  0.595695  0.108796  0.541510  1.494819  0.864747  1.438759   \n",
       "24782  0.611907  0.599925  0.105920  0.467849  1.076475  0.793000  1.439990   \n",
       "\n",
       "            mel       mel        mel        mel        mel        mel  \\\n",
       "24778  0.009032  0.059345   0.094872   0.042270   0.020747   0.055137   \n",
       "24779  0.013358  0.015413   0.063964   0.060538   0.070826   0.159765   \n",
       "24780  0.033299  0.194070   0.129483   0.045789   0.003905   0.001021   \n",
       "24781  2.012480  3.114951  21.496714  31.904784  25.889306  36.432177   \n",
       "24782  2.131597  9.019579  32.419324  35.022377  20.609777  13.506990   \n",
       "\n",
       "             mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778   0.036292  0.003616  0.011298  0.057143  0.090732  0.046198  0.043124   \n",
       "24779   0.199303  0.054368  0.009223  0.063817  0.053219  0.120010  0.245420   \n",
       "24780   0.002371  0.014030  0.046403  0.106909  0.164091  0.090521  0.086963   \n",
       "24781  13.280785  1.619520  0.770733  3.219379  7.050670  6.812521  4.209394   \n",
       "24782   2.347880  0.966765  1.660041  3.588389  4.513102  3.573969  2.361245   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.039419  0.044442  0.026989  0.021998  0.088756  0.081188  0.095127   \n",
       "24779  0.271008  0.344488  0.079667  0.072464  0.166110  0.132764  0.255801   \n",
       "24780  0.045958  0.074889  0.110980  0.049130  0.049725  0.061341  0.051585   \n",
       "24781  7.169079  5.664094  1.260896  0.852129  0.549142  0.194683  0.094934   \n",
       "24782  1.280359  0.467326  0.134435  0.144517  0.577176  1.394501  1.915011   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel        mel  \\\n",
       "24778  0.057110  0.063391  0.052137  0.061667  0.052075  0.037905   0.012853   \n",
       "24779  0.269304  0.199609  0.077808  0.054156  0.045447  0.033343   0.043260   \n",
       "24780  0.066949  0.188229  0.188903  0.157806  0.096955  0.022311   0.068481   \n",
       "24781  0.271491  0.580219  5.189417  8.306912  4.077939  2.129786   2.072622   \n",
       "24782  8.929031  9.435276  2.215534  4.382540  5.678197  7.023208  12.068122   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel        mel  \\\n",
       "24778  0.021025  0.008507  0.006129  0.012864  0.035194  0.050898   0.052126   \n",
       "24779  0.080957  0.161356  0.161980  0.165373  0.261900  0.237476   0.177378   \n",
       "24780  0.096177  0.057816  0.095490  0.054344  0.041118  0.021336   0.008077   \n",
       "24781  5.182110  6.749469  3.834243  3.135310  2.685067  4.049170   1.950059   \n",
       "24782  7.764633  5.108208  6.621411  8.702986  7.613229  9.046283  10.816843   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.017207  0.024111  0.013910  0.011910  0.016019  0.018280  0.013507   \n",
       "24779  0.128297  0.091435  0.041402  0.071076  0.108694  0.132639  0.028384   \n",
       "24780  0.032519  0.128685  0.061266  0.018746  0.029230  0.013624  0.015305   \n",
       "24781  1.223609  2.247702  3.884009  1.774382  1.170372  1.111732  1.442600   \n",
       "24782  9.685508  5.519378  3.727723  4.177566  6.447048  5.834067  3.144169   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.003757  0.002855  0.003046  0.001857  0.002147  0.002103  0.003201   \n",
       "24779  0.018485  0.010136  0.008793  0.010192  0.014159  0.006357  0.041276   \n",
       "24780  0.034468  0.014040  0.009790  0.006121  0.004840  0.015512  0.008388   \n",
       "24781  1.909349  2.164764  1.883826  0.609570  0.370517  0.463228  0.441120   \n",
       "24782  2.718003  1.080829  0.877817  0.646774  0.726231  0.339065  0.121204   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.008896  0.005241  0.003217  0.005578  0.004855  0.004800  0.001446   \n",
       "24779  0.035302  0.019856  0.015073  0.012102  0.008323  0.010512  0.018413   \n",
       "24780  0.017746  0.013110  0.005659  0.008049  0.007609  0.011967  0.004412   \n",
       "24781  0.462427  0.293852  0.194079  0.171825  0.133636  1.031969  0.540905   \n",
       "24782  0.057631  0.046279  0.045617  0.040016  0.020995  0.007697  0.031670   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.000582  0.001034  0.001469  0.000948  0.000737  0.001306  0.002862   \n",
       "24779  0.007697  0.002625  0.002576  0.002031  0.004848  0.004092  0.004719   \n",
       "24780  0.001054  0.001223  0.001630  0.001185  0.000486  0.000696  0.000943   \n",
       "24781  0.091394  0.020530  0.002251  0.012173  0.062106  0.123772  0.368276   \n",
       "24782  0.093844  0.167326  0.213018  0.240931  0.270418  0.334137  0.375846   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.001120  0.000777  0.000475  0.000245  0.000148  0.000372  0.001103   \n",
       "24779  0.003144  0.001334  0.000752  0.000601  0.000957  0.002569  0.003178   \n",
       "24780  0.000466  0.000151  0.000140  0.000110  0.000067  0.000088  0.000222   \n",
       "24781  0.797135  0.357009  0.314608  0.526035  0.464763  0.494539  0.858060   \n",
       "24782  0.398335  0.218672  0.513310  0.353635  0.145692  0.174510  0.272515   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.001979  0.001484  0.001756  0.000432  0.000250  0.000339  0.000319   \n",
       "24779  0.002321  0.001084  0.000905  0.000736  0.000702  0.000594  0.000371   \n",
       "24780  0.000344  0.000324  0.000386  0.000257  0.000242  0.000322  0.000309   \n",
       "24781  1.988180  1.387661  0.792763  0.673456  1.040225  0.848740  0.204700   \n",
       "24782  0.142709  0.051990  0.039068  0.091555  0.136489  0.160709  0.172265   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.000238  0.000163  0.000097  0.000072  0.000047  0.000091  0.000105   \n",
       "24779  0.000377  0.000200  0.000216  0.000227  0.000139  0.000093  0.000149   \n",
       "24780  0.000192  0.000149  0.000057  0.000029  0.000041  0.000033  0.000033   \n",
       "24781  0.034160  0.019386  0.024073  0.043195  0.051860  0.063158  0.088456   \n",
       "24782  0.203239  0.118917  0.078777  0.130462  0.046902  0.027368  0.012811   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.000041  0.000033  0.000038  0.000029  0.000019  0.000019  0.000017   \n",
       "24779  0.000151  0.000127  0.000171  0.000140  0.000114  0.000106  0.000077   \n",
       "24780  0.000015  0.000012  0.000009  0.000011  0.000006  0.000005  0.000003   \n",
       "24781  0.097171  0.160235  0.138793  0.066637  0.031959  0.007413  0.002262   \n",
       "24782  0.012970  0.006990  0.009978  0.027250  0.025636  0.025410  0.027141   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.000017  0.000008  0.000005  0.000009  0.000015  0.000020  0.000010   \n",
       "24779  0.000030  0.000022  0.000032  0.000041  0.000038  0.000034  0.000099   \n",
       "24780  0.000004  0.000003  0.000004  0.000007  0.000006  0.000006  0.000002   \n",
       "24781  0.002185  0.009627  0.038512  0.076637  0.118891  0.109286  0.052654   \n",
       "24782  0.020168  0.021944  0.011948  0.008220  0.026534  0.040798  0.040358   \n",
       "\n",
       "            mel       mel       mel       mel       mel       mel       mel  \\\n",
       "24778  0.000006  0.000007  0.000009  0.000006  0.000008  0.000008  0.000007   \n",
       "24779  0.000090  0.000032  0.000026  0.000026  0.000018  0.000023  0.000032   \n",
       "24780  0.000003  0.000003  0.000002  0.000002  0.000003  0.000003  0.000003   \n",
       "24781  0.018820  0.005708  0.004651  0.005241  0.001435  0.000033  0.000002   \n",
       "24782  0.039717  0.028370  0.018373  0.005395  0.000444  0.000008  0.000001   \n",
       "\n",
       "                mel           mel           mel           mel           mel  \\\n",
       "24778  7.056818e-06  9.596689e-06  6.661124e-06  5.487163e-06  3.610811e-06   \n",
       "24779  1.272312e-05  8.955307e-06  1.367225e-05  7.171851e-06  4.043730e-06   \n",
       "24780  2.407622e-06  1.514996e-06  1.679450e-06  1.593449e-06  1.080508e-06   \n",
       "24781  4.148847e-07  3.101894e-07  3.152449e-07  2.569581e-07  1.733312e-07   \n",
       "24782  5.600791e-07  4.436708e-07  3.612102e-07  3.039560e-07  1.974624e-07   \n",
       "\n",
       "                mel           mel           mel           mel           mel  \\\n",
       "24778  1.066116e-06  3.940568e-08  4.386587e-09  3.934327e-09  3.616842e-09   \n",
       "24779  1.242139e-06  8.010924e-08  1.141345e-08  9.478378e-09  8.492636e-09   \n",
       "24780  2.516526e-07  7.536460e-08  4.531382e-08  3.777022e-08  3.345398e-08   \n",
       "24781  8.675330e-08  4.705479e-08  3.916420e-08  3.369339e-08  2.982289e-08   \n",
       "24782  7.464940e-08  3.639231e-08  3.299891e-08  3.035463e-08  2.815176e-08   \n",
       "\n",
       "                mel           mel           mel           mel           mel  \\\n",
       "24778  3.364994e-09  3.155534e-09  2.986337e-09  2.844588e-09  2.727968e-09   \n",
       "24779  7.798572e-09  7.266692e-09  6.860868e-09  6.536293e-09  6.277775e-09   \n",
       "24780  3.045017e-08  2.817581e-08  2.645287e-08  2.508716e-08  2.400698e-08   \n",
       "24781  2.693547e-08  2.465684e-08  2.287610e-08  2.143892e-08  2.028743e-08   \n",
       "24782  2.636237e-08  2.484636e-08  2.361095e-08  2.257019e-08  2.170813e-08   \n",
       "\n",
       "                mel           mel           mel           mel   contrast  \\\n",
       "24778  2.635663e-09  2.563692e-09  2.513333e-09  2.481457e-09  18.123065   \n",
       "24779  6.078889e-09  5.926311e-09  5.821557e-09  5.755594e-09  16.209617   \n",
       "24780  2.317983e-08  2.254846e-08  2.211556e-08  2.184430e-08  19.568966   \n",
       "24781  1.939513e-08  1.871170e-08  1.823823e-08  1.794401e-08  14.076762   \n",
       "24782  2.102621e-08  2.049209e-08  2.011938e-08  1.988239e-08  14.235088   \n",
       "\n",
       "        contrast   contrast   contrast   contrast   contrast   contrast  \\\n",
       "24778  16.864310  20.108288  18.542917  19.385362  19.506113  60.697950   \n",
       "24779  15.989443  18.808640  18.342743  19.651828  19.176303  58.457264   \n",
       "24780  17.647668  21.506072  21.893552  20.946511  20.155151  57.492762   \n",
       "24781  17.334168  23.895092  16.602850  24.084574  30.465436  46.417020   \n",
       "24782  16.730982  22.263361  20.216165  21.885407  29.649005  47.240011   \n",
       "\n",
       "        tonnetz   tonnetz   tonnetz   tonnetz   tonnetz   tonnetz  \n",
       "24778  0.007471  0.020963  0.050210 -0.102916 -0.026035  0.005766  \n",
       "24779  0.010165  0.010087  0.037996 -0.079704 -0.029291 -0.001756  \n",
       "24780 -0.009299  0.001061 -0.017164 -0.085846 -0.004636  0.017903  \n",
       "24781  0.000386  0.020902  0.035148 -0.054016 -0.017736  0.006993  \n",
       "24782  0.014806 -0.003009 -0.001559 -0.052127 -0.002941 -0.016477  "
      ]
     },
     "execution_count": 17,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape # (64483, 194)\n",
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 610,
     "status": "ok",
     "timestamp": 1549506181390,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "_xRFGg5fNFw5",
    "outputId": "79df278b-b6d7-4c98-a803-dbe0a923be67"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(537,)"
      ]
     },
     "execution_count": 18,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check how many ids were loaded into dataframe...not going to load all of them\n",
    "df.id.unique().shape # (537,)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "EBtL8R_khAdk"
   },
   "outputs": [],
   "source": [
    "# Split the dataframe into x and y\n",
    "y = df.id\n",
    "x = df.iloc[:,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZsfssuE_ggdc"
   },
   "outputs": [],
   "source": [
    "# Train, test, split into training and testing\n",
    "x_train1, x_test, y_train1, y_test = train_test_split(x,y, test_size = 0.1, random_state = 121)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "bi7PGiDcIzGw"
   },
   "outputs": [],
   "source": [
    "# Split the training data into training and validation\n",
    "x_train, x_val, y_train, y_val = train_test_split(x_train1, y_train1, test_size = 0.2, random_state = 121)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 522,
     "status": "ok",
     "timestamp": 1549506183606,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "OYiV2M-blnJR",
    "outputId": "38f8d663-7e27-426a-b9eb-cf0599cb556c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6449,)"
      ]
     },
     "execution_count": 22,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at the shape of each portion of data\n",
    "x_train.shape # (46427, 193)\n",
    "x_val.shape # (11607, 193)\n",
    "x_test.shape # (6449, 193)\n",
    "\n",
    "y_train.shape # (46427,)\n",
    "y_val.shape # (11607,)\n",
    "y_test.shape # (6449,) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 279,
     "status": "ok",
     "timestamp": 1549506184490,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "6aL1SuKXl1Wy",
    "outputId": "b2f11ef6-7f40-40a6-ee2d-3ec2d2db4737"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6449, 1, 193)"
      ]
     },
     "execution_count": 23,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# reshape x datasets from 2D into 3D arrays\n",
    "x_train_array = x_train.values # convert df to array\n",
    "x_train_ = np.reshape(x_train_array, newshape = (46427, 1,193))\n",
    "x_train_.shape # (46427, 1, 193)\n",
    "\n",
    "x_val_array = x_val.values # convert df to array\n",
    "x_val_ = np.reshape(x_val_array, newshape = (11607, 1,193))\n",
    "x_val_.shape # (11607, 1,193)\n",
    "\n",
    "x_test_array = x_test.values # convert df to array\n",
    "x_test_ = np.reshape(x_test_array, newshape = (6449, 1, 193))\n",
    "x_test_.shape # (6449, 1, 193)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 946,
     "status": "ok",
     "timestamp": 1549506187227,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "AYiUjW4gm1ZY",
    "outputId": "e76f8ce4-9845-4248-8b1f-55fbe7c120c6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6449, 537)"
      ]
     },
     "execution_count": 24,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# reshape y data for last layer in models (softmax activation function)\n",
    "y_train_cat = pd.get_dummies(y_train) # converts variables in one column to have their own column\n",
    "y_train_ = y_train_cat.values \n",
    "y_train_.shape # (46427, 537)\n",
    "\n",
    "y_val_cat = pd.get_dummies(y_val) # converts variables in one column to have their own column\n",
    "y_val_ = y_val_cat.values \n",
    "y_val_.shape # (11607, 537)\n",
    "\n",
    "y_test_cat = pd.get_dummies(y_test) # converts variables in one column to have their own column\n",
    "y_test_ = y_test_cat.values \n",
    "y_test_.shape # (6449, 537)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "7wXEGyHyZjlc"
   },
   "source": [
    "# <font color = 'red' >Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_wXa-j-ChlHe"
   },
   "source": [
    "## PReLU Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "FfiiqzAhhkj8"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(123)\n",
    "# Define model as Sequential class\n",
    "model = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 139
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 914,
     "status": "ok",
     "timestamp": 1549490724165,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "uNNOSFXOcFlF",
    "outputId": "ccd03e52-71ad-4bee-f5b5-26b66a0acdf5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n",
      "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n"
     ]
    }
   ],
   "source": [
    "model.add(layers.Dense(128, input_shape = (1,193)))\n",
    "model.add(PReLU(alpha_initializer='zeros'))\n",
    "\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "model.add(Dropout(rate = 0.1))\n",
    "\n",
    "model.add(layers.Dense(256))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "\n",
    "model.add(layers.Dense(542))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "model.add(Dropout(rate = 0.1))\n",
    "\n",
    "model.add(layers.Dense(542))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "\n",
    "model.add(layers.Dense(256))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "model.add(Dropout(rate = 0.1))\n",
    "\n",
    "model.add(layers.Dense(256))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "\n",
    "model.add(layers.Dense(128))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "model.add(Dropout(rate = 0.1))\n",
    "\n",
    "model.add(layers.Dense(128))\n",
    "model.add(PReLU())\n",
    "model.add(PReLU())\n",
    "\n",
    "model.add(layers.Dense(128))\n",
    "model.add(PReLU())\n",
    "\n",
    "model.add(layers.Flatten())\n",
    "model.add(Dense(537, activation = 'softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "YiAoWE0riZHe"
   },
   "outputs": [],
   "source": [
    "opt = Adamax(lr = 1e-3, decay = 1e-5)\n",
    "model.compile(loss = 'categorical_crossentropy', \n",
    "              optimizer = opt,\n",
    "             metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GXm0dYDxQuhj"
   },
   "outputs": [],
   "source": [
    "earlystop = EarlyStopping(monitor='val_loss',\n",
    "                              min_delta=0.001,\n",
    "                              patience=3,\n",
    "                              verbose=0, mode='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 14534,
     "status": "ok",
     "timestamp": 1549492439667,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "XByf38p2iY_-",
    "outputId": "cc597890-bc82-4254-a58a-bdfd1d21b12e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "46427/46427 [==============================] - 28s 598us/step - loss: 0.9288 - acc: 0.7560 - val_loss: 1.5074 - val_acc: 0.7010\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_train_, \n",
    "                    y_train_, \n",
    "                    epochs = 100, \n",
    "                    batch_size = 25,\n",
    "                    validation_data= (x_val_,y_val_), callbacks = [earlystop])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1XFNbiw2nQ19"
   },
   "source": [
    "### Errors & Accuracy Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1202,
     "status": "ok",
     "timestamp": 1549492525980,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "xulF3tQ4nNkH",
    "outputId": "2d00ef42-4d33-4f0a-f6dc-76efe03a5015"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA68AAAGCCAYAAAAPNfUoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiIiIhT8WriIiIiIiIhLz/D5SUPcsuK96yAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Look at the history of the model\n",
    "hist_dense = model.history.history # contains information from fitting model\n",
    "hist_dense.keys() # returns: dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n",
    "\n",
    "# Call each key to get the values to plot\n",
    "loss_prelu = hist_dense['loss'] \n",
    "val_loss_prelu = hist_dense['val_loss']\n",
    "acc_prelu = hist_dense['acc']\n",
    "val_acc_prelu = hist_dense['val_acc']\n",
    "\n",
    "epochs = range(1, len(loss_prelu) + 1)\n",
    "\n",
    "# Plot Losses of training and validation sets v. epochs\n",
    "plt.figure(figsize=(25,6))\n",
    "plt.subplot(131)\n",
    "plt.plot(epochs, loss_prelu, 'k', label = 'Training loss')\n",
    "plt.plot(epochs, val_loss_prelu, 'r', label = 'Validation loss')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation loss over epochs')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot accuracy scores over epochs for training and validation sets                      \n",
    "plt.subplot(132)\n",
    "plt.plot(epochs, acc_prelu, 'k', label='Training acc')\n",
    "plt.plot(epochs, val_acc_prelu, 'r', label= 'Validation acc')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation accuracy over epochs [PReLU]')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4add3biujqcD"
   },
   "source": [
    "### Basic Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "MRleeFQfx9Ta"
   },
   "outputs": [],
   "source": [
    "# Use model to predict on testing set\n",
    "y_pred = model.predict(x_test_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "gdk4zTOyZQR0"
   },
   "outputs": [],
   "source": [
    "# Convert y_pred from continuous variables to discreet binary (0,1)\n",
    "yy = []\n",
    "for elem in y_pred: # each element is an array\n",
    "  row = []\n",
    "  for e in elem: # each number in the array\n",
    "    if e == elem.max(): # since softmax outputs probability...\n",
    "      row.append(1) # assign highest prob as 1\n",
    "    else:\n",
    "      row.append(0) # assign all other probs as 0\n",
    "  yy.append(row)\n",
    "y_pred_ = np.array(yy)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 380,
     "status": "ok",
     "timestamp": 1549492616061,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "E_h4VLuBoZl-",
    "outputId": "127cc3d5-0dc2-4632-c354-5b487d8b19f6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7021173008041286"
      ]
     },
     "execution_count": 30,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_test_, y_pred_, average= 'weighted') # 0.7021173008041286\n",
    "precision_score(y_test_, y_pred_, average= 'weighted') # 0.7242833973660131\n",
    "accuracy_score(y_test_, y_pred_) # 0.7066211815785393\n",
    "recall_score(y_test_, y_pred_, average= 'weighted') # 0.7066211815785393"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "7UEY58Mzn2VA"
   },
   "source": [
    "### AUC/ ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "_yInNNPxn1aF"
   },
   "outputs": [],
   "source": [
    "# Iterate through y_test_ and y_pred_ to get predictions for each of the 537 IDs\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(y_test_.shape[1]):\n",
    "  # use y_pred not y_pred_, because the roc_curve function needs probability for threshold metric\n",
    "  fpr[i], tpr[i], thresholds = roc_curve(y_test_[:,i], y_pred[:,i]) # fpr = false positive rate, tpr = true positive rate\n",
    "  roc_auc[i] = auc(fpr[i], tpr[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 512
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1355,
     "status": "ok",
     "timestamp": 1549492633078,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "f4Rb16F1Itpp",
    "outputId": "89065826-80ed-4ce4-833b-3f545e41f96f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAHvCAYAAADtrFPzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot some of 537 celebs ROC curve \n",
    "#Seaborns Beautiful Styling\n",
    "sns.set_style(\"darkgrid\", {\"axes.facecolor\": \".9\"})\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "lw = 2 # linewidth\n",
    "for i in range(50):\n",
    "  plt.plot(fpr[i], tpr[i], lw = lw)\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "  plt.yticks([ii/20.0 for ii in range(21)])\n",
    "  plt.xticks([ii/20.0 for ii in range(21)])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for PReLU Model')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Feo69Bi0H-kY"
   },
   "source": [
    "## SeLU Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "2iRZE-whLRW8"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(123)\n",
    "# Define model as Sequential class\n",
    "model_selu = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NGjQGM-gLRUy"
