{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xfNT-mlFwxVM"
   },
   "source": [
    "# Intro to Autoencoders"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ITZuApL56Mny"
   },
   "source": [
    "This tutorial introduces autoencoders with three examples: the basics, image denoising, and anomaly detection.\n",
    "\n",
    "An autoencoder is a special type of neural network that is trained to copy its input to its output. For example, given an image of a handwritten digit, an autoencoder first encodes the image into a lower dimensional latent representation, then decodes the latent representation back to an image. An autoencoder learns to compress the data while minimizing the reconstruction error. \n",
    "\n",
    "To learn more about autoencoders, please consider reading chapter 14 from [Deep Learning](https://www.deeplearningbook.org/) by Ian Goodfellow, Yoshua Bengio, and Aaron Courville."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "e1_Y75QXJS6h"
   },
   "source": [
    "## Import TensorFlow and other libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:12:40.668592Z",
     "iopub.status.busy": "2022-12-14T06:12:40.668147Z",
     "iopub.status.idle": "2022-12-14T06:12:43.222802Z",
     "shell.execute_reply": "2022-12-14T06:12:43.221987Z"
    },
    "id": "YfIk2es3hJEd"
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras import layers, losses\n",
    "from tensorflow.keras.datasets import fashion_mnist\n",
    "from tensorflow.keras.models import Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import requests\n",
    "import zipfile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "basic_autoencoder_model_url = \"https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/basic_autoencoder_model.zip\"\n",
    "convolutional_autoencoder_model_url = \"https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/convolutional_autoencoder_model.zip\"\n",
    "ecg_autoencoder_model_url = \"https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/ecg_autoencoder_model.zip\"\n",
    "\n",
    "notebook_path = os.getcwd()\n",
    "\n",
    "tmp_folder_path = os.path.join(notebook_path, \"tmp\")\n",
    "\n",
    "if not os.path.exists(tmp_folder_path):\n",
    "    os.makedirs(tmp_folder_path)\n",
    "\n",
    "file_path = os.path.join(tmp_folder_path,\"autoencoder\")\n",
    "\n",
    "if not os.path.exists(file_path):\n",
    "    os.makedirs(file_path)\n",
    "\n",
    "zip_store_path = os.path.join(file_path, \"zip-store\")\n",
    "\n",
    "if not os.path.exists(zip_store_path):\n",
    "    os.makedirs(zip_store_path)\n",
    "\n",
    "basic_autoencoder_model_response = requests.get(basic_autoencoder_model_url)\n",
    "convolutional_autoencoder_model_response = requests.get(convolutional_autoencoder_model_url)\n",
    "ecg_autoencoder_model_response = requests.get(ecg_autoencoder_model_url)\n",
    "\n",
    "basic_autoencoder_model_name = os.path.basename(basic_autoencoder_model_url)\n",
    "convolutional_autoencoder_model_name = os.path.basename(convolutional_autoencoder_model_url)\n",
    "ecg_autoencoder_model_name = os.path.basename(ecg_autoencoder_model_url)\n",
    "\n",
    "basic_autoencoder_model_save_path = os.path.join(zip_store_path, basic_autoencoder_model_name)\n",
    "convolutional_autoencoder_model_save_path = os.path.join(zip_store_path, convolutional_autoencoder_model_name)\n",
    "ecg_autoencoder_model_save_path = os.path.join(zip_store_path, ecg_autoencoder_model_name)\n",
    "\n",
    "with open(basic_autoencoder_model_save_path, \"wb\") as file:\n",
    "    file.write(basic_autoencoder_model_response.content)\n",
    "    \n",
    "with open(convolutional_autoencoder_model_save_path, \"wb\") as file:\n",
    "    file.write(convolutional_autoencoder_model_response.content)\n",
    "\n",
    "with open(ecg_autoencoder_model_save_path, \"wb\") as file:\n",
    "    file.write(ecg_autoencoder_model_response.content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "bam_zip_file_path = f\"./tmp/autoencoder/zip-store/{basic_autoencoder_model_name}\"\n",
    "bam_extract_path = \"./tmp/autoencoder/basic_autoencoder_model\"\n",
    "\n",
    "cam_zip_file_path = f\"./tmp/autoencoder/zip-store/{convolutional_autoencoder_model_name}\"\n",
    "cam_extract_path = \"./tmp/autoencoder/convolutional_autoencoder_model\"\n",
    "\n",
    "ecg_zip_file_path = f\"./tmp/autoencoder/zip-store/{ecg_autoencoder_model_name}\"\n",
    "ecg_extract_path = \"./tmp/autoencoder/ecg_autoencoder_model\"\n",
    "\n",
    "zip_ref = zipfile.ZipFile(bam_zip_file_path, 'r')\n",
    "zip_ref.extractall(bam_extract_path)\n",
    "zip_ref.close()\n",
    "\n",
    "zip_ref = zipfile.ZipFile(cam_zip_file_path, 'r')\n",
    "zip_ref.extractall(cam_extract_path)\n",
    "zip_ref.close()\n",
    "\n",
    "zip_ref = zipfile.ZipFile(ecg_zip_file_path, 'r')\n",
    "zip_ref.extractall(ecg_extract_path)\n",
    "zip_ref.close()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "id": "iYn4MdZnKCey"
   },
   "source": [
    "## Load the dataset\n",
    "To start, you will train the basic autoencoder using the Fashion MNIST dataset. Each image in this dataset is 28x28 pixels. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:12:43.227459Z",
     "iopub.status.busy": "2022-12-14T06:12:43.226760Z",
     "iopub.status.idle": "2022-12-14T06:12:43.742908Z",
     "shell.execute_reply": "2022-12-14T06:12:43.742171Z"
    },
    "id": "YZm503-I_tji"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n",
      "(10000, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "(x_train, _), (x_test, _) = fashion_mnist.load_data()\n",
    "\n",
    "x_train = x_train.astype('float32') / 255.\n",
    "x_test = x_test.astype('float32') / 255.\n",
    "\n",
    "print (x_train.shape)\n",
    "print (x_test.shape)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "id": "VEdCXSwCoKok"
   },
   "source": [
    "## First example: Basic autoencoder\n",
    "![Basic autoencoder results](https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/intro_autoencoder_result.png)\n",
    "\n",
    "Define an autoencoder with two Dense layers: an `encoder`, which compresses the images into a 64 dimensional latent vector, and a `decoder`, that reconstructs the original image from the latent space.\n",
    "\n",
    "To define your model, use the [Keras Model Subclassing API](https://www.tensorflow.org/guide/keras/custom_layers_and_models).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:12:43.746890Z",
     "iopub.status.busy": "2022-12-14T06:12:43.746271Z",
     "iopub.status.idle": "2022-12-14T06:12:47.174336Z",
     "shell.execute_reply": "2022-12-14T06:12:47.173619Z"
    },
    "id": "0MUxidpyChjX"
   },
   "outputs": [],
   "source": [
    "latent_dim = 64 \n",
    "\n",
    "class Autoencoder(Model):\n",
    "  def __init__(self, latent_dim):\n",
    "    super(Autoencoder, self).__init__()\n",
    "    self.latent_dim = latent_dim   \n",
    "    self.encoder = tf.keras.Sequential([\n",
    "      layers.Flatten(),\n",
    "      layers.Dense(latent_dim, activation='relu'),\n",
    "    ])\n",
    "    self.decoder = tf.keras.Sequential([\n",