   },
   "outputs": [],
   "source": [
    "model_selu.add(layers.Dense(128, input_shape = (1,193), activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(512, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(512, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "\n",
    "model_selu.add(layers.Dense(512, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(512, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(layers.Dense(256, activation = 'selu'))\n",
    "model_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "model_selu.add(layers.Flatten())\n",
    "model_selu.add(Dense(537, activation = 'softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "P1AB7FkWLRSu"
   },
   "outputs": [],
   "source": [
    "opt = Adamax(lr = 1e-3, decay = 1e-5)\n",
    "model_selu.compile(loss = 'categorical_crossentropy', \n",
    "              optimizer = opt,\n",
    "             metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "nxJrviG1LROl"
   },
   "outputs": [],
   "source": [
    "earlystop = EarlyStopping(monitor='val_loss',\n",
    "                              min_delta=0.001,\n",
    "                              patience=3,\n",
    "                              verbose=0, mode='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 88
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 32952,
     "status": "ok",
     "timestamp": 1549493814591,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "qugsopXKLRFn",
    "outputId": "18707b94-e212-4a35-8e74-b3036201746a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "46427/46427 [==============================] - 26s 564us/step - loss: 0.9875 - acc: 0.7398 - val_loss: 1.6399 - val_acc: 0.6798\n",
      "Epoch 43/100\n",
      "46427/46427 [==============================] - 26s 563us/step - loss: 0.9592 - acc: 0.7484 - val_loss: 1.6330 - val_acc: 0.6842\n"
     ]
    }
   ],
   "source": [
    "history = model_selu.fit(x_train_, \n",
    "                    y_train_, \n",
    "                    epochs = 100, \n",
    "                    batch_size = 25,\n",
    "                    validation_data= (x_val_,y_val_), callbacks = [earlystop])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_wroWXnBJvEu"
   },
   "source": [
    "### Errors & Accuracy Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1270,
     "status": "ok",
     "timestamp": 1549493844855,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "sZUr1RrFJuBd",
    "outputId": "be6e0cc8-e8df-4ac4-a280-43337c75a7fb"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA68AAAGCCAYAAAAPNfUoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Look at the history of the model\n",
    "hist_dense = model_selu.history.history # contains information from fitting model\n",
    "hist_dense.keys() # returns: dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n",
    "\n",
    "# Call each key to get the values to plot\n",
    "loss_selu = hist_dense['loss'] \n",
    "val_loss_selu = hist_dense['val_loss']\n",
    "acc_selu = hist_dense['acc']\n",
    "val_acc_selu = hist_dense['val_acc']\n",
    "\n",
    "epochs = range(1, len(loss_selu) + 1)\n",
    "\n",
    "# Plot Losses of training and validation sets v. epochs\n",
    "plt.figure(figsize=(25,6))\n",
    "plt.subplot(131)\n",
    "plt.plot(epochs, loss_selu, 'k', label = 'Training loss')\n",
    "plt.plot(epochs, val_loss_selu, 'r', label = 'Validation loss')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation loss over epochs')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot accuracy scores over epochs for training and validation sets                      \n",
    "plt.subplot(132)\n",
    "plt.plot(epochs, acc_selu, 'k', label='Training acc')\n",
    "plt.plot(epochs, val_acc_selu, 'r', label= 'Validation acc')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation accuracy over epochs [SeLU]')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4ZE03wihjhce"
   },
   "source": [
    "### Basic Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "nRYzEglWLbCg"
   },
   "outputs": [],
   "source": [
    "# Use model to predict on testing set\n",
    "y_pred = model_selu.predict(x_test_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "x8g_UELSLegc"
   },
   "outputs": [],
   "source": [
    "# Convert y_pred from continuous variables to discreet binary (0,1)\n",
    "yy = []\n",
    "for elem in y_pred: # each element is an array\n",
    "  row = []\n",
    "  for e in elem: # each number in the array\n",
    "    if e == elem.max(): # since softmax outputs probability...\n",
    "      row.append(1) # assign highest prob as 1\n",
    "    else:\n",
    "      row.append(0) # assign all other probs as 0\n",
    "  yy.append(row)\n",
    "y_pred_ = np.array(yy)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1154,
     "status": "ok",
     "timestamp": 1549493916102,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "oDo0jWbIRXEf",
    "outputId": "eb975fe3-1411-4850-aec3-feec08e4842f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6962319739494496"
      ]
     },
     "execution_count": 44,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_test_, y_pred_, average= 'weighted') # 0.6930043492601609\n",
    "precision_score(y_test_, y_pred_, average= 'weighted') # 0.7148270141132256\n",
    "accuracy_score(y_test_, y_pred_) # 0.6962319739494496\n",
    "recall_score(y_test_, y_pred_, average= 'weighted') # 0.6962319739494496"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qnxWs4tuRaoN"
   },
   "source": [
    "### AUC/ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "jp9xrdGYRh2F"
   },
   "outputs": [],
   "source": [
    "# Iterate through y_test_ and y_pred_ to get predictions for each of the 537 IDs\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(y_test_.shape[1]):\n",
    "  # use y_pred not y_pred_, because the roc_curve function needs probability for threshold metric\n",
    "  fpr[i], tpr[i], thresholds = roc_curve(y_test_[:,i], y_pred[:,i]) # fpr = false positive rate, tpr = true positive rate\n",
    "  roc_auc[i] = auc(fpr[i], tpr[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 512
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1590,
     "status": "ok",
     "timestamp": 1549493941437,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "-hzx6VxHReF6",
    "outputId": "9f9d697d-99b3-4c80-9092-e4102a7cee3e"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAHvCAYAAADtrFPzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot some of 537 celebs ROC curve \n",
    "#Seaborns Beautiful Styling\n",
    "sns.set_style(\"darkgrid\", {\"axes.facecolor\": \".9\"})\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "lw = 2 # linewidth\n",
    "for i in range(50):\n",
    "  plt.plot(fpr[i], tpr[i], lw = lw)\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "  plt.yticks([ii/20.0 for ii in range(21)])\n",
    "  plt.xticks([ii/20.0 for ii in range(21)])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for SeLU Model')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-QeBKSGsqb4v"
   },
   "source": [
    "## Talos (Disregard)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "67fSzqWpqpbl"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "wAu_tU9RqeWG"
   },
   "outputs": [],
   "source": [
    "# set parameter space\n",
    "p = {'first_neuron':[64],\n",
    "     'num_neurons': [64],\n",
    "     'hidden_layers':[10,12,15],\n",
    "     'batch_size': [20],\n",
    "     'epochs': [25],\n",
    "     'dropout': [0],\n",
    "     'alpha': [0.001],\n",
    "     'advanced_act': [PReLU()],\n",
    "     'optimizer': [\"sgd\"],\n",
    "     'losses': [\"categorical_crossentropy\"],\n",
    "     'activation':[relu,elu ],\n",
    "     'last_activation': [softmax]}\n",
    "\n",
    "# # set parameter space\n",
    "# p = {'first_neuron':[64,128],\n",
    "#      'num_neurons': [64,128,256, 512],\n",
    "#      'hidden_layers':[10, 11, 12],\n",
    "#      'batch_size': [20,25,30],\n",
    "#      'epochs': [25],\n",
    "#      'dropout': [0],\n",
    "#      'alpha': [0.001, 0.01],\n",
    "#      'advanced_act': [PReLU()],\n",
    "#      'optimizer': [\"sgd\",\"Adam\"],\n",
    "#      'losses': [\"categorical_crossentropy\"],\n",
    "#      'activation':[relu, elu],\n",
    "#      'last_activation': [softmax]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lIdyLgRRqrf9"
   },
   "outputs": [],
   "source": [
    "def best_model(x_train, y_train, x_val, y_val, params):\n",
    "  # Define model as Sequential class\n",
    "  model = Sequential()\n",
    "    \n",
    "  # first layer\n",
    "  model.add(layers.Dense(params['first_neuron'], \n",
    "                           input_shape = (1,193)))\n",
    "  model.add(PReLU(alpha_initializer='zeros', ))\n",
    "    \n",
    "  # possibly add dropout\n",
    "  model.add(Dropout(params['dropout']))\n",
    "    \n",
    "  # add hidden layer\n",
    "  hidden_layers(model, params, y_train.shape[1])\n",
    "            \n",
    "  # add flatten layer for last output layer\n",
    "  model.add(layers.Flatten())\n",
    "    \n",
    "  # add last layer\n",
    "  model.add(Dense(343, activation = params['last_activation']))\n",
    "    \n",
    "  # Compile the model\n",
    "  model.compile(loss = params['losses'],\n",
    "                  optimizer = params['optimizer'],\n",
    "                  metrics = ['acc'])\n",
    "    \n",
    "  # Fit the model\n",
    "  history = model.fit(x_train, y_train, \n",
    "                        validation_data = (x_val, y_val),\n",
    "                        batch_size = params['batch_size'],\n",
    "                        epochs = params['epochs'],\n",
    "                        callbacks=[live()],\n",
    "                        verbose=0)\n",
    "  return history, model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "uziHzKXqxH1_"
   },
   "outputs": [],
   "source": [
    "# run the experiment\n",
    "t = ta.Scan(x = x_train_,\n",
    "            y = y_train_,\n",
    "            x_val = x_test_,\n",
    "            y_val = y_test_,\n",
    "            model = best_model,\n",
    "            params = p,\n",
    "            experiment_no='1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1086,
     "status": "ok",
     "timestamp": 1549314416261,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "sSNWClzhUgFs",
    "outputId": "aa90d35b-d213-463f-e2ca-261b130139a4"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "talos.scan.Scan.Scan"
      ]
     },
     "execution_count": 239,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(t)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gMte_XcUhJov"
   },
   "source": [
    "## CNN w/ LeakyReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "eOkjMdjfb9e0"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(123)\n",
    "# Define model as Sequential class\n",
    "model_cnn = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "vYsUSuaphIwn"
   },
   "outputs": [],
   "source": [
    "# Block 1\n",
    "model_cnn.add(Conv1D(128, (2), strides=(2), padding='same', input_shape = (1,193)))\n",
    "model_cnn.add(LeakyReLU(alpha=.001)) \n",
    "model_cnn.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 2\n",
    "model_cnn.add(Conv1D(256, (2), strides=(2), padding='same'))\n",
    "model_cnn.add(LeakyReLU(alpha=.001)) \n",
    "model_cnn.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 3\n",
    "model_cnn.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnn.add(LeakyReLU(alpha=.001)) \n",
    "model_cnn.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 4\n",
    "model_cnn.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnn.add(LeakyReLU(alpha=.001)) \n",
    "model_cnn.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 5\n",
    "model_cnn.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnn.add(LeakyReLU(alpha=.001)) \n",
    "model_cnn.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Output block\n",
    "model_cnn.add(layers.Flatten())\n",
    "model_cnn.add(Dense(537, activation = 'softmax'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "l40voR15cHge"
   },
   "outputs": [],
   "source": [
    "opt = Adam(lr = 1e-3, decay = 1e-5)\n",
    "model_cnn.compile(loss = 'categorical_crossentropy', \n",
    "              optimizer = opt,\n",
    "             metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "sTkzYXZycHbO"
   },
   "outputs": [],
   "source": [
    "earlystop = EarlyStopping(monitor='val_loss',\n",
    "                              min_delta=0.001,\n",
    "                              patience=5,\n",
    "                              verbose=0, mode='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9Ip-gjidcHUG"
   },
   "outputs": [],
   "source": [
    "history = model_cnn.fit(x_train_, \n",
    "                    y_train_, \n",
    "                    epochs = 100, \n",
    "                    batch_size = 25,\n",
    "                    validation_data= (x_test_,y_test_), callbacks= [earlystop])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gbTX5IoNSEFQ"
   },
   "source": [
    "### Errors & Accuracy Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 712863,
     "status": "ok",
     "timestamp": 1549494684142,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "PYBvaqOXSGSx",
    "outputId": "aec6f05c-08ce-4bed-9ff2-057a04b317d9"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA68AAAGCCAYAAAAPNfUoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Look at the history of the model\n",
    "hist_dense = model_cnn.history.history # contains information from fitting model\n",
    "hist_dense.keys() # returns: dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n",
    "\n",
    "# Call each key to get the values to plot\n",
    "loss_cnn = hist_dense['loss'] \n",
    "val_loss_cnn = hist_dense['val_loss']\n",
    "acc_cnn = hist_dense['acc']\n",
    "val_acc_cnn = hist_dense['val_acc']\n",
    "\n",
    "epochs = range(1, len(loss_cnn) + 1)\n",
    "\n",
    "# Plot Losses of training and validation sets v. epochs\n",
    "plt.figure(figsize=(25,6))\n",
    "plt.subplot(131)\n",
    "plt.plot(epochs, loss_cnn, 'k', label = 'Training loss')\n",
    "plt.plot(epochs, val_loss_cnn, 'r', label = 'Validation loss')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation loss over epochs')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot accuracy scores over epochs for training and validation sets                      \n",
    "plt.subplot(132)\n",
    "plt.plot(epochs, acc_cnn, 'k', label='Training acc')\n",
    "plt.plot(epochs, val_acc_cnn, 'r', label= 'Validation acc')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation accuracy over epochs [Leaky ReLU CNN]')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Qtgm69fVj4s6"
   },
   "source": [
    "### Basic Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4J3DbrkHizH_"
   },
   "outputs": [],
   "source": [
    "# Use model to predict on testing set\n",
    "y_pred = model_cnn.predict(x_test_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "vdfBar3oi2rJ"
   },
   "outputs": [],
   "source": [
    "# Convert y_pred from continuous variables to discreet binary (0,1)\n",
    "yy = []\n",
    "for elem in y_pred: # each element is an array\n",
    "  row = []\n",
    "  for e in elem: # each number in the array\n",
    "    if e == elem.max(): # since softmax outputs probability...\n",
    "      row.append(1) # assign highest prob as 1\n",
    "    else:\n",
    "      row.append(0) # assign all other probs as 0\n",
    "  yy.append(row)\n",
    "y_pred_ = np.array(yy)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1075,
     "status": "ok",
     "timestamp": 1549501433152,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "NtDXR6v1i55q",
    "outputId": "a1417dc3-1641-476a-9b0e-41adafc8176f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6399441773918437"
      ]
     },
     "execution_count": 61,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_test_, y_pred_, average= 'weighted') # 0.6385767271678208\n",
    "precision_score(y_test_, y_pred_, average= 'weighted') # 0.6672402298965948\n",
    "accuracy_score(y_test_, y_pred_) # 0.6399441773918437\n",
    "recall_score(y_test_, y_pred_, average= 'weighted') # 0.6399441773918437"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4Fcm5uLeTCHj"
   },
   "source": [
    "### AUC/ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "vcPSNhFMWylO"
   },
   "outputs": [],
   "source": [
    "# Iterate through y_test_ and y_pred_ to get predictions for each of the 537 IDs\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(y_test_.shape[1]):\n",
    "  # use y_pred not y_pred_, because the roc_curve function needs probability for threshold metric\n",
    "  fpr[i], tpr[i], thresholds = roc_curve(y_test_[:,i], y_pred[:,i]) # fpr = false positive rate, tpr = true positive rate\n",
    "  roc_auc[i] = auc(fpr[i], tpr[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 512
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 715736,
     "status": "ok",
     "timestamp": 1549494695488,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "TAWH1QuIWzoC",
    "outputId": "1f3d4623-058c-4baa-eff6-7226496fcec4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAHvCAYAAADtrFPzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIb4Bj7qibC42sBUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot some of 537 celebs ROC curve \n",
    "#Seaborns Beautiful Styling\n",
    "sns.set_style(\"darkgrid\", {\"axes.facecolor\": \".9\"})\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "lw = 2 # linewidth\n",
    "for i in range(50):\n",
    "  plt.plot(fpr[i], tpr[i], lw = lw)\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "  plt.yticks([ii/20.0 for ii in range(21)])\n",
    "  plt.xticks([ii/20.0 for ii in range(21)])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for Leaky ReLU CNN Model')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "N0E8Jp4_jKgY"
   },
   "source": [
    "## CNN  w/ PReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XxI9onASjJgY"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(121)\n",
    "# Define model as Sequential class\n",
    "model_cnnpl = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7wMW81jMjTDY"
   },
   "outputs": [],
   "source": [
    "# Block 1\n",
    "model_cnnpl.add(Conv1D(128, (2), strides=(2), padding='same', input_shape = (1,193)))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 2\n",
    "model_cnnpl.add(Conv1D(256, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 3\n",
    "model_cnnpl.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 4\n",
    "model_cnnpl.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 5\n",
    "model_cnnpl.add(Conv1D(512, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 6\n",
    "model_cnnpl.add(Conv1D(256, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 7\n",
    "model_cnnpl.add(Conv1D(256, (2), strides=(2), padding='same'))\n",
    "model_cnnpl.add(PReLU())\n",
    "model_cnnpl.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# FC block\n",
    "model_cnnpl.add(layers.Flatten())\n",
    "model_cnnpl.add(Dense(537, activation = 'softmax'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "FGthn29XjS92"
   },
   "outputs": [],
   "source": [
    "opt = Adamax(lr = 1e-3, decay = 1e-5)\n",
    "model_cnnpl.compile(loss = 'categorical_crossentropy', \n",
    "              optimizer = opt,\n",
    "             metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "MQ39bcwqjS1P"
   },
   "outputs": [],
   "source": [
    "earlystop = EarlyStopping(monitor='val_loss',\n",
    "                              min_delta=0.001,\n",
    "                              patience=3,\n",
    "                              verbose=0, mode='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "77kPc9wUjSs8"
   },
   "outputs": [],
   "source": [
    "history = model_cnnpl.fit(x_train_, \n",
    "                    y_train_, \n",
    "                    epochs = 100, \n",
    "                    batch_size = 25,\n",
    "                    validation_data= (x_test_,y_test_), callbacks = [earlystop])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jQPsVMTNXZcq"
   },
   "source": [
    "### Errors & Accuracy Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 608443,
     "status": "ok",
     "timestamp": 1549502104314,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "H2awcUa5Xd_f",
    "outputId": "fbf5f725-2a4d-4b2d-c3cf-37c4e9ae1ea0"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA68AAAGCCAYAAAAPNfUoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRURERERExO79P4tBIT9X/tPVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Look at the history of the model\n",
    "hist_dense = model_cnnpl.history.history # contains information from fitting model\n",
    "hist_dense.keys() # returns: dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n",
    "\n",
    "# Call each key to get the values to plot\n",
    "loss_cnnpl = hist_dense['loss'] \n",
    "val_loss_cnnpl = hist_dense['val_loss']\n",
    "acc_cnnpl = hist_dense['acc']\n",
    "val_acc_cnnpl = hist_dense['val_acc']\n",
    "\n",
    "epochs = range(1, len(loss_cnnpl) + 1)\n",
    "\n",
    "# Plot Losses of training and validation sets v. epochs\n",
    "plt.figure(figsize=(25,6))\n",
    "plt.subplot(131)\n",
    "plt.plot(epochs, loss_cnnpl, 'k', label = 'Training loss')\n",
    "plt.plot(epochs, val_loss_cnnpl, 'r', label = 'Validation loss')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation loss over epochs')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot accuracy scores over epochs for training and validation sets                      \n",
    "plt.subplot(132)\n",
    "plt.plot(epochs, acc_cnnpl, 'k', label='Training acc')\n",
    "plt.plot(epochs, val_acc_cnnpl, 'r', label= 'Validation acc')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation accuracy over epochs [PReLU CNN]')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "mLNCBXhWkPAg"
   },
   "source": [
    "### Basic Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1xIxQYohlU7F"
   },
   "outputs": [],
   "source": [
    "# Use model to predict on testing set\n",
    "y_pred = model_cnnpl.predict(x_test_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "icarjTH8lXxJ"
   },
   "outputs": [],
   "source": [
    "# Convert y_pred from continuous variables to discreet binary (0,1)\n",
    "yy = []\n",
    "for elem in y_pred: # each element is an array\n",
    "  row = []\n",
    "  for e in elem: # each number in the array\n",
    "    if e == elem.max(): # since softmax outputs probability...\n",
    "      row.append(1) # assign highest prob as 1\n",
    "    else:\n",
    "      row.append(0) # assign all other probs as 0\n",
    "  yy.append(row)\n",
    "y_pred_ = np.array(yy)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1551,
     "status": "ok",
     "timestamp": 1549503561005,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "7YOB1I7_lbpr",
    "outputId": "f67fc687-3c9c-41ce-cb56-7077a78d61ac"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.678089626298651"
      ]
     },
     "execution_count": 75,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_test_, y_pred_, average= 'weighted') # 0.6753212975691832\n",
    "precision_score(y_test_, y_pred_, average= 'weighted') # 0.708424003115032\n",
    "accuracy_score(y_test_, y_pred_) # 0.678089626298651\n",
    "recall_score(y_test_, y_pred_, average= 'weighted') # 0.678089626298651"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3S4crnn4YDQ8"
   },
   "source": [
    "### AUC/ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "jchYzyq6imoy"
   },
   "outputs": [],
   "source": [
    "# Iterate through y_test_ and y_pred_ to get predictions for each of the 537 IDs\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(y_test_.shape[1]):\n",
    "  # use y_pred not y_pred_, because the roc_curve function needs probability for threshold metric\n",
    "  fpr[i], tpr[i], thresholds = roc_curve(y_test_[:,i], y_pred[:,i]) # fpr = false positive rate, tpr = true positive rate\n",
    "  roc_auc[i] = auc(fpr[i], tpr[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 512
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1185,
     "status": "ok",
     "timestamp": 1549503639052,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "EKpVfWFkirlE",
    "outputId": "c2025c6b-2aee-4a2d-ab74-17c9db797703"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAHvCAYAAADtrFPzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wF2tOhTmvkgQAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot some of 537 celebs ROC curve \n",
    "#Seaborns Beautiful Styling\n",
    "sns.set_style(\"darkgrid\", {\"axes.facecolor\": \".9\"})\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "lw = 2 # linewidth\n",
    "for i in range(50):\n",
    "  plt.plot(fpr[i], tpr[i], lw = lw)\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "  plt.yticks([ii/20.0 for ii in range(21)])\n",
    "  plt.xticks([ii/20.0 for ii in range(21)])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for PReLU CNN Model')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "oB2sMPhX9_8a"
   },
   "source": [
    "## CNN w/ SeLU (Best Model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "b1dpx56c98QP"
   },
   "outputs": [],
   "source": [
    "# Set seed to get same results\n",
    "np.random.seed(121)\n",
    "# Define model as Sequential class\n",
    "model_cnn_selu = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "W9HQQ0Dt953h"
   },
   "outputs": [],
   "source": [
    "# Block 1\n",
    "\n",
    "# first layer: convolutional layer with filter size of 2 and strid of 2, due to \n",
    "# the features having shape divisable by 2. selu shown to work better than relu\n",
    "# and leaky relu without CNN.\n",
    "# Max Pool to get only the important features\n",
    "# add dropout to combat overfitting. rate = 10% of neurons to drop\n",
    "model_cnn_selu.add(Conv1D(128, (2), strides=(2), padding='same', \n",
    "                     input_shape = (1,193), activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 2\n",
    "model_cnn_selu.add(Conv1D(256, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 3\n",
    "model_cnn_selu.add(Conv1D(256, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(Conv1D(512, (2), strides=(2), padding='same',activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 4\n",
    "model_cnn_selu.add(Conv1D(512, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 5\n",
    "model_cnn_selu.add(Conv1D(512, (2), strides=(2), padding='same',activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 7\n",
    "model_cnn_selu.add(Conv1D(256, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Block 8 \n",
    "model_cnn_selu.add(Conv1D(256, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "model_cnn_selu.add(Dropout(rate = 0.1))\n",
    "\n",
    "# Block 9\n",
    "model_cnn_selu.add(Conv1D(256, (2), strides=(2), padding='same', activation= 'selu'))\n",
    "model_cnn_selu.add(MaxPooling1D((2), strides=(2), padding='same'))\n",
    "\n",
    "# Output Block\n",
    "model_cnn_selu.add(layers.Flatten())\n",
    "model_cnn_selu.add(Dense(537, activation = 'softmax'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1Qr_MVAu-n9F"
   },
   "outputs": [],
   "source": [
    "# Compile the model with a specified Adamax optimizer\n",
    "opt = Adamax(lr = 1e-3, decay = 1e-5) # Adamax has shown to yield faster learning than Adam and SGD\n",
    "model_cnn_selu.compile(loss = 'categorical_crossentropy', \n",
    "              optimizer = opt,\n",
    "             metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-AswpKqz-oi_"
   },
   "outputs": [],
   "source": [
    "# Add automated stopping after val_loss difference from epoch t and t-1 is \n",
    "# more than 0.001; give it three more epochs to try and get back on track (patience)\n",
    "earlystop = EarlyStopping(monitor='val_loss',\n",
    "                              min_delta=0.001,\n",
    "                              patience=3,\n",
    "                              verbose=0, mode='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 224
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 181596,
     "status": "ok",
     "timestamp": 1549504472573,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "eSCbQi4j-tOD",
    "outputId": "a6da53b4-6289-457f-bfb0-8997379934f3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "46427/46427 [==============================] - 33s 708us/step - loss: 0.7676 - acc: 0.7881 - val_loss: 1.4352 - val_acc: 0.7118\n",
      "Epoch 20/100\n",
      "46427/46427 [==============================] - 33s 711us/step - loss: 0.7049 - acc: 0.8018 - val_loss: 1.4556 - val_acc: 0.7138\n",
      "Epoch 21/100\n",
      "46427/46427 [==============================] - 33s 708us/step - loss: 0.6802 - acc: 0.8077 - val_loss: 1.4141 - val_acc: 0.7190\n",
      "Epoch 22/100\n",
      "46427/46427 [==============================] - 33s 711us/step - loss: 0.6433 - acc: 0.8178 - val_loss: 1.4173 - val_acc: 0.7233\n",
      "Epoch 23/100\n",
      "46427/46427 [==============================] - 33s 710us/step - loss: 0.6072 - acc: 0.8285 - val_loss: 1.4263 - val_acc: 0.7281\n",
      "Epoch 24/100\n",
      "46427/46427 [==============================] - 33s 716us/step - loss: 0.5658 - acc: 0.8350 - val_loss: 1.4368 - val_acc: 0.7281\n"
     ]
    }
   ],
   "source": [
    "# fit the data and save it with history variable. \n",
    "history = model_cnn_selu.fit(x_train_, \n",
    "                    y_train_, \n",
    "                    epochs = 100, \n",
    "                    batch_size = 25,\n",
    "                    validation_data= (x_val_,y_val_), callbacks = [earlystop])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 966,
     "status": "ok",
     "timestamp": 1549465237811,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "LMi8cQmzPfb8",
    "outputId": "c7cb5275-e11d-4795-ad2a-24d514d9f906"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved model to drive\n"
     ]
    }
   ],
   "source": [
    "# Save best model so don't have to rerun in scenario of crash\n",
    "model_json = model_cnn_selu.to_json()\n",
    "with open(\"model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "    \n",
    "model_cnn.save_weights(\"model_weights.h5\")\n",
    "print(\"Saved model to drive\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "CbhXaYbWjXK7"
   },
   "outputs": [],
   "source": [
    "# # Read in model\n",
    "# json_file = open('model.json', 'r')\n",
    "# loaded_model_json = json_file.read()\n",
    "\n",
    "# json_file.close()\n",
    "# loaded_model = model_from_json(loaded_model_json)\n",
    "\n",
    "# # load weights into new model\n",
    "# loaded_model.load_weights(\"model_weights.h5\")\n",
    "# print(\"Loaded model into notebook\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lkIFV80ljdnk"
   },
   "outputs": [],
   "source": [
    "# # Compile the model with a specified Adamax optimizer\n",
    "# opt = Adamax(lr = 1e-3, decay = 1e-5) # Adamax has shown to yield faster learning than Adam and SGD\n",
    "# loaded_model.compile(loss = 'categorical_crossentropy', \n",
    "#               optimizer = opt,\n",
    "#              metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "CjQ14EJwjIPA"
   },
   "source": [
    "### Errors & Accuracy Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1218,
     "status": "ok",
     "timestamp": 1549504548784,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "FEGpDTtGjQnM",
    "outputId": "5cf34b98-05b7-4e61-dea1-205fc72abde3"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7IAAAGCCAYAAADdZf9wAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3XlcFdX7wPHP5bIJKAqIIuICLiyG\nWqaZGIIgoFIZppSWS99vi2maWpZmi5VZZlpmi/XLJSuXr1pmuWDuu7gvbOGGyiqK7HDvnd8fxCSx\nqAFeluf9et2Xl5m5Z545Dpz7zDlzRqMoioIQQgghhBBCCFFLmBg7ACGEEEIIIYQQ4k5IIiuEEEII\nIYQQolaRRFYIIYQQQgghRK0iiawQQgghhBBCiFpFElkhhBBCCCGEELWKJLJCCCGEEEIIIWoVSWRF\nCW+99RbBwcEEBwfj5eWFn5+f+nNWVtYdlRUcHExaWlqF28yZM4effvqpMiFXuZEjR7JmzZoqKatj\nx44kJSURERHB66+/Xqn9rVy5Un1/O3V7u1577TW++OKLKimrvjhw4ACBgYHGDkMIUQnS3tW/9k7U\nbJVtWy9duoSXlxfBwcGcPn2avLw83nnnHYKCgggODiYoKIgvv/zytsrx9PQsc11535n8/f2JjIws\n8zPx8fG8+OKLBAQEEBgYyNChQ9m2bVuJz44cObJUDP7+/ur7jh07smDBghLbrFmzhtdee43Tp0+r\nf8cuXbp0y+OrS0yNHYCoWd555x31vb+/Px999BHdunX7V2Vt3LjxlttMmjTpX5Vd2wQGBlbqj3Nq\nairffvstQ4YMAW6vboUQQpRP2rvqIe2dMKZmzZqp58zHH3/MjRs3+PXXXzE3NyctLY1hw4bRqlUr\nBgwYcFfiSU5OZvjw4YwfP57PP/8cjUbD0aNHeeGFF/j444/x8fEB4OLFi2zZsoWAgIAyy7G1tWX5\n8uUMHjyYZs2alVjn5eXFxo0b1cS3PpEeWXFHnnrqKebOnUtISAhHjhwhLS2NZ555huDgYPz9/Vm0\naJG6bfHV2QMHDjB06FDmzJlDSEgI/v7+HDx4ECh5Zcvf31/9JfXx8WHWrFlqWV999RU9e/YkLCyM\nH374odxf1lWrVhESEkK/fv0YNmwYly9fBoquWr300ktMnTqVoKAg+vfvT1xcHAAJCQk8/vjjBAQE\nMGnSJPR6falyd+zYQWhoaIlljzzyCDt37qywDoqtWbNGvdpW0f7++OMPQkNDCQoK4rHHHiMqKgqA\n8PBwrly5QnBwMAUFBWrdAixdupT+/fsTHBzMCy+8QHp6ulq3n332GaNGjcLPz49Ro0aRm5tb3n8t\nANHR0YSHhxMcHMwjjzzCrl27AMjOzubFF18kJCSEvn378sYbb1BYWFju8tsp12Aw4OPjw6lTp9Tt\nFi9ezMsvvwzAihUr1DqdOHEieXl56nF98MEHhIaGsmHDhlL72rJlC6GhofTt25fRo0eXqI+ZM2fy\n1FNP0bt3b55//nm1Pso7boCFCxfSt29fgoKC+OCDD1AURV335ZdfEhISQkBAAPv37wcgNjaWoUOH\nMmDAAPr168eyZcsqrHMhRM0k7d3f6kp7l5uby4QJEwgKCsLf358PP/xQXZeQkMCwYcMIDAwkLCyM\n06dPV7j8nz1wxT9funQJHx8fZs6cyfDhwys8Vijdxuj1enr16sXJkyfVbZYtW8aYMWNKHY+0rRWL\njY3Fy8sLc3NzABwcHPjxxx/VCy1//vknw4cPJygoiNDQ0BJ1XlUWL17Mgw8+SHh4OBqNBoCuXbvy\nxRdf4Orqqm43efJkZs+eTUFBQZnl2NjYMGLECObMmVPlMdZqihDl8PPzUw4dOlRi2fDhw5XRo0cr\ner1eURRFmTFjhvLmm28qiqIoFy9eVLy8vJQrV64oiqIoHTp0UBITE5X9+/crnTp1UiIiIhRFUZRv\nvvlGGTlypKIoijJlyhRlwYIF6v4mTpyo6HQ6JSkpSfHy8lISExOV2NhY5b777lOSk5OVvLw8Zfjw\n4Yqfn1+peNPS0pROnTopiYmJiqIoymuvvaZMnTpVURRFWb16tdK5c2fl5MmTiqIoyttvv61MmzZN\nURRFeemll5Q5c+YoiqIox48fVzw9PZXVq1eXKDs/P1/p1q2bcvHiRfVYu3fvrhQWFt5WHaxevVoZ\nMWJEhfsrLCxUunXrphw9elRRFEWZP3+++pn9+/crAQEBajzF5R49elR56KGHlLS0NPX/o/iYp0yZ\nooSEhCjXrl1TCgsLlYcfflj55ZdfStVb8f+BXq9XQkJClF9//VVRFEU5ceKEcv/99yuZmZnKsmXL\nlNdee01RFEUpLCxU3nzzTeXMmTPlLr9ZReW+9dZbyqeffqpuO2zYMGXTpk3KoUOHlJ49eypJSUmK\noijK9OnTlVmzZqnxhoaGKnl5eaWO5eLFi0rXrl2VmJgYRVEU5auvvlLGjRunfs7Pz09JT09X9Hq9\nMmzYMGXx4sUVxnfo0CElMDBQyczMVPLz85WwsDDl999/V8/pLVu2KIqiKN9++63y9NNPK4qiKOPG\njVPWrFmjKIqiXL16VXnhhReU/Pz8UrEKIWoOae/+Vpfbu//7v/9T/vOf/ygGg0G5fv260r17d/X/\nfcSIEcoPP/ygKIqiREREKP37969w+T/PmeKfExISFC8vL7UdqOhYy2tj3n33XWXmzJlq2U8//bSy\nfv36EscibWvptjUhIaHE78uyZcuULl26KJ988oly5MgRpbCwsET99evXT1m5cqWiKIoSGRmp+Pj4\nKIWFhUpCQoLi4eFRqh6Kj7f49/hmZf0NURRFCQsLK/Nc/OdnExISlMmTJyvffvttqWMpfp+fn68E\nBgYqx48fVxSl6Hd9ypQppcqpT6RHVtwxX19fTEyKTp033niD6dOnA+Di4kLTpk3LHJ9vbW2tDpfw\n8vLiypUrZZYdGhqKVqulWbNm2Nvbk5iYyKFDh+jevTuOjo5YWFgQFhZW5mft7e05fPgwzZs3B6Bb\nt24kJCSo693c3OjUqRMAnp6eJCYmAhAZGUn//v0B8Pb2LnGFrJi5uTl+fn5s3boVQB3+YWpqett1\nUKy8/ZmamrJ37166dOlSZvxl2b59O0FBQdjb2wPw+OOPs2fPHnW9r68vjRs3xtTUlA4dOqjHXJZL\nly6RlpamDre55557aNGiBSdPnsTOzo6jR4+ye/duDAYD77zzDh4eHuUuv91yg4KC1DpNT08nOjoa\nX19ftm7dSv/+/dXhM0888QSbN29Wy+zZsycWFhaljmHnzp10796dDh06AEVX9rdu3ar2Avj7+9Ok\nSRNMTEwICAjg6NGjFca3c+dOfH19sbGxwdzcnO+//55+/foBRVdH+/btCxSdT8U9Bvb29mzatInT\np0/TpEkTvvjiC/VqsBCidpH2rm61d6NHj+aLL75Ao9Fga2tL+/btuXTpEvn5+Rw4cICBAwcC0Ldv\nX1auXFnu8lspLCxUe/0qOtby2pgBAwbw+++/YzAYuH79OqdOncLPz6/EPqRtvXXbOmzYMGbOnMnp\n06cZOXIkDzzwADNnziQ/P5+zZ89y9epVBg8eDMB9992nfqepShkZGTg4ONzWtpMnT2bRokVcvXq1\nzPXm5ua88sorzJw5sypDrNXkHllxx2xtbdX3J0+eZM6cOSQmJmJiYkJqaioGg6HUZxo2bKi+NzEx\nKXMbKPoDVkyr1aLX67lx40aJff7z3oBier2ezz77TP3jmp2dTdu2bcuMobhsKPojc/N+GzVqVGb5\nQUFBLF26lBEjRrBlyxZ1mM/t1kGxivb3/fffs3btWgoKCigoKFCHoZQnPT0dR0fHEmXd/AewvGMu\nr6yGDRuW2GejRo1IT09nwIABZGRk8Omnn3L27FkefvhhXn/9dUJCQspcfnPjUlG5wcHBJCcnc+XK\nFfbu3Yuvry8WFhZkZmYSERHB7t27AVAUpcSQ5ZvPh5tlZmYSGRlJcHCwuszGxobr168D0Lhx4xIx\n3Lhxo8L4rl27VqJ+GzRoUKLcYjef05MnT+brr79mwoQJ5Ofn89xzzzFs2LBy610IUXNJe1e32rvz\n588za9Yszp49i4mJCUlJSTz22GNcv34dg8GglqHRaLC2tiY5ObnM5bei1WpLHHd5x1peG9O1a1fM\nzMw4ePAgSUlJ+Pj4YGVlVao+pG29ddsaEhJCSEgIBQUF7Nu3j/feew8LCwv8/PzIy8sjJCRE3TYr\nK4vr16/j5ORUbnkajabMc16v16PVakstb9KkCcnJybeME4p+38PDw5k3bx7PPfdcmdsEBgaydOlS\nfv3119sqs66TRFZUyiuvvMKIESN44okn0Gg09O7du8r3YWNjQ05OjvpzSkpKmdv9/vvvbN26lWXL\nlmFnZ8fKlStv6xe9UaNGJWaoLL7v45969+7N1KlTOX/+POfPn+eBBx4A7rwOytvfkSNH+Oabb1i1\nahUtW7Zkz5496pXv8jg4OKgNCcD169dv+8rfP9nb25ORkYGiKGrDc/36dfXqd3h4OOHh4SQnJzNu\n3Dh+/vlnhgwZUu7y2ylXq9USEBDAtm3b2LVrl3pl1NHRkUGDBjFlypQ7OgZHR0cefPBBPvvsszLX\nX7t2TX2fkZGBra1thfE1adKkxGdufl8ea2trJk6cyMSJEzlx4gT//e9/efDBB0t8yRRC1D7S3tX+\n9m7GjBl4eXmxYMECtFot4eHhQFGyodFouHbtGnZ2diiKwsWLF3FycipzeatWrUpdpMjIyChznxUd\na3ltTJMmTRgwYAAbN24kKSmJQYMGlSpX2taK29bCwkJ27NiBn58fWq0Wc3NzfH19uXjxIrt372bo\n0KFYW1uXOZlYRaMMmjZtqt6PXiwnJ4eUlJQyE+AePXqwadOmUv+Hf/zxBxYWFupkT8WeeeYZBgwY\ngK+vb7kxTJs2jTFjxvCf//yn3G3qCxlaLCrl6tWrdOrUCY1Gw9q1a8nNzS3RCFcFb29vDhw4QHp6\nOgUFBfz888/lxuLs7IydnR3Xrl1jw4YNZGdn37L8Ll26EBERARQ1OBcvXixzO3Nzc3x8fJg9ezZ9\n+/ZVr7zdaR2Ut7/09HTs7e1p0aIFubm5rF27lpycHBRFwdTUlJycHHQ6XYmy+vTpQ0REhNoILF++\nvMI/fhVp2bIlzZs35/fff1djS0tLw9vbmwULFvC///0PKLpi2LJlSzQaTbnLb7dcQB0CdfLkSR56\n6CGgaJjS5s2b1S89W7ZsYeHChbc8Bh8fHyIjI9VhWydOnOC9995T1+/atYsbN26g1+vZsmUL3bp1\nqzA+f39/tm7dSkZGBjqdjhdffFG9kl2e559/Xp1YpUOHDtjY2Nyyp0EIUfNJe1f727urV6/i4eGB\nVqtlz549XLhwgZycHMzNzenVqxdr164FitqKZ599ttzlGo2Gpk2bEh0dDRRdWMjPzy9znxUda0Vt\nzMCBA9myZQtHjx4t8zilba24bTU1NWXu3Ll89dVXau98VlYWW7du5f7778fZ2ZnmzZuriWx6ejoT\nJ0685e/0o48+yvbt29VJvwoKCpg1axbBwcHqUP+bjRgxgpMnT7Jw4UL1wsfhw4d56623sLS0LLV9\ngwYNmDBhArNnzy43Bnd3dx588EGWLFlSYaz1gSSyolLGjx/Piy++SGhoKDk5OQwdOpTp06eX2zj+\nG97e3gwaNIhBgwbx9NNPl7pPpNjAgQO5fv06gYGBTJo0iQkTJpCUlFRiNsiyvPLKK2zbto2AgAB+\n+OEHHnzwwXK3DQoKYsuWLSWGotxpHZS3v969e+Po6EhAQACjR49mxIgRNGzYkJdeeomOHTtia2tL\nr169Stxv5e3tzbPPPsuwYcMIDg4mMzNTnZnwTmk0Gj755BOWLVtGSEgI7733Hp9++ilWVlY88sgj\n/PLLL+qz2MzMzHjkkUfKXX675QI88MADnDp1igcffFAdkuzl5cXzzz/PU089RUhICIsXL1bvmamI\no6Mj7777rjqT8owZM9T7s4r3NXbsWHx9fWnUqBFhYWEVxtelSxeeeeYZHn30UQYMGICnp6d6r1R5\nhg8fzqRJkwgJCWHQoEE8+eSTtGnT5g7/N4QQNY20d7W/vXvhhRf48MMPGThwIAcPHmTs2LHMnz+f\nw4cP8/7777Nt2zb69u3LvHnz+PjjjwHKXT5mzBgWL17MwIEDiY+Pp127dmXus6JjraiN6dixI40b\nN8bHx6fMhEfa1orbVo1GwzfffENcXBwhISEEBQURFhZGt27dGDVqlBrfDz/8QHBwMMOHD6dnz55q\n/en1evW50sWvhIQEXF1dmTdvHu+99x4hISE88sgjWFpa8u6775YZR/FMycePHycgIICQkBA+/fRT\n5s2bV+7jvkJDQ8sd5l1swoQJpKamVrhNfaBRlJvmuxaihrp5aMr27duZN29euVeqhSjLa6+9RqtW\nrcp8hIEQQtQU0t6JYv/9738ZPnz4vx5pdTfUpLb10qVLPP300+pEV/WNv78/S5cupWXLlsYO5a6R\nHllR46Wnp/PAAw9w+fJlFEVhw4YN6ux/QgghRF0h7Z0odvjwYS5fvlwt92ILUVfIZE+ixrOzs2PC\nhAmMHDkSjUaDq6srr776qrHDEkIIIaqUtHcC4PXXX+fIkSPMnj1bffyTuD3JyckEBwczZ84cvLy8\njB3OXXH69GkmTZp027Mj1yUytFgIIYQQQgghRK0il3mEEEIIIYQQQtQqksgKIYQQQgghhKhVasw9\nsomJiUDRNNVpaWlGjqZ2kzqsPKnDqiH1WHlSh5VXXIdlPaxe3Lni9hrk/KwKUoeVJ3VYeVKHlSd1\nWHl32l7XuB5ZMzMzY4dQ60kdVp7UYdWQeqw8qcPKkzqsPlK3lSd1WHlSh5UndVh5UoeVd6d1WOMS\nWSGEEEIIIYQQoiKSyAohhBBCCCGEqFUkkRVCCCGEEEIIUatIIiuEEEIIIYQQolaRRFYIIYQQQggh\nRK0iiawQQgghhBBCiFpFElkhhBBCCCGEELWKqbEDEEIIUXlffPEFMTExpKenk5eXR4sWLWjUqBHv\nvvvuLT+7YcMGbGxs6N27d5nr58+fz+DBg2/7AeX/NH78eMaPH4+rq+u/+rwQt1KTz38hhBDVQxJZ\nIYSoA8aMGQMUfSk/d+6c+vPtCAkJqXD9uHHjKhWbENVNzn8hhKh/JJEVQog67OjRo6xYsYLc3FzG\njBnDsWPH2LFjB4qi0KNHD0aOHMmiRYuwtbWlbdu2rF27FhMTEy5cuICvry8jR45Ue1R37NhBdnY2\nCQkJXLlyhbFjx9KjRw9+/PFH/vjjD1q0aIFOp2PIkCF07dq1VCxZWVnMmjWLrKwsdDodL730Eh06\ndOCzzz4jJiYGg8HAww8/TEhISJnLhLhTNeH8j4yM5LvvvsPMzAwbGxvefvttzMzMmD9/PmfOnKFB\ngwaMHTsWV1dXdZlWq2XixIkyikEIISogiawQQlSxGTNmsH79+iopS6vVotfrGThwIG+++ea/KuPc\nuXN8//33mJubc+zYMebPn4+JiQlPPPEEjz/+eIlto6OjWbp0KQaDgfDwcEaOHFlifUpKCh9++CEH\nDhxg3bp1eHh4sHbtWpYtW0Z2djbDhw9nyJAhZcaxevVqPD09efLJJ4mOjmbBggW8++677N+/nx9/\n/BGdTsfGjRu5ceNGqWWi9qjK879YbT7/s7KymD59Ok5OTsycOZNDhw5hbm5OSkoKX375JZcuXWLT\npk2kp6ery44fP862bdskkRVCiArUqUQ2Ly+Pbdu2ERgYiKlpnTo0IYT419zc3DA3NwfA0tKS8ePH\no9VqycjI4MaNGyW2bd++PZaWluWWdc899wDQtGlTsrOzuXz5Mq6urlhYWGBhYYG7u3u5n42JiWH4\n8OEAuLu7c/nyZRo1akTLli2ZNm0avr6+9OvXD3Nz81LLhPi3jH3+29raMnv2bPR6PVeuXKFr165c\nv36dTp06AXD//ffTsmVLfvrpJ3VZ586d6dy5c5UcvxBCVKfMzEzi4uI4f/48PXr0wNnZ+a7tu05l\nexERETz33HMsWLCAQYMGGTscIUQ99eabb/7r3qN/cnJyIjExsVJlFF/YS0pKYuXKlXzzzTdYWVmV\n6m2Coh7gity8XlEUFEVBo9Goy25+/08ajQZFUdSf9Xo9AB999BGxsbFs2bKFzZs38/HHH5e5TNQO\nVXn+VwVjn/8fffQRs2bNonXr1sybNw8AExOTEr8L5S0TQoiaQFEUUlJSiIuLIy4ujj///FP9Nykp\nSd1u2LBhzJ49+67FVacS2WbNmgFw4sQJSWSFEOIfMjIyaNKkCVZWVsTGxpKcnIxOp6tUmc2bN+fc\nuXPodDqysrKIiYkpd9uOHTty7NgxvLy8OH36NG3btiUxMZG9e/cSFhZGhw4dePbZZ8tcJkRlGev8\nz87OxtHRkczMTI4dO4abmxvu7u78+OOPhIeHc+bMGZYsWYKfn5+6LC4ujt9++40JEyZUKj4hhLgT\ner2eixcvlkhYi5PWf45gAXB2dqZPnz60a9eO9u3b079//7sab51KZIuH9ERFRRk5EiGEqHnatWun\nTixzzz33EBoayty5c9Xhkv+GnZ0dAQEBPP/887Ru3RoPD49ye7UGDx7MrFmzePnllzEYDEyYMAEH\nBwdOnTrF1q1bMTMzIyQkpMxlQlSWsc7/Rx99lLFjx+Li4kJ4eDiLFy9mwYIFtG7dmnHjxmFubs6L\nL76Iq6sre/bsUWdJfvnllyt1vEIIcSsZGRns27eP3bt3s3//fuLj48nPzy+xjampKW3btqVXr160\nb99efbm5uWFtbW2kyItolBoyjqV46Fxlh9F1796d/Px8jh8/XlWh1TpVMRSxvpM6rBpSj5VXG+pw\nw4YNBAQEoNVqGT16NB999BGOjo7GDktVXIfyHNCqcfP5WBvOz+pW2fNf6rDypA4rT+qw8mpDHebm\n5nLo0CF2797N7t27OXHiBAaDASiaQ6BDhw4lktV27drRpk0bzMzM7kp8d9pe16keWQAPDw8iIiJI\nS0vDwcHB2OEIIUSdl56ezgsvvICZmRkBAQE1KokVorrJ+S+EqKl0Oh3Hjh1TE9fIyEgKCgqAop7W\n++67Dx8fH3r37k3Xrl2xsLAwcsR3ps4lsp6enkRERBAVFUXv3r2NHY4QQtR5w4YNY9iwYcYOQwij\nkPNfCFFTGAwGoqOj1cR1//79ZGVlqeu9vLzUxLV79+7Y2NgYMdrKq3OJrIeHB4AkskIIIYQQQog6\nS1EU4uLi2LdvH3v37mXv3r1cvXpVXe/q6spjjz2Gj48PPXv2xN7e3ojRVr06m8ieOXPGyJEIIYQQ\nQgghRNXQ6/VERUWxf/9+9ZWenq6ub968OYMHD8bHx4devXrd1We6GkOdS2Tbtm2LhYUF0dHRxg5F\nCCGEEEIIIf6VwsJCTp48qSatBw8eLPEYHCcnJx577DF69uxJjx49cHNzq/B57nVNnUtkTU1N6dCh\nA7Gxseh0OvVB6EIIIYQQQghRU+Xn53Ps2DH279/Pvn37iIyMJCcnR13fpk0bQkJC6NmzJw888AAu\nLi71KnH9JxNjB1AdPDw8yMvL49y5c8YORQgh7ooxY8YQExNTYtnChQtZsWJFmdsfPXqUN998E4Bp\n06aVWr9mzRoWLVpU7v7i4+NJSEgA4J133in13Lk7MXTo0BINtRB3qjaf/0KI+sdgMHDp0iV27drF\nkiVLePvttwkLC8Pd3Z1Bgwbx4YcfsnPnTpydnXn66adZsGABhw8fZu/evcydO5chQ4bQqlWrep3E\nQjX2yB44cIDx48fTvn17ADp06MD06dOra3cl3DzhU/H+hRCiLgsICGD79u107NhRXbZz507mzp17\ny8++//77d7y/nTt30rFjR1xcXHjrrbfu+PPi7po5cybHjx9Ho9EwdepUvL291XU//PAD69atw8TE\nhE6dOpWZ2NV0cv4LIWoaRVFISUnh3LlznD17lrNnz3Lu3DnOnTvH+fPnycvLK7G9RqPBw8ND7W3t\n0aOHPEr0Fqp13G337t357LPPqnMXZSpOZKOjo3n44Yfv+v6FEOJu8/PzY+zYsTz33HMAxMTE4ODg\nQNOmTYmMjOS7777DzMwMGxsb3n777RKfffjhh1m3bh2HDx/m888/x87ODnt7e5ycnNDpdMyaNYvU\n1FRyc3MZOXIkzZs3Z926dTRu3JgmTZrw9ttvs2jRIrKysvjoo48oLCzExMSEV199FYBZs2bRokUL\n4uPjadeunbr8n1JSUkp9vmnTprz//vtcvXqVwsJCRo4cyX333VdqWY8ePaq1fmuzgwcPcuHCBVas\nWEF8fDxTp05VeyqzsrL4v//7PzZv3oypqSmjR4/m2LFjdOnSxchR35nafP6vW7eO7777Dq1WS5s2\nbZg8eTI6nY4PPviA5ORkzM3Nef3112nSpEmpZU2bNr0r9SuEKJ+iKMTGxhIREcHRo0eJj49XE9bs\n7OxS29vY2NC+fXtcXV1p27at+m+7du2wtbU1whHUXnXyBlJPT09AZi4WQhhHwxkzsFy/vmoK02pp\nqteTN3AgmX8NhSxLkyZNaNGiBVFRUXh4eLB9+3YCAgKAomRl+vTpODk5MXPmTA4dOkSDBg1KlfHN\nN98wbdo02rVrx5QpU3ByciIzM5Nu3boRHBzMlStXePvtt1m4cCHdu3fH19dXvXAIsGjRIvr374+/\nvz/bt29n8eLFjBw5kpiYGN58802aNGnC448/TmZmJg0bNiy1/7I+HxYWRkZGBp999hmZmZkcOHCA\ns2fPllomyrdv3z71XHBzcyMjI4OsrCxsbGwwMzPDzMyMnJwcrKysyM3NrfQXqSo9//9Sl8//3Nxc\nPvroIxo2bMhLL73E2bNnOXPmDHZ2dkyfPp0//viDvXv3otVqSy175JFHqrCWhRC3S1EUoqKiWL9+\nPb/++ivx8fEl1ltaWtK2bdsSiaqrqyuurq44ODjU+yHBVaVaE9k///yT559/noyMDMaOHUuvXr2q\nc3eq4quwUVFRd2V/QghRE/Tt25dt27bh4eHBnj17WLBgAQC2trbMnj0bvV7PlStX6Nq1a5lf5JOS\nkmjXrh0AnTt3Jj8/n4YNGxKerEqvAAAgAElEQVQdHc369evRaDRkZGSUu/+YmBj++9//AtC1a1eW\nLl0KgLOzs/rsOnt7e7Kzs8tMZMv6fKtWrcjJyeH999+nd+/e+Pv7U1BQUGqZKF9aWhpeXl7qz3Z2\ndqSmpmJjY4OFhQUvvvgiAQEBWFhYMGDAANq2bVtheQ4ODpiZmak/Ozk5ldzAxga02io9BhsbG2z+\nuZ9/CAsL4+DBg/j7+3PgwAGWL1+Ora0trq6ufPrpp+j1ehISEvDz88Pe3p4GDRrg5OSEiYkJTk5O\npKSkqM+f9/HxIT8/n/bt2/O///2Pl19+GRMTE7Kzs3FycsLKygo7OzucnJwwNTWlefPmxMfH88Yb\nb+Dg4EBwcDA//vgjjo6OtGnThk6dOgFFj8awtrYuUWe2trbMmDEDgISEBExNTbl8+TK9e/fGycmJ\n4cOHA/D222+XWib+Vuo8FHdM6rBiiqJw4sQJVq1axapVq4iNjQWgQYMGhIWFERAQQIcOHejQoQMt\nWrTAxKROTkVU7e7kPKy2RLZNmzaMHTuWkJAQEhISePrpp9m8eTPm5uZlbn9zw1gVv0hdunQhIiIC\nKyuretlNL3+MKk/qsGrUy3r88suiVxUxBWz+elXk8ccfJywsjCFDhtCuXTs6dOgAwPDhw1m4cCFu\nbm7MmDGDxo0bl/lF3tTUVP3/srGxwdzcnMjISHQ6HatWreL69esMHjy43C/yZmZmODo60rRpU7Ra\nLebm5jg6OmJpaamWa2ZmRtOmTUucFxV9vm3btqxdu5YjR46wdu1ajh07xgcffFDmsorUy/OwHIqi\nqO+zsrL4+uuv2bhxIzY2NowYMYLo6Gjc3d3L/XxaWpr63snJicTExJIbTJxY9Kpq/9zPP3h7e7Ng\nwQJ69OiBk5MTOTk55OTkMGXKFGbNmkXr1q2ZN28e169fx9LSktzcXBITEzEYDCQmJqIoinosN27c\noKCggGXLlpGUlMScOXO4ceMGzz33HImJieTk5JCenk5iYiI6nY6kpCT0ej1JSUkUFhZy9epV9Ho9\nKSkpavlQ9CiN5ORk9QtuYWEhM2bMYOHChdjb2/Paa69x9epV8vLy1PKLlbVMFCnzPBR3ROqwbIqi\ncPr0adavX8/69es5e/YsUNTjOnDgQAYOHEhAQABWVlYl6jA5OdmYYddaxXV4u212tSWyzZo1o3//\n/gC0atUKBwcHkpOTcXFxKXP74oaxqn6RXF1dAdi+fTvdu3evdHm1ifwxqjypw6oh9Vh5d1qHrVu3\n5tNPP8XPz6/El3KNRkNsbCx79uyhefPmZX6Rt7Oz48CBA7i4uLBr1y48PT25cOECjRs3Jjk5mV9/\n/ZW8vDwSExPJy8sjNTW1xBd5Nzc3Nm3aRN++ffnjjz9wdXUlJSWFwsLCEl/kU1JS0N7UY1fR53fs\n2MH58+fp168fzz//POPGjStzWUV1dKcNY13j6OhYIvlMSUlR762Mj4/HxcUFOzs7ALp168apU6cq\nTGRrKisrK1xdXfnhhx/o27evujw7OxtHR0cyMzM5duwYbm5uZX6+adOmXLx4ERcXF44dO4anpycZ\nGRnqxZ5du3ah0+kAMDExQa/Xl/h8x44dOXbsGH379uXYsWMlJp4qT05ODlqtFnt7e1JSUoiJiUGn\n0+Hu7s6RI0fo06cPe/fu5ezZs2Uuk55ZIaqeoiicOnVKTV6Ln4LSoEEDQkNDGThwIH379sXKysrI\nkYpqS2TXrVtHamoqzzzzDKmpqVy9epVmzZpV1+5KuXnCp/qWyAoh6q+AgABmzpzJG2+8oS579NFH\nGTt2LC4uLoSHh7N48WJ1CO/NnnnmGd566y2aNWumJjoPPfQQ06ZN48yZM4SEhNC0aVOWLFmCt7c3\n8+fPL9GQjx49mo8++oj169djamrKq6++qn7xvx1lfd7S0pJvvvmGX3/9FRMTE8LDw3Fyciq1TJSv\nV69ezJ8/n/DwcE6fPo2joyM2NkX9+87OzsTHx5OXl4elpSWnTp3C19fXyBH/e7Xt/Le1taVXr148\n99xzuLm58cQTT7BgwQIWLlzI4cOHGT9+PFqtltdff53GjRuXWiaEqBqKonDy5Ek1eT1//jxQdIHs\n4YcfJjQ0FD8/P0leaxiNcvMYoyqUlZXF5MmTuXHjBoWFhYwdO7bCxrH4anpV9eCcPHmSoKAgRowY\nccshZ3WN9IJVntRh1ZB6rDypw8qr7z2yAB9//DGRkZFoNBreeustzpw5Q8OGDQkMDGT58uWsWbMG\nrVZL165dy51VutjN56Ocn5UndVh5UoeVV9/qUK/XExMTQ2RkJJGRkezfv59Lly4BRclrYGAgoaGh\n9OnT57aT1/pWh9WhxgwttrGx4auvvqqu4m+pffv2aLVamblYCCFEvTd58uQSP988dDg8PFx6tYUQ\ndVpWVhZHjhzh0KFDREZGcuTIETIzM9X1jRs35tFHH1WT17ImRBQ1T518/A4U3YTt6upKdHQ0iqLI\nNNdCCCGEEELUcYqikJCQQGRkJIcOHeLQoUNER0djMBjUbdzc3Ojfvz/3338/999/P25ubjLLcC1U\nZxNZKLpPNi4ujsuXL9OyZUtjhyOEEEIIIYSoQgUFBZw6dUpNXCMjI0vMGmxpaakmrN26deO+++5T\nHwknarc6nch6enqybt06zpw5I4msEEIIIYQQtVxqaqp6b+vhw4c5ceIEeXl56npHR0cGDBigJq6d\nOnUq9/Gfonar04ls8czFUVFR9OvXz8jRCCGEEEIIIW6XXq8nOjpaTVojIyPVGYWh6FFYHh4edOvW\nTe11bdmypdxSWE/Ui0RWJnwSQgghhBCiZsvIyFAnZTp8+DBHjhwhOztbXW9ra4u/vz/dunWjW7du\ndOnSRX2cmKh/6nQi6+zsTKNGjYiKijJ2KEIIIYQQQoib6HQ6tm/fzqZNm4iMjCQ2Npabnwzavn17\n7rvvPrXHVSZlEjer04msRqPB3d2dyMhI9WHvQgghhBBCCOOJi4tj+fLlrF69mpSUFKDo+a29evVS\nJ2S69957adKkiZEjFTVZnU5koWh48cGDB4mNjcXb29vY4QghhBBCCFHvZGRk8Msvv7By5UqOHDkC\nFD2/deTIkQwePBhvb29MTet8aiKqUJ0/Wzw9PYGiCZ8kkRVCCCGEEOLu0Ov17N69mxUrVrBx40by\n8vIwMTHB39+foUOHEhgYKCMmxb9W5xNZd3d3ALlPVgghhBBCiLvg3LlzrFy5klWrVnHlyhUAXF1d\nGTp0KIMHD8bJycnIEYq6QBJZIYQQQgghRKVkZWWxfv16VqxYwYEDBwCwsbHhySefZOjQoXTr1k0e\niyOqVJ1PZBs2bEirVq0kkRVCCCGEEKIKFRYWsm/fPtasWcP69evJyckBwMfHh6FDhxISEoKVlZWR\noxR1VZ1PZKFowqdNmzaRmppK06ZNjR2OEEIIIYQQtVJ2djbbtm1j48aN/PHHH2RkZADg4uLCCy+8\nwJAhQ3BxcTFylKI+qFeJ7JkzZ/D19TV2OEIIIYQQQtQaqampbN68mU2bNrFr1y7y8/MBaNGiBWFh\nYQwYMIAePXrIM17FXVVvElkouk9WElkhhBBCCCEqdvbsWTZu3MjGjRs5fPgwiqIARU8E6devH8HB\nwdxzzz1y36swmnqVyEZHRxs5EiGEEEIIIWoeg8HAiRMn1OQ1NjYWABMTE3r06EFwcDBBQUG0bt3a\nyJEKUaReJLJt27bF0tKSM2fOGDsUIYQQQgghaoS8vDz279/P7