    "      layers.Dense(784, activation='sigmoid'),\n",
    "      layers.Reshape((28, 28))\n",
    "    ])\n",
    "\n",
    "  def call(self, x):\n",
    "    encoded = self.encoder(x)\n",
    "    decoded = self.decoder(encoded)\n",
    "    return decoded\n",
    "  \n",
    "autoencoder = Autoencoder(latent_dim) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:12:47.178636Z",
     "iopub.status.busy": "2022-12-14T06:12:47.178075Z",
     "iopub.status.idle": "2022-12-14T06:12:47.194568Z",
     "shell.execute_reply": "2022-12-14T06:12:47.193991Z"
    },
    "id": "9I1JlqEIDCI4"
   },
   "outputs": [],
   "source": [
    "autoencoder.compile(optimizer='adam', loss=losses.MeanSquaredError())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7oJSeMTroABs"
   },
   "source": [
    "Train the model using `x_train` as both the input and the target. The `encoder` will learn to compress the dataset from 784 dimensions to the latent space, and the `decoder` will learn to reconstruct the original images.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:12:47.198144Z",
     "iopub.status.busy": "2022-12-14T06:12:47.197587Z",
     "iopub.status.idle": "2022-12-14T06:13:28.605951Z",
     "shell.execute_reply": "2022-12-14T06:13:28.605167Z"
    },
    "id": "h1RI9OfHDBsK"
   },
   "outputs": [],
   "source": [
    "# autoencoder.fit(x_train, x_train,\n",
    "#                 epochs=10,\n",
    "#                 shuffle=True,\n",
    "#                 validation_data=(x_test, x_test))\n",
    "# autoencoder.save('basic_autoencoder_model', save_format='tf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To save time, just load the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder = tf.keras.models.load_model('./tmp/autoencoder/basic_autoencoder_model')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wAM1QBhtoC-n"
   },
   "source": [
    "Now that the model is trained, let's test it by encoding and decoding images from the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:28.609455Z",
     "iopub.status.busy": "2022-12-14T06:13:28.609161Z",
     "iopub.status.idle": "2022-12-14T06:13:28.709188Z",
     "shell.execute_reply": "2022-12-14T06:13:28.708405Z"
    },
    "id": "Pbr5WCj7FQUi"
   },
   "outputs": [],
   "source": [
    "encoded_imgs = autoencoder.encoder(x_test).numpy()\n",
    "decoded_imgs = autoencoder.decoder(encoded_imgs).numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:28.712776Z",
     "iopub.status.busy": "2022-12-14T06:13:28.712545Z",
     "iopub.status.idle": "2022-12-14T06:13:29.351321Z",
     "shell.execute_reply": "2022-12-14T06:13:29.350685Z"
    },
    "id": "s4LlDOS6FUA1"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "  # display original\n",
    "  ax = plt.subplot(2, n, i + 1)\n",
    "  plt.imshow(x_test[i])\n",
    "  plt.title(\"original\")\n",
    "  plt.gray()\n",
    "  ax.get_xaxis().set_visible(False)\n",
    "  ax.get_yaxis().set_visible(False)\n",
    "\n",
    "  # display reconstruction\n",
    "  ax = plt.subplot(2, n, i + 1 + n)\n",
    "  plt.imshow(decoded_imgs[i])\n",
    "  plt.title(\"reconstructed\")\n",
    "  plt.gray()\n",
    "  ax.get_xaxis().set_visible(False)\n",
    "  ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "id": "r4gv6G8PoRQE"
   },
   "source": [
    "## Second example: Image denoising\n",
    "\n",
    "\n",
    "![Image denoising results](https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/image_denoise_fmnist_results.png)\n",
    "\n",
    "An autoencoder can also be trained to remove noise from images. In the following section, you will create a noisy version of the Fashion MNIST dataset by applying random noise to each image. You will then train an autoencoder using the noisy image as input, and the original image as the target.\n",
    "\n",
    "Let's reimport the dataset to omit the modifications made earlier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:29.355365Z",
     "iopub.status.busy": "2022-12-14T06:13:29.355077Z",
     "iopub.status.idle": "2022-12-14T06:13:29.712231Z",
     "shell.execute_reply": "2022-12-14T06:13:29.711398Z"
    },
    "id": "gDYHJA2PCQ3m"
   },
   "outputs": [],
   "source": [
    "(x_train, _), (x_test, _) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:29.716601Z",
     "iopub.status.busy": "2022-12-14T06:13:29.715941Z",
     "iopub.status.idle": "2022-12-14T06:13:29.805548Z",
     "shell.execute_reply": "2022-12-14T06:13:29.804779Z"
    },
    "id": "uJZ-TcaqDBr5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n"
     ]
    }
   ],
   "source": [
    "x_train = x_train.astype('float32') / 255.\n",
    "x_test = x_test.astype('float32') / 255.\n",
    "\n",
    "x_train = x_train[..., tf.newaxis]\n",
    "x_test = x_test[..., tf.newaxis]\n",
    "\n",
    "print(x_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aPZl_6P65_8R"
   },
   "source": [
    "Adding random noise to the images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:29.809481Z",
     "iopub.status.busy": "2022-12-14T06:13:29.808752Z",
     "iopub.status.idle": "2022-12-14T06:13:30.037924Z",
     "shell.execute_reply": "2022-12-14T06:13:30.037145Z"
    },
    "id": "axSMyxC354fc"
   },
   "outputs": [],
   "source": [
    "noise_factor = 0.2\n",
    "x_train_noisy = x_train + noise_factor * tf.random.normal(shape=x_train.shape) \n",
    "x_test_noisy = x_test + noise_factor * tf.random.normal(shape=x_test.shape) \n",
    "\n",
    "x_train_noisy = tf.clip_by_value(x_train_noisy, clip_value_min=0., clip_value_max=1.)\n",
    "x_test_noisy = tf.clip_by_value(x_test_noisy, clip_value_min=0., clip_value_max=1.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wRxHe4XXltNd"
   },
   "source": [
    "Plot the noisy images.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:30.042035Z",
     "iopub.status.busy": "2022-12-14T06:13:30.041417Z",
     "iopub.status.idle": "2022-12-14T06:13:30.754051Z",
     "shell.execute_reply": "2022-12-14T06:13:30.753330Z"
    },
    "id": "thKUmbVVCQpt"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x200 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "plt.figure(figsize=(20, 2))\n",
    "for i in range(n):\n",
    "    ax = plt.subplot(1, n, i + 1)\n",
    "    plt.title(\"original + noise\")\n",
    "    plt.imshow(tf.squeeze(x_test_noisy[i]))\n",
    "    plt.gray()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Sy9SY8jGl5aP"
   },
   "source": [
    "### Define a convolutional autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vT_BhZngWMwp"
   },
   "source": [
    "In this example, you will train a convolutional autoencoder using  [Conv2D](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Conv2D) layers in the `encoder`, and [Conv2DTranspose](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Conv2DTranspose) layers in the `decoder`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:30.758019Z",
     "iopub.status.busy": "2022-12-14T06:13:30.757779Z",
     "iopub.status.idle": "2022-12-14T06:13:30.801068Z",
     "shell.execute_reply": "2022-12-14T06:13:30.800464Z"
    },
    "id": "R5KjoIlYCQko"
   },
   "outputs": [],
   "source": [
    "class Denoise(Model):\n",
    "  def __init__(self):\n",
    "    super(Denoise, self).__init__()\n",
    "    self.encoder = tf.keras.Sequential([\n",
    "      layers.Input(shape=(28, 28, 1)),\n",
    "      layers.Conv2D(16, (3, 3), activation='relu', padding='same', strides=2),\n",
    "      layers.Conv2D(8, (3, 3), activation='relu', padding='same', strides=2)])\n",
    "\n",
    "    self.decoder = tf.keras.Sequential([\n",
    "      layers.Conv2DTranspose(8, kernel_size=3, strides=2, activation='relu', padding='same'),\n",
    "      layers.Conv2DTranspose(16, kernel_size=3, strides=2, activation='relu', padding='same'),\n",
    "      layers.Conv2D(1, kernel_size=(3, 3), activation='sigmoid', padding='same')])\n",
    "\n",
    "  def call(self, x):\n",
    "    encoded = self.encoder(x)\n",
    "    decoded = self.decoder(encoded)\n",
    "    return decoded\n",
    "\n",
    "autoencoder = Denoise()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:30.804670Z",
     "iopub.status.busy": "2022-12-14T06:13:30.804089Z",
     "iopub.status.idle": "2022-12-14T06:13:30.812259Z",
     "shell.execute_reply": "2022-12-14T06:13:30.811706Z"
    },
    "id": "QYKbiDFYCQfj"
   },
   "outputs": [],
   "source": [
    "autoencoder.compile(optimizer='adam', loss=losses.MeanSquaredError())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:13:30.815294Z",
     "iopub.status.busy": "2022-12-14T06:13:30.814943Z",
     "iopub.status.idle": "2022-12-14T06:14:49.408243Z",
     "shell.execute_reply": "2022-12-14T06:14:49.407505Z"
    },
    "id": "IssFr1BNCQX3"
   },
   "outputs": [],
   "source": [
    "# autoencoder.fit(x_train_noisy, x_train,\n",
    "#                 epochs=10,\n",
    "#                 shuffle=True,\n",
    "#                 validation_data=(x_test_noisy, x_test))\n",
    "\n",
    "# autoencoder.save('convolutional_autoencoder_model', save_format='tf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder = tf.keras.models.load_model('./tmp/autoencoder/convolutional_autoencoder_model')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "G85xUVBGTAKp"
   },
   "source": [
    "Let's take a look at a summary of the encoder. Notice how the images are downsampled from 28x28 to 7x7."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:49.411804Z",
     "iopub.status.busy": "2022-12-14T06:14:49.411532Z",
     "iopub.status.idle": "2022-12-14T06:14:49.421768Z",
     "shell.execute_reply": "2022-12-14T06:14:49.421061Z"
    },
    "id": "oEpxlX6sTEQz"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 14, 14, 16)        160       \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 7, 7, 8)           1160      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1320 (5.16 KB)\n",
      "Trainable params: 1320 (5.16 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder.encoder.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DDZBfMx1UtXx"
   },
   "source": [
    "The decoder upsamples the images back from 7x7 to 28x28."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:49.428121Z",
     "iopub.status.busy": "2022-12-14T06:14:49.427521Z",
     "iopub.status.idle": "2022-12-14T06:14:49.439404Z",
     "shell.execute_reply": "2022-12-14T06:14:49.438852Z"
    },
    "id": "pbeQtYMaUpro"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_transpose (Conv2DTr  (None, 14, 14, 8)         584       \n",
      " anspose)                                                        \n",
      "                                                                 \n",
      " conv2d_transpose_1 (Conv2D  (None, 28, 28, 16)        1168      \n",
      " Transpose)                                                      \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 28, 28, 1)         145       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1897 (7.41 KB)\n",
      "Trainable params: 1897 (7.41 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder.decoder.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A7-VAuEy_N6M"
   },
   "source": [
    "Plotting both the noisy images and the denoised images produced by the autoencoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:49.446418Z",
     "iopub.status.busy": "2022-12-14T06:14:49.445949Z",
     "iopub.status.idle": "2022-12-14T06:14:49.760335Z",
     "shell.execute_reply": "2022-12-14T06:14:49.759624Z"
    },
    "id": "t5IyPi1fCQQz"
   },
   "outputs": [],
   "source": [
    "encoded_imgs = autoencoder.encoder(x_test_noisy).numpy()\n",
    "decoded_imgs = autoencoder.decoder(encoded_imgs).numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:49.764066Z",
     "iopub.status.busy": "2022-12-14T06:14:49.763833Z",
     "iopub.status.idle": "2022-12-14T06:14:50.490419Z",
     "shell.execute_reply": "2022-12-14T06:14:50.489657Z"
    },
    "id": "sfxr9NdBCP_x"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x400 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "\n",
    "    # display original + noise\n",
    "    ax = plt.subplot(2, n, i + 1)\n",
    "    plt.title(\"original + noise\")\n",
    "    plt.imshow(tf.squeeze(x_test_noisy[i]))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display reconstruction\n",
    "    bx = plt.subplot(2, n, i + n + 1)\n",
    "    plt.title(\"reconstructed\")\n",
    "    plt.imshow(tf.squeeze(decoded_imgs[i]))\n",
    "    plt.gray()\n",
    "    bx.get_xaxis().set_visible(False)\n",
    "    bx.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ErGrTnWHoUYl"
   },
   "source": [
    "## Third example: Anomaly detection\n",
    "\n",
    "## Overview\n",
    "\n",
    "\n",
    "In this example, you will train an autoencoder to detect anomalies on the [ECG5000 dataset](http://www.timeseriesclassification.com/description.php?Dataset=ECG5000). This dataset contains 5,000 [Electrocardiograms](https://en.wikipedia.org/wiki/Electrocardiography), each with 140 data points. You will use a simplified version of the dataset, where each example has been labeled either `0` (corresponding to an abnormal rhythm), or `1` (corresponding to a normal rhythm). You are interested in identifying the abnormal rhythms.\n",
    "\n",
    "Note: This is a labeled dataset, so you could phrase this as a supervised learning problem. The goal of this example is to illustrate anomaly detection concepts you can apply to larger datasets, where you do not have labels available (for example, if you had many thousands of normal rhythms, and only a small number of abnormal rhythms).\n",
    "\n",
    "How will you detect anomalies using an autoencoder? Recall that an autoencoder is trained to minimize reconstruction error. You will train an autoencoder on the normal rhythms only, then use it to reconstruct all the data. Our hypothesis is that the abnormal rhythms will have higher reconstruction error. You will then classify a rhythm as an anomaly if the reconstruction error surpasses a fixed threshold."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "i5estNaur_Mh"
   },
   "source": [
    "### Load ECG data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "y35nsXLPsDNX"
   },
   "source": [