t27Wbt2LYmJiQBYWlqqva6BgYHY\n29sbOVIhSqsXiaxWq6Vjx45ERUWh0+kwNa0Xhy2EEEIIIUQJ58+fZ+vWrWzbto09e/aQl5cHQJMm\nTRgyZAhBQUH4+vrKbMOixqs3GZ27uzvHjx/n7NmzdOjQwdjhCCGEEEIIUe1yc3PZv3+/mryePXtW\nXdehQwf8/PwYOnQo7dq1k84eUavUm7PV09MTKJrwSRJZIYQQQghRV507d05NXPfu3av2ulpbWxMU\nFIS/vz9+fn60bNkSACcnJ3VYsRC1Rb1JZG+eufiRRx4xcjRCCCGEEEJUjZycHPbt28e2bdvYtm0b\n586dU9e5u7vj5+eHv78/999/P+bm5kaMVIiqU+8SWZnwSQghhBBC1HYGg4GtW7eyZMmSEve62tjY\nEBISgr+/P3369MHZ2dnIkYoaTVHQZGSgTUoCne7fl6PRoOvQAczMqi62W6g3iay9vT3NmjUjKirK\n2KEIIYQQQgjxr+Tl5bF69WoWLlxIXFwcUNRhU9zr2q1bN+l1FX8rLESbmIj28mX1ZXLTe+3ly5hk\nZ1fJrrJHjuTGzJlVUtbtqDeJLBQNrdixYwcZGRnY2toaOxwhhBBCCCFuS3p6OkuXLuW7774jLS0N\nU1NTBg8ezHPPPYeXl5exwxPGYDBgkpaGSVJSqWRVe/ky2itXMElKQqMoZX+8cWP0rVtT4OyM3skJ\nKnEBRNFoyA0L+9ef/zfqVSLr4eHBjh07iI6OpkePHsYORwghhBBCiAqdO3eOhQsXsmLFCvLy8mjU\nqBFjxoxh9OjRtGjRwtjhiWqiycnBJDERbVIS2qSkovfJyUVJ61+Jq0lKCppyhgMrpqbonZwo6NED\nvbMzBmdn9M7O6Fu2LPq3RQsUG5u7fFRVq14lsjfPXCyJrBBCCCGEqKkOHTrEV199xcaNG1EUBWdn\nZ5599lmeeOIJbGp5AlLvFQ/3TUj4+3X5clHCWpy43rhR7scVU1P0zZpR2LkzeicnDM2bo2/W7O8k\n1dkZg6MjaLV38aDuvnqVyN48c7EQQgghhBA1iV6vZ+PGjXz11VccPnwYgM6dO/P8888zYMAAec5r\nbaHTFfWaXrz4d6J66dLf7xMT0ej1ZX7U0Lgx+hYtKLj33qIEtXlzDE5ORYnqX+8N9vZgYnKXD6rm\nqVe/De3atUOr1crMxUIIIYQQosbIyclh+fLlfPPNN1y4cAGAwMBAXnjhBXr06IFGozFyhPWQokBu\nLibZ2WiKX1lZmOTkoMnK+ntZdnbRZEmZmdjFxhYlqleulJmoKhoNhubNKbzvPnQuLuhdXIp6UVu1\nKupJbdYMrKyMcLC1U71KZC0sLGjXrh0xMTEYDAZM5EqGEEIIIYQwkrS0NL799luWLl3K9evXsbCw\nYNiwYTz77LO0b9/e2OHVbYqCaWwsFtu2Yb57N9rU1KLE9OZEtZxJkspjAeibN6ewa1f0Li7oWrUq\nSlSLk9YWLcDConqOpx6qV4ksFA0vjomJ4dKlS7Rq1crY4QghhBBCiHomJSWFL7/8kiVLlpCXl0eT\nJk14+eWXGTVqFA4ODsYOr87SZGRgsXs3Ftu2YbFtG9rERHWdwcoKxdoaxcYGg709irU1BhubomVW\nVih/vTdYWxctu+llsLbGwcODRDMzsLQ04hHWL/Uykf355585c+aMJLJCCCHqhZkzZ3L8+HE0Gg1T\np07F29sbgOTkZCZPnqxul5CQwKRJkwgNDTVWqELUaUlJSXzxxRcsW7aMvLw8nJycGDNmDE888QRW\nMqS06hkMmJ46heVfiavZ4cPqkF9DkybkPvII+X5+5Pv6YmjWrHL7cnKCmxJjUf3qXSJ788zFwcHB\nRo5GCCGEqF4HDx7kwoULrFixgvj4eKZOncqKFSsAaNasGd9//z0AOp2Op556Cn9/f2OGK0SddOXK\nFT7//HN++ukn8vPzcXZ2Zty4cQwdOhQLGWpapUyuXsV8x46iXtft29FevQqAYmJCYdeu5PfpQ76f\nH4WdO9f5WX3runqXyLq7uwMQHR1t5EiEEEKI6rdv3z4CAgIAcHNzIyMjg6ysrFKP71i7di1BQUFY\nW1sbI0wh6qRLly7x+eefs3z5cgoKCnBxceGll17i8ccfx9zc3Njh1QmajAxMo6Ox2LEDi+3bMTt+\nXL23Ve/oSM6QIUW9rr17o9jZGTlaUZXqXSLbokULbG1tZeZiIYQQ9UJaWhpeXl7qz3Z2dqSmppZK\nZFetWsV33313t8MTok66ePEi8+fPZ+XKlRQWFtKmTRteeuklwsLCMDMzM3Z4lacomMbEYPnbb1hu\n2IBJSkrRzLsuLuhbty6a5Kh166KXk1Plez5zcjA9fx7Tc+fQnj2LaXy8+r64xxWKnq9a0LOn2uuq\n8/QEmfG5zqp3iaxGo8HDw4ODBw+Sk5Mj9yMIIYSoV5QyZuE8evQorq6upZLbsjg4OJT4Iu7k5FSl\n8dVHUoeVV1PqMD4+npkzZ7J06VJ0Oh3t27fnjTfe4Mknn6zxz4C9ZR0qChw+DKtXF73i4oqWm5uD\niwvakyfhyJHSnzM1hTZtoG1bcHUt/WrcuGi7ggI4d66o3NjYv/+NjYVLl0qXq9UWldm9O3TsCH36\noPH3x6JhQ4w1WLumnIe12Z3UYc3+jaomHh4e7N+/n7i4ODp37mzscIQQQohq4+joSFpamvpzSkoK\nTZs2LbHN9u3b6dmz522Vd3NZTk5OJMrkJpUidVh5NaEO4+PjmT9/PqtXr0av19OuXTsmTJjAww8/\njKmpKampqUaN71bKrUO9HrPISCx//x3L33/H9PJlAAwNGpA/YAB5/fuTHxCA0rAh6PWYJCZimpCA\n9sIFtBcvYvrXv9oLF9BGRJS5b4OtLYZGjcp99qq+RQt0Pj7oXF3Rt22LztW16L2LS1ESfbOsrKKX\nEdSE87C2K67D201m62UiWzzh05kzZySRFUIIUaf16tWL+fPnEx4ezunTp3F0dCzV83ry5En69+9v\npAiFqL3i4uL49NNP+fnnnzEYDHTs2JEJEyYwcOBAtLV1IqHCQsz37StKXjduRJuSAoChUSNywsKK\nkldfX/jnqEatFkPLlhS0bAllXBjTZGejLSvJvXgRk4wMCu+99+8k9a9/da1bl96PEH+pl4msTPgk\nhBCivrj33nvx8vIiPDwcjUbDW2+9xZo1a2jYsCGBgYEApKamYm9vb+RIhag94uLimDt3Lr/88guK\nouDh4cHLL79M//79MTExMXZ4dy4/H4uIiKJ7XiMiMLl2DSh6RE3OE0+QN2AA+T4+pXtA74BibY3O\n3R3dX9/Dhaisep3IRkVFGTkSIYQQovrd/KxY+LsdLPbrr7/ezXCEqLX+mcB6enoyceJEgoODjZvA\nKgrk56PJzsYkOxtN8SsrC5OcHDRZWX8vu3mbrCw0mZlw+DB2mZkA6Js1I3vkSPIGDKCgR4+ie1yF\nqIHq5ZlpbW1NmzZtOH36NIqioJHZzIQQQgghRDliY2OZO3cu69atQ1EUvLy8mDhxIkFBQUZJYDU3\nbmCxfTsWERFY7NqFydWrZd5fetvatCHrySfJGzCAwnvvhdrYqyzqnXqZyELRhE8bNmwgJSWFZs2a\nGTscIYQQQghRw8TExDB37lx+/fVXNYGdNGkSQUFBd70jRHv+fNHw34gIzPfvR6PTAUU9qIVdu6JY\nW2OwsUGxtkaxskL5673B2rpo2V8vg41NifWKtTXN3dzITEq6q8cjRGXV20TW3d2dDRs2EBUVJYms\nEEIIIYRQxcTE8Mknn7B+/XoURaFTp05MmjSJfv363b0EVqfD7PBhLCMisIiIwKz4cTdAQefO5AcG\nkhcQgO6eeyr/rFQZnShqoXqbyBbPXBwVFUWfPn2MG4wQQgghhDC66Oho5s6dqyaw99xzD5MmTSIw\nMPCuJLA3Dxm23LpVnXRJsbQkLzCQvH79yO/bF0Pz5tUeixA1Xb1NZD08PACZ8EkIIYQQor4rTmCL\nJz7z9vZm0qRJBAQEVHsCq71woShx3by55JBhJyeyn3qK/MBA8nv1ggYNqjUOIWqbepvItm7dmgYN\nGnDmzBljhyKEEEIIIYwgOjpaHUIM0LlzZyZOnFj9CWxhIZYbNmC1aBEWBw6oi9Uhw4GB6Dp1kiG/\nQlSg3iayWq0Wd3d3Tp06RWFhIWZmZsYOSQghhBBC3AWXLl1i9uzZ/O9//0NRFLp06cLEiRPp27dv\ntSawJikpWC1bhtWyZWj/mlwp/6GHyA0NlSHDQtyhepvIQtGET0ePHuXs2bN07NjR2OEIIYQQQohq\ndPXqVebPn8/ixYspKCjA09OTKVOmVG8PrKJgFhmJ9aJFWP72G5rCQgwNG5L9n/+QPWIEeje36tmv\nEHVcvU5ki++TPXPmjCSyQgghhBB1VE5ODt988w1ffPEFmZmZtGzZkilTpjBo0KDqew5sbi4Nfv4Z\n60WLMDt1CoDCjh3JGTWK3LAwFGvr6tmvEPVEvU5kb565eNCgQUaORgghhBBCVKXCwkJ++uknPvnk\nE1JSUrCzs2PGjBk89dRTWFhYVMs+tRcvYrVkCVbLl2Ny7RqKVkvugAHkjBpFQc+ect+rEFWkTiWy\npnFxNJoyhYw5c9C3bXvL7d3d3QGZuVgIIYQQoi5RFIX169fz4YcfcvbsWaysrJgwYQIvvPACDRs2\nrPodGgyY79yJ9aJFWGzZgkZR0Ds4kDl+PDnDh2Nwdq76fQpRz9WpRFZ76RIW+/dj9f33ZL755i23\nt7Ozo3nz5pLICiGEEELUEbt37+b999/n+PHjmJqaMmLECF5++WUcHR2rdkeKgsmVK1hu2ID14sWY\nnj0LQMF995E9ahR5AwZANfX6CiHqWCKb37MnBmtrLDdsIHP69NsauuHh4cG2bdu4fv06jRs3vgtR\nCiGEEEKIqnby5ElmzpzJjh07AHj44YeZMmUKbW9jlN4tKQomly9jdvIkZidOqC/t1atFqy0syBky\nhJxRoyjs3Lny+xNC3FKdSmSxtCS/b18arFuHaVQUur/uga1IcSIbHR3NAw88cBeCFEIIIYQQVSU+\nPp7Jkyfz888/A9C7d2+mTZuGt7f3vytQUdBevqwmq6bFSWt6eonNdC1bktu/P4X330/O4MEo9vaV\nPRQhxB2oW4kskNe/Pw3WrcPy99/Juo1EtnjCpzNnzkgiK4QQQghRS1y/fp2PP/6YpUuXotPpuOee\ne5g6dSq+vr63X4iioL10qUQvq+nJk6WTVhcXcnv2pNDbm8J77qHwnnskcRXCyOpcIpvv749iYYHl\nhg1kTZ58y+2LH8Ej98kKIYQQQtR8iqKwevVqZsyYQVpaGm5ubkyePJnQ0NDbepSOSVoa5jt3YrFz\nJxY7dqBNTi6xXteqFbkPPliUsP6VuCp2dtV1OEKIf6laE9m8vDwGDhzImDFjeOyxx6pzVyrFxob8\n3r2x3LIF7dmz6F1dK9zezc0NU1NToqOj70p8QgghhBDi34mNjeX1119n3759WFpa8vrrr/P2229z\n9a97VctUUID5oUNY7NiBxfbt6jNdAfQODuSGhpZMWps0uQtHIoSorGpNZL/88ktsbW2rcxdlyuvf\nH8stW7DcsIHsF1+scFtzc3Pat29PVFQUBoOh+h6KLYQQQggh/pWcnBzmzZvHV199hU6no1+/frz7\n7ru4uLhgbm5ecmNFQRsfj8X27Vjs3In53r2Y5OQUrTI3J9/Hh3xfX/J9fYvmU5HvfkLUStWWyMbH\nx/Pnn3/Sp0+f6tpFufICA1G0Wiw3brxlIgtFw4ujoqK4ePEibdq0qf4AhRBCCCHEbdm0aRPTp0/n\n0qVLODs789577xEUFFRiG83161js3o3F9u2Y79iB6eXL6rrC9u3J/StxLejZE8XK6m4fghCiGlRb\nIvvhhx8yffp0dQa5W3FwcMDMzAwAJyenyu3cyQl8fTHfuhUngwFu8RDqHj16sGbNGpKSkujZs2fl\n9l1DVLoOhdRhFZF6rDypw8qTOhSi9klISGD69Ols3rwZU1NTxo4dy4QJE7D6KxHVZGVhtWgRbN1K\ns0OH0BgMABgaNyY3NLSo1/WhhzC0bGnMwxBCVJNqSWR//vlnunTpgouLy21/Ji0tDSj6spGYmFjp\nGKz69sV261YyliwhZ9SoCrd1/ivR3bt3b52Yubiq6rA+kzqsGlKPlSd1WHnFdSjJrBC1Q0FBAV9/\n/TVz584lLy+Pnj178sEHH9ChQ4eiDXQ6rH76CZvZs9GmpYFWS2G3bkWJa58+FHp7g1Zr3IMQQlS7\naklkt2/fTkJCAtu3bycpKQlzc3OaN2/Ogw8+WB27K1NecDC206ZhuWHDLRNZmblYCCGEEML49u7d\ny+uvv05cXBwODg589NFHhIWFodFoQFGw2LaNhjNmYBYbi8HKisxXXqHhtGlc/eseWCFE/VEtiey8\nefPU9/Pnz8fZ2fmuJrEABicnCu69F/N9+9BcvVrhs76aN29OkyZNJJEVQgghhDCC1NRUZsyYwerV\nq9FoNIwYMYIpU6bQuHFjAEzPnKHRO+9gsWsXiokJOcOGkTl5MoZmzWhoawuSyApR79TpadryQkLQ\n6PVYRkRUuJ1Go8HDw4Nz586RI38IhRBCCCHuCr1ez+LFi+nduzerV6/G29ub3377jQ8++IDGjRtj\nkpSE7aRJOAQGYrFrF3l9+pAWEUHG7NkYmjUzdvhCCCOq9kR23Lhxd+0Zsv+UFxICgOWGDbfc1sPD\nA0VRiImJqe6whBBCCCHqvbi4OAYOHMjUqVNRFIX333+f3377jS5duqDJycFmzhya9uqF1U8/oevY\nkas//si1H39E99ctYUKI+q1O98jqXV0p9PDAYudONFlZFW7r7u4OyH2yQgghhBDVbfXq1YSEhHD8\n+HEGDRrErl27GDVqFFqgwfLlNPXxoeGcOSg2NlyfPZu0zZspMMIjHYUQNVedTmThr+HF+flYbN1a\n4Xaenp6AJLJCCCGEENUlNzeXSZMmMW7cOExMTPj6669ZsGABjo6OmO/ahUNQEI0nTsTk+nUyJ0wg\ndc8ecocNA9Nqe2KkEKKWqheJLIDl779XuF3Hjh3RaDSSyAohhBBCVIPiocQ//fQTnTp1YtOmTYSG\nhmIaG0uTp5/GfuhQTKOiyHn8cVJ27ybr1VdRbGyMHbYQooaq84msztMTXevWWPzxB+TllbudlZUV\nbdu2JSoqCkVR7mKEQsIeCyYAACAASURBVAghhBB125o1awgJCSEqKooRI0awbt062hcU0Oi113Do\n2xfLLVvI79WLtA0byPj0UwwtWhg7ZCFEDVfnE1k0GvJCQjDJzsZi164KN3V3d+fatWskJyffpeCE\nEEIIIequ3NxcXnnlFcaOHYuJiQlLP/iAz1q1omVoKE39/LBeuhR9mzakL15M+sqV6Ly9jR2yEKKW\nqPuJLJDXvz9w69mLPf6aBU+GFwshhBBCVM6ff/5JaGgo6374gaktWpDQsSPDp06l0bvvYhobS15Q\nENe+/prUrVvJ79cPNBpjhyyEqEXqxZ3zhffei75ZMyw3bSJDpyt3woDiCZ9Onz6Nn5/f3QxRCCGE\nqDYzZ87k+PHjaDQapk6divdNvV6JiYlMnDiRwsJCPD09mTFjhhEjFXXFL6tW8cerrzI9P5/HtFrM\nr1yBK1couP9+csPCyB04EMXOzthhCiFqsXrRI4uJCXnBwZhcu4b5gQPlbta5c2cAtt5ihmMhhBCi\ntjh48CAXLlxgxYoVvP/++7z//vsl1s+aNYvRo0fzv//9D61Wy5UrV4wUqaj1FAXDnj2c7tOHwePH\nsyo/n3DApE0bMl99lZR9+7j6yy/kPP20JLFCiEqrFz2yUDR7sfWSJVj+/jsFvXqVuU2LFi3w8fFh\n9+7dnD9/njZt2tzdIIUQQogqtm/fPgICAgBwc3MjIyODrKwsbGxsMBgMHD58mE/+n737Dqu67v84\n/jzAYW8UxW1WzvLOkXZbTix3Gg7ce5QDLTNHiZqr1LAsU3NmOTDRtDTTNFvepVa3plba8OdWZCgb\nzjm/P0huTRkG5xzG63FdXHDg/eW8zuei4O33M157DYDw8HB7RpUiyvHUKdyionCKjMTt/HnKA9GO\njpwPCcE4aBAZDzygacMiUuBKTCOb9sgjmH19cf3kE669/DI43PlmdPfu3fnqq6+IjIxkwoQJNk4p\nIiJSsKKjo6ldu3bWY39/f65cuYKnpycxMTF4eHgwZ84cjh07RoMGDXjuuedy/H6lSpXCaDRmPQ4K\nCrJa9pKiyI7hzp0QHg4HDwKQCKwFYtu3Z9iGDbja8OicIjuGhYjGMP80hvl3N2NYYhpZjEZSHn8c\n98hIjD/+SHq9encsa9euHZMnTyYyMpLx48fjkE3DKyIiUhTdfMScxWLh0qVL9OvXj/LlyzNs2DA+\n//xzmjdvnu310dHRWR8HBQVx4cIFa8Yt9oriGBoSE/GaPh2P997D4uDA0XLleOX8efa4uzNt/ny6\nde5M7PXrcP26TfIUxTEsbDSG+acxzL8bY5jXZrZEdWkpbdsCOe9e7O7uTqdOnTh//jxfffWVraKJ\niIhYRWBg4C3N5+XLlyldujQAfn5+lCtXjkqVKuHo6MgjjzzCyZMn7RVVigDjd99RKjgYj/feI7Fa\nNbpWrUrd8+f5sVYtNu/aRefOne0dUURKiBLVyKY2bYrZ3R3XHTvgpn+R/rsePXoAsHHjRltFExER\nsYomTZqwa9cuIHNX/sDAQDz/mvLp5ORExYoV+fPPP7O+XrVqVXtFlcIsNRWvWbMI6NIFx//7P048\n+ST3REcT9dtv9O3bl+3bt1OtWjV7pxSREqTkTC0GcHMjtVUr3LZvx+mXX8ioUeOOZQ0aNOCee+5h\n586dxMfH4+PjY+OgIiIiBaNevXrUrl2b0NBQDAYD4eHhREVF4eXlRevWrZk8eTITJ07EYrFw//33\n07JlS3tHlkLG6dgxfMeMwXjiBBmVK/N+69YMXrUKR0dHXnvtNUJDQ+0dUURKoJLVyJI5vdht+3Zc\nd+wgIZtG1mAw0KNHD+bMmcP27dvp06ePjVOKiIgUnPHjx9/yuMZNv/8qV67M+vXrbR1JigKTCY/F\ni/GaPx9DejrXe/UiLD2dVcuXU7p0aZYvX07Dhg3tnVJESqgSNbUYILVVKyzOzpnTi3PQtWtXHBwc\nNL1YREREShzHP/8koEsXvOfMweznx59vvUXwqVOs2rSJBx54gB07dqiJFRG7KnGNrMXLi9THHsN4\n/DiOp09nWxcUFETTpk05fPiwNr4QERGRksFiwX3tWkoFB+N86BDJHTvy5eLFPDp7Nt999x0dO3Zk\ny5YtlC9f3t5JRaSEK3GNLORt92LQpk8iIiJScjhcvIhfnz74vPACGI3EvvUW73fsSLu+fTl37hwv\nvPACS5Yswd3d3d5RRURKZiOb+sQTWBwccp1e/MQTT+Dj48MHH3xARkaGjdKJiIiI2Jbrhx9SumVL\nXPftI7VZMy7t2cPM335j6LBhGAwGVqxYQVhYGAaDwd5RRUSAEtrImgMCSGvcGOdDh3C4dCnbOldX\nVzp37szly5fZv3+/DROKiIiIWJ8hNhbfp5/G7+mnMaSkED97NmfeeYch4eG89tprVKxYke3bt9P2\nr9lsIiKFRYlsZOGm6cWffJJjnaYXi4iISHHksncvpVu1wu3DD0mrX58re/bwa3AwT3buzI4dO3jk\nkUfYsWMHNWvWtHdUEZHblNxGtk0bgFynF9etW5fq1avz6aefEhMTY4toIiIiIlZjuHoVn9Gj8e/T\nB4erV7k2aRJXt2zhmytXaNu2LcePH6dfv35s2LCBgIAAe8cVEbmjEtvImsuXJ+1f/8L5wAEMsbHZ\n1t04UzYtLY0tW7bYMKGIiIhIAbJYcI2KonSzZrhv3kxa3bpE79hB4ujRvL9xI927dyc+Pp45c+Yw\nd+5cjEajvROLiGSrxDaykDm92JCRgeuePTnWhYSE4OjoSGRkpI2SiYiIiBQch3Pn8OvXD79Ro3BI\nSuLa1Klc3b6d5PvvZ8qUKTz//PN4enqyfv16+vfvb++4IiK5KtmNbLt2QO7Ti0uXLk3Lli05evQo\nx48ft0U0ERERkfwzm3FftYrSzZvj+tlnpD76KFf27iVxxAhirl2jd+/erFq1iho1arBjxw6aNGli\n78QiInlSohtZU7VqpFevjsv+/RgSE3OsDQ0NBbTpk4iIiBQNjidPEtClCz5TpoCTE3GvvUbMxo2Y\nqlTh4sWLdOnSha+++orHH3+cbdu2UblyZXtHFhHJsxLdyMJf04tTUnDZty/HulatWuHv78/mzZtJ\nS0uzUToRERGRu5SWhufChZRu3RrngwdJ7tCBK/v3kxwaCgYDZ8+e5amnnuLkyZMMGzaMlStX4unp\nae/UIiJ3RY3sjWN4cple7OzsTEhICDExMXz22We2iCYiIiJyV4w//kiptm3xevVVzH5+xKxYQdyy\nZZgDAwH4888/6dKlC3/++Sdjx44lPDwcB4cS/+egiBRBJf7/XBl16pBRsSIun30Gqak51nbv3h3Q\n9GIREREpXAxJSXhNm0ZAhw4YT5wgsU8frnz+Oal//YM9wMmTJ3nqqac4d+4cL7zwAhMmTMBgMNgx\ntYjIP1fiG1kMBlLatcPh+nVcvv46x9LatWtTp04dPvvsM65cuWKjgCIiIiLZc/7iC0q1bInnsmWY\nKlfm6gcfcO3VV7H4+GTVnDhxgpCQEC5evEh4eDhhYWF2TCwikn9qZMn79GKAHj16YDKZ2Lx5s7Vj\niYiIiGTLEBuLz9ixBISG4njuHAkjR3Jlzx7S/v3vW+qOHDlC165diY6OZvbs2QwfPtxOiUVECo4a\nWSC9QQNMgYG4fPIJmEw51nbp0gWj0cjGjRuxWCw2SigiIiLyPw4XL1LqiSdwj4wkvU4donfs4PqU\nKeDmdkvdoUOH6N69O3FxcSxYsIABAwbYJ7CISAFTIwvg4EBKmzY4xsTg/N13OZb6+/vz+OOP88sv\nv3DkyBEbBRQRERHJZEhIwL9vX5zOniVh9Giid+wg44EHbqv7z3/+Q8+ePUlMTGTRokX07NnTDmlF\nRKxDjexfUtq0AfI+vRhgw4YNVs0kIiIicouMDHyffhrjsWMk9u3L9YkTwcnptrL9+/fTq1cv0tLS\nWLJkCU899ZQdwoqIWI8a2b+k/fvfmH18cN25E3KZMty8eXPKlCnD1q1bSUlJsVFCERERKdEsFrxf\nfBHXzz4jpUULrs2aBXfYdXj37t0MGDAAi8XCihUraN++vR3CiohYlxrZG5ydSWndGsfz5zHmMmXY\nycmJkJAQ4uPj2bVrl40CioiISEnm8fbbeLz7Lum1ahG3dOkd78R+/PHHDB48GIPBwJo1awgODrZD\nUhER61Mje5Os3Yu3bMm19sb04sjISKtmEhEREXHdvh3vmTMxBQURs3YtFk/P22qioqIYMWIELi4u\nrFu3jqZNm9ohqYiIbaiRvUlqy5aYgoJwX7sWh6tXc6y97777qFevHvv37+fChQs2SigiIiIljfHg\nQXzHjMHs6UnM2rWYg4Juq1m/fj2jR4/Gw8ODDRs20LhxYzskFRGxHTWyN3NxIWHkSBySk/FYsiTX\n8h49emA2m/nggw9sEE5ERERKGsfff8d/wADIyCB22TIyatW6rWbVqlU899xz+Pr6EhkZSf369W0f\nVETExtTI/k1Sr16YypbFfdUqDLncle3UqROurq46U1ZEREQKnOHqVfz79sUhNpb4uXNJa978tpol\nS5YwZcoUSpUqxebNm3nwwQdtH1RExA7UyP6dqysJzzyDQ1ISnkuX5ljq4+NDmzZt+P333zl06JCN\nAoqIiEixl5KC/8CBOP3xBwljxpDcu/dtJYsWLWLGjBkEBQURFRVFjRo17BBURMQ+1MjeQVLv3pjK\nlMm8KxsTk2PtjU2fNm7caItoIiIiUtyZzfiGheF86BDJXbpw/YUXbivZvHkzc+bMoUKFCkRFRXHv\nvffaIaiIiP2okb0TN7fMtbKJiXguW5Zj6aOPPkq5cuXYtm0bSUlJNgooIiIixZXX7Nm4bd9OaqNG\nxL322m1nxR4+fJjx48fj7e3NunXrqFy5sp2SiojYjxrZbCT17o0pMBD3FStyvCvr6OhI9+7dSUhI\nYMeOHTZMKCIiIsWN+7vv4rl4MRnVqhG7ciW4uNzy9fPnzzNo0CDS09NZsmSJ7sSKSImlRjY7N92V\n9XjnnRxLu3XrBmh6sYiIWJc2FizeXD77DO/JkzEFBBDz3ntY/Pxu+XpSUhIDBw7kypUrTJs2jeZ3\n2PxJRKSkUCObg6Q+fTCVLo3HihUYYmOzratatSqNGjXi66+/5syZMzZMKCIiJUmLFi2IiIi46981\ns2fPpkePHoSGhnLkyJFbvtayZUt69epF37596du3L5cuXSrIyJJHTkeP4jt8ODg7E7tmDaa/TRe2\nWCyMGzeOo0eP0qtXLwYPHmynpCIihYMa2Zy4uZE4ciQOCQm53pW9selTZGSkLZKJiEgJtGnTJkqX\nLs3kyZMZOHAg27dvJy0tLcdrvvvuO06fPs3GjRuZNWsWs2bNuq3mnXfeYe3ataxdu5YyZcpYK75k\nw+HcOfz79cOQnEzcm2+SXq/ebTURERFs376dRo0aMXv2bAx/WzcrIlLSqJHNRWIe78p27NgRd3d3\nIiMjMZvNNkwoIiIlRenSpenTpw9r165l2rRprF+/nscee4yIiAhSU1PveM2BAwcIDg4GoFq1asTH\nx5OQkGDL2JKT+Hj8+/bF8dIlroeHk9Ku3W0lH330EfPnz6dixYosX74cZ2dnOwQVESlc1Mjmxt2d\nxKefxuH69Rzvynp4eNChQwfOnDnDgQMHbBhQRERKkoMHDzJp0iSGDh1KvXr1WLduHd7e3oSFhd2x\nPjo6Gr+b1lr6+/tz5cqVW2rCw8Pp2bMn8+fP1zpcW0pPh27dMP78M4mDBpE4dOhtJUePHmXMmDF4\neHiwevVqAgIC7BBURKTwcbJ3gKIgqV8/PBYvxmPFChKHDcPi63vHuh49ehAZGcn69etp0qSJjVOK\niEhx17p1a8qXL0/37t2ZMWMGRqMRyLzTumfPnjx9j783qmPGjOGxxx7Dx8eHkSNHsmvXLtq0aZPt\n9aVKlcp6XoCgoKB/8EoEiwWGDIHdu6FjRzyWLcPD0fGWkosXLzJkyBBSU1PZunUrLVu2tFPYwk8/\nh/mnMcw/jWH+3c0YqpHNA8tfd2W9X34Zj+XLSRg//o51jRs3pnr16mzdupWnn36a2rVr2zipiIgU\nZ8uXL8disVClShUAjh8/Tq1atQBYt27dHa8JDAwkOjo66/Hly5cpXbp01uPOnTtnfdy0aVN+/fXX\nHBvZm79XUFAQFy5c+EevpUTLyMDn+edx37gR6tfnYkQElsuXbylJSUmhW7dunDlzhkmTJtGwYUON\ndTb0c5h/GsP80xjm340xzGszq6nFeZTUvz+mgAA8li/HEB9/xxqDwUB4eDhms5nw8HBNzxIRkQIV\nFRXF0qVLsx4vW7aM+fPnA2S7+U+TJk3YtWsXAMeOHSMwMBBPT08Arl+/zuDBg7M2jDp48CD33Xef\nNV+CJCXhN3gw7hs3kla3LuzcicXd/ZYSi8XCCy+8wOHDh3nqqacYNWqUncKKiBReamTz6MZdWYdr\n1/BYvjzbuubNmxMcHMw333zDzp07bZhQRESKu2+//ZY5c+ZkPV64cCGHDx/O8Zp69epRu3ZtQkND\nmTlzJuHh4URFRbF79268vLxo2rRp1tE8/v7+Od6NlfwxxMUR0LMnrrt3k9q0KTGbNsFNd8dvePvt\nt9m0aRMPPfQQ8+fP1w7FIiJ3oKnFdyFpwIDMtbLvvEPikCFYfHzuWBceHs7+/fuZPn06LVu2xNXV\n1cZJRUSkOEpPTyctLS1r19rExEQyMjJyvW7835bE1KhRI+vj/v37079//4INKrdxuHAB/169MP7y\nC8mdOxO3cCHcYffh3bt3M2vWLIKCgli5cqX+hhARyYbuyN4Fi7s7iSNGZN6VXbEi27pq1aoxaNAg\nzpw5w7Jly2yYUEREirPQ0FDatWvH2LFjGTNmDB06dCA0NNTesSQXjidPEtCpE8ZffiFx8GDi3nzz\njk3sL7/8wjPPPIOLiwurVq3Smb4iIjmwWiObnJxMWFgYffr0oVu3buzbt89aT2VTSQMHYvbzw+Od\ndzBcu5Zt3bhx4wgICOCNN97g4sWLNkwoIiLFVbdu3Xj33Xdp27YtHTp0YN26dTz++OP2jiU5MH7/\nPaU6d8bp3DmuTZrEtRkzwOH2P7+uXr3KgAEDSExMZOHChTz44IN2SCsiUnTkqZH96aefshrRiIgI\n+vfvz6FDh3K8Zt++fdSpU4f33nuPhQsXMnfu3PynLQQsHh4kjBiBQ3x8jndlvb29mThxIklJSbes\nZxIREcmPpKQk/P398fPz4/fff6d79+72jiTZcNm3D/9u3TDExxM3fz6Jo0fDHda7pqWlMWzYME6f\nPs24cePo1KmTHdKKiBQteWpkZ86cSdWqVTl06BBHjx7lpZde4o033sjxmnbt2jH0r4O9L1y4UKym\nx9xyV/b69WzrQkNDqV27Nps2beKHH36wYUIRESmOZs6cyejRo3nmmWeYO3cu48aN48knn7R3LLkD\n16go/Pr3x2CxELtiBcm9et2xzmKx8OKLL3LgwAHat2/Pc889Z+OkIiJFU542e3JxcaFKlSps3LiR\n7t27c++99+Jwh2kxdxIaGsrFixdZsmRJjnU3H7BeJA4Tfv55mDyZsps2wZQp2ZYtXryYZs2aMWPG\nDL755ps8j1t+FYkxLOQ0hgVD45h/GsP8Ky5jePToUXbu3Enfvn1Zu3YtP/30E7t377Z3LPkbj2XL\n8J42DbOPDzGrV5PeqFG2tW+99RbvvfcetWvX5vXXX7fZ3wkiIkVdnhrZ5ORkdu7cyZ49exg5ciRx\ncXFcy2F96M02bNjAiRMneP7559m2bVu2W8jfOGC9qBwmbOjalcB582D+fC5364bFy+uOdffddx8d\nO3Zk+/btLF68mJCQEKtnKypjWJhpDAuGxjH/NIb5d7cHrBdmN3YrTk9Px2KxUKdOHV555RU7p5Is\nFgtes2bhuXgxprJliXn/fTJq1sy2/IsvvmDs2LGUKlWKVatW4f6382RFRCR7efpnv2effZbt27cz\nbtw4PD09Wbt2LQMGDMjxmp9++inrj6+aNWtiMpmIiYnJd+DCwuLpSeKwYTjExeG+alWOtS+++CKu\nrq7MmjWLxMREGyUUEZHipmrVqrz//vs0aNCAgQMHMn36dK7nsMRFbCg9HZ9x4/BcvJiMatW4um1b\njk3s2bNnGT58OI6OjqxYsYIKFSrYMKyISNGXp0a2cePGvPrqq7Rr147o6GgeeeQROnTokOM1hw4d\nYuXKlUDm3dakpCT8/Pzyn7gQSRw0CLOvL55Ll2JISMi2rmLFiowYMYKLFy/y1ltv2TChiIgUJ9On\nT6d9+/Y8++yzhISEULly5VyX7ogNJCXhN3gw7pGRpD30ENFbt2LKoTE1m808++yzxMfH89Zbb9Gw\nYUMbhhURKR7y1Mi+/PLL7Ny5k7i4OEJDQ3nvvfeYNm1ajteEhoYSExNDr169GDZsGFOnTi126z4s\nXl6Zd2VjY3O9Kztq1CiCgoJ4++23OXPmjI0SiohIcTJ79mx8fX1xcHCgY8eODBgwgLJly9o7Volm\niI0lIDQU1z17SGnenJjISCwBATles2bNGr766iuCg4MZPHiwjZKKiBQveeosjx8/Trdu3di5cydd\nunRh4cKFnD59OsdrXF1dWbBgAevWrSMqKoqWLVsWSODCJnHQIMw+PnguWZLjXVl3d3emTJlCamoq\nM2fOtGFCEREpLhwdHTlw4ACpqamYzeasN7EPh/PnCejSBedDh0ju0oXY1auxeHjkeM0ff/zBzJkz\n8fPzY968ednuHSIiIjnLUyNrsVgA+Pzzz7Ma0rS0NOulKkIs3t55vivbpUsX6tevz/bt2zlw4ICN\nEoqISHGxadMmBg0aRN26dalVqxa1atWidu3a9o5VIjn+3/9RqlMnjL/+SsLQocQtWgR/bcaVHZPJ\nxNixY0lOTmbWrFnF6mhCERFby1MjW7VqVdq1a0diYiI1a9Zk69at+Pj4WDtbkZE4ePD/7srmsJmT\nwWBgxowZALz00kuYTCZbRRQRkWLg8OHDnDhxgp9//jnr7cSJE/aOVfKYTPiOHo3j+fNcmzSJ69Om\nQR6WT73zzjscPHiQDh066PxfEZF8ytPxOzNnzuTXX3+lWrVqANx77728+uqrVg1WlFi8vUkcMgSv\nBQtwX72axJEjs6196KGH6N69O5GRkWzYsIHevXvbMKmIiBRlr7/++h0/HxYWZuMkJZvHihU4HzxI\ncseOJI4enadrTp48ySuvvEJAQABz5szRlGIRkXzK0x3ZlJQU9u7dy5gxY3j66af5+uuvs86yk0yJ\nQ4Zg9vbG4+23c7wrCzBx4kTc3d2ZO3duns/jFRERcXR0zHozm818++23On7HxhxPncJr7lxMAQFc\nmz07T9dkZGQQFhZGamoqr776KgG5bAYlIiK5y1Mj+9JLL5GQkEBoaCjdu3cnOjqaF1980drZihSL\njw+JQ4bgGBOD++rVOdaWLVuWsLAwrl69SkREhG0CiohIkTdq1Kist3HjxrF27VrOnz9v71glh8mE\n77hxGFJSuDZ3LuY8NqSLFy/mxx9/JCQkhLZt21o5pIhIyZCnRjY6OpoXXniB5s2b06JFC6ZMmcKl\nS5esna3ISRw6NHOtbEQEjidP5lg7dOhQKlWqxIoVKzh16pSNEoqISHGSkZHB//3f/9k7RonhsXQp\nzocPk9y5Mynt2+fpmmPHjrFgwQLKli3Lyy+/bOWEIiIlR54a2eTkZJKTk7MeJyUlkZqaarVQRZXF\nx4f4V1/FISkJv+HDISkp21pXV1emTp1KRkZG1gZQIiIiOWnWrBnNmzfPemvcuDEPP/ywvWOVCE6/\n/orXvHmYSpcmPo/H6KWlpREWFkZ6ejrz58/H19fXyilFREqOPG321KNHD9q2bUudOnWAzH9d1MYS\nd5bSsSOJBw7gsXo1Pi++SPxrr2Vb27ZtW5o0acKePXvYt28fLVq0sGFSEREpatatW5f1scFgwNPT\nE29vbzsmKiEyMvAZOxZDairxr7yCxd8/T5ctXLiQ48eP07Nnz6zjC0VEpGDk6Y5s165dWb9+PZ07\nd6ZLly5s2LBB02FzcC08nPQHHsB9wwbcNm3Kts5gMDB9+nQcHBwIDw8nPT3dhilFRKSoSU5OZsOG\nDZQvX55y5coxZ84cTuaylEXyz2PxYpx//JGkrl1JbdMmT9f897//ZdGiRZQvX55p06ZZN6CISAmU\np0YWICgoiODgYFq1akWZMmU4cuSINXMVbS4uxC5ditnLC++JE3H69ddsS2vVqkWfPn04deoUa9as\nsWFIEREpaqZPn06zZs2yHoeEhGh5ipU5nTiB14IFmMqW5VoexzolJYUxY8ZgMpmIiIjAy8vLyilF\nREqePDeyf2exWAoyR7FjqlKF+NdewyE5Gd9hwzDksF72+eefx9vbmwULFnD16lUbphQRkaLEZDLR\noEGDrMcNGjTQ72NrSk/HNywMQ3o68fPmYcnjGtd58+Zx8uRJBgwYwKOPPmrlkCIiJdM/bmR1kHfu\nUtq3J3HwYIy//or35MnZ1gUEBPDcc88RHx/P/PnzbZhQRESKEi8vL9atW8dvv/3GyZMnWblyJR4e\nHvaOVWx5LlqE8aefSAoNJbVVqzxdc/DgQZYsWUKVKlV0VKGIiBXluNlTs2bN7tiwWiwWYmNjrRaq\nOLn24osYDx/GPTKStMaNSQ4NvWPdgAEDWLt2LWvXrqVfv37UrFnTxklFRKSwmzNnDgsWLGD9+vUA\n1KtXjzlz5tg5VfHk9NNPeC5ciCkoiGt5XOOalJTE2LFjAYiIiMDd3d2KCUVESrYcG9mbd0eUf8jF\nhbglSyj1xBP4TJ5M+r/+RUaNGreVGY1Gpk2bRp8+fXjppZeIjIzEweEf3zAXEZFiyN/fn6FDh1Kl\nShUAjh8/jn8ed9CVu5CWhu/YsRgyMohbsABLHneGnj17Nn/88QfDhw+nUaNGVg4pIlKy5dgplS9f\nPsc3yRtTpUrEHH4p5AAAIABJREFUvfYahpQU/IYNw5CYeMe6li1bEhwczDfffMPSpUttnFJERAq7\niIiIW34/LFu2TEtSrMBz4UKMx4+T2KcPac2b5+mar7/+mpUrV3LvvfcyYcIE6wYUEZF/vkZW7k5q\n27YkDB2K06lTeE+cCNlszrFgwQICAwOZPXs2Bw8etHFKEREpzL799ttbphIvXLiQw4cP2zFR8WP8\n73/xXLSIjAoVuD51ap6uSUhI4Nlnn8XBwYHXX38dNzc3K6cUERE1sjZ0fcoU0h56CPfNm3HLZtp2\n6dKleeutt7BYLDz99NPExMTYOKWIiBRW6enppKWlZT1OTEwkIyPDjomKmdRUfMLCMJhMxC9YgMXT\nM0+XzZgxgzNnzjBq1CgeeughK4cUERFQI2tbzs7ELV2K2dcXn5dewun48TuWNWnShOeee47z588z\nduxYHa0gIiIAhIaG0q5dO8aOHcuYMWPo0KED7du3t3esYsNrwQKMv/5KYv/+pD32WJ6u+fzzz3nv\nvfeoVasW48aNs3JCERG5QY2sjZkqVCBu4cL/rZdNSLhj3ZgxY3jsscfYs2eP1suKiAgA3bp1Y86c\nObRr146OHTsSFhbGxo0b7R2rWDB+/z0eixeTUakS1/N4bE58fDzPPfccTk5OLFy4EBcXFyunFBGR\nG3LctVisI/Xxx0kYPhzPpUvxmTCBuLfegr8dc+To6Mibb75J69atmT17Ng0aNKBBgwZ2SiwiIoXB\nrFmz+Oqrr4iOjqZSpUqcOXOGQYMG2TtW0ZecnLlLsdlMfEQEljyezTt16lQuXLjA888/T506dawc\nUkREbqY7snZyffJk0urXx23rVtzfe++ONTfWy5rNZkaMGKGze0VESrgjR46wc+dOatSowebNm1m5\nciXJycn2jlXkec2bh9OpUyQOGULaI4/k6Zq9e/eyadMmHnzwQUaNGmXlhCIi8ndqZO3FaCT27bcx\n+/nhPXUqTkeP3rHs5vWyYWFhWi8rIlKCOTs7A5mbPlksFurUqcP333+f63WzZ8+mR48ehIaGcuTI\nkTvWLFiwgL59+xZo3qLAePAgHkuXklG1KtcmTszTNSaTiZkzZ+Lg4EBERARGo9HKKUVE5O/UyNqR\nuUIF4l5/HUNqKn4jRmC4fv2OdVovKyIiAFWrVuX999+nQYMGDBw4kOnTp3M9m98dN3z33XecPn2a\njRs3MmvWLGbNmnVbzalTp0rmkW9JSfj+tUFTXEQEuLvn6bItW7bw888/07VrV2rWrGnNhCIikg01\nsnaWGhxMwjPP4PTHH/iMH3/H82VvrJe9cb7soUOH7JBURETsbfr06bRv355nn32WkJAQKleuzJIl\nS3K85sCBAwQHBwNQrVo14uPjSfjbRoNz584tkTvues+di9Pvv5M4bBjpDz+cp2tSU1OZN28ezs7O\njB8/3soJRUQkO9rsqRC4/sILOB88iNv27aQ98ghJAwbcVnNjvWyPHj0YMWIEu3fvxs/Pz/ZhRUTE\nbgwGA76+vgB07NgxT9dER0dTu3btrMf+/v5cuXIFz7/OSI2KiuLhhx+mfPnyefp+pUqVumUqbVBQ\nUF7jFy5ffAHLl0P16nhGRODp5pany9544w3OnDnDuHHjaNiwYYFEKbJjWIhoDPNPY5h/GsP8u5sx\nVCNbGPy1XrZ069Z4T5tGWr16ZDz44G1lN9bLzps3j7CwMNasWYPhb7sdi4iI5OTmvRbi4uKIiopi\n1apVXLp0KU/XR0dHZ30cFBTEhQsXCjyjLQSMH48zED1/PulxcRAXl+s1CQkJzJgxA09PTwYNGlQg\nr70oj2FhoTHMP41h/mkM8+/GGOa1mdXU4kLCXK4ccYsWYUhLw3/QIJyy2YxD62VFRORuBAYG3tJ8\nXr58mdKlSwPwn//8h5iYGHr37s2oUaM4duwYs2fPtldUm3E6cgTngwdJadWK9Pr183zdsmXLuHr1\nKiNGjCAgIMCKCUVEJDdqZAuR1JYtuTZlCo7nz1OqY0fc33nntjWzWi8rIiJ3o0mTJuzatQuAY8eO\nERgYmDWtuE2bNuzYsYPIyEjefPNNateuzeTJk+0Z1yY8li8HIHHw4Dxfc/XqVd5++20CAgIYPny4\ntaKJiEgeqZEtZBJHjuTqunWYvb3xCQ/Hb+BADDExt9TofFkREcmrevXqUbt2bUJDQ5k5cybh4eFE\nRUWxe/due0ezC4fLl3H78EPS77uPtGbN8nzdG2+8QWJiImPHjsXDw8OKCUVEJC+0RrYQSmvenOg9\ne/AdPRrXTz+ldOvWxC5eTHqjRlk1Wi8rIiJ59ffddWvUqHFbTYUKFVi7dq2tItmN+7vvYkhPJ2nw\nYMjj782zZ8+yZs0aKlasSJ8+faycUERE8kJ3ZAspc5kyxKxfz/UJE3C4dImArl3xfP11MJmyarRe\nVkRE5C6kpuL+7ruYfXxI7to1z5fNnz+ftLQ0JkyYgIuLixUDiohIXqmRLcwcHUkYO5armzdjDgzE\n65VX8O/VC4fLl//68q3rZQ8fPmznwCIiIoWX24cf4hgdTVKvXljc3fN0zc8//8ymTZuoWbMmnTt3\ntnJCERHJKzWyRUB6o0Zc2b2blOBgXL78klLBwTjv3w9ovayIiEieWCx4LF+OxcGBpIED83zZK6+8\ngsViYeLEiTg6OloxoIiI3A01skWExd+f2DVriJ8+HYf4ePx79cJrzhzIyMhaL3vu3DnGjh17yxmB\nIiIiAsbvvsP400+ktG2LqUKFPF1z8OBBdu3axcMPP0xwcLCVE4qIyN1QI1uUGAwkDR3K1W3bMFWq\nhOeiRQQ89RQOZ89mrZfdvXs3ERER9k4qIiJSqGQduTNkSJ7qLRZL1pm6kyZN0oaKIiKFjBrZIii9\nbl2id+0iuVMnnA8dovTjj+O+e3fWetkXXniBvXv32jumiIhIoeB49iyuO3eS/sADpD/8cJ6u2bdv\nH99++y3BwcE0uunUABERKRzUyBZRFm9v4t5+m7h58zCkpOA/aBDV3niDd958E6PRyODBg9n/1zpa\nERGRksx91SoMZnPm3dg83Fk1m83Mnj0bg8HApEmTbJBQRETulhrZosxgILl3b6J37CD9/vvxWLGC\nti+/zO7FiwEYOHAgX3/9tZ1DioiI2I8hKQn3deswlS5NcqdOebpm69atHD9+nJCQEGrWrGnlhCIi\n8k+okS0GMmrU4OqOHST17Inx6FGajB7NrpEjMZvN9OvXj2+//dbeEUVEROzCbdMmHOLjSerbF/Jw\nBmxaWhrz5s3DaDQyfvx4GyQUEZF/Qo1sMWFxdyd+wQJi33wTTCaaRkTwTadOpKel0adPHw4dOmTv\niCIiIrZlNuOxYgUWo5Gkfv3ydMn777/P6dOn6devH5UqVbJyQBER+afUyBYzKU89BV99hblsWep/\n8AEnGjTAkpxM7969+eGHH+wdT0RExGacv/gCp1OnSH7yScyBgbnWJyYmsnDhQtzd3QkLC7NBQhER\n+afUyBZH9eoRvXMnafXrc++33/J75cp4JSTQq1cvjhw5Yu90IiIiNnG3R+4sX76cK1euMHz4cEqV\nKmXNaCIikk9qZIspc2AgVzdtIqlrV8r88Qe/+Phwb3w8PXv25NixY/aOJyIiYlWOp07huncvaQ0b\nkvHgg7nWx8TEsHjxYvz9/RkxYoQNEoqISH6okS3OXF2Jf/11rr34Ih7x8RwwGmkdG0uPHj345Zdf\n7J1ORETEajxWrgQgcejQPNUvWrSI69evExYWhpeXlzWjiYhIAVAjW9wZDCQ+8wyxq1fj6OzMRmBM\nTAw9unXj5MmT9k4nIiJS4Azx8bhFRpJRvjwpbdrkWn/u3DlWr15N+fLl6du3rw0SiohIfqmRLSFS\nW7fm6vbtZFSqRDjwVnQ0A7p14/fff7d3NBERkQLlvn49DklJJA0YAE5OudYvWLCA1NRUnn/+eVxd\nXa0fUERE8k2NbAmSUaMG0R9/TOojj9AV2Hz5MmOfeorTp0/bO5qIiEjBMJlwX7UKi6srSb165Vp+\n8uRJIiMjuf/++wkJCbFBQBERKQhqZEsYS0AAMevXk9inDw8BH12+zJyOHTl79qy9o4mIiOSby6ef\n4nTmDEndumHx88u1/pVXXsFsNjNx4kQcHR1tkFBERAqCGtmSyNmZa6+8QvzMmZQyGIiMjmZ9mzac\nO3fO3slERETy5caRO0mDB+da+/3337Njxw7q16/PE088Ye1oIiJSgNTIllQGA0mDBhG3bh0mFxde\ni4nh++BgLqqZFRGRIsrpp59wOXCA1KZNybj//hxrLRYLs2fPBmDKlCkYDAZbRBQRkQKiRraES2vW\njOu7d3PJ15dh8fHENWtGtDaAEhGRIshjxQoAEocMybV2//79fPPNN7Rs2ZLGjRtbO5qIiBQwNbKC\n+d57MX/zDccrVqRFUhKuLVuStGMHWCz2jiYiIpInDtHRuG3ZQsY995DasmWu9YsWLQJg4sSJ1o4m\nIiJWoEZWMvn64vvVV+yuU4f70tKoNmQIno89htu6dZCcbO90IiIiOXJfuxZDWhqJgwaBQ85/3pw4\ncYIDBw7QtGlT6tSpY6OEIiJSkNTIShaD0UjtXbtY3Ls3GwG333/Hd/x4yjRogNecOTicP2/viCIi\nIrdLS8N9zRrMXl4kd++ea/mqVasAGDhwoLWTiYiIlVi1kX311Vfp0aMHISEhfPrpp9Z8KikgBoOB\nLvPmEb90KTVdXJgDpKSm4rloEYGNGuE7fDjG777TtGMRESk0XD/6CMfLl0nq2ROLp2eOtXFxcWze\nvJmKFSsSHBxso4QiIlLQrNbI/uc//+HkyZNs3LiR5cuXZ+0MKEVDx44deWPLFhYGBhKQlMTyxo1J\nv/9+3LZvp1TnzgS0bYvbpk2QmmrvqCIiUpJZLHgsX47FYCApD3dYN27cSHJyMv3799e5sSIiRZjV\nGtmGDRvy+uuvA+Dt7U1ycjImk8laTydW8K9//YsdO3ZQrXZthv7nP7Ty9+f06tWktG2L8aef8A0L\nI7BhQzznzcPh0iV7xxURkRLIePgwzj/+SOrjj2OqXDnHWrPZzOrVq3F1daVnz542SigiItZgtUbW\n0dERd3d3AD744AOaNm2qf/ksgsqVK8fWrVtp06YNX339NS1nzODwlClcOXCAhBEjMKSn4xURQeDD\nD+M7ahTGH3+0d2QRESlBPN55B8jbkTt79+7l9OnTdOnSBT8/P2tHExERKzJYLNZd7Lhnzx6WLl3K\nypUr8fLyyrYuPT0do9FozSiSD2azmcmTJ/PKK6/g5+dHVFQUzZs3h8REWLsW3ngDTpzILG7cGMaO\nhZAQcHKya24RESlYFy5cyPo4KCjolse25nDuHIGNG5NRvTrRu3eDwZBjfe/evdm3bx+ffvppodmt\n2N5jWBxoDPNPY5h/GsP8uzGGQUFBeaq3apfx5ZdfsmTJEpYvX55jEwsQHR0N6IegIFhrDMPCwihb\ntiwTJkygdevWzJkzh969e8OTT0KnTjh/8QUeK1bgumcPhIZiKleOxMGDSerVC4uPT4HnsSb9HBYM\njWP+aQzz725/MUrR4bFmDQaTKfPInVya2N9++419+/bx8MMPF5omVkRE/jmrTS2+fv06r776KkuX\nLsXX19daTyM21qNHDzZu3IiXlxfPP/8806dPz1z7bDCQ1qwZse++y+WvviJxwAAMsbF4v/wygfXr\n4/3SSzj++ae944uIlEizZ8+mR48ehIaGcuTIkVu+FhkZSffu3QkNDWXatGlYeaJWwUlKwv399zH5\n+5PcpUuu5WvWrAF05I6ISHFhtUZ2x44dxMbGMnbsWPr27Uvfvn05r3NIi4XGjRvz8ccfc++997J0\n6VIGDRpEQkJC1tdN99zDtdmzuXz4MNcmT8bi44PHihWUbtIEv8GDMX77rY7vERGxke+++47Tp0+z\nceNGZs2axaxZs7K+lpyczMcff8z777/Phg0b+P333/nhhx/smDbv3LZswSE2lqS+fcHNLcfaxMRE\nNm7cSJkyZWjXrp2NEoqIiDVZbWpxjx496NGjh7W+vdhZlSpV2L59O8OHD2f37t107tyZ1atXU6FC\nhawai68viaNGkTh8OK4ffYTHsmW47tyJ686dpNWtS+LQoaR07AhaGy0iYjUHDhzIOi+1WrVqxMfH\nk5CQgKenJ25ubll3KpOTk0lISKB06dL2jJtnHqtWYXFyIql//1xrP/jgA65fv87w4cO1H4eISDFh\ntTuyUvz5+Pjw3nvv0b9/f44fP0779u35/vvvby80Gknp0oWrO3YQvWVL5vE9R47gN2oUgY0b4/Hm\nmxji4mz/AkRESoDo6Ohbduj19/fnypUrt9QsW7aM1q1b06ZNGypWrGjriHfN6cQJjMePk9qqFeay\nZXOstVgsrFq1CqPRSJ8+fWyUUERErE1bykq+ODk5MWfOHO677z6mTp1KSEgIERERdO7c+fZig4H0\nRo2IbdQIxz//xGPFCtw2bMB79mw8IyJI7tGDxCFDMN1zj+1fiIhICXGnNbDDhg2jX79+DB06lPr1\n61O/fv1sry9VqtQtdzXtsonWX+fUuw4enOvz7927l19//ZVevXpRt25dW6S7a9qILP80hvmnMcw/\njWH+3c0YqpGVAjFo0CCqVq3K8OHDeeaZZ/jvf//LhAkTcMtm3ZKpShWuvfwy18ePx339etxXrMBj\n9Wrc16whtXVrkvr1I/Xf/wZXVxu/EhGR4iUwMDDrZACAy5cvZ00fjouL4+TJkzRs2BBXV1eaNm3K\n999/n2Mje/P3ssuu2mYzgWvXYvDy4lL9+pDL88+fPx+Anj17FsodwLUzef5pDPNPY5h/GsP8u9tT\nBjS1WApMixYt2L59O1WqVGHp0qUEBwfzzTff5HiNxceHxBEjuHLgALFLlpD+0EO4fvop/n36UKZ2\nbfz698d97VoctFGYiMg/0qRJE3bt2gXAsWPHCAwMxNPTE4CMjAwmTpxIYmIiAEePHqVq1ap2y5oX\nzt99h+P586S0a5frJk9nz55l165dPPjgg9SrV89GCUVExBZ0R1YKVPXq1dmzZw/z5s3jnXfeoWvX\nrvTt25cpU6bg7e2d/YVOTqR06kRKp04YDx/G9aOPcP3sM1x378Z19258gPRatUht1YqU4GDS69UD\nR0ebvS4RkaKqXr161K5dm9DQUAwGA+Hh4URFReHl5UXr1q0ZOXIk/fr1w8nJierVq9OqVSt7R86R\n2+bNACQ/9VSute+++y5ms5mBAwdiyOWcWRERKVoMlkJyYNyNW/G6LZ9/hWUMf/jhB5577jl+/vln\ngoKCmDNnDo8//vhdfQ/H06dx2bMHl88+w+WbbzCkpQFg9vMjtUULUlq1IrV5cyw3bWRSEArLGBZ1\nGsf80xjm391OVZKc3fzzaPOfz9RUyvzrX1hcXbl86FCO/6CZnJxMgwYNADh8+DCuhXSpiv4bzz+N\nYf5pDPNPY5h/mloshcZDDz3EJ598wvjx44mOjmbAgAGMGDHilvVVuTFVrkzS4MHErlvHpePHiVm1\nisQ+fbC4uuIWFYXfyJGUeeABAjp3xmPRIpyOH9cZtSIixZTL3r04xMeT3LlzrrNytm3bRmxsLL17\n9y60TayIiPxzamTFqpydnXn22Wf59NNPqV+/Ptu2baNZs2Zs3rz5jjtn5sTi7k7qE09w7dVXuXzo\nEFd27+baxImk16uH8dAhvOfMoXRwMIENG+I9aRLGb78Fs9lKr0xERGzNbcsWIPdpxRaLhZUrV+Lg\n4EC/fv1sEU1ERGxMjazYRPXq1dm6dSszZswgJSWF0aNH07dvX86ePfvPvqHBQEbt2iSOGcPVbdu4\ndOQIsW++SXLnzhiSkvBYs4ZSXbpQunFjvObOxenXXwv2BYmIiE0Zrl3Ddfdu0u+7j4w6dXKsPXz4\nMEePHuWJJ56gQoUKNkooIiK2pEZWbMbR0ZEhQ4awb98+mjZtyt69e2nRogWrVq3CnM87pxZ/f1Ke\neoq4xYu5dOQIV9evJ6l7dxzi4vB84w1KN29Oqdat8Xj7bRy0fkFEpMhx/fhjDKmppDz1FOSycdPK\nlSuBzKPhRESkeFIjKzZXqVIl1q9fz8KFCzEajUyZMoUuXbpw8uTJgnkCJyfSmjUjfuFCLv34I7Fv\nv01K69Y4/fIL3i+/TGCDBvh3747b+vUYrl0rmOcUERGryppW3KVLjnWXLl3io48+onr16vz73/+2\nRTQREbEDNbJiFwaDge7du/P555/ToUMHDh48SOvWrXnjjTdIT08vuCdydyflySeJXbOGSz/+SPzc\nuaQ3bIjLV1/h+9xzlKlbF9+hQ3HZuRNSUwvueUVEpMA4XLiA89dfk9awIaZKlXKsfe+998jIyNCR\nOyIixZwaWbGrwMBAli1bxooVK/D19WXu3Lm0bduW//73vwX+XBZ/f5L69ePq1q1c/vZbrk2cSEbl\nyrh9/DH+gwdT5l//wuf553E+cECbRImIFCJuH36IwWLJdZOntLQ01q5di5eXFyEhITZKJyIi9qBG\nVgqFtm3b8vnnn9OzZ0+OHz9O+/btmTx5MvHx8VZ5PlPFiiSOGUP0vn1c+fRTEkaMwOLmhvv77xMQ\nEgIVK+I3aBCer76K60cf4fjbb2AyWSWLiIjkzC0qCouTE8kdOuRYt2PHDi5fvkyPHj3w8PCwUToR\nEbEHJ3sHELnB19eXBQsW0KVLFyZNmsTq1av56KOPmDp1KiEhIdaZImYwkFGnDtfr1OH6lCk4HziA\nW1QU7p99husnn+D6ySdZpRZXV9KrVyejRg3Sa9Uio0YNMmrVwhwQUPC5REQEAKdff8X400+kBAdj\nyeX/t6tWrQJgwIABNkgmIiL2pEZWCp1HH32UPXv2sGzZMiIiIhgzZgzr1q1jzpw5VK9e3XpP7OhI\n2qOPkvboo7iXLculI0dwOn4c488/43TiBMbjxzGeOIHz36Y9m0qXJqNmTdJr1sx8X6MGGfffD66u\n1ssqIlJCuEVFAbmfHXv06FEOHjxIixYtuOeee2wRTURE7EiNrBRKLi4ujB49ms6dOzN16lR27dpF\n69atGTJkCM8995z1p4wZDJgDA0kLDCStefP/fT4jA6fff89sbE+cwOmvN5cvvsDliy+yyiyOjpiq\nVs1saqtXz2xwq1fHVKUKODpaN7uISHFhseC6ZQtmDw9SHn88x9Ibd2N15I6ISMmgRlYKtYoVK7Jq\n1Sp2797Niy++yJIlS/jwww+ZMWMG7dq1s/2OlE5OZNx/Pxn330/Kk09mfdpw7RpOP//8v7u3J07g\n9PPPuJ06BR99lFVncXUl4777bmtwzUFBuZ6LKCJS0hgPHsTpzBmSunUDd/ds62JiYti6dStVqlSh\nRYsWNkwoIiL2okZWioTWrVvTpEkT3nzzTRYvXszQoUNp3rw5M2fOLBRTyCze3qQ//DDpDz980yct\nOFy4gPGXX3D6+ef/NbonT2I8evSW680+PmRUr35bg2vx87PxKxERKTzyenbs+vXrSUlJYcCAATg4\naB9LEZGSQI2sFBnu7u5MmDCBkJAQpkyZwueff07Lli0ZOXIko0aNws3Nzd4Rb2UwYC5XjtRy5Ui9\n+Q6ByYTjn3/e1uAaDx3C+bvvbvkWprJlM9fe1qjxv/W3990HLi42fjEiIjaWlobbtm2YSpcm7dFH\nsy0zmUysWbMGNzc3unfvbsOAIiJiT2pkpcipVq0a69ev56OPPiI8PJyIiAiioqKYOXMmrVq1sne8\n3Dk6YqpWDVO1atCu3f8+n5KC06lTmQ3uX1OTjSdO4LpvH+zbl1VmcXQko1q1W3dPrlkTU4UKmp4s\nIsWGy+ef4xAbS+KQIeCU/Z8ru3fv5uzZs/Tp0wdfX18bJhQREXtSIytFksFgoGPHjrRo0YIFCxaw\nfPly+vbtS5s2bZgxYwYVKlSwd8S75+pKRp06ZNSpc8unDXFxOP3yC8bjx7OaW6eff8b466+4bduW\nVWf29Mxsbv+6g5teuzbpDzwAhe1OtYhIHmTtVhwSkmPdjU2eBg4caPVMIiJSeKiRlSLN09OT8PBw\nunfvzqRJk/jkk0/Yv38/YWFhDBkyBPccNgcpKiy+vqQ3akR6o0Y3fdKC49mzt9y5dTpxAuMPP+B8\n6ND/ypycyKhZk7R69Uh/6CHSH3qIjGrVQGvIRKQQMyQk4Prpp2Tccw/pDz6Ybd3Jkyf58ssveeSR\nR6hZs6YNE4qIiL2pkZVioWbNmmzZsoVNmzbx8ssvM3fuXJYvX84zzzxDv379ikVDewuDAVPFipgq\nViT15iMpUlMzpyf//DNOR47g/P33GH/6KXNzqTVrADB7e5Net+7/mtt69TCXKmWnFyIicjvXnTsx\npKRknh2bw5IJHbkjIlJyqZGVYsNgMNC9e3cef/xxli1bxvLly5kxYwZvv/02I0eOpG/fvoVvQ6iC\n5uJCRu3aZNSuDTem46WlYTxxAuP332fesf3+e1y+/BKXL7/MuiyjYsWsO7ZpDz105ynJGRkYkpIy\n3xITsz52uPnxze/T0qBaNVy9vDCVK4epXDnMgYG6GywiuXLbvBnIebfi69evs2nTJoKCgnjiiSds\nFU1ERAoJNbJS7Pj6+jJhwgSGDh3K0qVLWbFiBdOmTWPx4sWMGjWK3r17F/+G9mbOzqTXrUt63brw\n1xoyQ1wcxh9/zLxj+1dz67ZtW9aaW4uTE6ZKlSA1NbNRTUrCkJr6j57+5gOELEYjprJlM5vacuUw\nlS+f1eTe+Nji65v9HZjkZBzi4nCIjc18HxeH4cbHf703/PV5h7g4LE5OmMuUwVS2bOb7MmUwly2b\n9d7s76/GWqSQcbh8GeevviKtXj1MVatmWxcZGUliYiKjR4/GKYfNoEREpHjS//ml2PLz82PixIkM\nGzYsq6GdOnUqb731VlZD6+rqau+YdmHx9SWteXPSmjf/6xMWHE+fzmxqf/gB4w8/4Pjnn1jc3DCV\nLo3FwwM+EypiAAAgAElEQVSLu/v/3v56bM7m8xYPDyxOTpTKyCD+2DEcz53D8fz5rDfngwcxmM13\nzGZ2c8NcvjymoCDIyLi1aU1JyfNrNHt7Y0hLw3DkSPbj4OSEOTDwlkb3lo8rVMj8Q9rR8W6GV0Ty\nwfXDDzGYzZnTinOwbt06nJ2d6dWrl42SiYhIYaJGVoo9f39/Jk2axLBhw1iyZAmrVq3ipZdeyrpD\n26tXL1xK+rmsBgOmKlUwValCSg5T+e5aUBBJjzxy++fT03G8dAmHGw3u3xpdx3PncDp1CgCzry9m\nX1/Sa9TA7OuLxdcXs59f5uf9/DIf3/w5X18sPj6Zx3VYLBiuXct8rosXb31/6RKOFy/icOkSxqNH\nMXz//R1fgsXVNfP83lq1MneCrlWLjFq1sHh5Fdw4iUgWt6goLI6OpHTqlG3NhQsXOHHiBC1atKCU\n1viLiJRIamSlxAgICGDKlCmMGDGCxYsXs3r1aqZMmcKbb77JmDFjCA0NVUNrK0Zj5t3OChVIz64m\nJQWMxvzdDTUYsPj4kOHjA/ffn32d2YwhNhbHS5duaXYd//gD47FjGI8dw/nHH2+5JKNSpczm9q8G\nN6N2bUwVK/6zs3xTU3GIicl8u3r1f++vXYNatTAGBpJx331YPD3v/nuLFCGOp07h/N//ktKyZY6b\n0H3xxRcANGvWzFbRRESkkFEjKyVOQEAAL730Ek8//XRWQztp0iQWLVqU1dBKIWDLad8ODlgCAsgI\nCCCjVq3bv56Wlrkb9PHjOB07lvXe9ZNPcP3kk6wys5cXGTVrZjW3pqpVMVy7dmuTGht7a7MaE4ND\nQkKO8W78OW8KCiLj/vtJv/9+Mu67L+vN4u9fgINhYxYLpKZiSE7OXIt98/vkZDAaSXvkEa1lLiGy\nzo7NZVrx/v37ATWyIiIlmcFisVjsHQIypwkBBAUFZX0s/4zG8O5cuXKFxYsXs2bNGlJSUihfvjwv\nvfQSrVu3LlmbQllBsf5ZtFgypyXf3NweP47Tb79lu/73lsudnTEHBGD29898u9PHXl4EXL/O/7d3\n70FR1/sfx18Ly1UJ5I5XQPGuKYqJCipqaeUx0dRT1nHm1GhON39jjTkq0ziaWtMprXOOmU0N2Mio\n1LHpeE0FQVQUBLHT4eLdSMALaqJHYH9/mJvkDVngu6vPxwyz+/3u7nff38984ePLz+f7/V7at08u\nBQUyFxTI+TbtWe3vr6qbw23Hjqrq2FE1AQF1HyG2WKTq6utXp66puf68ulqma9dkunLleqj87dF0\n5codf3Tz8m9htNbPzWH1xvN7tFf5N9/Uvo/yfbpxHIaEhNR7G/jdzb/TDfo7brEoYMAAOZWWqjQv\nT5ZmzW77tpqaGvXs2VOurq7av3+/TPWZBWFHHui/k02ENrQdbWg72tB299tfMyKLh15AQIASEhL0\nyiuv6NNPP1ViYqKmTZsmb29vjR8/XpMnT1anTp2MLhP2xmRSTXCwrgYH62pc3O/rKyvl8t//ynzo\nkJxPnpTF27t2OP3t0dKsWd1CZkiILo4e/fvXXrwoc1GRzL8FW3NhocyFhXLNzJTbrl21Plrj7a2a\n5s1/D6Z/DKk1NVJV1fXnTfB/mhY3N1k8PK7/eHmpJijo+nNPz1qPNZ6e0m+PNf7+utarV6PXBuO5\nZGfLfOyYKuPj7xhiJSk/P19nz57VhAkTHD7EAgDqjyAL/CYwMFDvvvuupk+frjVr1ujzzz/XypUr\ntXLlSvXt21cvvPCCnn76aUZpcXceHrrWq1ejhS+Ll5f1nr+1XL4sc3GxNdi6FBTIuahIpspKWVxd\nJWdnWZydr59z7OR0fdls/v35jdduPL+x3tVVFnd36488PGot3/LzW1DVzes8Pa9/lluk4C6YVgwA\nuB/8qwL4g6CgIC1YsEBTp07Vli1blJSUpNTUVO3bt0/z5s1jlBb2ydNTVT16qKpHD6MrAe7ftWty\nX79e1X5+uhobe9e33rjQU0xMTFNUBgCwU1w9A7gDFxcXPfnkk/r666+VmZmp119/XW5ublq5cqWG\nDh2qP/3pT1qzZo0qKyuNLhUAHJrbzp1yPnPm+i137jJyf/nyZe3du1fdu3fntjsA8JAjyAJ10LZt\nW82aNUtZWVn6/PPPNWTIEO3fv19vvPGGIiMjNXfuXP33v/81ukwAcEge69ZJuve04szMTF27do1p\nxQAAgixwPxilBYCGZfr1V7lt3Kiq0FBdi4y863s5PxYAcANBFqinuozSFhQUGF0mAGjhwoWaOHGi\nJk2apLy8vFqv7d69WxMmTNCkSZP0zjvvqKYOt49qSG6bNsmpslKVY8fe80reqampcnd3V1RUVBNV\nBwCwVwRZwEZ3G6UdMmSInnnmGa1bt05XrlwxulQAD6G9e/fq2LFjSk5O1oIFC7RgwYJar8+bN09L\nly7V6tWr9euvv2rnzp1NWl9dr1Z86tQpFRYWasCAAXJzc2uK0gAAdowgCzSgm0dpV6xYodjYWO3d\nu1evvfaa+vTpo3fffVdFRUVGlwngIZKZmanhw4dLktq3b6+KigpdunTJ+npKSoqCg4MlSb6+vjp3\n7lyT1eZUXi631FT979FHVd2+/V3fe+NqxUwrBgBI3H4HaBQuLi566qmn9NRTT+nIkSNatWqVkpOT\ntXz5ci1fvlwDBw7U5MmTNWrUKLm6uhpdLoAHWHl5ubp162Zd9vX1VVlZmZo3by5J1sfS0lJlZGTo\njTfeuOv2/P395eLiYl0OCQmpf3Fr10rV1XKdMuWe28nKypIkjR8/3rbvtEMP2v4YgTa0HW1oO9rQ\ndvfThgRZoJGFhYVpzpw5euutt7Rx40YlJSUpIyNDGRkZ8vPz08SJEzV58mSFhoYaXSqAh4DFYrll\n3ZkzZzRt2jQlJCSoRYsWd/18eXm59XlISIhKSkrqXYvfl1/KxclJpUOHquYu26mpqdGmTZsUEhIi\nHx8fm77T3tjahqANGwJtaDva0HY32rCuYZapxUATcXNz05gxY7RmzRqlpaVp6tSpqqmp0d///ncN\nGDBAkyZN0vfff69r164ZXSqAB0hgYGCt8FlaWqqAgADr8qVLl/Tyyy/rzTff1KBBg5qsLucjR+Sa\nna3/xcSoJjDwru/Nz8/XuXPnFBsbK9M9LggFAHg4EGQBA3To0EEJCQnav3+/PvnkEz322GNKS0vT\nyy+/rKioKC1atEjHjh0zukwAD4CBAwdq06ZNkqRDhw4pMDDQOp1YkhYtWqS//OUvio2NbdK6PL75\nRtK9L/IkcdsdAMCtmFoMGMjd3V3x8fGKj49XQUGBEhMTtXbtWi1dulRLly5VVFSUxo0bp9GjR99z\nuh8A3E5kZKS6deumSZMmyWQyKSEhQSkpKfLy8tKgQYP07bff6tixY1q7dq0k6emnn9bEiRMbvS6P\nlBRZ3N11ZdSoe743NTVVJpNJMTExjV4XAMAxEGQBO9GxY0fNnz9f77zzjr7//nutWbNGGRkZysrK\n0ty5czVs2DCNHz9ew4YN49YTAO7LzJkzay137tzZ+jw/P7+py7nu6lVdnjBBlptGh2/n119/VVZW\nlnr06CE/P78mKg4AYO8IsoCd8fT01LPPPqtnn31WJSUl+uabb5SSkqKNGzdq48aN8vb21ujRozVu\n3DhFRUXJyYkzBAA4nrLMTMnZ+Z7vy8zM1LVr15hWDACohX8BA3YsJCRE06dP19atW7V161a98sor\n8vDwUFJSksaOHavo6GgtXrxYhYWFRpcKAPfHbJbqcOEmzo8FANwOQRZwEF27dtXcuXOVlZWl5ORk\nTZgwQWfPntXHH3+swYMHa9SoUfr8889VVlZmdKkA0GBSU1Pl6empPn36GF0KAMCOEGQBB+Ps7KyY\nmBh99NFHys3N1aeffqphw4YpPz9f8+bNU2RkpCZPnqx169bp/PnzRpcLAPV28uRJFRUVacCAAVwb\nAABQC+fIAg7M09NTY8eO1dixY1VWVqZ//etfWrdunbZt26Zt27bJbDarf//+GjlypJ544gm1atXK\n6JIBoM7S0tIkqclvDQQAsH8EWeABERAQoJdeekkvvfSSCgsLtWHDBm3cuFHp6elKT0/XnDlz1LNn\nT40cOVIjR45Up06dZKrD+WkAYBTOjwUA3AlBFngARUREKCIiQq+//rp+/vlnbd68WRs3btSuXbuU\nl5enJUuWKDQ0VE888YRGjhypvn37yrkOVw8FgKZSXV2t9PR0tWzZUh06dDC6HACAnSHIAg+4li1b\nasqUKZoyZYoqKir0ww8/aNOmTdq2bZuWL1+u5cuXy8/PTyNGjNDIkSMVExMjDw8Po8sG8JA7ePCg\nzp07p5EjRzJ7BABwC4Is8BDx9vZWfHy84uPjdfXqVWVkZGjjxo3atGmTVq9erdWrV8vDw0NDhw7V\n448/rtjYWAUHBxtdNoCHENOKAQB306hXLS4oKNDw4cOVlJTUmF8DoB7c3NwUFxenJUuWKCcnR+vX\nr9f06dMVEhKif//733rzzTcVGRmp2NhYzZ49Wxs2bFBFRYXRZQN4SKSmpspkMmnQoEFGlwIAsEON\nNiJ7+fJlzZ8/X9HR0Y31FQAaiJOTk/r27au+fftqzpw5Kiws1A8//KD09HRlZmbqyy+/1Jdffikn\nJyf17NlTMTExGjRokKKiouTu7m50+QAeMJcuXdK+ffv06KOPytfX1+hyAAB2qNGCrKurq1asWKEV\nK1Y01lcAaCQ3LhY1bdo0/e9//1NOTo527typ9PR0ZWdn68CBA1q2bJnc3NwUFRVlDbY9e/bkolEA\nbLZr1y5VVVUxrRgAcEeNFmTNZrPM5rpv3t/fXy4uLpKkkJCQxirroUEb2o42/F27du30zDPPSJIu\nXryonTt36ocfftDWrVutt/eRJB8fHw0ZMkTDhg3TsGHDFBwcTDs2ANrQdrShY+H8WADAvdjNxZ7K\ny8slXf/HRklJicHVODba0Ha04d317t1bvXv31syZM1VeXq6MjAylp6dr586d+vbbb/Xtt99Kklq0\naKFevXqpd+/eioyMVK9evZgmeJ84Fm13ow0Js44jNTVVzZo1U2RkpNGlAADslN0EWQCOyd/fX2PG\njNGYMWMkScePH9fOnTut96zdvn27tm/fbn1/WFiYIiMjreG2a9eucnV1Nap8AHbmxIkTOnz4sEaM\nGMHfBgDAHRFkATSotm3b6vnnn9fzzz+vkJAQ5efnKycnR9nZ2crJydGBAwe0bt06rVu3TtL1qyd3\n7969Vrht06YN940EHlJpaWmSmFYMALi7Rguy+fn5Wrx4sU6dOiWz2axNmzZp2bJl8vHxaayvBGCH\n/Pz8NHz4cA0fPlySVFNTo+LiYuXk5FgDbm5urvbv32/9jL+/v3r37q0BAwYoLi5OHTp0INgCDwnO\njwUA1EWjBdnu3bsrMTGxsTYPwEE5OTlZr4o8YcIESVJlZaUOHjxoHbXNzs7Wli1btGXLFr377rtq\n3bq1hg4dqqFDh2rQoEFq3ry5wXsBoDFUV1crPT1drVu3Vnh4uNHlAADsGFOLARjOw8ND/fr1U79+\n/azrfvnlF6Wmpmr79u1KS0tTYmKiEhMT5eLioqioKGuw7dKlC6O1wAMiLy9P58+f11NPPcXvNQDg\nrgiyAOxScHCwJk6cqIkTJ6qqqko5OTnWC0ft2rVLu3bt0oIFCxQcHKwhQ4YoLi5OMTEx8vb2Nrp0\nAPXEtGIAQF0RZAHYPbPZrKioKEVFRentt99WeXm5dbR2x44dWr16tVavXi1nZ2f16dPHOlrbvXt3\nOTk5GV0+gDpKTU2Vk5OTBg4caHQpAAA7R5AF4HD8/f01btw4jRs3TjU1Ndbb/Gzbtk379u3T3r17\ntXjxYjVr1kydO3dW586d1aVLF3Xp0kWdO3dWixYtjN4FAH9w8eJF7d+/X7169eJ3FABwTwRZAA7N\nyclJvXr1Uq9evTRjxgydO3dOaWlp2r59uw4cOKADBw7UuiKydH3a8o2A27VrV3Xu3FkRERFyc3Mz\naC8A7Nq1S1VVVYqNjTW6FACAAyDIAnigtGjRQmPGjNGYMWMkSVevXlVRUZF++ukn/ec//7E+7tix\nQzt27LB+ztnZWeHh4bVGbzt16qQ2bdrI2dnZoL0BHh6cHwsAuB8EWQAPNDc3N3Xr1k3dunWrtf78\n+fO3hNuffvpJhYWF+u6776zvc3V1Vbt27dS+fXuFh4fXevTz8+PKqkADSUtLU7NmzRQZGWl0KQAA\nB0CQBfBQ8vHxUf/+/dW/f3/rOovFolOnTllDbUFBgQ4fPqzi4mIVFhbesg1vb+9a4fbG87CwMHl6\nejbl7gAO7fjx4zp8+LAef/xxubi4GF0OAMABEGQB4Dcmk0mtW7dW69atNWLECOt6i8WiM2fOqLi4\nWMXFxdZwe/jwYeXn5ysnJ+eWbbVs2VIdOnRQVFSUoqOjFRkZKXd396bcHcBhpKWlSWJaMQCg7giy\nAHAPJpNJ/v7+8vf312OPPVbrtaqqKp08edIabG+E3OLiYqWlpVn/ge7m5qbevXurf//+io6OVp8+\nfRi1BX7D+bEAgPtFkAUAG5jNZoWGhio0NFTDhg2r9drZs2e1Z88e7d69W5mZmdbnH330kVxcXPTo\no48qOjpa0dHRioqKUrNmzQzaC8A4VVVVSk9PV5s2bRQWFmZ0OQAAB0GQBYBG4uvrq1GjRmnUqFGS\npIqKCu3du1eZmZnavXu3srOztW/fPi1btkzOzs7q2bOndcS2X79+euSRRwzeA6Dx5ebmqqKiQqNH\nj+biaQCAOiPIAkAT8fb21ogRI6zn3168eFFZWVnWYJubm6ucnBz94x//kJOTk7p27aouXbrIy8tL\nQUFBCgwMVFBQkAICAhQUFCRfX185OTkZvFdwBAsXLlRubq5MJpNmz56tnj17Wl+7evWq5s2bp8LC\nQqWkpDR5bUwrBgDUB0EWAAzi5eWluLg4xcXFSZIuX76srKws61TkAwcOKD8//46fN5vNCggIUEBA\ngDXkBgYGWn+CgoLUpk0b+fv7M9L1ENu7d6+OHTum5ORkFRcXa/bs2UpOTra+vmTJEnXp0uW2V+Zu\nCqmpqXJyctLAgQMN+X4AgGMiyAKAnfD09NTgwYOtI1PV1dUym83Ky8tTaWmpTp8+rbKyMp0+fbrW\nckFBgfLy8u643ebNmyssLExhYWEKDw+3Pg8LC5Ovry8h9wGXmZmp4cOHS5Lat2+viooKXbp0Sc2b\nN5ckzZgxQ+fPn9f69eubvLYLFy4oOztbvXv3lo+PT5N/PwDAcRFkAcBOOTs7Kzg4WBaL5a7vs1gs\nunDhgkpLS60B98bj8ePHdeTIERUWFurgwYO3fNbb27tWsL056BIsHgzl5eXq1q2bddnX11dlZWXW\nINu8eXOdP3/ekNoyMjJUXV2t2NhYQ74fAOC4CLIA4OBMJpO8vb3l7e2tiIiI276npqZGJSUlOnLk\niI4cOaLDhw9bH3/88UcdOHDgls/4+vpaQ21oaGit54Rcx3Wv/xi5F39/f7m4uFiXQ0JC6r2tffv2\nSZLGjRtn03Yc3cO87w2FNrQdbWg72tB299OGBFkAeAg4OTmpVatWatWqlQYNGlTrterqap06dcoa\nbm/+yc3N1f79+2/ZXosWLW4bcENDQ+Xr69tUu4U6CAwMVHl5uXW5tLRUAQEB9d7ezdsKCQlRSUlJ\nvbe1YcMGeXl5qU2bNjZtx5HZ2oagDRsCbWg72tB2N9qwrmGWIAsADzlnZ2e1bdtWbdu21ZAhQ2q9\nVlVVpVOnTtUKt0ePHtWRI0d08OBBZWdn37I9Hx8fa6gNDw9XRESEIiIiFB4eLnd39ybaK9wwcOBA\nLVu2TJMmTdKhQ4cUGBhonVZspKNHj+ro0aMaOXJkrRFeAADqgiALALgjs9msdu3aqV27drcNuT//\n/PMtAffIkSO3na7s5OSkdu3aWYNtRESEOnbsqA4dOthFsHpQRUZGqlu3bpo0aZJMJpMSEhKUkpIi\nLy8vjRgxQq+//rp++eUXHTlyRC+88IImTJig0aNHN3pdaWlpkrjtDgCgfgiyAIB6MZvN1pHcP4aR\n6upqlZSUqKioSEVFRSooKFBhYaEKCgq0efNmbd68udb7W7ZsaQ22Nwddpik3jJkzZ9Za7ty5s/X5\n0qVLm7ocSdw/FgBgG4IsAKDBOTs7q3Xr1mrduvUtI7lnzpypFWwLCwtVWFio1NRUa7i5oVWrVkpO\nTlZ4eHgTVo/GZrFYlJGRoXbt2ik0NNTocgAADoggCwBoUn5+foqOjlZ0dHSt9RcuXLCG2oKCAhUV\nFencuXOcP/kAMplMGjx4sGJiYowuBQDgoAiyAAC78Mgjj6hPnz7q06eP0aWgCSxfvtzoEgAADszJ\n6AIAAAAAALgfBFkAAAAAgEMhyAIAAAAAHApBFgAAAADgUAiyAAAAAACHQpAFAAAAADgUgiwAAAAA\nwKEQZAEAAAAADoUgCwAAAABwKARZAAAAAIBDIcgCAAAAABwKQRYAAAAA4FAIsgAAAAAAh2KyWCwW\no4sAAAAAAKCuGJEFAAAAADgUgiwAAAAAwKEQZAEAAAAADoUgCwAAAABwKARZAAAAAIBDIcgCAAAA\nAByK2egCbrZw4ULl5ubKZDJp9uzZ6tmzp9ElOZQ9e/bojTfeUEREhCSpY8eOmjt3rsFVOY6CggJN\nnz5dU6ZM0eTJk1VSUqK3335b1dXVCggI0Pvvvy9XV1ejy7Rrf2zDWbNm6dChQ/Lx8ZEk/fWvf9WQ\nIUOMLdLOLVmyRPv371dVVZWmTp2qHj16cBzepz+24bZt2zgOGxj9tW3or21Df207+mvb0V/bztb+\n2m6C7N69e3Xs2DElJyeruLhYs2fPVnJystFlOZx+/fpp6dKlRpfhcC5fvqz58+crOjraum7p0qV6\n7rnnNGrUKH344Ydau3atnnvuOQOrtG+3a0NJ+r//+z8NHTrUoKocy+7du1VYWKjk5GSdO3dOY8eO\nVXR0NMfhfbhdG/bv35/jsAHRXzcM+uv6ob+2Hf217eivbdcQ/bXdTC3OzMzU8OHDJUnt27dXRUWF\nLl26ZHBVeFi4urpqxYoVCgwMtK7bs2ePhg0bJkkaOnSoMjMzjSrPIdyuDXF/oqKi9PHHH0uSHnnk\nEVVWVnIc3qfbtWF1dbXBVT1Y6K9hJPpr29Ff247+2nYN0V/bTZAtLy9XixYtrMu+vr4qKyszsCLH\nVFRUpGnTpunPf/6zMjIyjC7HYZjNZrm7u9daV1lZaZ0S4ufnx/F4D7drQ0lKSkrSiy++qBkzZujs\n2bMGVOY4nJ2d5enpKUlau3atYmNjOQ7v0+3a0NnZmeOwAdFfNwz66/qhv7Yd/bXt6K9t1xD9td1M\nLf4ji8VidAkOJzQ0VK+++qpGjRqlEydO6MUXX9TmzZuZn98AOB7rZ8yYMfLx8VGXLl302Wef6ZNP\nPtG8efOMLsvubd26VWvXrtUXX3yhxx9/3Lqe47Dubm7D/Px8jsNGxHF5/+ivGw/HY/3QX9cP/bXt\nbOmv7WZENjAwUOXl5dbl0tJSBQQEGFiR4wkKCtKTTz4pk8mktm3byt/fX6dPnza6LIfl6empK1eu\nSJJOnz7NFJx6iI6OVpcuXSRJcXFxKigoMLgi+7dz507985//1IoVK+Tl5cVxWA9/bEOOw4ZFf207\n+uuGxd9J2/F38v7RX9vO1v7aboLswIEDtWnTJknSoUOHFBgYqObNmxtclWNZv369Vq5cKUkqKyvT\nmTNnFBQUZHBVjmvAgAHWY3Lz5s2KiYkxuCLH89prr+nEiROSrp/DdOMKnbi9ixcvasmSJVq+fLn1\nin0ch/fndm3Icdiw6K9tR3/dsPg7aTv+Tt4f+mvbNUR/bbLY0dj3Bx98oH379slkMikhIUGdO3c2\nuiSHcunSJc2cOVMXLlzQtWvX9Oqrr2rw4MFGl+UQ8vPztXjxYp06dUpms1lBQUH64IMPNGvWLF29\nelUtW7bUe++9JxcXF6NLtVu3a8PJkyfrs88+k4eHhzw9PfXee+/Jz8/P6FLtVnJyspYtW6awsDDr\nukWLFmnOnDkch3V0uzaMj49XUlISx2EDor+2Df11/dFf247+2nb017ZriP7aroIsAAAAAAD3YjdT\niwEAAAAAqAuCLAAAAADAoRBkAQAAAAAOhSALAAAAAHAoBFkAAAAAgEMhyAIPmJSUFM2cOdPoMgAA\nwF3QXwO2IcgCAAAAAByK2egCgIdVYmKiNmzYoOrqaoWHh+ull17S1KlTFRsbq59++kmS9Le//U1B\nQUHasWOHPv30U7m7u8vDw0Pz589XUFCQcnNztXDhQrm4uMjb21uLFy+WJF26dEkzZ85UcXGxWrZs\nqU8++USlpaXW//m9cuWKJk6cqPHjxxu2/wAAOAL6a8A+MSILGCAvL09btmzRqlWrlJycLC8vL+3a\ntUsnTpxQfHy8vv76a/Xr109ffPGFKisrNWfOHC1btkyJiYmKjY3VRx99JEl66623NH/+fCUlJSkq\nKkqpqamSpKKiIs2fP18pKSkqLCzUoUOHtGHDBoWHhysxMVFJSUm6cuWKkU0AAIDdo78G7BcjsoAB\n9uzZo+PHj+vFF1+UJF2+fFmnT5+Wj4+PunfvLkmKjIzUV199paNHj8rPz0/BwcGSpH79+mn16tU6\ne/asLly4oI4dO0qSpkyZIun6OTc9evSQh4eHJCkoKEgXL15UTEyMvv76a82aNUuDBw/WxIkTm3iv\nAQBwLPTXgP0iyAIGcHV1VVxcnObNm2ddd/LkScXHx1uXLRaLTCaTTCZTrc/evN5isdx2+87Ozrd8\npn379vr++++VlZWljRs36quvvtLq1asbcK8AAHiw0F8D9oupxYABIiMjlZaWpl9//VWStGrVKpWV\nlamiokI//vijJCk7O1udOnVSaGiozpw5o59//lmSlJmZqUcffVQtWrSQj4+P8vLyJElffPGFVq1a\ndd10IKYAAAEFSURBVMfv/O6773Tw4EENGDBACQkJKikpUVVVVSPvKQAAjov+GrBfjMgCBujRo4ee\nf/55vfDCC3Jzc1NgYKAee+wxBQUFKSUlRYsWLZLFYtGHH34od3d3LViwQDNmzJCrq6s8PT21YMEC\nSdL777+vhQsXymw2y8vLS++//742b9582+/s0KGDEhIS5OrqKovFopdffllmM38CAAC4E/prwH6Z\nLHea6wCgSZ08eVLPPfec0tLSjC4FAADcAf01YB+YWgwAAAAAcCiMyAIAAAAAHAojsgAAAAAAh0KQ\nBQAAAAA4FIIsAAAAAMChEGQBAAAAAA6FIAsAAAAAcCgEWQAAAACAQ/l/J2HAeFvO+roAAAAASUVO\nRK5CYII=\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Look at the history of the model\n",
    "hist_dense = model_cnn_selu.history.history # contains information from fitting model\n",
    "hist_dense.keys() # returns: dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n",
    "\n",
    "# Call each key to get the values to plot\n",
    "loss_cnn_selu = hist_dense['loss'] \n",
    "val_loss_cnn_selu = hist_dense['val_loss']\n",
    "acc_cnn_selu = hist_dense['acc']\n",
    "val_acc_cnn_selu = hist_dense['val_acc']\n",
    "\n",
    "epochs = range(1, len(loss_cnn_selu) + 1)\n",
    "\n",
    "# Plot Losses of training and validation sets v. epochs\n",
    "plt.figure(figsize=(25,6))\n",
    "plt.subplot(131)\n",
    "plt.plot(epochs, loss_cnn_selu, 'k', label = 'Training loss')\n",
    "plt.plot(epochs, val_loss_cnn_selu, 'r', label = 'Validation loss')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation loss over epochs')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot accuracy scores over epochs for training and validation sets                      \n",
    "plt.subplot(132)\n",
    "plt.plot(epochs, acc_cnn_selu, 'k', label='Training acc')\n",
    "plt.plot(epochs, val_acc_cnn_selu, 'r', label= 'Validation acc')\n",
    "\n",
    "# Add the good stuff                      \n",
    "plt.title('Training and validation accuracy over epochs [SeLU CNN]')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "uhRiOEYXPi-8"
   },
   "source": [
    "### Basic Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "SxSMW6z6AkHm"
   },
   "outputs": [],
   "source": [
    "# Use model to predict on testing set\n",
    "y_pred = model_cnn_selu.predict(x_test_)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lhJ8EY3aAnJt"
   },
   "outputs": [],
   "source": [
    "# Convert y_pred from continuous variables to discreet binary (0,1)\n",
    "yy = []\n",
    "for elem in y_pred: # each element is an array\n",
    "  row = []\n",
    "  for e in elem: # each number in the array\n",
    "    if e == elem.max(): # since softmax outputs probability...\n",
    "      row.append(1) # assign highest prob as 1\n",
    "    else:\n",
    "      row.append(0) # assign all other probs as 0\n",
    "  yy.append(row)\n",
    "y_pred_ = np.array(yy)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1088,
     "status": "ok",
     "timestamp": 1549504652950,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "da92kxivAsCd",
    "outputId": "2fba2c62-c550-443e-cc20-b1dd2f456002"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.732826794851915"
      ]
     },
     "execution_count": 91,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_test_, y_pred_, average= 'weighted') # 0.7299842196935593\n",
    "precision_score(y_test_, y_pred_, average= 'weighted') # 0.7503546303263359\n",
    "accuracy_score(y_test_, y_pred_) # 0.732826794851915\n",
    "recall_score(y_test_, y_pred_, average= 'weighted') # 0.732826794851915"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "a08P_-J5j6XK"
   },
   "source": [
    "### AUC/ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "p2DGNnFVj8ft"
   },
   "outputs": [],
   "source": [
    "# Iterate through y_test_ and y_pred_ to get auc value for each of the 537 IDs\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(y_test_.shape[1]):\n",
    "  # use y_pred not y_pred_, because the roc_curve function needs probability for threshold metric\n",
    "  fpr[i], tpr[i], thresholds = roc_curve(y_test_[:,i], y_pred[:,i]) # fpr = false positive rate, tpr = true positive rate\n",
    "  roc_auc[i] = auc(fpr[i], tpr[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 512
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1174,
     "status": "ok",
     "timestamp": 1549504680185,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "rOX_WD3oj_C3",
    "outputId": "17b4e81c-d3bd-4f23-d223-c8587f6f9090"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm4AAAHvCAYAAADtrFPzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nXG8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot some of 537 celebs ROC curve \n",
    "#Seaborns Beautiful Styling\n",
    "sns.set_style(\"darkgrid\", {\"axes.facecolor\": \".9\"})\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "lw = 2 # linewidth\n",
    "for i in range(50):\n",
    "  plt.plot(fpr[i], tpr[i], lw = lw)\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "  plt.yticks([ii/20.0 for ii in range(21)])\n",
    "  plt.xticks([ii/20.0 for ii in range(21)])\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for SeLU CNN Model')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "J1QEPAR7avAU"
   },
   "outputs": [],
   "source": [
    "# # Confusion matrix for multiple labels:\n",
    "# def plot_confusion_matrix(cm, classes, normalize=False, title='Confusion matrix',cmap = plt.cm.Blues):\n",
    "    \n",
    "#     #Add Normalization Option\n",
    "#     '''prints pretty confusion metric with normalization option '''\n",
    "#     if normalize:\n",
    "#         cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "#         print(\"Normalized confusion matrix\")\n",
    "#     else:\n",
    "#         print('Confusion matrix, without normalization')\n",
    "    \n",
    "#     # Plot\n",
    "#     plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "#     plt.title(title)\n",
    "#     plt.colorbar()\n",
    "#     tick_marks = np.arange(len(classes))\n",
    "#     plt.xticks(tick_marks, classes, rotation=45)\n",
    "#     plt.yticks(tick_marks, classes)\n",
    "    \n",
    "#     fmt = '.2f' if normalize else 'd'\n",