    "The dataset you will use is based on one from [timeseriesclassification.com](http://www.timeseriesclassification.com/description.php?Dataset=ECG5000).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "datasets_url = \"https://static-1300131294.cos.ap-shanghai.myqcloud.com/data/deep-learning/autoencoder/autoencoder/autoencoder_ecg.zip\"\n",
    "\n",
    "notebook_path = os.getcwd()\n",
    "\n",
    "tmp_folder_path = os.path.join(notebook_path, \"tmp\")\n",
    "\n",
    "if not os.path.exists(tmp_folder_path):\n",
    "    os.makedirs(tmp_folder_path)\n",
    "    \n",
    "file_path = os.path.join(tmp_folder_path,\"autoencoder\")\n",
    "\n",
    "if not os.path.exists(file_path):\n",
    "    os.makedirs(file_path)\n",
    "\n",
    "zip_store_path = os.path.join(file_path, \"zip-store\")\n",
    "\n",
    "if not os.path.exists(zip_store_path):\n",
    "    os.makedirs(zip_store_path)\n",
    "    \n",
    "datasets_response = requests.get(datasets_url)\n",
    "\n",
    "datasets_name = os.path.basename(datasets_url)\n",
    "\n",
    "datasets_save_path = os.path.join(zip_store_path, datasets_name)\n",
    "\n",
    "with open(datasets_save_path, \"wb\") as file:\n",
    "    file.write(datasets_response.content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "zip_file_path = f\"./tmp/autoencoder/zip-store/{datasets_name}\"\n",
    "extract_path = \"./tmp/autoencoder\"\n",
    "\n",
    "zip_ref = zipfile.ZipFile(zip_file_path, 'r')\n",
    "zip_ref.extractall(extract_path)\n",
    "zip_ref.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>131</th>\n",
       "      <th>132</th>\n",
       "      <th>133</th>\n",
       "      <th>134</th>\n",
       "      <th>135</th>\n",
       "      <th>136</th>\n",
       "      <th>137</th>\n",
       "      <th>138</th>\n",
       "      <th>139</th>\n",
       "      <th>140</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.112522</td>\n",
       "      <td>-2.827204</td>\n",
       "      <td>-3.773897</td>\n",
       "      <td>-4.349751</td>\n",
       "      <td>-4.376041</td>\n",
       "      <td>-3.474986</td>\n",
       "      <td>-2.181408</td>\n",
       "      <td>-1.818286</td>\n",
       "      <td>-1.250522</td>\n",
       "      <td>-0.477492</td>\n",
       "      <td>...</td>\n",
       "      <td>0.792168</td>\n",
       "      <td>0.933541</td>\n",
       "      <td>0.796958</td>\n",
       "      <td>0.578621</td>\n",
       "      <td>0.257740</td>\n",
       "      <td>0.228077</td>\n",
       "      <td>0.123431</td>\n",
       "      <td>0.925286</td>\n",
       "      <td>0.193137</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.100878</td>\n",
       "      <td>-3.996840</td>\n",
       "      <td>-4.285843</td>\n",
       "      <td>-4.506579</td>\n",
       "      <td>-4.022377</td>\n",
       "      <td>-3.234368</td>\n",
       "      <td>-1.566126</td>\n",
       "      <td>-0.992258</td>\n",
       "      <td>-0.754680</td>\n",
       "      <td>0.042321</td>\n",
       "      <td>...</td>\n",
       "      <td>0.538356</td>\n",
       "      <td>0.656881</td>\n",
       "      <td>0.787490</td>\n",
       "      <td>0.724046</td>\n",
       "      <td>0.555784</td>\n",
       "      <td>0.476333</td>\n",
       "      <td>0.773820</td>\n",
       "      <td>1.119621</td>\n",
       "      <td>-1.436250</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.567088</td>\n",
       "      <td>-2.593450</td>\n",
       "      <td>-3.874230</td>\n",
       "      <td>-4.584095</td>\n",
       "      <td>-4.187449</td>\n",
       "      <td>-3.151462</td>\n",
       "      <td>-1.742940</td>\n",
       "      <td>-1.490659</td>\n",
       "      <td>-1.183580</td>\n",
       "      <td>-0.394229</td>\n",
       "      <td>...</td>\n",
       "      <td>0.886073</td>\n",
       "      <td>0.531452</td>\n",
       "      <td>0.311377</td>\n",
       "      <td>-0.021919</td>\n",
       "      <td>-0.713683</td>\n",
       "      <td>-0.532197</td>\n",
       "      <td>0.321097</td>\n",
       "      <td>0.904227</td>\n",
       "      <td>-0.421797</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.490473</td>\n",
       "      <td>-1.914407</td>\n",
       "      <td>-3.616364</td>\n",
       "      <td>-4.318823</td>\n",
       "      <td>-4.268016</td>\n",
       "      <td>-3.881110</td>\n",
       "      <td>-2.993280</td>\n",
       "      <td>-1.671131</td>\n",
       "      <td>-1.333884</td>\n",
       "      <td>-0.965629</td>\n",
       "      <td>...</td>\n",
       "      <td>0.350816</td>\n",
       "      <td>0.499111</td>\n",
       "      <td>0.600345</td>\n",
       "      <td>0.842069</td>\n",
       "      <td>0.952074</td>\n",
       "      <td>0.990133</td>\n",
       "      <td>1.086798</td>\n",
       "      <td>1.403011</td>\n",
       "      <td>-0.383564</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.800232</td>\n",
       "      <td>-0.874252</td>\n",
       "      <td>-2.384761</td>\n",
       "      <td>-3.973292</td>\n",
       "      <td>-4.338224</td>\n",
       "      <td>-3.802422</td>\n",
       "      <td>-2.534510</td>\n",
       "      <td>-1.783423</td>\n",
       "      <td>-1.594450</td>\n",
       "      <td>-0.753199</td>\n",
       "      <td>...</td>\n",
       "      <td>1.148884</td>\n",
       "      <td>0.958434</td>\n",
       "      <td>1.059025</td>\n",
       "      <td>1.371682</td>\n",
       "      <td>1.277392</td>\n",
       "      <td>0.960304</td>\n",
       "      <td>0.971020</td>\n",
       "      <td>1.614392</td>\n",
       "      <td>1.421456</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 141 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0         1         2         3         4         5         6    \\\n",
       "0 -0.112522 -2.827204 -3.773897 -4.349751 -4.376041 -3.474986 -2.181408   \n",
       "1 -1.100878 -3.996840 -4.285843 -4.506579 -4.022377 -3.234368 -1.566126   \n",
       "2 -0.567088 -2.593450 -3.874230 -4.584095 -4.187449 -3.151462 -1.742940   \n",
       "3  0.490473 -1.914407 -3.616364 -4.318823 -4.268016 -3.881110 -2.993280   \n",
       "4  0.800232 -0.874252 -2.384761 -3.973292 -4.338224 -3.802422 -2.534510   \n",
       "\n",
       "        7         8         9    ...       131       132       133       134  \\\n",
       "0 -1.818286 -1.250522 -0.477492  ...  0.792168  0.933541  0.796958  0.578621   \n",
       "1 -0.992258 -0.754680  0.042321  ...  0.538356  0.656881  0.787490  0.724046   \n",
       "2 -1.490659 -1.183580 -0.394229  ...  0.886073  0.531452  0.311377 -0.021919   \n",
       "3 -1.671131 -1.333884 -0.965629  ...  0.350816  0.499111  0.600345  0.842069   \n",
       "4 -1.783423 -1.594450 -0.753199  ...  1.148884  0.958434  1.059025  1.371682   \n",
       "\n",
       "        135       136       137       138       139  140  \n",
       "0  0.257740  0.228077  0.123431  0.925286  0.193137  1.0  \n",
       "1  0.555784  0.476333  0.773820  1.119621 -1.436250  1.0  \n",
       "2 -0.713683 -0.532197  0.321097  0.904227 -0.421797  1.0  \n",
       "3  0.952074  0.990133  1.086798  1.403011 -0.383564  1.0  \n",
       "4  1.277392  0.960304  0.971020  1.614392  1.421456  1.0  \n",
       "\n",
       "[5 rows x 141 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe = pd.read_csv(\"./tmp/autoencoder/autoencoder_ecg.csv\", header=None)\n",