    "#     thresh = cm.max() / 2.\n",
    "#     for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "#         plt.text(j, i, format(cm[i, j], fmt), horizontalalignment=\"center\", color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "    \n",
    "#     plt.tight_layout()\n",
    "#     plt.ylabel('True label')\n",
    "#     plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "mrQVX0bZcHfL",
    "outputId": "75a4fe04-0f27-4567-a920-b31efe004732"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized confusion matrix\n"
     ]
    }
   ],
   "source": [
    "# # Get Confusion matrix\n",
    "# cm_ = confusion_matrix(y_test_.argmax(axis = 1), y_pred_.argmax(axis = 1))\n",
    "# classes_ = df.id\n",
    "# plot_confusion_matrix(cm_, classes = classes_, normalize=True, \\\n",
    "#                       title='Non-normalized confusion matrix')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "oRwEJ3jkZWa5"
   },
   "source": [
    "# Sum Up"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "hRYaE44Vs5sF"
   },
   "source": [
    "## EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "b3YB3zVflDdo"
   },
   "outputs": [],
   "source": [
    "# Convert meta_df id column into integers like df columns\n",
    "meta_df.id = meta_df.id.apply(lambda x: int(x[2:]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 872,
     "status": "ok",
     "timestamp": 1549505650530,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "vGuulpp1mC_f",
    "outputId": "0e0ad85d-31a9-48a2-f7c4-f82041b04bad"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>name</th>\n",
       "      <th>gender</th>\n",
       "      <th>nationality</th>\n",
       "      <th>set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10001</td>\n",
       "      <td>A.J._Buckley</td>\n",
       "      <td>m</td>\n",
       "      <td>Ireland</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10002</td>\n",
       "      <td>A.R._Rahman</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10003</td>\n",
       "      <td>Aamir_Khan</td>\n",
       "      <td>m</td>\n",
       "      <td>India</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10004</td>\n",
       "      <td>Aaron_Tveit</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10005</td>\n",
       "      <td>Aaron_Yoo</td>\n",
       "      <td>m</td>\n",
       "      <td>USA</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id          name gender nationality  set\n",
       "0  10001  A.J._Buckley      m     Ireland  dev\n",
       "1  10002   A.R._Rahman      m       India  dev\n",
       "2  10003    Aamir_Khan      m       India  dev\n",
       "3  10004   Aaron_Tveit      m         USA  dev\n",
       "4  10005     Aaron_Yoo      m         USA  dev"
      ]
     },
     "execution_count": 27,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meta_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1053,
     "status": "ok",
     "timestamp": 1549505769525,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "D9E0VSP8vD_h",
    "outputId": "54e92f03-1282-4083-cf6c-1dd1d706521d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64483, 198)"
      ]
     },
     "execution_count": 33,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape # (64483, 194)\n",
    "full_df = df.merge(meta_df, how = 'left', on = 'id')\n",
    "full_df.shape # (64483, 198)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dWfWUaX7svXM"
   },
   "outputs": [],
   "source": [
    "# Save merged dataframes to pickle file\n",
    "# full_df.to_pickle(\"full_df.pkl\")\n",
    "\n",
    "# Read in pickled file\n",
    "full_df = pd.read_pickle('full_df.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 253
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 635,
     "status": "ok",
     "timestamp": 1549508860629,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "GC1Am7UkvvXo",
    "outputId": "6032614b-dbb9-48a3-c81e-096d25a895f9"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>...</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>name</th>\n",
       "      <th>gender</th>\n",
       "      <th>nationality</th>\n",
       "      <th>set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10332</td>\n",
       "      <td>-274.717176</td>\n",
       "      <td>124.096613</td>\n",
       "      <td>-57.174884</td>\n",
       "      <td>47.091800</td>\n",
       "      <td>-31.190586</td>\n",
       "      <td>-3.631755</td>\n",
       "      <td>-34.891319</td>\n",
       "      <td>-4.590279</td>\n",
       "      <td>-0.825808</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004705</td>\n",
       "      <td>0.003553</td>\n",
       "      <td>-0.009748</td>\n",
       "      <td>0.026007</td>\n",
       "      <td>0.001958</td>\n",
       "      <td>-0.009467</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>f</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10332</td>\n",
       "      <td>-263.282737</td>\n",
       "      <td>110.169695</td>\n",
       "      <td>-37.532975</td>\n",
       "      <td>49.808548</td>\n",
       "      <td>-30.783113</td>\n",
       "      <td>4.530054</td>\n",
       "      <td>-34.272742</td>\n",
       "      <td>-0.396505</td>\n",
       "      <td>2.088190</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008559</td>\n",
       "      <td>0.001090</td>\n",
       "      <td>-0.052647</td>\n",
       "      <td>0.053715</td>\n",
       "      <td>-0.014433</td>\n",
       "      <td>-0.001219</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>f</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10332</td>\n",
       "      <td>-255.330134</td>\n",
       "      <td>107.140581</td>\n",
       "      <td>-37.526152</td>\n",
       "      <td>66.718413</td>\n",
       "      <td>-25.090227</td>\n",
       "      <td>5.835736</td>\n",
       "      <td>-34.682863</td>\n",
       "      <td>0.156228</td>\n",
       "      <td>0.196809</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000381</td>\n",
       "      <td>0.004534</td>\n",
       "      <td>0.003975</td>\n",
       "      <td>0.063643</td>\n",
       "      <td>-0.003111</td>\n",
       "      <td>-0.009005</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>f</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10332</td>\n",
       "      <td>-238.647293</td>\n",
       "      <td>116.777418</td>\n",
       "      <td>-43.337226</td>\n",
       "      <td>57.847068</td>\n",
       "      <td>-29.790134</td>\n",
       "      <td>4.556106</td>\n",
       "      <td>-29.883005</td>\n",
       "      <td>-1.273415</td>\n",
       "      <td>0.166903</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.015756</td>\n",
       "      <td>-0.000640</td>\n",
       "      <td>0.002584</td>\n",
       "      <td>0.037222</td>\n",
       "      <td>-0.006243</td>\n",
       "      <td>0.002152</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>f</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10332</td>\n",
       "      <td>-253.325462</td>\n",
       "      <td>110.840374</td>\n",
       "      <td>-47.403615</td>\n",
       "      <td>46.822241</td>\n",
       "      <td>-28.057003</td>\n",
       "      <td>5.674004</td>\n",
       "      <td>-26.317193</td>\n",
       "      <td>-0.719588</td>\n",
       "      <td>1.632264</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.006521</td>\n",
       "      <td>0.009226</td>\n",
       "      <td>0.008907</td>\n",
       "      <td>0.020127</td>\n",
       "      <td>0.001455</td>\n",
       "      <td>0.002741</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>f</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 198 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id        mfcc        mfcc       mfcc       mfcc       mfcc      mfcc  \\\n",
       "0  10332 -274.717176  124.096613 -57.174884  47.091800 -31.190586 -3.631755   \n",
       "1  10332 -263.282737  110.169695 -37.532975  49.808548 -30.783113  4.530054   \n",
       "2  10332 -255.330134  107.140581 -37.526152  66.718413 -25.090227  5.835736   \n",
       "3  10332 -238.647293  116.777418 -43.337226  57.847068 -29.790134  4.556106   \n",
       "4  10332 -253.325462  110.840374 -47.403615  46.822241 -28.057003  5.674004   \n",
       "\n",
       "        mfcc      mfcc      mfcc ...    tonnetz   tonnetz   tonnetz   tonnetz  \\\n",
       "0 -34.891319 -4.590279 -0.825808 ...   0.004705  0.003553 -0.009748  0.026007   \n",
       "1 -34.272742 -0.396505  2.088190 ...  -0.008559  0.001090 -0.052647  0.053715   \n",
       "2 -34.682863  0.156228  0.196809 ...  -0.000381  0.004534  0.003975  0.063643   \n",
       "3 -29.883005 -1.273415  0.166903 ...  -0.015756 -0.000640  0.002584  0.037222   \n",
       "4 -26.317193 -0.719588  1.632264 ...  -0.006521  0.009226  0.008907  0.020127   \n",
       "\n",
       "    tonnetz   tonnetz            name  gender  nationality  set  \n",
       "0  0.001958 -0.009467  Gemma_Atkinson       f           UK  dev  \n",
       "1 -0.014433 -0.001219  Gemma_Atkinson       f           UK  dev  \n",
       "2 -0.003111 -0.009005  Gemma_Atkinson       f           UK  dev  \n",
       "3 -0.006243  0.002152  Gemma_Atkinson       f           UK  dev  \n",
       "4  0.001455  0.002741  Gemma_Atkinson       f           UK  dev  \n",
       "\n",
       "[5 rows x 198 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 253
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 290,
     "status": "ok",
     "timestamp": 1549508864103,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "5n3A0f_nwQ8f",
    "outputId": "484a04db-1bde-44a0-88ff-3940f406f5b5"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>...</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>name</th>\n",
       "      <th>gender</th>\n",
       "      <th>nationality</th>\n",
       "      <th>set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10332</td>\n",
       "      <td>-274.717176</td>\n",
       "      <td>124.096613</td>\n",
       "      <td>-57.174884</td>\n",
       "      <td>47.091800</td>\n",
       "      <td>-31.190586</td>\n",
       "      <td>-3.631755</td>\n",
       "      <td>-34.891319</td>\n",
       "      <td>-4.590279</td>\n",
       "      <td>-0.825808</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004705</td>\n",
       "      <td>0.003553</td>\n",
       "      <td>-0.009748</td>\n",
       "      <td>0.026007</td>\n",
       "      <td>0.001958</td>\n",
       "      <td>-0.009467</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10332</td>\n",
       "      <td>-263.282737</td>\n",
       "      <td>110.169695</td>\n",
       "      <td>-37.532975</td>\n",
       "      <td>49.808548</td>\n",
       "      <td>-30.783113</td>\n",
       "      <td>4.530054</td>\n",
       "      <td>-34.272742</td>\n",
       "      <td>-0.396505</td>\n",
       "      <td>2.088190</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008559</td>\n",
       "      <td>0.001090</td>\n",
       "      <td>-0.052647</td>\n",
       "      <td>0.053715</td>\n",
       "      <td>-0.014433</td>\n",
       "      <td>-0.001219</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10332</td>\n",
       "      <td>-255.330134</td>\n",
       "      <td>107.140581</td>\n",
       "      <td>-37.526152</td>\n",
       "      <td>66.718413</td>\n",
       "      <td>-25.090227</td>\n",
       "      <td>5.835736</td>\n",
       "      <td>-34.682863</td>\n",
       "      <td>0.156228</td>\n",
       "      <td>0.196809</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000381</td>\n",
       "      <td>0.004534</td>\n",
       "      <td>0.003975</td>\n",
       "      <td>0.063643</td>\n",
       "      <td>-0.003111</td>\n",
       "      <td>-0.009005</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10332</td>\n",
       "      <td>-238.647293</td>\n",
       "      <td>116.777418</td>\n",
       "      <td>-43.337226</td>\n",
       "      <td>57.847068</td>\n",
       "      <td>-29.790134</td>\n",
       "      <td>4.556106</td>\n",
       "      <td>-29.883005</td>\n",
       "      <td>-1.273415</td>\n",
       "      <td>0.166903</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.015756</td>\n",
       "      <td>-0.000640</td>\n",
       "      <td>0.002584</td>\n",
       "      <td>0.037222</td>\n",
       "      <td>-0.006243</td>\n",
       "      <td>0.002152</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10332</td>\n",
       "      <td>-253.325462</td>\n",
       "      <td>110.840374</td>\n",
       "      <td>-47.403615</td>\n",
       "      <td>46.822241</td>\n",
       "      <td>-28.057003</td>\n",
       "      <td>5.674004</td>\n",
       "      <td>-26.317193</td>\n",
       "      <td>-0.719588</td>\n",
       "      <td>1.632264</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.006521</td>\n",
       "      <td>0.009226</td>\n",
       "      <td>0.008907</td>\n",
       "      <td>0.020127</td>\n",
       "      <td>0.001455</td>\n",
       "      <td>0.002741</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>UK</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 198 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id        mfcc        mfcc       mfcc       mfcc       mfcc      mfcc  \\\n",
       "0  10332 -274.717176  124.096613 -57.174884  47.091800 -31.190586 -3.631755   \n",
       "1  10332 -263.282737  110.169695 -37.532975  49.808548 -30.783113  4.530054   \n",
       "2  10332 -255.330134  107.140581 -37.526152  66.718413 -25.090227  5.835736   \n",
       "3  10332 -238.647293  116.777418 -43.337226  57.847068 -29.790134  4.556106   \n",
       "4  10332 -253.325462  110.840374 -47.403615  46.822241 -28.057003  5.674004   \n",
       "\n",
       "        mfcc      mfcc      mfcc ...    tonnetz   tonnetz   tonnetz   tonnetz  \\\n",
       "0 -34.891319 -4.590279 -0.825808 ...   0.004705  0.003553 -0.009748  0.026007   \n",
       "1 -34.272742 -0.396505  2.088190 ...  -0.008559  0.001090 -0.052647  0.053715   \n",
       "2 -34.682863  0.156228  0.196809 ...  -0.000381  0.004534  0.003975  0.063643   \n",
       "3 -29.883005 -1.273415  0.166903 ...  -0.015756 -0.000640  0.002584  0.037222   \n",
       "4 -26.317193 -0.719588  1.632264 ...  -0.006521  0.009226  0.008907  0.020127   \n",
       "\n",
       "    tonnetz   tonnetz            name  gender  nationality  set  \n",
       "0  0.001958 -0.009467  Gemma_Atkinson       0           UK  dev  \n",
       "1 -0.014433 -0.001219  Gemma_Atkinson       0           UK  dev  \n",
       "2 -0.003111 -0.009005  Gemma_Atkinson       0           UK  dev  \n",
       "3 -0.006243  0.002152  Gemma_Atkinson       0           UK  dev  \n",
       "4  0.001455  0.002741  Gemma_Atkinson       0           UK  dev  \n",
       "\n",
       "[5 rows x 198 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Binarize gender column\n",
    "le_gender = LabelEncoder()\n",
    "full_df.gender = le_gender.fit_transform(full_df.gender)\n",
    "full_df.gender.unique() # array([0, 1])\n",
    "full_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 253
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 916,
     "status": "ok",
     "timestamp": 1549508866785,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "l5CJQGaEyOzl",
    "outputId": "12d1b475-74e7-40fa-a3a0-80734c19bf66"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>...</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>tonnetz</th>\n",
       "      <th>name</th>\n",
       "      <th>gender</th>\n",
       "      <th>nationality</th>\n",
       "      <th>set</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10332</td>\n",
       "      <td>-274.717176</td>\n",
       "      <td>124.096613</td>\n",
       "      <td>-57.174884</td>\n",
       "      <td>47.091800</td>\n",
       "      <td>-31.190586</td>\n",
       "      <td>-3.631755</td>\n",
       "      <td>-34.891319</td>\n",
       "      <td>-4.590279</td>\n",
       "      <td>-0.825808</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004705</td>\n",
       "      <td>0.003553</td>\n",
       "      <td>-0.009748</td>\n",
       "      <td>0.026007</td>\n",
       "      <td>0.001958</td>\n",
       "      <td>-0.009467</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10332</td>\n",
       "      <td>-263.282737</td>\n",
       "      <td>110.169695</td>\n",
       "      <td>-37.532975</td>\n",
       "      <td>49.808548</td>\n",
       "      <td>-30.783113</td>\n",
       "      <td>4.530054</td>\n",
       "      <td>-34.272742</td>\n",
       "      <td>-0.396505</td>\n",
       "      <td>2.088190</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008559</td>\n",
       "      <td>0.001090</td>\n",
       "      <td>-0.052647</td>\n",
       "      <td>0.053715</td>\n",
       "      <td>-0.014433</td>\n",
       "      <td>-0.001219</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10332</td>\n",
       "      <td>-255.330134</td>\n",
       "      <td>107.140581</td>\n",
       "      <td>-37.526152</td>\n",
       "      <td>66.718413</td>\n",
       "      <td>-25.090227</td>\n",
       "      <td>5.835736</td>\n",
       "      <td>-34.682863</td>\n",
       "      <td>0.156228</td>\n",
       "      <td>0.196809</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000381</td>\n",
       "      <td>0.004534</td>\n",
       "      <td>0.003975</td>\n",
       "      <td>0.063643</td>\n",
       "      <td>-0.003111</td>\n",
       "      <td>-0.009005</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10332</td>\n",
       "      <td>-238.647293</td>\n",
       "      <td>116.777418</td>\n",
       "      <td>-43.337226</td>\n",
       "      <td>57.847068</td>\n",
       "      <td>-29.790134</td>\n",
       "      <td>4.556106</td>\n",
       "      <td>-29.883005</td>\n",
       "      <td>-1.273415</td>\n",
       "      <td>0.166903</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.015756</td>\n",
       "      <td>-0.000640</td>\n",
       "      <td>0.002584</td>\n",
       "      <td>0.037222</td>\n",
       "      <td>-0.006243</td>\n",
       "      <td>0.002152</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10332</td>\n",
       "      <td>-253.325462</td>\n",
       "      <td>110.840374</td>\n",
       "      <td>-47.403615</td>\n",
       "      <td>46.822241</td>\n",
       "      <td>-28.057003</td>\n",
       "      <td>5.674004</td>\n",
       "      <td>-26.317193</td>\n",
       "      <td>-0.719588</td>\n",
       "      <td>1.632264</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.006521</td>\n",
       "      <td>0.009226</td>\n",
       "      <td>0.008907</td>\n",
       "      <td>0.020127</td>\n",
       "      <td>0.001455</td>\n",
       "      <td>0.002741</td>\n",
       "      <td>Gemma_Atkinson</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 198 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id        mfcc        mfcc       mfcc       mfcc       mfcc      mfcc  \\\n",
       "0  10332 -274.717176  124.096613 -57.174884  47.091800 -31.190586 -3.631755   \n",
       "1  10332 -263.282737  110.169695 -37.532975  49.808548 -30.783113  4.530054   \n",
       "2  10332 -255.330134  107.140581 -37.526152  66.718413 -25.090227  5.835736   \n",
       "3  10332 -238.647293  116.777418 -43.337226  57.847068 -29.790134  4.556106   \n",
       "4  10332 -253.325462  110.840374 -47.403615  46.822241 -28.057003  5.674004   \n",
       "\n",
       "        mfcc      mfcc      mfcc ...    tonnetz   tonnetz   tonnetz   tonnetz  \\\n",
       "0 -34.891319 -4.590279 -0.825808 ...   0.004705  0.003553 -0.009748  0.026007   \n",
       "1 -34.272742 -0.396505  2.088190 ...  -0.008559  0.001090 -0.052647  0.053715   \n",
       "2 -34.682863  0.156228  0.196809 ...  -0.000381  0.004534  0.003975  0.063643   \n",
       "3 -29.883005 -1.273415  0.166903 ...  -0.015756 -0.000640  0.002584  0.037222   \n",
       "4 -26.317193 -0.719588  1.632264 ...  -0.006521  0.009226  0.008907  0.020127   \n",
       "\n",
       "    tonnetz   tonnetz            name  gender  nationality  set  \n",
       "0  0.001958 -0.009467  Gemma_Atkinson       0           22  dev  \n",
       "1 -0.014433 -0.001219  Gemma_Atkinson       0           22  dev  \n",
       "2 -0.003111 -0.009005  Gemma_Atkinson       0           22  dev  \n",
       "3 -0.006243  0.002152  Gemma_Atkinson       0           22  dev  \n",
       "4  0.001455  0.002741  Gemma_Atkinson       0           22  dev  \n",
       "\n",
       "[5 rows x 198 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Label Encode nationality column\n",
    "le_nationality = LabelEncoder()\n",
    "full_df.nationality = le_nationality.fit_transform(full_df.nationality)\n",
    "full_df.nationality.unique()\n",
    "full_df.head()\n",
    "# le_nationality.classes_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 198
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 322,
     "status": "error",
     "timestamp": 1549508871905,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "v6zw5Jdi4Zm-",
    "outputId": "5187ada3-64ed-4a5c-f19a-5b593a6edfb0"
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "ignored",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-14-d8169bf99b3d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mid10332_mfcc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mid10332_df\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmfcc\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mid10332_mfcc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0max\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatterplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mid10332_mfcc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mid10332_mfcc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m: module 'seaborn' has no attribute 'scatterplot'"
     ]
    }
   ],
   "source": [
    "id10332_df = full_df[full_df.id == 10332]\n",
    "id10332_mfcc = id10332_df.mfcc\n",
    "id10332_mfcc.iloc[1,:]\n",
    "ax = sns.scatterplot(x = len(id10332_mfcc.columns), y = id10332_mfcc.iloc[1,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 314
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1088,
     "status": "ok",
     "timestamp": 1549507338107,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "PnieeA86v9zg",
    "outputId": "a7c7e851-b180-4ff1-877d-741214662d7e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show>"
      ]
     },