    "raw_data = dataframe.values\n",
    "dataframe.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:50.721788Z",
     "iopub.status.busy": "2022-12-14T06:14:50.721160Z",
     "iopub.status.idle": "2022-12-14T06:14:50.728196Z",
     "shell.execute_reply": "2022-12-14T06:14:50.727587Z"
    },
    "id": "UmuCPVYKsKKx"
   },
   "outputs": [],
   "source": [
    "# The last element contains the labels\n",
    "labels = raw_data[:, -1]\n",
    "\n",
    "# The other data points are the electrocadriogram data\n",
    "data = raw_data[:, 0:-1]\n",
    "\n",
    "train_data, test_data, train_labels, test_labels = train_test_split(\n",
    "    data, labels, test_size=0.2, random_state=21\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "byK2vP7hsMbz"
   },
   "source": [
    "Normalize the data to `[0,1]`.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:50.731859Z",
     "iopub.status.busy": "2022-12-14T06:14:50.731401Z",
     "iopub.status.idle": "2022-12-14T06:14:50.759412Z",
     "shell.execute_reply": "2022-12-14T06:14:50.758824Z"
    },
    "id": "tgMZVWRKsPx6"
   },
   "outputs": [],
   "source": [
    "min_val = tf.reduce_min(train_data)\n",
    "max_val = tf.reduce_max(train_data)\n",
    "\n",
    "train_data = (train_data - min_val) / (max_val - min_val)\n",
    "test_data = (test_data - min_val) / (max_val - min_val)\n",
    "\n",
    "train_data = tf.cast(train_data, tf.float32)\n",
    "test_data = tf.cast(test_data, tf.float32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BdSYr2IPsTiz"
   },
   "source": [
    "You will train the autoencoder using only the normal rhythms, which are labeled in this dataset as `1`. Separate the normal rhythms from the abnormal rhythms."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:50.762928Z",
     "iopub.status.busy": "2022-12-14T06:14:50.762445Z",
     "iopub.status.idle": "2022-12-14T06:14:50.795166Z",
     "shell.execute_reply": "2022-12-14T06:14:50.794591Z"
    },
    "id": "VvK4NRe8sVhE"
   },
   "outputs": [],
   "source": [
    "train_labels = train_labels.astype(bool)\n",
    "test_labels = test_labels.astype(bool)\n",
    "\n",
    "normal_train_data = train_data[train_labels]\n",
    "normal_test_data = test_data[test_labels]\n",
    "\n",
    "anomalous_train_data = train_data[~train_labels]\n",
    "anomalous_test_data = test_data[~test_labels]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wVcTBDo-CqFS"
   },
   "source": [
    "Plot a normal ECG. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:50.798929Z",
     "iopub.status.busy": "2022-12-14T06:14:50.798330Z",
     "iopub.status.idle": "2022-12-14T06:14:50.929609Z",
     "shell.execute_reply": "2022-12-14T06:14:50.928740Z"
    },
    "id": "ZTlMIrpmseYe"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.grid()\n",
    "plt.plot(np.arange(140), normal_train_data[0])\n",
    "plt.title(\"A Normal ECG\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QpI9by2ZA0NN"
   },
   "source": [
    "Plot an anomalous ECG."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:50.932822Z",
     "iopub.status.busy": "2022-12-14T06:14:50.932544Z",
     "iopub.status.idle": "2022-12-14T06:14:51.049383Z",
     "shell.execute_reply": "2022-12-14T06:14:51.048565Z"
    },
    "id": "zrpXREF2siBr"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.grid()\n",
    "plt.plot(np.arange(140), anomalous_train_data[0])\n",
    "plt.title(\"An Anomalous ECG\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0DS6QKZJslZz"
   },
   "source": [
    "### Build the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:51.053160Z",
     "iopub.status.busy": "2022-12-14T06:14:51.052452Z",
     "iopub.status.idle": "2022-12-14T06:14:51.068520Z",
     "shell.execute_reply": "2022-12-14T06:14:51.067812Z"
    },
    "id": "bf6owZQDsp9y"
   },
   "outputs": [],
   "source": [
    "class AnomalyDetector(Model):\n",
    "  def __init__(self):\n",
    "    super(AnomalyDetector, self).__init__()\n",
    "    self.encoder = tf.keras.Sequential([\n",
    "      layers.Dense(32, activation=\"relu\"),\n",
    "      layers.Dense(16, activation=\"relu\"),\n",
    "      layers.Dense(8, activation=\"relu\")])\n",
    "    \n",
    "    self.decoder = tf.keras.Sequential([\n",
    "      layers.Dense(16, activation=\"relu\"),\n",
    "      layers.Dense(32, activation=\"relu\"),\n",
    "      layers.Dense(140, activation=\"sigmoid\")])\n",
    "    \n",
    "  def call(self, x):\n",
    "    encoded = self.encoder(x)\n",
    "    decoded = self.decoder(encoded)\n",
    "    return decoded\n",
    "\n",
    "autoencoder = AnomalyDetector()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:51.072186Z",
     "iopub.status.busy": "2022-12-14T06:14:51.071554Z",
     "iopub.status.idle": "2022-12-14T06:14:51.079644Z",
     "shell.execute_reply": "2022-12-14T06:14:51.079100Z"
    },
    "id": "gwRpBBbg463S"
   },
   "outputs": [],
   "source": [
    "autoencoder.compile(optimizer='adam', loss='mae')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zuTy60STBEy4"
   },
   "source": [
    "Notice that the autoencoder is trained using only the normal ECGs, but is evaluated using the full test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:51.082689Z",
     "iopub.status.busy": "2022-12-14T06:14:51.082465Z",
     "iopub.status.idle": "2022-12-14T06:14:54.154111Z",
     "shell.execute_reply": "2022-12-14T06:14:54.153134Z"
    },
    "id": "V6NFSs-jsty2"
   },
   "outputs": [],
   "source": [
    "# history = autoencoder.fit(normal_train_data, normal_train_data, \n",
    "#           epochs=20, \n",
    "#           batch_size=512,\n",
    "#           validation_data=(test_data, test_data),\n",
    "#           shuffle=True)\n",
    "\n",
    "# autoencoder.save('ecg_autoencoder_model', save_format='tf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder = tf.keras.models.load_model('./tmp/autoencoder/ecg_autoencoder_model')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ceI5lKv1BT-A"
   },
   "source": [
    "You will soon classify an ECG as anomalous if the reconstruction error is greater than one standard deviation from the normal training examples. First, let's plot a normal ECG from the training set, the reconstruction after it's encoded and decoded by the autoencoder, and the reconstruction error."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:54.320314Z",
     "iopub.status.busy": "2022-12-14T06:14:54.319653Z",
     "iopub.status.idle": "2022-12-14T06:14:54.462620Z",
     "shell.execute_reply": "2022-12-14T06:14:54.461789Z"
    },