     "execution_count": 18,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAADRFJREFUeJzt3W+MpWdZx/HfsmNctgxkLMcAii1G\nvIip0QgqiPwpVMBKgkKFFyASiUggaoIGX2ikIIlGaQpSjZAgClGMf1AwIBCKgIgk0hiMQW8F+aN0\n0UFWHShbumV9MWfLFHe72+c8017T8/kkzZ55es59rvTF8+3zzNn7HDp16lQAoJu73dkDAMCZCBQA\nLQkUAC0JFAAtCRQALW3cEW+yvb0zy0cFt7aO5vjxG+ZYCoAVzXVOXiw2D53p+IG6gtrYOHxnjwDA\n0n6fkw9UoABYHwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVAS3fIXnwA9HLkn8fq\niywesvoat+G8AlVVlyR5U5KrxxjXVNX9k7w2yVcluSnJM8YYn96/MQFYN+e8xVdVFyR5ZZJr9xx+\naZJXjzEeleRPk7xgf8YDYF2dz++gbkxyeZLr9xx7XpI/WT7eTnLhzHMBsObOeYtvjHEyycmq2nvs\n80lSVYeTPD/JS25rja2to7Nty75YbM6yDsBaO3ZklmX285w8+UMSyzi9Psm7xhjX3tZz5/qSwcVi\nM9vbO7OsBbDOjuycWHmNzWSWc/LZIrfKx8xfm+RfxhgvXmENADijSYGqqqcn+eIY40UzzwMASc7j\nFl9VPTjJVUkuTnJTVV2R5GuTnKiqdy+f9uExxvP2a0gA1s/5fEjiuiSP3v9RAODLbHUEQEsCBUBL\nAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsC\nBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsb5/OkqrokyZuS\nXD3GuKaq7p/k9UkOJzmW5EfGGDfu35gArJtzXkFV1QVJXpnk2j2HX5LkN8YYj0jykSQ/tj/jAbCu\nzucW341JLk9y/Z5jj07y5uXjP09y2bxjAbDuznmLb4xxMsnJqtp7+II9t/T+M8l9b2uNra2j2dg4\nPHnIvRaLzVnWAVhrx47Mssx+npPP63dQ53DoXE84fvyGGd5m9z/E9vbOLGsBrLMjOydWXmMzmeWc\nfLbITf0U3+eq6u7Lx1+XW9/+A4CVTQ3UO5M8Zfn4KUneNs84ALDrnLf4qurBSa5KcnGSm6rqiiRP\nT/I7VfUTST6R5Hf3c0gA1s/5fEjiuux+au8rfd/s0wDAkp0kAGhJoABoSaAAaEmgAGhJoABoSaAA\naEmgAGhJoABoSaAAaEmgAGhJoABoSaAAaEmgAGhJoABoSaAAaEmgAGhJoABo6ZzfqNvK3/9Djuyc\nWGmJE99cMw0DwH5yBQVASwIFQEsCBUBLAgVASwIFQEsCBUBLAgVASwIFQEsCBUBLk3aSqKp7JHld\nkq0kX53kxWOMt885GADrbeoV1LOSjDHGpUmuSPKK2SYCgEwP1GeSXLh8vLX8GQBmMylQY4w/SPIN\nVfWRJO9N8rOzTgXA2pv6O6hnJPnkGOMJVfVtSV6T5CFne/7W1tFsbByeOOIex5LNzSMrLbG52Fx9\nDoCD7thq59LTFvt4Tp36dRsPT/L2JBljfKiq7ldVh8cYN5/pyceP3zB1vltZJNlZ9es2tndmmQXg\nIFv1q4uSZDPJ9gzn1LNFburvoD6S5LuTpKouSvK5s8UJAKaYegX1qiS/XVXvWa7x3PlGAoCJgRpj\nfC7JU2eeBQBuYScJAFoSKABaEigAWhIoAFoSKABaEigAWhIoAFoSKABaEigAWhIoAFoSKABaEigA\nWhIoAFoSKABaEigAWhIoAFoSKABaEigAWhIoAFoSKABaEigAWhIoAFoSKABaEigAWhIoAFoSKABa\nEigAWhIoAFramPrCqnp6khcmOZnkF8cYb5ltKgDW3qQrqKq6MMmLknxvkicmedKcQwHA1Cuoy5K8\nc4yxk2QnyXPmGwkApgfq4iRHq+rNSbaSXDnGuPZsT97aOpqNjcMT32qPY8nm5pGVlthcbK4+B8BB\nd2y1c+lpi308p04N1KEkFyb5oSQXJfnLqrpojHHqTE8+fvyGiW9za4skOzsnVlrjxPbOLLMAHGRH\nVjyXJslmku0Zzqlni9zUT/H9R5L3jzFOjjE+mt3bfIuJawHA/zM1UO9I8piqutvyAxP3SPKZ+cYC\nYN1NCtQY41NJ/jjJB5L8RZKfHGN8ac7BAFhvk/8e1BjjVUleNeMsAHALO0kA0JJAAdCSQAHQkkAB\n0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQ\nkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQ0kqBqqq7V9VHq+pZM80DAElWv4L6\nhSSfnWMQANhrcqCq6kFJviXJW+YbBwB2rXIFdVWSF8w1CADstTHlRVX1zCR/M8b4WFWd8/lbW0ez\nsXF4ylvd2rFkc/PISktsLjZXnwPgoDu22rn0tMU+nlMnBSrJDyT5xqp6YpKvT3JjVf37GOOdZ3ry\n8eM3TJ3vVhZJdnZOrLTGie2dWWYBOMiOrHguTZLNJNsznFPPFrlJgRpjPO3046q6MsnHzxYnAJjC\n34MCoKWpt/huMca4coY5AOBWXEEB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB\n0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQkkAB0JJAAdCSQAHQ\nkkAB0JJAAdCSQAHQkkAB0NLG1BdW1a8mecRyjV8eY7xxtqkAWHuTrqCq6tIkl4wxHpbkCUlePutU\nAKy9qbf43pvkh5eP/zvJBVV1eJ6RAGDiLb4xxs1JPr/88dlJ3ro8dkZbW0ezsTFDv44lm5tHVlpi\nc7G5+hwAB92x1c6lpy328Zw6+XdQSVJVT8puoB53W887fvyGVd7mFoskOzsnVlrjxPbOLLMAHGRH\nVjyXJslmku0Zzqlni9wqH5J4fJKfT/KEMcb/TF0HAM5kUqCq6l5Jfi3JZWOMz847EgBMv4J6WpJ7\nJ/nDqjp97JljjE/OMhUAa2/qhyReneTVM88CALewkwQALQkUAC0JFAAtCRQALQkUAC0JFAAtCRQA\nLQkUAC0JFAAtCRQALQkUAC0JFAAtCRQALQkUAC0JFAAtCRQALQkUAC0JFAAtCRQALQkUAC0JFAAt\nCRQALQkUAC0JFAAtCRQALQkUAC0JFAAtCRQALW1MfWFVXZ3koUlOJfnpMcbfzjYVAGtv0hVUVT0q\nyQPHGA9L8uwkvz7rVACsvam3+B6b5M+SZIzxj0m2quqes00FwNqbeovvPkmu2/Pz9vLY/57pyYvF\n5qGJ7/MVC12SzRWXWPX1AHcJi4fMs8xi/86qc31IYp4AAcDS1EBdn90rptPul+TY6uMAwK6pgXpH\nkiuSpKq+I8n1Y4yd2aYCYO0dOnXq1KQXVtWvJHlkki8lef4Y40NzDgbAepscKADYT3aSAKAlgQKg\npclbHd3RbK0E0EdVXZLkTUmuHmNcsx/vcSCuoGytBNBHVV2Q5JVJrt3P9zkQgYqtlQA6uTHJ5dn9\nO7H75qAE6j7Z3U7ptNNbKwFwBxtjnBxjfGG/3+egBOor2VoJ4C7uoATK1koAa+agBMrWSgBr5sDs\nJGFrJYAequrBSa5KcnGSm5J8KsmTxxifnfN9DkygAFgvB+UWHwBrRqAAaEmgAGhJoABoSaAAaEmg\noIGq+qaq+vidPQd0IlAAtHRgvg8KuqiqQ0muye73k306yb8l+Ux2v3rgRdndK/KmJD8+xvjY8sro\nFUm+P8kDkjx3jHFtVX1Pkt/K7ubH1+1Zf2t5fJHkXkmuGmP8flVduXz9RUl+Zoxxy2vgrsgVFNx+\nj03yXct/nrr8+Wh2o/LkMcajsvtdOS/b85ovjDEel+SlSX5qeexlSX5ujPHY7IbutJcmedsY4zHZ\n3T3lJVW1WP67ByS5VJxYB66g4Pb79iR/Nca4Ocnnq+ptSR6U5L5J3lhVSXI4u9/+fNq7l39+IsnX\nLB9/a5L3LR+/K18O16VJvrOqfnT5803ZDVOSfGCMYfsX1oJAwe13t+zuCXnazdn9ArdPjjEefZbX\nnNzz+NCeP0+vc3jPv78xyfPGGB/cu0BVXZ7kixNnhgPHLT64/f4pyUOr6lBVHU3y+CT/muTeVXVJ\nklTVI6vqOedY58NJHrZ8fNme4+/L7q3DVNXdq+o3q8r/TLJ2BApuv7dm94MRH0zye0nen2QnyTOS\nvKaq3pPkl5K85xzrvDDJy6vqrUnuuef4lUkeWFXvS/LeJH83xjh5htfDXZrdzOF2qqp7JfnBJK8b\nY5yqqjcnecMY4w138mhwl+IKCm6/nSQPT3JdVf11kv9K8kd37khw1+MKCoCWXEEB0JJAAdCSQAHQ\nkkAB0JJAAdDS/wFVxG73lcx2HQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "ax = sns.distplot(full_df[\"gender\"], hist = True, kde = False, color = '#ff8c91',\n",
    "             norm_hist = True, label = 'female')\n",
    "ax.set(xticks = full_df.gender.unique())\n",
    "plt.tight_layout()\n",
    "plt.show"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 2635
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 867,
     "status": "ok",
     "timestamp": 1549505843876,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "TVHrEWZrjgTP",
    "outputId": "2a4f6c81-15ff-4fc7-d320-bcc4de4ce5a7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Gemma_Atkinson', 'Keri_Hilson', 'Dyan_Cannon', 'Carrie_Fisher',\n",
       "       'Jeremy_Renner', 'Daniel_Craig', 'Brett_Davern',\n",
       "       'Katheryn_Winnick', 'Katee_Sackhoff', 'Chris_Hemsworth',\n",
       "       'Josh_Henderson', 'Donna_Murphy', 'Luke_Evans', 'Dreama_Walker',\n",
       "       'Daniel_Auteuil', 'Carrie_Ann_Inaba', 'Irrfan_Khan',\n",
       "       'Kristen_Wiig', 'Hannah_Spearritt', 'Lou_Diamond_Phillips',\n",
       "       'Kenya_Moore', 'Keeley_Hawes', 'Mathilda_May', 'Jon_Lovitz',\n",
       "       'Drea_de_Matteo', 'Jessie_J', 'Brenda_Blethyn', 'Hannah_Simone',\n",
       "       'Jessica_Lowndes', 'Lotte_Verbeek', 'Carolyn_Hennesy',\n",
       "       'Jon_Huertas', 'Diego_Klattenhoff', 'Hal_Sparks', 'Ingrid_Bergman',\n",
       "       'Dana_Delany', 'Janice_Dickinson', 'Chloe_Bennet', 'Armie_Hammer',\n",
       "       'Kate_Mulgrew', 'Bradley_Steven_Perry', 'Geena_Davis',\n",
       "       'Hailee_Steinfeld', 'Janet_Montgomery', 'Jeremy_Irons',\n",
       "       'Kristen_Stewart', 'Josh_Groban', 'Ben_McKenzie', 'Keegan_Allen',\n",
       "       'Drake_Bell', 'Jon_Favreau', 'Indira_Varma', 'Armand_Assante',\n",
       "       'Mathieu_Kassovitz', 'Dan_Stevens', 'Ludacris', 'Ben_Feldman',\n",
       "       'America_Ferrera', 'Damon_Wayans', 'John_Goodman', 'Margot_Robbie',\n",
       "       'Kayvan_Novak', 'Jennifer_Saunders', 'Johnny_Galecki',\n",
       "       'Brad_Paisley', 'Loretta_Swit', 'Lily_Cole', 'Janet_McTeer',\n",
       "       'Mark_Webber', 'Dakota_Fanning', 'Amber_Riley', 'Mathieu_Amalric',\n",
       "       'Guy_Pearce', 'Derek_Theler', 'Lucy_Punch', 'Dr._Dre',\n",
       "       'Kristen_Johnston', 'Josh_Duhamel', 'Janet_Jackson',\n",
       "       'Imelda_Staunton', 'Antonio_Cupo', 'James_Cosmo',\n",
       "       'Jesse_McCartney', 'Loretta_Devine', 'Derek_Mears', 'Max_Adler',\n",
       "       'Jennifer_Coolidge', 'Margo_Martindale', 'Bonnie_Wright',\n",
       "       'John_Terry', 'Janel_Parrish', 'Josh_Brolin', 'Kate_Flannery',\n",
       "       'Candice_Accola', 'Lucy_Hale', 'Jesse_Eisenberg', 'Damon_Lindelof',\n",
       "       'Derek_Jacobi', 'Jemaine_Clement', 'Guillaume_Canet',\n",
       "       'Mary_McDonnell', 'Jerry_Trainor', 'John_Slattery',\n",
       "       'Janeane_Garofalo', 'Lorenza_Izzo', 'Kenny_Rogers', 'Jane_Badler',\n",
       "       'John_Simm', 'Jennette_McCurdy', 'John_DiMaggio',\n",
       "       'Marcia_Gay_Harden', 'Goran_Visnjic', 'Jenna_Fischer',\n",
       "       'Denise_Crosby', 'Garry_Marshall', 'Michelle_Keegan',\n",
       "       'Joseph_Gordon-Levitt', 'Melissa_Leo', 'Jana_Kramer',\n",
       "       'Kris_Jenner', 'Michael_Cudlitz', 'Jenna_Coleman',\n",
       "       'James_Badge_Dale', 'Idina_Menzel', 'Matthew_Modine',\n",
       "       'John_Oliver', 'Lorde', 'Marcia_Cross', 'Jenifer_Lewis',\n",
       "       'Candace_Cameron_Bure', 'Mary_Lynn_Rajskub', 'Mark_Sheppard',\n",
       "       'Doutzen_Kroes', 'Allison_Williams', 'Anthony_Rapp',\n",
       "       'Agyness_Deyn', 'Kat_Dennings', 'Amy_Poehler', 'Melissa_George',\n",
       "       'Jordin_Sparks', 'Cote_de_Pablo', 'Moon_Bloodgood',\n",
       "       'Miranda_Cosgrove', 'John_Noble', 'Denis_Leary', 'Glenn_Close',\n",
       "       'Dougray_Scott', 'Colin_Donnell', 'Ian_Somerhalder',\n",
       "       'Jamel_Debbouze', 'Kodi_Smit-McPhee', 'Jamie_Foxx', 'John_Nettles',\n",
       "       'Mary_Elizabeth_Winstead', 'Alison_Pill', 'Jordana_Brewster',\n",
       "       'Costas_Mandylor', 'Mira_Furlan', 'Adrianne_Palicki',\n",
       "       'John_Lloyd_Young', 'Jamie_Chung', 'Chelsea_Handler',\n",
       "       'Karoline_Herfurth', 'Adrianne_Curry', 'Demian_Bichir',\n",
       "       'Jeffrey_Tambor', 'Michael_Copon', 'Jamie-Lynn_Sigler',\n",
       "       'Jeffrey_Donovan', 'Meghan_Trainor', 'Anthony_Mackie',\n",
       "       'Cliff_Curtis', 'Glen_Hansard', 'Adam_Driver', 'Callan_McAuliffe',\n",
       "       'Gillian_Jacobs', 'Matthew_Lillard', 'Mindy_Kaling',\n",
       "       'Mark_Ruffalo', 'Gemma_Arterton', 'Natalie_Dormer',\n",
       "       'Hunter_Parrish', 'Douglas_Booth', 'Mary-Louise_Parker',\n",
       "       'Michaela_Conlin', 'Della_Reese', 'Michael_Cera', 'Claudia_Lee',\n",
       "       'Amaury_Nolasco', 'Karla_Souza', 'Chazz_Palminteri', 'Jeff_Garlin',\n",
       "       'Deidre_Hall', 'Gates_McFadden', 'Alison_Arngrim',\n",
       "       'Alicja_Bachleda', 'Dot-Marie_Jones', 'Natalia_Tena',\n",
       "       'Alexandra_Daddario', 'Freddie_Highmore', 'Hugh_Laurie',\n",
       "       'Debra_Jo_Rupp', 'Gary_Busey', 'Anthony_Anderson', 'John_Cho',\n",
       "       'Jon_Seda', 'Martine_McCutcheon', 'Nicola_Peltz', 'Debi_Mazar',\n",
       "       'John_Hurt', 'Jon_Pertwee', 'Dean_Cain', 'Matthew_Goode',\n",
       "       'David_Lyons', 'Mark_Harmon', 'CCH_Pounder', 'Adam_Brody',\n",
       "       'Bob_Barker', 'Karan_Brar', 'Gregg_Sulkin', 'Garrett_Hedlund',\n",
       "       'Neil_Gaiman', 'Hugh_Dancy', 'Jeff_Daniels', 'Olivia_Munn',\n",
       "       'Milla_Jovovich', 'Burt_Reynolds', 'James_Woods', 'Julie_Walters',\n",
       "       'Megan_Fox', 'Michael_Trevino', 'Julie_Gonzalo', 'Dawn_French',\n",
       "       'Jake_Abel', 'Michael_C._Hall', 'John_Gallagher_Jr.',\n",
       "       'Naseeruddin_Shah', 'Freema_Agyeman', 'Blake_Shelton',\n",
       "       'Neal_McDonough', 'James_Roday', 'Jean-Paul_Belmondo',\n",
       "       'Greg_Grunberg', 'Jeannie_Mai', 'Olivia_Holt', 'Jeanine_Mason',\n",
       "       'David_Wenham', 'Corbin_Bleu', 'Michael_Socha', 'John_Barrowman',\n",
       "       'John_Cleese', 'Nick_Thune', 'Matthew_Fox', 'Jaime_Winstone',\n",
       "       'James_Purefoy', 'Alexz_Johnson', 'David_Tennant', 'Paul_Scheer',\n",
       "       'Blair_Underwood', 'Jean_Smart', 'David_Suchet', 'Abigail_Spencer',\n",
       "       'Greg_Finley', 'Martin_Shaw', 'NeNe_Leakes', 'James_Nesbitt',\n",
       "       'Pablo_Schreiber', 'Olivia_Cooke', 'Peter_Coyote',\n",
       "       'Rade_Serbedzija', 'Hugh_Bonneville', 'James_Maslow',\n",
       "       'Cindy_Williams', 'David_Oyelowo', 'Raoul_Bova', 'Jean_Reno',\n",
       "       'Naomie_Harris', 'Francia_Raisa', 'David_Mazouz', 'Mark_Gatiss',\n",
       "       'Missi_Pyle', 'Joel_Edgerton', 'David_Mamet', 'Nazanin_Boniadi',\n",
       "       'Forest_Whitaker', 'Abigail_Breslin', 'Martin_Kove',\n",
       "       'Michael_Shannon', 'Peter_Capaldi', 'Rene_Auberjonois',\n",
       "       'Miley_Cyrus', 'David_Kross', 'Greer_Garson', 'Patrick_Duffy',\n",
       "       'Grant_Bowler', 'Olivia_Colman', 'David_Koechner', 'Paul_Reubens',\n",
       "       'Naturi_Naughton', 'Chris_Colfer', 'John_Amos', 'David_Henrie',\n",
       "       'Nancy_Cartwright', 'Rachel_McAdams', 'Joe_Flanigan', 'Reid_Scott',\n",
       "       'Oscar_Isaac', 'Joaquim_de_Almeida', 'Gordon_Ramsay',\n",
       "       'Abbie_Cornish', 'Billie_Joe_Armstrong', 'Marisa_Miller',\n",
       "       'Matt_Lanter', 'Donna_Mills', 'Hope_Solo', 'Randolph_Mantooth',\n",
       "       'Paul_McGann', 'Chloe_Dykstra', 'Oliver_Platt', 'Richard_E._Grant',\n",
       "       'Mia_Wasikowska', 'Patrick_Dempsey', 'James_Marsden',\n",
       "       'Pierre_Niney', 'John_Abraham', 'Nana_Visitor', 'Rachel_Dratch',\n",
       "       'David_Harewood', 'Florence_Henderson', 'Miles_Teller',\n",
       "       'Martin_Henderson', 'Mischa_Barton', 'Aaron_Yoo', 'Joel_McHale',\n",
       "       'Regina_Hall', 'Matt_Dallas', 'Bill_Pullman', 'Robbie_Kay',\n",
       "       'Nadia_Bjorlin', 'Robert_Vaughn', 'RZA', 'Pierfrancesco_Favino',\n",
       "       'David_Attenborough', 'Bill_Nighy', 'Richard_Dean_Anderson',\n",
       "       'Paul_Gross', 'Nathan_Fillion', 'Maya_Rudolph', 'David_Giuntoli',\n",
       "       'Dave_Foley', 'James_Lafferty', 'Danielle_Panabaker', 'Peter_Berg',\n",
       "       'Dan_Fogler', 'Russell_Tovey', 'Dan_Aykroyd', 'Pat_Monahan',\n",
       "       'Miriam_Margolyes', 'Matt_Bomer', 'David_Faustino',\n",
       "       'Jacqueline_Emerson', 'Cyndi_Lauper', 'Mena_Suvari',\n",
       "       'Martin_Clunes', 'Cassandra_Peterson', 'Rik_Mayall',\n",
       "       'Ryan_McPartlin', 'Reggie_Lee', 'Max_Thieriot', 'Robbie_Amell',\n",
       "       'Michael_Kelly', 'Olga_Kurylenko', 'Mila_Kunis', 'Fiona_Shaw',\n",
       "       'Omari_Hardwick', 'Betty_White', 'James_Frain', 'Neve_Campbell',\n",
       "       'Nathalie_Emmanuel', 'Jack_Whitehall', 'Michael_K._Williams',\n",
       "       'Rico_Rodriguez', 'Miranda_Richardson', 'Reece_Ritchie',\n",
       "       'Melora_Hardin', 'Sandra_Bernhard', 'Aamir_Khan', 'Finola_Hughes',\n",
       "       'Sarah_Bolger', 'Sam_Worthington', 'Phyllis_Diller', 'Arden_Cho',\n",
       "       'Miranda_Lambert', 'Odeya_Rush', 'Rooney_Mara', 'Rob_Riggle',\n",
       "       'Michelle_Williams', 'Parineeti_Chopra', 'James_Earl_Jones',\n",
       "       'Om_Puri', 'Nate_Parker', 'Odette_Annable', 'Ansel_Elgort',\n",
       "       'Michael_Jai_White', 'Sally_Hawkins', 'Quentin_Tarantino',\n",
       "       'Sanaa_Lathan', 'Philip_Glenister', 'Fergie', 'Hilary_Swank',\n",
       "       'Michelle_Stafford', 'Hilarie_Burton', 'Robert_Picardo',\n",
       "       'Paolo_Sorrentino', 'Penn_Jillette', 'A.R._Rahman',\n",
       "       'Rebecca_Mader', 'Melissa_Rycroft', 'Max_Irons',\n",
       "       'Moritz_Bleibtreu', 'Prince', 'Anne_Hathaway', 'Ruby_Dee',\n",
       "       'Ronda_Rousey', 'Sam_Trammell', 'Olivia_Williams', 'Nelly',\n",
       "       'A.J._Buckley', 'Retta', 'Shannen_Doherty', 'Sasha_Pieterse',\n",
       "       'Octavia_Spencer', 'Natalie_Zea', 'Felicity_Jones', 'Sara_Gilbert',\n",
       "       'Reba_McEntire', 'Shemar_Moore', 'Ricky_Gervais', 'Pamela_Adlon',\n",
       "       'Ron_Perlman', 'Melissa_Rauch', 'Sarah_Wayne_Callies',\n",
       "       'Morena_Baccarin', 'Seann_William_Scott', 'Norman_Reedus',\n",
       "       'Santiago_Cabrera', 'Sammy_Davis_Jr.', 'Paul_Wesley', 'Rami_Malek',\n",
       "       'Ruth_Jones', 'Sharon_Gless', 'Michelle_Krusiec', 'Noomi_Rapace',\n",
       "       'Rowan_Atkinson', 'Moran_Atias', 'Shane_West', 'Paul_Sorvino',\n",
       "       'Shiri_Appleby', 'Olivia_Wilde', 'Ron_Livingston',\n",
       "       'Sammi_Hanratty', 'Stephen_Colbert', 'Ralph_Macchio',\n",
       "       'Richard_Madden', 'Stana_Katic', 'Melissa_Peterman',\n",
       "       'Preity_Zinta', 'Sofia_Coppola', 'Sandra_Oh', 'Noel_Fielding',\n",
       "       'Robert_Carlyle', 'Sarah_Palin', 'Steven_Soderbergh',\n",
       "       'Seth_MacFarlane', 'Sidse_Babett_Knudsen', 'Sharni_Vinson',\n",
       "       'Paloma_Faith', 'Sean_Patrick_Flanery', 'Noel_Clarke',\n",
       "       'Richard_Kind', 'Rutger_Hauer', 'Samira_Wiley', 'Shiloh_Fernandez',\n",
       "       'Ralf_Little', 'Natalie_Martinez', 'Ryan_Potter', 'Sonja_Sohn',\n",
       "       'Peter_Jackson', 'Patton_Oswalt', 'Tabu', 'Melissa_McCarthy',\n",
       "       'Pooja_Kumar', 'Suzanne_Pleshette', 'Simon_Callow',\n",
       "       'Rosie_Huntington-Whiteley', 'Stephen_Mangan', 'Ruta_Lee',\n",
       "       'Steve_Burton', 'Tamala_Jones', 'Sarah_Hyland', 'Nina_Hoss',\n",
       "       'Sunny_Leone', 'Nina_Arianda', 'Sterling_Knight',\n",
       "       'Stefanie_Powers', 'Samantha_Barks', 'Richard_Hammond',\n",
       "       'Ryan_Phillippe', 'Stacy_Keach', 'Sidharth_Malhotra',\n",
       "       'Rainn_Wilson', 'Ted_Danson', 'Rob_Reiner', 'Simon_Baker',\n",
       "       'Tamera_Mowry-Housley', 'Sam_Palladio', 'Tahmoh_Penikett',\n",
       "       'Taryn_Manning', 'Nikolaj_Coster-Waldau', 'Sherri_Shepherd',\n",
       "       'Stephen_Merchant', 'Smokey_Robinson', 'Roshon_Fegan',\n",
       "       'Nicole_Scherzinger', 'Peter_Facinelli', 'Sharlto_Copley',\n",
       "       'Nicole_Richie', 'Serinda_Swan', 'Stanley_Tucci',\n",
       "       'Silas_Weir_Mitchell', 'Nicki_Minaj', 'Spencer_Boldman',\n",
       "       'Steven_Moffat', 'Nick_Kroll', 'Sung_Kang', 'Shahid_Kapoor',\n",
       "       'Sonam_Kapoor'], dtype=object)"
      ]
     },
     "execution_count": 37,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_df.name.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Kj-xnF4HjgRm"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dWtWQPA2jgNr"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "2tqDL1KEjgKS"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_qhqxihs35fb"
   },
   "source": [
    "#Stuff below is irrelevant to this project"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Dld2qPgRsWcg"
   },
   "source": [
    "## Concatenate the file clips into full videos"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EJ8x_XbqsWch"
   },
   "source": [
    "According to the article the final dataset consists of 121 videos including 61 deceptive and 60 truthful trial clips. The videos are written in the format 'trial_lie_001_000.mp4' - 'trial_lie_001_004.mp4' and 'trial_truth_041_000.mp4' - 'trial_truth_041_008.mp4'. The exact number of chunks per video is unknown."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "34N66mRqsWci"
   },
   "outputs": [],
   "source": [
    "# Get the filenames in the directory: video_chunks\n",
    "files = [f for f in listdir('video_chunks/') if isfile(join('video_chunks/',f))]\n",
    "files.pop(0) # .DS_Store is first element in the list, don't need this"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "i-Jx0rPlsWck"
   },
   "outputs": [],
   "source": [
    "print(len(files)) # 829. matches how many video files I have in my directory \n",
    "files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Sl9waUKGsWcn"
   },
   "outputs": [],
   "source": [
    "# Split files list into truth list and lie list\n",
    "truth = 'trial_truth_'\n",
    "lie = 'trial_lie_'\n",
    "truth_list = list(filter(lambda file: file.startswith(truth),files))\n",
    "lie_list = list(filter(lambda file: file.startswith(lie),files))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "mP46pfPosWcq"
   },
   "outputs": [],
   "source": [
    "# Format list into dictionary\n",
    "lie_dict = {}\n",
    "for file in lie_list:\n",
    "    try: \n",
    "        lie_dict[file[10:13]]: lie_dict[file[10:13]].append(file) # append filename to appropriate key\n",
    "    except:\n",
    "        lie_dict[file[10:13]] = [] # create key with value as empty list\n",
    "        lie_dict[file[10:13]]: lie_dict[file[10:13]].append(file) # append filename to appropriate key\n",
    "\n",
    "truth_dict = {}\n",
    "for file in truth_list:\n",
    "    try: \n",
    "        truth_dict[file[12:15]]: truth_dict[file[12:15]].append(file)\n",
    "    except:\n",
    "        truth_dict[file[12:15]] = []\n",
    "        truth_dict[file[12:15]]: truth_dict[file[12:15]].append(file)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "x2qXC7_zsWct",
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Concat truth videos and save them in a new folder\n",
    "# Had to restart kernel and start at stopping point because the error was OSError: [Errno 24] Too many open files'\n",
    "\n",
    "try:\n",
    "    for key in list(truth_dict.keys())[49:]: # There are 50 keys in the list. some numbers are skipped\n",
    "        vids = [VideoFileClip(\"Video_chunks/{}\".format(clip)) for clip in truth_dict[key]] # read each clip, save as list of clips\n",
    "        full_video = concatenate_videoclips(vids) # put all the chunks together to one video\n",
    "        name = truth_dict[key][0][:15] # call it the appropriate name\n",
    "        full_video.write_videofile('videos/'+name+'.mp4') # write the full video and save in a diff. folder\n",
    "except OSError:\n",
    "    print('stopped at index: {}'.format(list(truth_dict.keys()).index(key)))\n",
    "    # Clear the kernel and change the first line of try to start at the index it failed on"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "FhQkwMcGsWcw"
   },
   "outputs": [],
   "source": [
    "# Concat lie videos and save them in a new folder\n",
    "# Had to restart kernel and start at stopping point because the error was OSError: [Errno 24] Too many open files'\n",
    "\n",
    "try:\n",
    "    for key in list(lie_dict.keys())[55:]: # There are 61 keys in the list. some numbers are skipped\n",
    "        vids = [VideoFileClip(\"Video_chunks/{}\".format(clip)) for clip in lie_dict[key]] # read each clip, save as list of clips\n",
    "        full_video = concatenate_videoclips(vids) # put all the chunks together to one video\n",
    "        name = lie_dict[key][0][:13] # call it the appropriate name\n",
    "        full_video.write_videofile('videos/'+name+'.mp4') # write the full video and save in a diff. folder\n",
    "except OSError:\n",
    "    print('stopped at index: {}'.format(list(lie_dict.keys()).index(key)))\n",
    "    # Clear the kernel and change the first line of try to start at the index it failed on"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "W8xdt-DosWcz"
   },
   "source": [
    "## Get audio from each video"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "wuP4N74DsWcz",
    "outputId": "ad2ef2e9-517a-4638-a5f1-075bb1585581"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'.DS_Store'"
      ]
     },
     "execution_count": 6,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the filenames in the directory: video_chunks\n",
    "videos = [f for f in listdir('videos/')]\n",
    "videos.pop(0)# .DS_Store is first element in the list, don't need this. 109 videos in total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "collapsed": true,
    "id": "JdJ2EAQPsWc5",
    "outputId": "1f03e60f-b90d-48c7-d593-c314e50aeadf"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_001.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 377/377 [00:00<00:00, 2556.18it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_002.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1381/1381 [00:00<00:00, 2868.92it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_003.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 158/158 [00:00<00:00, 2634.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_004.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "100%|██████████| 258/258 [00:00<00:00, 2719.71it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_005.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1181/1181 [00:00<00:00, 2720.25it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_006.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 404/404 [00:00<00:00, 2714.65it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_007.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1034/1034 [00:00<00:00, 2861.60it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_008.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 163/163 [00:00<00:00, 2695.55it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_009.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 457/457 [00:00<00:00, 2391.07it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_010.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 646/646 [00:00<00:00, 2867.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_011.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 745/745 [00:00<00:00, 2855.76it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_012.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 176/176 [00:00<00:00, 2625.41it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_013.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 422/422 [00:00<00:00, 2685.66it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_014.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 335/335 [00:00<00:00, 2467.03it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_015.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 778/778 [00:00<00:00, 2831.39it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_016.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 833/833 [00:00<00:00, 2721.75it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_017.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:00<00:00, 2502.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_018.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 773/773 [00:00<00:00, 2681.49it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_019.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 823/823 [00:00<00:00, 2785.83it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_020.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 271/271 [00:00<00:00, 2758.04it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_021.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 421/421 [00:00<00:00, 2740.55it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_022.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 939/939 [00:00<00:00, 2810.95it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_023.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 994/994 [00:00<00:00, 2861.38it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_024.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 564/564 [00:00<00:00, 2743.41it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_025.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 668/668 [00:00<00:00, 2808.90it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_026.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 642/642 [00:00<00:00, 2799.23it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_027.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 611/611 [00:00<00:00, 2835.29it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_028.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 559/559 [00:00<00:00, 2859.13it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_029.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 476/476 [00:00<00:00, 2827.53it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_030.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 839/839 [00:00<00:00, 2690.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_031.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 593/593 [00:00<00:00, 2863.21it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_033.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 997/997 [00:00<00:00, 2853.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_035.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 779/779 [00:00<00:00, 2923.37it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_036.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 864/864 [00:00<00:00, 2998.00it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_037.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 577/577 [00:00<00:00, 2977.31it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_038.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 511/511 [00:00<00:00, 2902.50it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_039.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 666/666 [00:00<00:00, 2993.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_040.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 533/533 [00:00<00:00, 2947.86it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_041.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 576/576 [00:00<00:00, 2907.48it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_042.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 532/532 [00:00<00:00, 3004.38it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_043.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 313/313 [00:00<00:00, 2812.11it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_044.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 202/202 [00:00<00:00, 2602.46it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_045.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 423/423 [00:00<00:00, 2702.84it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_046.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 732/732 [00:00<00:00, 2933.33it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_047.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 357/357 [00:00<00:00, 2607.00it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_048.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1063/1063 [00:00<00:00, 2890.64it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_049.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 511/511 [00:00<00:00, 2938.05it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_050.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 555/555 [00:00<00:00, 2816.27it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_051.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 158/158 [00:00<00:00, 3011.41it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_052.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "100%|██████████| 1062/1062 [00:00<00:00, 2928.09it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_053.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 801/801 [00:00<00:00, 2926.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_054.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 445/445 [00:00<00:00, 2847.16it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_055.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 709/709 [00:00<00:00, 2828.44it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_056.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 643/643 [00:00<00:00, 2856.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_057.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 423/423 [00:00<00:00, 2685.23it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_058.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [00:00<00:00, 2880.40it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_059.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 886/886 [00:00<00:00, 2894.20it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_lie_060.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 555/555 [00:00<00:00, 2836.39it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_lie_061.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 643/643 [00:00<00:00, 2859.00it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_001.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 317/317 [00:00<00:00, 2602.97it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_002.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 458/458 [00:00<00:00, 2650.65it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_003.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 283/283 [00:00<00:00, 2732.29it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_004.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1799/1799 [00:00<00:00, 2740.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_005.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 769/769 [00:00<00:00, 2829.13it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_006.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 677/677 [00:00<00:00, 2802.87it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_007.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1581/1581 [00:00<00:00, 2538.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_008.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1011/1011 [00:00<00:00, 2829.56it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_009.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 504/504 [00:00<00:00, 2551.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_truth_010.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1615/1615 [00:00<00:00, 2792.21it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_011.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 945/945 [00:00<00:00, 2904.07it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_012.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 635/635 [00:00<00:00, 2655.05it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_013.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 653/653 [00:00<00:00, 2801.32it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_014.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 304/304 [00:00<00:00, 2821.64it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_015.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 820/820 [00:00<00:00, 2874.90it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_016.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 169/169 [00:00<00:00, 2677.53it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_017.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 103/103 [00:00<00:00, 2623.54it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_018.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 157/157 [00:00<00:00, 3140.24it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_019.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 289/289 [00:00<00:00, 2676.39it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_020.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 156/156 [00:00<00:00, 2761.83it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_021.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 243/243 [00:00<00:00, 2830.16it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_022.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 633/633 [00:00<00:00, 2834.64it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_023.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 520/520 [00:00<00:00, 2772.05it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_024.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 536/536 [00:00<00:00, 2814.76it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_025.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 787/787 [00:00<00:00, 2660.04it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_truth_026.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 710/710 [00:00<00:00, 2603.16it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_027.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 557/557 [00:00<00:00, 2947.43it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_030.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1218/1218 [00:00<00:00, 2909.10it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_037.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [00:00<00:00, 2868.07it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_039.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 710/710 [00:00<00:00, 2818.64it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_041.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 757/757 [00:00<00:00, 2922.10it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_042.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 669/669 [00:00<00:00, 2979.75it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_043.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 492/492 [00:00<00:00, 2817.07it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_044.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 470/470 [00:00<00:00, 2812.15it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_045.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 735/735 [00:00<00:00, 2832.84it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_046.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 823/823 [00:00<00:00, 2927.37it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_047.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 580/580 [00:00<00:00, 2901.54it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_048.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 580/580 [00:00<00:00, 2806.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_049.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 669/669 [00:00<00:00, 2836.48it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_050.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 514/514 [00:00<00:00, 3039.52it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_051.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 643/643 [00:00<00:00, 2786.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_052.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 647/647 [00:00<00:00, 2996.47it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_053.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 666/666 [00:00<00:00, 2926.40it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_054.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 533/533 [00:00<00:00, 2872.03it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_055.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 643/643 [00:00<00:00, 2858.54it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_056.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 754/754 [00:00<00:00, 2848.12it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_057.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 842/842 [00:00<00:00, 2999.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_058.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 467/467 [00:00<00:00, 2869.59it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Writing audio in audio/trial_truth_059.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 599/599 [00:00<00:00, 3028.27it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n",
      "[MoviePy] Writing audio in audio/trial_truth_060.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 445/445 [00:00<00:00, 2859.97it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MoviePy] Done.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Write audio files from videos and save it in the empty audio file\n",
    "for video in videos:\n",
    "    vid = VideoFileClip('videos/'+video)\n",
    "    name = video[:-4]\n",
    "    vid.audio.write_audiofile(\"audio/\"+name+\".wav\") # name the audio file like its corresponding video, minus the .mp4\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "KWBimd_osWc8"
   },
   "source": [
    "## Extract features & store in dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7aEchr-GsWc_",
    "outputId": "e956776f-165f-4822-f853-54e2ea41854d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'.DS_Store'"
      ]
     },
     "execution_count": 16,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the name of the audio files in the audio folder\n",
    "audio_files = [f for f in listdir('audio/')]\n",
    "audio_files.pop(0) # '.DS_Store' is first element in list of audio_files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "orGY5uq1sWdC"
   },
   "outputs": [],
   "source": [
    "def extract_features(file_name):\n",
    "    y, samp_rate = librosa.load('audio/'+file_name, sr = None) # sr = None, keeps original sample_rate from audio\n",
    "\n",
    "    mfccs = librosa.feature.mfcc(y = y, sr = samp_rate, n_mfcc=40) # returns shape (num of mfcc, num of frames), each column is a frame of 40 mfcc values\n",
    "    mfccs_avg = np.mean(scale(mfccs, axis = 1), axis = 1) # average mfcc over the frames frame. shape = (40,). scale across columns, within a row\n",
    "\n",
    "    mel = librosa.feature.melspectrogram(y = y, sr = samp_rate)\n",
    "    mel_avg = np.mean(scale(mel, axis = 1), axis = 1)\n",
    "\n",
    "    stft = np.abs(librosa.stft(y))\n",
    "    chroma = librosa.feature.chroma_stft(S = stft, y = y, sr = samp_rate)\n",
    "    chroma_avg = np.mean(scale(chroma, axis = 1), axis = 1)\n",
    "\n",
    "    contrast = librosa.feature.spectral_contrast(S = stft, y = y, sr = samp_rate)\n",
    "    contrast_avg = np.mean(scale(contrast, axis = 1), axis = 1)\n",
    "\n",
    "    tonnetz = librosa.feature.tonnetz(y = librosa.effects.harmonic(y), sr = samp_rate)\n",
    "    tonnetz_avg = np.mean(scale(tonnetz, axis = 1), axis = 1)\n",
    "    \n",
    "    return mfccs_avg, mel_avg, chroma_avg, contrast_avg, tonnetz_avg # return shape: (40,), (128,), (12,), (7,), (6,) respectively\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 297,
     "status": "ok",
     "timestamp": 1548968356634,
     "user": {
      "displayName": "Patrina Bailey",
      "photoUrl": "https://lh6.googleusercontent.com/-_-IxxZOx6Gk/AAAAAAAAAAI/AAAAAAAAJ6M/E0SG_dIR-sM/s64/photo.jpg",
      "userId": "15476889788748924708"
     },
     "user_tz": 300
    },
    "id": "J6AAARKLtOHd",
    "outputId": "f960b2b4-17e8-43dc-cbea-7cb2a4a2b1c1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['mfcc', 'mfcc', 'mel', 'mel']"
      ]
     },
     "execution_count": 48,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "['mfcc']*40 + ['mel']*128 + ['chroma']*12 + ['contrast']*7 + ['tonnetz']*6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "UXu8ROz3sWdD",
    "outputId": "4a50251b-1bae-4328-a345-48b9c36fb66d"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>lie/truth</th>\n",
       "      <th>mfcc</th>\n",
       "      <th>mel</th>\n",
       "      <th>chroma</th>\n",
       "      <th>contrast</th>\n",
       "      <th>tonnetz</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [lie/truth, mfcc, mel, chroma, contrast, tonnetz]\n",
       "Index: []"
      ]
     },
     "execution_count": 501,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build Dataframe framework\n",
    "columns = ['lie/truth','mfcc','mel','chroma','contrast','tonnetz'] # make each feature its own column, each cell will have list of floats\n",
    "\n",
    "audio_df = pd.DataFrame(columns = columns)\n",
    "audio_df # 0 rows × 194 columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "aXhCXhHasWdH"
   },
   "outputs": [],
   "source": [
    "mfcc, mel, chroma, contrast, tonnetz = extract_features(audio_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Lbxd9IaxsWdK"
   },
   "outputs": [],
   "source": [
    "mfcc = np.reshape(np.append(mfcc,[[0]*(mel.shape[0] - mfcc.shape[0])]), newshape = (128,1)) \n",
    "mel = np.reshape(mel, newshape = (128,1))\n",
    "chroma = np.reshape(np.append(chroma,[[0]*(mel.shape[0] - chroma.shape[0])]), newshape = (128,1))\n",
    "contrast = np.reshape(np.append(contrast,[[0]*(mel.shape[0] - contrast.shape[0])]), newshape = (128,1))\n",
    "tonnetz = np.reshape(np.append(tonnetz,[[0]*(mel.shape[0] - tonnetz.shape[0])]), newshape = (128,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "brYhzCk8sWdM",
    "outputId": "e33636d9-f0f7-4569-849e-d9a297b38a59"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 128, 5)"
      ]
     },
     "execution_count": 590,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = np.concatenate((mfcc, mel, chroma, contrast, tonnetz), axis = 1)\n",
    "data = np.reshape(data, newshape = (1,128,5))\n",
    "data.shape # (128, 5)\n",
    "       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "aBXZq4obsWdS",
    "outputId": "95a989bd-854b-47cd-9435-022bbc312ce4"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'truth'"
      ]
     },
     "execution_count": 591,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "name = 'trial_lie_001.wav'\n",
    "try:\n",
    "    name = re.search('lie', audio_file).group()\n",
    "except AttributeError:\n",
    "    name = re.search('truth', audio_file).group()\n",
    "name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "pCCjX743sWdV"
   },
   "outputs": [],
   "source": [
    "panel = pd.Panel(data, items = [name], minor_axis= ['mfcc','mel','chroma','contrast','tonnetz'])\n",
    "panel2 = pd.Panel(data, items = [name], minor_axis= ['mfcc','mel','chroma','contrast','tonnetz'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zcH5NN4OsWdX",
    "outputId": "bf317ca9-f937-450a-e14f-8b7ca9591a58"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<class 'pandas.core.panel.Panel'>\n",
       "Dimensions: 2 (items) x 128 (major_axis) x 5 (minor_axis)\n",
       "Items axis: truth to truth\n",
       "Major_axis axis: 0 to 127\n",
       "Minor_axis axis: mfcc to tonnetz"
      ]
     },
     "execution_count": 600,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([panel, panel2], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "vlgkHdvHsWdb"
   },
   "outputs": [],
   "source": [
    "panels = []\n",
    "for audio_file in audio_files:\n",
    "    # Get the features for each audio file\n",
    "    mfcc, mel, chroma, contrast, tonnetz = extract_features(audio_file)\n",
    "    \n",
    "    # reshape np.array\n",
    "    mfcc = np.reshape(np.append(mfcc,[[0]*(mel.shape[0] - mfcc.shape[0])]), newshape = (128,1)) \n",
    "    mel = np.reshape(mel, newshape = (128,1))\n",
    "    chroma = np.reshape(np.append(chroma,[[0]*(mel.shape[0] - chroma.shape[0])]), newshape = (128,1))\n",
    "    contrast = np.reshape(np.append(contrast,[[0]*(mel.shape[0] - contrast.shape[0])]), newshape = (128,1))\n",
    "    tonnetz = np.reshape(np.append(tonnetz,[[0]*(mel.shape[0] - tonnetz.shape[0])]), newshape = (128,1))\n",
    "    \n",
    "    # concat all the columns together and reshape to give it 3D shape\n",
    "    data = np.concatenate((mfcc, mel, chroma, contrast, tonnetz), axis = 1)\n",
    "    data = np.reshape(data, newshape = (1,128,5))\n",
    "       \n",
    "    # horizontally stack each array to one another\n",
    "#     features = np.hstack(extracted_features) # horizontally stack each array next to each other to achieve one big array. shape = (193,)\n",
    "    \n",
    "    # get the label of the audio\n",
    "    try:\n",
    "        name = re.search('lie', audio_file).group()\n",
    "    except AttributeError:\n",
    "        name = re.search('truth', audio_file).group()\n",
    "    \n",
    "    panel = pd.Panel(data, items = [name], minor_axis = ['mfcc','mel','chroma','contrast','tonnetz'])\n",
    "    panels.append(panel)\n",
    "#     row_df = pd.DataFrame([[name, mfcc, mel, chroma, contrast, tonnetz]], columns = columns)\n",
    "#     audio_df = audio_df.append(row_df, ignore_index = True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dAbhgZSgsWde"
   },
   "outputs": [],
   "source": [
    "dataframe = pd.concat(panels, axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4r582jj2sWdh"
   },
   "outputs": [],
   "source": [
    "dataframe.to_pickle('./df.pkl') # pickle dataframe so we can run models in a diff. file\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "SvQDQ9sTsWdk"
   },
   "outputs": [],
   "source": [
    "dataframe.to_pickle('./df.pkl')\n",
    "pd.read_pickle('./audio_df.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "S4NpzYV5sWdn"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [
    "4o2mHJO6skA5",
    "7wXEGyHyZjlc",
    "-QeBKSGsqb4v",
    "_qhqxihs35fb",
    "Dld2qPgRsWcg",
    "W8xdt-DosWcz",
    "KWBimd_osWc8"
   ],
   "name": "AudioRecognition.ipynb",
   "provenance": [],
   "toc_visible": true,
   "version": "0.3.2"
  },
  "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}