    "id": "hmsk4DuktxJ2"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGdCAYAAADAAnMpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAACNJElEQVR4nOzdd3xT9f7H8dfJ7t7QQaHsvQRkiegVxT2voqICjuvChSiOq3hduPXnwnEv7r0XogiyFJkiyCh7U0op3W3WOb8/TpM2tEBHRpt+no9HH2mTk+SbQ2ne+Y7PV9E0TUMIIYQQIkQMoW6AEEIIIVo2CSNCCCGECCkJI0IIIYQIKQkjQgghhAgpCSNCCCGECCkJI0IIIYQIKQkjQgghhAgpCSNCCCGECClTqBtQF6qqsnfvXmJiYlAUJdTNEUIIIUQdaJpGcXEx6enpGAxH7v9oFmFk7969ZGZmhroZQgghhGiAXbt20aZNmyPe3izCSExMDKC/mNjY2BC3RgghhBB1UVRURGZmpvd9/EiaRRjxDM3ExsZKGBFCCCGamWNNsZAJrEIIIYQIKQkjQgghhAgpCSNCCCGECKlmMWekLtxuN06nM9TNEGHIbDZjNBpD3QwhhAhbYRFGSkpK2L17N5qmhbopIgwpikKbNm2Ijo4OdVOEECIsNfsw4na72b17N5GRkaSkpEhRNOFXmqZx4MABdu/eTefOnaWHRAghAqDZhxGn04mmaaSkpBARERHq5ogwlJKSwvbt23E6nRJGhBAiAMJmAqv0iIhAkd8tIYQIrLAJI0IIIYRoniSMCCGEECKkJIyEyPjx4zn//POD+pxvv/028fHxQX1OIYQQ4lgkjAghhBAipCSMNAEnnXQSt956K3fffTeJiYmkpqby0EMP+RyjKArTp0/njDPOICIigg4dOvD55597b583bx6KolBQUOC9btWqVSiKwvbt25k3bx4TJkygsLAQRVFQFKXGcwghRG3mz4eXXwapKykCJezCiKZBaWlovhpTc+2dd94hKiqKJUuW8NRTT/Hwww8ze/Zsn2MeeOABLrroIv766y/Gjh3LpZdeyvr16+v0+MOGDeOFF14gNjaWffv2sW/fPiZPntzwBgshWoyxYzVuuQVOPbGCvLxQt0aEo2ZfZ+RwZWUQqkKZJSUQFdWw+/bp04epU6cC0LlzZ15++WXmzJnDqaee6j3m4osv5tprrwXgkUceYfbs2bz00ku8+uqrx3x8i8VCXFwciqKQmprasEYKIVqcQ4dgzx59efv8P2wM7Gvnu1lWevcOccNEWAm7npHmqk+fPj4/p6WlkZub63Pd0KFDa/xc154RIYRoiOxs/TIhooyshHx27LUy9HgXK5e7Q9swEVbCrmckMlLvoQjVczeU2Wz2+VlRFFRVrfP9DQY9V1bfn0c2DhRCNJYnjPRO3c/bF3/G2I/HsHhnO566ay8f/5oe2saJsNGgnpFXXnmFrKwsbDYbgwcPZunSpUc9vqCggJtvvpm0tDSsVitdunRh5syZDWrwsSiKPlQSiq9AF+r8448/avzcvXt3QC9ZDrBv3z7v7atWrfI53mKx4HbLpxkhRN1t2KBfdk4+SGJkOY+P/gmArxe1Ij9XPvAI/6h3GPnkk0+YNGkSU6dOZeXKlfTt25fRo0fXGFLwcDgcnHrqqWzfvp3PP/+c7Oxs3nzzTTIyMhrd+Jbms88+Y8aMGWzcuJGpU6eydOlSJk6cCECnTp3IzMzkoYceYtOmTfzwww88++yzPvfPysqipKSEOXPmkJeXR1lZWShehhCiGfH0jHROOgBAv/R99Gi1H7vLxAcv1v53X4j6qncYee6557juuuuYMGECPXr04LXXXiMyMpIZM2bUevyMGTPIz8/n66+/Zvjw4WRlZTFy5Ej69u3b6Ma3NP/5z3/4+OOP6dOnD++++y4fffQRPXr0APRhno8++ogNGzbQp08fnnzySR599FGf+w8bNowbbriBMWPGkJKSwlNPPRWKlyGEaEaq94yA3gM8tv8qAN750BqiVolwo2ha3RekOhwOIiMj+fzzz32qh44bN46CggK++eabGvc588wzSUxMJDIykm+++YaUlBQuv/xypkyZcsQdUO12O3a73ftzUVERmZmZFBYWEhsb63NsRUUF27Zto3379thstrq+lGZHURS++uqroFdtFS3nd0yIw7lcEBmp4XQqrL79BdrGFwKQVxpJt2cn4VKN/PV7MX2GxoS4paKpKioqIi4urtb37+rq1TOSl5eH2+2mdevWPte3bt2anJycWu+zdetWPv/8c9xuNzNnzuSBBx7g2WefrfGpvbpp06YRFxfn/crMzKxPM4UQQvjBtm3gdCpEmJy0iS30Xp8cVcbpXTYC8L8XCkLUOhFOAr60V1VVWrVqxRtvvMGAAQMYM2YM999/P6+99toR73PvvfdSWFjo/dq1a1egmymEEOIwnvkiHZMOYjjs3cIzVPPhzCQcjkZUfBSCei7tTU5Oxmg0sn//fp/r9+/ff8RCWmlpaZjNZp8hme7du5OTk4PD4cBisdS4j9VqxWqVscjq6jGaJoQQfuGdvJpysMZtp3baRKuoEnJLovnurf1cdH3rGscIUVf16hmxWCwMGDCAOXPmeK9TVZU5c+bUKMjlMXz4cDZv3uxTM2Pjxo2kpaXVGkSEEEI0DZ7Jq52SaoYRk1Hj0r5/AfDuW65gNkuEoXoP00yaNIk333yTd955h/Xr13PjjTdSWlrKhAkTALjqqqu49957vcffeOON5Ofnc9ttt7Fx40Z++OEHHn/8cW6++Wb/vQohhBB+5+kZ6VK5rPdwF/X+G4DZK1tRWiSBRDRcvSuwjhkzhgMHDvDggw+Sk5NDv379mDVrlndS686dO73VQAEyMzP56aefuOOOO+jTpw8ZGRncdtttTJkyxX+vQgghhN9t2KABindZ7+H6pObQJq6A3YXx/PThXi68QSqyioZpUDn4iRMneottHW7evHk1rhs6dGiN6qFCCCGarvx8OHBALyvdKan2rXoVBc7sms0bSwfz9ecuLrwhmC0U4UQ2yhNCCFGDZ4gmPaaIaOuRy76f1U2fWDLzj2RcLploLxpGwogQQogaqlbS1N4r4jGs3U7ibeUcLI1k0ddSHl40jIQR0Wxt374dRVFqbAgohGi8w8vAH4nZqDK6sgDaVx9XBLpZIkxJGAmR8ePHoygKiqJgNptp3749d999NxUVzeM/87x581AUhYKCgqA83/jx42uUws/MzGTfvn306tUrKG0QoiWp2iDv6D0jUDVU8+28eKQkkmgICSMhdPrpp7Nv3z62bt3K888/z+uvv87UqVND3Sy/cjgcAXtso9FIamoqJlOD5mELIY7C2zNShzBySqct2ExOth+M46+FhwLcMhGOJIyEkNVqJTU1lczMTM4//3xGjRrF7NmzAb2Y3LRp02jfvj0RERH07duXzz//3Of+a9eu5eyzzyY2NpaYmBhGjBjBli1bvPd/+OGHadOmDVar1bsE28MzxPHll19y8sknExkZSd++fVm8eLH3mB07dnDOOeeQkJBAVFQUPXv2ZObMmWzfvp2TTz4ZgISEBBRFYfz48QCcdNJJTJw4kdtvv53k5GRGjx5d63BKQUEBiqL4rL460ut56KGHeOedd/jmm2+8vUnz5s2r9XHnz5/P8ccfj9VqJS0tjXvuuQeXq6r+wUknncStt97K3XffTWJiIqmpqTz00EON+WcUIuy4XLBli97F0ekYwzQAURYnJ3XYCsDX7xUFtG0iPIXfR0pNg7Ky0Dx3ZKS+1q0B/v77b37//XfatWsH6JsFvv/++7z22mt07tyZBQsWcMUVV5CSksLIkSPZs2cPJ554IieddBJz584lNjaW3377zfvG+3//9388++yzvP766/Tv358ZM2Zw7rnnsnbtWjp37ux93vvvv59nnnmGzp07c//993PZZZexefNmTCYTN998Mw6HgwULFhAVFcW6deuIjo4mMzOTL774gosuuojs7GxiY2OJiIjwPuY777zDjTfeyG+//Vbn13+01zN58mTWr19PUVERb731FgCJiYns3bu3xmOceeaZjB8/nnfffZcNGzZw3XXXYbPZfALHO++8w6RJk1iyZAmLFy9m/PjxDB8+nFNPPbXe/25ChKN16/QN8iLNDp8N8o7mrG4bmLWxK9/+HMlDgW2eCEPhF0bKyiA6OjTPXVICUVF1Pvz7778nOjoal8uF3W7HYDDw8ssvY7fbefzxx/nll1+8ZfY7dOjAokWLeP311xk5ciSvvPIKcXFxfPzxx5jNZgC6dOnifexnnnmGKVOmcOmllwLw5JNP8uuvv/LCCy/wyiuveI+bPHkyZ511FgD/+c9/6NmzJ5s3b6Zbt27s3LmTiy66iN69e3vb4JGYmAhAq1atiI+P93ldnTt35qmnnvL+vH379mOei2O9noiICOx2+xH3QAJ49dVXyczM5OWXX0ZRFLp168bevXuZMmUKDz74oLcYX58+fbzDYZ07d+bll19mzpw5EkaEqPTjj/rlCVk7amyQdySndt4MwF+7kijKLSO2VWSAWifCUfiFkWbk5JNPZvr06ZSWlvL8889jMpm46KKLWLt2LWVlZTXeHB0OB/379wdg1apVjBgxwvvGXV1RURF79+5l+PDhPtcPHz6cv/76y+e6Pn36eL9PS0sDIDc3l27dunHrrbdy44038vPPPzNq1Cguuugin+OPZMCAAXU7AdUc7fXU1fr16xk6dChKtd6p4cOHU1JSwu7du2nbti1AjdeQlpZGbq4sSRTCY+ZM/fLUzpvqfJ/UmBLSY4rYWxzLyrn5nHSphBFRd+EXRiIj9R6KUD13PURFRdGpUycAZsyYQd++ffnf//7nXR3yww8/kJGR4XMfz27G1YdFGqP6m7/nTdyzqeG1117L6NGj+eGHH/j555+ZNm0azz77LLfccssxX1d1nh6J6jsPO52+RZT89Xrq4vDAoyiKz0aOQrRkBQXw2296Gfj6hBGAful72Zsdy7Lf7Jx0aUCaJ8JU+E1gVRR9qCQUXw2cLwL6G/Z9993Hv//9b3r06IHVamXnzp106tTJ5yszMxPQP90vXLiwxps6QGxsLOnp6TXmbPz222/06NGjXu3KzMzkhhtu4Msvv+TOO+/kzTffBPDuuOx2u4/5GCkpKQDs27fPe93htUGO9no8z3es5+revTuLFy/2CT2//fYbMTExtGnT5pjtFELAzz+D263QNfkAWQkF9bpv/3R9HteKP40BaJkIZ+EXRpqxiy++GKPRyOuvv87kyZO54447eOedd9iyZQsrV67kpZde4p133gH0/YGKioq49NJLWb58OZs2beK9994ju7I4wF133cWTTz7JJ598QnZ2Nvfccw+rVq3itttuq3N7br/9dn766Se2bdvGypUr+fXXX+nevTsA7dq1Q1EUvv/+ew4cOEDJUXqjIiIiGDJkCE888QTr169n/vz5/Pvf//Y55livJysri9WrV5OdnU1eXl6toeWmm25i165d3HLLLWzYsIFvvvmGqVOnMmnSJJ/NG4UQR+YZojmta+29IqbNm4l66y2iZswg6u23iXzvPUyb9fkinjCycmNMUNoqwkf4DdM0YyaTiYkTJ/LUU0+xbds2UlJSmDZtGlu3biU+Pp7jjjuO++67D4CkpCTmzp3LXXfdxciRIzEajfTr1887T+TWW2+lsLCQO++8k9zcXHr06MG3337rs5LmWNxuNzfffDO7d+8mNjaW008/neeffx6AjIwM/vOf/3DPPfcwYcIErrrqKt5+++0jPtaMGTO45pprGDBgAF27duWpp57itNNO895+rNdz3XXXMW/ePAYOHEhJSQm//vorWVlZPs+RkZHBzJkzueuuu+jbty+JiYlcc801NYKPEKJ2qgo//lg5RNOpZhhRysqI+PJLDIetWDTu3UvJjTfSL13v/dx0IImCvSXEp4doMYFodhRNa/r18oqKioiLi6OwsJDY2Fif2yoqKti2bRvt27fHZrOFqIUinMnvmGgpli2D44+HGIudLXc/hcXkO5cq4uuvsaxahTslBdcNN6AYDJg+/BDDpk04O3ak7Ior6PXC7ewujGf2u7sYdWVmiF6JaCqO9v5dnfRdCyGEAKqGaE7utKVGEDFu3Yqlcq6Xc9IkrA8/jOWhhzB88w2axYJ5yxYsy5d7h2qWL7YHs+mimZMwIoQQAoAfftAvawzROJ1EfP89APYhQ7DcemvVbd27w7RpANh+/plT4pcBsGKVzAIQdSdhRAghBPv368M0UFXAzMO6YAHG/HzU6GiUZ57BcFgZA+X229FOOAHF6eTKjQ9jwc7KjXHBaroIAxJGhBBCMGeOftkndR+pMdVWxzmdWJcsAcA+fjzmYcNq3tlgQPngA7ToaBIPbuUVbmbrwXgO7iwOQstFOJAwIoQQgjVr9Mvj2+72ud60eTOKw4EaG4vlvvt8Khz7aNsWPv0UTVG4lv9xE6+yfG5BYBstwoaEESGEEKxbp192TT7gc7258gZn//4YjrI3FIByxhnw6KMAvMDt7P90nt/bKcKThBEhhBCsW6dXeeiaUi2MuFyYKwsPcuGFR+4VqUa5917W9rwYMy7O/fl2tEOHAtFcEWYkjAghRAtXXg5bt+rfd6sWRkxbtuhDNDExmP75z7o9mKJw4Km32Uhn4t35uF94wf8NFmFHwogQQrRwGzeCqiokRJSRElXqvd47RNOvH4bKXb3rot+wSP7HNfp93/vYv40VYSlsF4KrhYVoh5UsDiQlMhJDnCxlE0I0P575It1SDlTt9+lyYd6wQf++jkM0HvHxsKrrGMi+h4htG1E3bMDQrZtf2yzCS1iGEbWwkOKXXwaXK3hPajIRM3FinQPJ+PHjvZveVTd69GhmzZrl79YJIcQReSevtsrzXmfatg3FbkeNjq77EE01Z9+cxbxbR3IS83G/9DKGV172V3NFGArLYRqtrCy4QQTA5ap3T8zpp5/Ovn37fL4++uijWo+tbZdah8PRoKY29H5CiPBU20oazxCNq29fDBkZ9X7MK6+ET02XAVDx8beNb6QIa2EZRpoLq9VKamqqz1dCQgIAiqIwffp0zj33XKKionjsscd46KGH6NevH//97399Nm3buXMn5513HtHR0cTGxnLJJZewf/9+7/Mc6X5CCAG+wzQAuN2YPEM0F1xQryEaj/h44MJ/YsdCTP4u1N9+80tbRXiSMNKEPfTQQ1xwwQWsWbOGq6++GoDNmzfzxRdf8OWXX7Jq1SpUVeW8884jPz+f+fPnM3v2bLZu3cqYMWN8Huvw+wkhBIDDAZs2+S7rNW3bhqG8HDUyEuNhf0vq48rbk5jJmQCUPfdq4xsrwlZYzhlpLr7//nuio6N9rrvvvvu47777ALj88suZMGGCz+0Oh4N3332XlJQUAGbPns2aNWvYtm0bmZn6dt3vvvsuPXv2ZNmyZQwaNKjW+wkhBMCmTeB2K8RaK0iL0cu3Vx+iMTdgiMZjyBCY3OZSLtj9NerM2WhuN4rR6Jd2i/AiYSSETj75ZKZPn+5zXWJiovf7gQMH1rhPu3btfALF+vXryczM9AYRgB49ehAfH8/69eu9YeTw+wkhBFSbL+JZSVNtiEY777wGDdF4KAp0vvM8Cu+IJa7iAK6vvsb0z4v80GoRbmSYJoSioqLo1KmTz1f1MBIVFVXrfRr6XEIIcbjDV9IYd+zAUFaGGhHRqCEaj8sm2PjGeCEA+556t9GPJ8KThJFmrnv37uzatYtdu3Z5r1u3bh0FBQX06NEjhC0TQjQHh6+k8Q7R9OmDsV27Rj9+XBzsGaKHkbjVv6O53Y1+TBF+JIyEkN1uJycnx+crLy/v2HesZtSoUfTu3ZuxY8eycuVKli5dylVXXcXIkSNrHeYRQojqfFbSqCrm9esB0M45p1FDNNXZTziVImKIteeh/vijXx5ThBcJIyE0a9Ys0tLSfL5OOOGEej2Goih88803JCQkcOKJJzJq1Cg6dOjAJ598EqBWCyHChcsF2dlVK2mMO3diKC1FtdkwXnqp356nTScbP3AWANp77/ntcUX4UDRN00LdiGMpKioiLi6OwsJCYmNjfW6rqKhg27ZtPvUzmkMFVtF81PY7JkQ4yM6Gbt0gyuxg173TiJw1E+vSpTgGDsS8dKnfekZ++QVeO/VzPudi1ORkDLm54KfHFk3b0d6/qwvL1TSGuDhiJk6UvWmEEOIoPEM0XVIOYED1zhfRzjrLb0EEICsLZnE65diIyMvDPX8+xpNO8tvji+YvLMMI6IEECQdCCHFE3vkirfIw7tqFoaQEzWrFeNllfn2ezEwoU6L4SRvN+XyD9s47IGFEVCNzRoQQooWaPVu/7JGyH/Pq1QA4e/bE2LmzX5/HaoX0NI0v0VfVKD//7NfHF82fhBEhhGiBVqyA+fPBZHBzUbc/sfz9NwDaP/+JYvD/W0NWe4XvORtVMWLcuxd1+XK/P4doviSMCCHCWnY2/PVXqFvR9Dz7rH55Ya+1tNu3HMVuR42Lw3jllQF5vqwshUMksi2hLwDq//4XkOdpETRNr+NfURHqlvhN2ISRZrAoSDRT8rvVfO3bB4MGaQwYoLFssTPUzWkydu6ETz/Vf68nDv0dc+Xmmc4hQzA2Yi+ao8nK0i8XR48CQJkzJyDPE9by8uCFF6BPH+jSBa1HD1i6NNSt8otmH0aMlZsuORyOELdEhCvP75ZRNvhqdh58EIqLFdxuhQmXlSN/JnQvvqhvjndi+630jd6MacsWAJRx4/y6iqY6Txj52XAaAIYtW9DqWeSxRXv5ZbSMDLjjDqgcUlO2bUMbOhRt6tTglrIIgGa/msZkMhEZGcmBAwcwm80YAjDWKVouVVU5cOAAkZGRmEzN/r9Li7JmDcyYoQH6jrRrd8Ty6F2HePj/EkLdtJAqLIQ33tDPyy3DFmNevRpF03BlZmI6++yAPa8njCwt7YM7IQHjoUO4PvsM0403Buw5w8ZTT8GUKSiAOy0Nx3HH4ezUCducOfpcn4cfRv39dwyeGcnNULP/66ooCmlpaWzbto0dO3aEujkiDBkMBtq2bRuwT4wiMO66C1RV4dwe67igx1omfH4x016O5cIrnfQbaA5180Lmv//Ve4u6peQyquMmLLP1CTWuUaOwxcQE7Hk9YWTnoXhcfTtiXLEcfvwRJIwc3aOPwgMPAFAxciT2k07yFowrv+giXF26EPH11xh++QV1zRoMvXuHsLEN1+zDCIDFYqFz584yVCMCwmKxSI9bM/PTT/qX2ejmoVN+oX3iIb5cu47v1vdg/KUlLFsfh7kF5hFNg+nT9V6Rm4b+gTFnH8bcXDSjEcO11wb0uTMzQVE0yl1m8jO6k7ZiOcqSJQF9zmbviSeqgsgpp2AfMcL3dkXB2acPlj/+wLR3L+r330sYCTWDwSCluoUQuN16rwjAtccvo0PSIQCeOXMmi7Zn8deWOD6enseVtyaHsJWhsWEDbNmiYDW6uKjX31h+WgGAs0cPzIMGBfS5PbVG9uxV2Bw/gFTlfYy5uahr12Lo2TOgz90sZWejPfAAClB+6qk4hg8/4qHuzExMe/eiLVoUvPb5mXzcE0KEldde0+eLxNvKuXvEfO/1rWNKuWagXtviy4/CZ0lkffzwg355QvvtxBTnYPnzTwC0yy9HCUJXUVZ7fXhha3k67spVO+qnnwb8eZulSZNQXC6cXbocNYgAuNq2BcDQjNewSxgRQoSNnBy4/359yer9//iVhEjf0HFWtw0A/LIyhfJSd9DbF2qeMHJa501Yf/0VRVVxdu6MOUjzNrKy9DCysyAeV4cO+pW//BKU525WZs6EmTPRjEYqTjvtmIe7MzMBMOzdi5qbG+jWBYSEESFE2LjrLigsVOiXtperB9as8NkvfR/pMUWUOKzM/rh5/tFuqMJCWLRID2rnx+mrMDRFwX3HHUHb5NM7ibUgHlfHjoD+aV5zt7xgeEQOB0yapH87ZAhq8rGHE7XYWNT4eBRNQ/UkzmZGwogQIiz8+iu8/74+SfL5s7/HaKhZrE5R4Mxu2QB8/VnLKoL288/gcil0ST5A+2VfA+iTH8eNC1obqocRd5s2aBYLhtJSVOkdqfLKK5CdjRoVRcXhE1aPwlXZO8KvvwaoYYElYUQI0Wx89x088wxs2+Z7fVkZ3HST/v01A5fTP2PfER/DM1Tz/W9JuFwtp7qu5wPzxNYfY96yBc1oRLv3XgyRkUFrQ/UwgtGIq/IK9euvg9aGJs1uR3v4YUBfPUM9FmW4K+eNUIc9fzRN70V87rkGtTIgJIwIIZqFP/+ECy7QuOsu6NABRgxzc999cMopkJiosWEDpESV8O9/zD3q4wxvt51YawUHSqL4/YeWUQFUVeHHH/XgNTb3RUAfArBccEFQ21EVRuLQNLxDNcqCBUFtR5O1dStKQQGaxYKzX7963dXTM2LcsgXNbj/qsbt26aH+vikutCZSEkPCiBCiyXO74V//0kuYp8UUoaCxaLGRadNg7lyw2xXSY4p4/cKviY84+koZi0lldJeNAHz9YRkAS5bAccfBgw+oAX8tobB8OeTmKgww/0Xigc1oBgPcfz+KxRLUdnhqjZQ5LRwsi/ROYjVu3IhWXBzUtjRJW7cCoCYmQj1rG6mtWqFZrSgOB+65Rw/k+fn6pd1lwlXYNM572NQZEUKEr1de0d9QY60VzPvXG7g1A5+v6c36AykMSN/Die230Tn5IHUtkntWt2w+W9OHb3+NY9RMuPhijbIyhVWrFC4+r5zeAyMC+4KCzDNEc2/sS3AQXF26YPnHP4Lejuq1RnYWxJOcXooaFYWhtBTXDz9guvTSoLepSakeRurLYMDVpo0+BPfzz3DGGUc8tLCw6nt7qQtzSv2fzt+kZ0QIEXJr1sCUKbBokT6eXd2uXVXLdf9z6i+0jiklPbaYW4f/zvTzv+Ha45fTJaXuQQTglE6bsRhdbDkQzznn6EEkwuxE0xT+M7nIj6+safjhBzDgZnTpVwCo55yDYrWGpC2eWiM7C+JBUbzzRrSffw5Je5qUyg0L1YSG7Z/kmTdyrMq21cNIRXnT6A2UMCKECLlbbtH3AhsxAgb1d/HWW/q2JZ9+CldfDSUlCoMzdzLuuBV+eb4Yq4OTOlR+ClUVLumzmh8nvAXAlwtasXpZuV+epynYtAlWrIBTmEN0RT5qRATGCRNC1p7qtUYA3JVhRFm8OEQtakI8PSPx8Q26u7f42bp1aIen+moKCqq+t5c1jWXVMkwjhAipsjL4/Xd9vxSL0cWKv0xcfbXvMSaDm+fP/r6+w+hHdfXA5fy2PYt/DV7KA/+Yg8EA5/VYyzfrevLQnUV8uSA8hmoefFC/vCfmJSgGV//+mLt2DVl7PJNYdxyKB/D2jBi3bEErKUGJjg5Ju5oETxhpaM9IRgaaomAoLMT9++8Yj1C5tfCQyr08wR4ysJef3ODm+pOEESFESP32GzidCm1iC5l//eu8vWIAX6/riYJGjM1OjNXBRb3+pkfrA3593tO7bmLXvdN8As7dIxfwzbqefLWwNX8tKcNpjPSWl+/ZEwYMgCFD9MmuzWET51Wr4OOPIYYiRpb/BFSWfg/hxo+ewqvbDunzItTk5Kp5IzNnYrrkkpC1LaQ0DW3rVhQaOGcEwGLBnZWFads2lPPPR/v9d5TOnWsc1vnnV7iF+ynHxrai1Y1rt59IGBFChJSnRtMJHbaTFFXOnScu4s4Tg7Ph1+HvyT1b53J+j7V8va4nJ/7DQlFZ1W1Ll8Jb+kgOt15fwf+91vQ35rz/fv3yyYznMe5x4k5ODvmbvadTZmNekv6NouBq1w7LunX6vJGWGkZyclDKy9EUBbURFXHLLryQqHfewZiXhzZsGNpvv6F06VJ1wKZN/OPnKQBEUEHF3gNAzcASbDJnRAgRUp4wMqLdtqMfGCRTTpqPgkZRmQmL0cXFvdfwxoVfMvnEBYzqtBlF0XjxdRvTn69KKpqmb7Hy228hbPhhFi3StzgxGdxcqb0HgGvECIwpoV064QkjuwvjKXPon4dl3gjeIRotLg6MxgY/jBYTQ+n48bhTUlDy8mDoULT5lRtGut0wfjwWV9WcKFfeoUY121+kZ0QIETLFxbBsmT5fZET77aFuDgDdWx3gnUs+ZU9RHP/stYaU6DKf259dcAKPzD2FWydb6dbLSa9+Zm64Ab78Eswmlc0bXLTtGNz6HYfTNLj3Xv37Kd2/IHrtFjRFgWuuCWm7AJKTISlR5WC+gc0Hk+iTtr9q3simTWhlZShBrArbZHhW0jR0iKYaLTqa0nHjiHr3XYy5uXDSSWhnnonSrRv8/jtlhiiMqhMrDtwHm8bqMekZEUKEzMKFeiGzrIR82sYXHvsOQXJujw3cOGRJjSACMGnEIv7Zaw0u1cg/L1Tp1Uvjyy/125wuA/97KvQb8M2bp/eM2ExOJqtPAeDq3h1zCGqL1KZbN33CzeaD+iZwakoKamQkitOJ+8cfQ9m00Gnk5NXDadHRlI4fj33gQDSDAWXmTG/991eS7icfPfS4DjWN/3cSRkJI0+Cnn+DgwVC3RIjgWLoUduyo+tk7RNN+R+13aIIUBV4671v6p+8hv8RKbq5Cj1b7mXTCQgDe/jwWtzu0e958951+eWPXWcRtWAmA+8YbMUQ0jRVCXSvDyMa8yh1pFQV3u3YAaLNmhapZodXIZb210SIjqTj7bEpuvBFn5fiYs1s3PjZdSTEx+jGFJX57vsaQYZoQ+uUXOP106NGxgpVrbYSoBpEQQbFkCQwdqpEU52bV30YyMhQ8VatPzNoa2sbVU4TZxQeXfsKdP5xJr9RcJo9YgKop/G/5QHbmx/LzJ3mccfmxt34PlDlz9Mt/lb+Eomk4O3XCctVVIWvP4bp10y83eSaxoi/xNa9f33LnjVQO07j9MExzODUlhbLLLkMpKkKLjqboFVtVGCluGmFEekZCyLO54rotNh6enB/axggRYI88ApqmkFdgYuy5RRw8CH/+qfcgNJX5IvWRHlvMR5d9wv0n/4rV5CbC7OKSPmsAePOV0G0+lpsLq1dDKvvoukNPe+6rr8YQGxuyNh3OM4l108GqwOYzb6Q8fIrO1ZXm52GaWp8jNhYMBooqJIyIajZtqvr+qVfjWLXcGbrGCBFAf/5ZWZJcUYk0O5i/Mo6LzrGjaQqdk/JIjWkafxAby1Mh9vslrdm/OzSBxDP0NS3yYRS3G1fbtlhuuCEkbTkST8/I5rwk1Mpq5GpKCmpEhL7Rm2cznZaitBQlJwcIbBjxKKwWRpTymvOiQqFBYeSVV14hKysLm83G4MGDWbp06RGPffvtt1EUxefLZmv66/ODYfNm/TI5shSXamTCpWW4XKFtkxCB8Pjj+uWFvdby9JkzAZi/WB+XPLHD9hC1yv96peYyIGM3TreRt571b5G2upozBxI5yGUV7wDguuIKDEF4g6uP9u3BZNJ3791bXNljYzB4d/H1TnppKbbpy9pVmw0CPK+nwmnE7jZ5w4jBbg/o89VVvcPIJ598wqRJk5g6dSorV66kb9++jB49mtzcI88gj42NZd++fd6vHTuaz2S1QNq0Se+ifvm8b4i3lbNqSxxP3d801nwL4S/r1sEXX+i/65NOWMjl/f7i4t5VVR9HZDWN+iL+ctVxfwIw46OoGpv+BcOcOfBvHsWqluNOTcV8663Bb8QxmM3QqaN+cjblVRuq6dgRAGXhwpC0K2SCMETjUWTXOwO8YcTZNMJIvSewPvfcc1x33XVMqNxo6bXXXuOHH35gxowZ3HPPPbXeR1EUUlNTG9fSMFNcDDk5+ozyoe128vjpP3HT1+fzwNNxpHd2MP7aqjoFGzfqxYvGjYMm9gFHCC+XC0y1/EWZNk2fK3J29/Xeku7PnvUDa3JSyS2JZmT7xk9eVUpLMa9eDQYDWlQUamQkhvx8TLt2Ydy9G8VuR42LQ42LQ4uJ0Y8zGMBkwtWhA+7MzJrlWBvowl5/c9+s0WzaH8/87w5x0rnB+0+7fTuYt25gIi8D4Lj2WiJatw7a89dH124GNmTrk1hP7qj/DnjCiGH7drS9e1HS00PZxODx1BhJSjrGgY1XVKH3SHrDiKMi4M9ZF/UKIw6HgxUrVnCvp5oOYDAYGDVqFIuPMgO6pKSEdu3aoaoqxx13HI8//jg9e/Y84vF2ux17ta6joqKmUZTFn6oP0cTZ7FzW9y/+2NmWd1cex4TrLBQesnPzHVaeew4efFDDblf44L9l/PpHJC15HynR9CxZAjfcoO+Dkpykkp6h0KqVgtWqfwL+9lu9qNnkEVWfdmNtDub96w1UTSHS0rixSdOmTUR8/TWG0tKjHmcoKYE9e2reMH8+anw8jj59cAwYoFfAbIQYq4Pze67lg1X9efvV0qCGkTlz4DkmYcaFs2tXrJMnB+2566tbN/jmG9+eES0uDndyMsa8PFyffYbptttC2MIgCsCy3iPx9IwUoQ+PmVzNsGckLy8Pt9tN68OSduvWrdmwYUOt9+natSszZsygT58+FBYW8swzzzBs2DDWrl1LmzZtar3PtGnT+M9//lOfpjU7nsmrHRL1VTSKAi+c/R1RFgfT/xjC7XdbeWm6ypZtBkDBqKgsXxvJ+aeVMHNeNJZ6FHjMy4M77oDx4+GUU/z+UkQLVVIC//43vPiihqbpvXx5Bw3k1aibo3Bqp030S9/nc63NXM+ty91ucDpRnE4UhwOcTiwrV2KtnLPmTk5GzcxEKSzUlzDGxKD26IE2dCiGTp1g5059bH7/fjSnE8XlgoICTCtWYCgowLZgAdalSym78EJc1ffyaIDL+63ig1X9+XJeCq8Uu4mKaXh57/o4+N5MruFHXIoJ9f77MTcyWAWSd4+ag75LoF0dO+r7qsyaBS0tjARp8ipU9YyYnE1j5VLA64wMHTqUoUOHen8eNmwY3bt35/XXX+eRRx6p9T733nsvkyZN8v5cVFREZmZmoJsaVN4wkly1pNdggMdH/0S8rZxp805myzYDsdYKHjv9Z3qk7Ofcd8YxZ3E0V15YwkffRte5V/nZZ+H992HOD+Vs2WMjIqIZbDcqmqSdO/X6OHPnws8/axw4oAAKY/r+xb0j51HqtLC/OJq8sigcbiNOtxEFjbO71/5h5ajsdiwrVmBZtgxDYSGKZ9lFbYcOH45h+nTMvXt7r9M0DVMdttbViotx/fe/KC+/jHHrVqI+/JCKE0/EMWQIpo0bMa9bh1JaimPQIJx9+tRpOGdo2520iz/EjoIEPn99H+Mmp9XtNTeCZndw4aI7ANjR41Q6jBkT8OdsjOoraqpzdeyIdckSjEuXoqlqSHcYDhrPME0w5owcNkxjaY5hJDk5GaPRyP79+32u379/f53nhJjNZvr3789mzzhFLaxWK9YwrwDmefkdE3w/RioKTDlpAW3jC1ixpw2TRiwkPbYYgPcv/YRLPrycT3+IZnGmi649jHTurOB06uFm0yaN+FiNX+fr3eSgf5h8/329m3zfoQheezyXOx5pFcyXKsLEnDlw2mkaqup5g1fIjCvg+bO/Z1TnLd7jerZuZDl0hwPrwoVYly1Dqag5nq0pCpjNaGYzWkwMznHjsDzwAIaoKJ/jlDoEEQAlJgbTHXfAzTejXnsthvfew7ZgAbYFC3yOM+3Zg3vJEspHj/Zu7HYkBgNc2vcvnpx/Eu++pzAuCKMlB25/jE7ujeynFYlTxqPUp/s0BDw9I3uK4iixm4m26qUNXO3aoRkMGPLzca9YgXHQoBC2MghUFW3bNhRC0zNidjWNMFKvyGmxWBgwYABzPOX9AFVVmTNnjk/vx9G43W7WrFlDWlrgPyk0ZZ6ekY5JtRc7u6zfap45a6Y3iACc3HErr1/wFVaTi117Tfzyi8L06fDf/8L8+bB3r8K6DQaevy/He59582D37qo/yk++GEVZWWhLVYvm6YsvQFX1uiCTT1zAt+PeYfktL/sEkUaz24n64ANsCxeiVFTgTkqiYvx4HL/8gmvVKlybNqHu3o26dy/k5GDYsQPbE0/UCCINYrFgePdd1DfeQDObAXCnpGA/6yyc11yDZrVi3LeP6LffJuaFF4j45hvMf/0FRyjQdVnfvwD4dU1rdm4K8B/8V1+l1WsPAzAj5S7iL70gsM/nB4mJkJKs93ZtPlitd8Rqxd22LQDqZ5+FomnBtXcvisOBZjDoRckCrPCw1TRWd9MII/Ueppk0aRLjxo1j4MCBHH/88bzwwguUlpZ6V9dcddVVZGRkMG3aNAAefvhhhgwZQqdOnSgoKODpp59mx44dXHvttf59Jc2MvqxXoUNi/TamubDXWk7usIUNB1LYkp/EtvxEjAaVjokHOVAaxb9/Hs1rHyVw33MqMbEG3n1Xv98Vx/3JvC3t2V0Yz/RHc7nzcekdEfXjKaY1ddQvnN092/9PYLcT9eGHmHbsQLNaqbjuOsxTpmA7wtyyQDFcdx3a6NE4167FOGQI1spPq9qjj+KeOBHDV19hKCjA8uefWP78E81qxT5kCPYhQ3xqRGQlFjC07Q4W72zHOy8c5IFXAvQ63n8fbr4ZgEf4N9qQoSiVYaqp69Zd4cBCfRJrv/SqD1GuDh0wbd+O4tkvIJx5hmji48EY+LlFhw/TWN1No+hZvcPImDFjOHDgAA8++CA5OTn069ePWbNmeSe17ty5E0O1Mb5Dhw5x3XXXkZOTQ0JCAgMGDOD333+nR48e/nsVzUxREeTm6r0Vngms9ZEQWcHQdrsY2m6Xz/VuVeGt5QPZkp/Em0/k8K/7UivrOyhc1X8lA9N3c/v35/D0K1Hc+G+NyEiZOyLqJicHNmwABY3h7QJQJ+jwIDJ1KrYpU0I2X0Bp2xZz5adz73WpqRg//xwtLw/XV1+h/fwzhkWLMObkYJs/H+sff+Ds0QMtIgLNakWNieGa7lks3jmO976I5t8v68OwfvXpp2jjx6MA0w038aD6MEsuWOXnJwmcrl0VFi70LQsPlUt8587F+PffaBUVKOFcKLNy8YcagD1pauMZpnGYIsClhxHN5UKpbV1+EDXof/rEiRPZsWMHdrudJUuWMHjwYO9t8+bN4+233/b+/Pzzz3uPzcnJ4YcffqB///6Nbnhz5hmiSYkqIdbmv5LRRoPGLcN+B+CF16P45BON0lK992VQm91c3m8VmXEF7C+K4tVHQlMdUjRP8+frl71Sc0iI9GNdgooKLIsWEfPSS00miByLkpyM6brrMH/2GcY9e3C/8QZqmzYodjuWP//E+vvv2H79lchvv+Xqn6/mZ+U0hu3/ml/ePPI8uXrLz9cLD40Zg+J2s7PTCdysvkSrqFL6j24+Q+BVG+b5hhF3WppeGt5ux/399yFoWRCt0fczUutRD0bTYM7mjuwvrv/wZJFd7xmxxOqXNncpmiN0eyl5yK69IeCdL9KAXpFjubTvXzz268nsyo/hzjtUQOHSvqtRFLCYVO46cQG3fncuT7wcw1V3aN6JrkIcjWeI5oT2fuoVcbmwLlqEdfFilMqaQmpsLI577mnSQaQGgwHjddfBNdfgfvtt1Dlz9K7P0lIMW7di3LGDU5nNqcyG62Hj7X1wjjyVLuf3wNw2DdLSIC5O7543mfRqiOvX61979kBWFvTsCd27Q0UF7N0LGzei/ec/KDk5aIqCY9gwXuQRtM0GTu6yA1Na8+l19i7vPSyMYDDgbt8ew7p1aD/9BP/8Z/AbFyx//w2Au1Xdh87fWHo8U348gzO6buCjyz6p19N5ekZscTbIhwitHLW8HCIj6/U4/iZhJAQ8K2mqL+v1F5vZzfWDl/Lo3H9QWKz/Qb+kT1Xp7cv6/cVrSwazLrc1N1yWzxe/JPq/61iEnXnz9MsR7Rpfut2Qm0vkl19irNwYzJ2cjPP88zFNnoy1S5c6r4JpUgwGjFdfjfHqq32uVv/8k+LnXufg5/PpULGBLuWrYdZqmNW4p1PQz1v5eefhzsxk9mv6u/qpI8qa1fnzlHPZlp+ApvkOY7kyMvRl1X/+GZrGBYOmof39t/7vWccwklMczWNzTwZg1d76V6gtqgwjUck2qPzvrB06BEGo/no0EkZCwNszklC/yat1dc3AZTy/8ARKnRaGtdtOVkKB9zazUWX6BV9zypvX8tXcRD78bwljr5OSruLI9u2D7Gx9vsiwdjsb/kCqimXJEmy//ILidqNGROC48krMDz6ILSPDfw1uQgz9+xP33mvEvQd7F27kt8mfU7hsOxnaLgbGZpPkztULuKmq/mUyoSYl4W7VCjUmBkNBAcYDBzDk5YHZjBoTgxYbi6tdO+zDhoHZTG5JFKtz9KGZ0873w6qiIMrKAkXRN8w7UBpFq+iqKrruyt8Jw+bNaJrWrEJWne3fj3LwIJqioCYnH/t44IGfT/VWUd1XHEux3UKMte7DLJ6ekfgUMy6MmHCj5fv/g3F9SRgJgcOrr/pbQmQFNw9bzFPzR3LD4CU1bu+blsPkExfyxLyTuOV2MyefpZGYqDB7tt62668Hf6yUFOHB0yvSOzWH+IiGzRdRysqI+PprzBs3AuDs3Bn12Wexnn12eL7J1CJ9RBcuXnIfzz2rct1kA7H2Cv64+VWf5ftHdHi3QTXztuo73fZO3UfGoLa1HtNUWSyQ2UZj5y6FbfkJvmEkLQ1NUTAUFuJevx5jOC568MwXSUzU9044hvlbs/hsTR8MiorVrFLuMLHlYFKN6sZH45kzkpKiUUwMCRQ0iTDSTAZmw4tnt94OR6gx4g/3njSPzXc9zbk9aq98eeeIhfRN28uhMisnHG8nJUXj3HPhzjvhzusKAtYuEVhlZdC3L5x2iqtRO8ZWv693iKbD9gY9lnH3bqJffx3zxo1oJhMVY8ZgWLQI6znntJggUt1ttxs4vr+DIruNSd+fVbd/p6Ocpzmb9c3lRvXeh6EZblzVsZP+2rYfOqzgl9Xq7S1Qf/kl2M0Kjsr5InWZvOpwGZj8w1kAXDvkTwYM1M+bT42WOvAM07RqVbW8l4KCej1GIEgYCbKCAsjLa/iy3rpSFEiOOvL6cbNRZfr5X2Mxuti2x0ZJiUJajL4h4Zsfx7JqSdPYyVHUz9y5sHo1zJ5rYvncQ/W+//z5MHgwZLVT+f13/V3SE0ZOaLe9fg/mdGL99VeiZszQP90mJFDxzDNY338fYz0m64UboxFmvGfBbFKZtbErX62tuWnouyv6c8L065m/Neuoj6Wq8OsWPYyMPq15FjPs0OEIYYSqoRr++COYTQqeekxe/e+yQWw6mEyrqBIeedJC1+56TZJNeXUPI5pW1TPSOlXxbpbXFMKIDNMEmWeIpnV0cb3G+QKhR+sDfHjpx6zcm+HdyOzqzy/iq7W9uPW6Mub/ZZPJrc3Mzz9Xff/ZjCIGnVJ7eWlVhe+/B4cDYmP1HuKXXoKvvvIcYeCUk9w8+RRs3GjEoKgMbVv3lTSmjRuJmDkTQ+UfOWfPnmivv07E8OENe2FhpmdPuP8+lYceNnDbd+egaXBR77UAPL9wOP+ZMwqAG766gD9ufpU4W+07q67NbU1uaTRRZgcnnB2cOhX+1kEfZWLbkcLIqlUoq1fXuC0seMJISsoxD/12nT5Mddc5q0k5YShdFuvX16dnpMRhQdX0PojWqYaqnpHCwno0OjBadBh58EF9Yt7dd8OAAcF5Tu9KmgD2itTHqM5bfMp5P3zqbGZld2XhmkQ+nVHImGua7q6foqaffqr6/ovZcTx5hKkG998PTzxR83qjQWX8gBXsKYxl1sau3Kbvu0af1BziI4681bhxxw5MGzdi3L8f4/79GIr1eRBqbCz2q67CMnUqxjpO0Gsp7r3fxNyfylmwJIJrvvgnszd3olVUKS/+rge2GKudfcWxPPzLKTx79sxaH2Nu5RDNCR12ENmpfdDa7k+eMLLjUM0w5U7XV4sYtm5Fc7tRglChNGhUFW3tWn1PmmP0jBSUW1m2W6/ge+7l0SiK4l0WfXjBuKPxTF41G9zEp5gprAwjWhPoGWnRwzS//AKffgrbl9V98k9jeSevJtW/Cz0YMuOLuO2E3wCYfLeJ0tLm2fXbEm3fDhs3glFRsRpdbD0Qz5/za/6effVVVRAZ1GY3vVJzaBd/iLO6ree3G6bz7Fkz+eDST/jX8VWTn49UX8Rw8CCRH39M9FtvYfvtN8ybN2MoLkYzGLCPHIn7l1+wvfiiBJFaWCwwZ1EE/761AIOi8vFf/bxB5OFzF/Llp/rGcf9bPog/dtbctXx9bgrv/akXkBw1uCDkFTQbqqOep9h+KL7Gbe7WrdGMRgzl5ajLlwe3YYG2fTtKaSma0XjM6qvzt3XArRnonJRHpxH6JGVPGNmSl1Tn+WGeUvCxtgpsMRbpGWkqPJOXSxYugRvOD8pz7qqs4N42rmmGEYBbh/3G+3/2Y3d+PN07Oencw0S7dgqnnAJjxui1mUTT4+kVOT5zF4mRZfywoTufzijmuJOqur83boRx4/QtAm4a+gePj/6p1scyGjSeOnMWXZLz+GR1X67qv6LqRlXFuHcv5tWrsSxfjqKqaAYDzr59YdAgOO44lMGDsfTo0eR3jg01kwke+b94TjvXzthLXew5GMFzYxZwy5sDMURHM/4KJ2+/b+a2b89hwQ2vYzW5KXOYeHrBSF76fSgu1UicrZwLL2kee9HUxtMzsq84lnKniQizq+pGkwl3aiqmPXvQ5s7VJzSFC8/k1ZSUY+5J4+kBO6Xnbgzx/QD9vBmNGqVOC/uKY+q0KsuzJDjOVoEtJlbCSFPh+TvpcAZvYoRnBVVSZNPYnKg2kRYXz5w5k3GfXcKuHDO7KveveusteGSqk4ceNXPJJfo26aLp8ISRf3TaSlZ8Pj9s6M4XP8UwrXKopqQELrwQiosVhrbdwX9GzT7mY157/HKuPX45uN0Yt+7AvH495g0bvMMwAM5OnXBPmYL1iivCew+RABpxipXsnRYOrN5NZt8h3vP47P+ZmTnTSXZeCr2fvx2naqDYbsWl6m9eZ3TL5oVHi2h/RpDGmQMgMRFiY1SKig3sOBRPt1Z5Pre709P1MLJ0aYhaGCCe+SLHWEmjaTCncpLyaadUBTWLBdpnaWzeorApL6lOYcQzTBNrs2ONMlWFkeI6LC8PMAkjgNN19OP8yRNGEiKaxrbNR3J61038ffvzbMxLZldhPNkHUnh7xXFkb4nkssvguacczJlvISYm1C0VAC4XzJmj93ic0nEznZLysBpdbM5N4K9FBXTqH8+FF8LatZAaXczbF3+G2age/UGdTkxbt2Jevx5TdjaG8qrfWc1iwdW9O+7zz8dy222YE2qfKCvqLiJCoe1g3+GYxER4abqJMWMgt7Rq2W6b2EKenvAnl0ztiyGha7Cb6leKAh07Kvy5Sl9RUyOMZGTAsmUolTU5wkbl6znW5NVNeUnsLozHYnRx8nnxPrd17WZg8xbYcjCJkXVYeu8dpolwYLMp3jCiSRgJLc8wjdMZvI/4zSWMAKREl5ESvRPQq27eccJCXlsyhJd/H8qyP208NjmPJ16XuQBNwZIlUFSkkBhRRt+0fRgNGv/otJkfs7vx2vNl/LknnqVLIcri4J1LPqV1TOkRH8u4dSuWFSswb9qkVwetpEZG4urTB849F8Mll2Dq0AGzLLcKuEsuUejWoYLiDTuJjlGISTCRkWXFkjkybOq0dKgWRg7nWd5r3LkzvHbw9QzTHGPy6i+bOwEwrN1O4rr7FrXr0gV++KHuk1g9PSNxkU6sVqqFkSP/PQiWFh1GvD0javBmaOfn659em0MYOVyszcHdIxfQo9V+rvjkUl6YEc8Nd7vI6tiif42aBM8Qzckdt2I06LPZzu+xjh+zu/H6V/qKhMSIMj4b+wED2uxFKS7GvG4dakKCvl270QgVFUT8/DOWlSu9j6vGxupzQc47D9OYMZgzMsLmDbA56TPQBgO7hLoZAVO1vLfmRE41KQnNYkFxOHAtXIjp1FOD3LoAcDjQNmyo0540cyuHaEYNykexdPC5zTOJta7Le71zRqJdPmFELQ19XakW/S7i6RlxuIMZRvTL5hhGPM7qls0JWdtYtL09d19/gE9/OfYaeRFYVWGkapn26V03YjG6cLhNZMQW8uWV79MtaieW2b9hXbIExaWPT6rR0Th79MCcnY2hsBBNUXAOHow2fjym887D0rq1BBARUFUramoZ7jMY9Hkj27ejzZ8P4RBGNm5EcbnQLBa0uCOXT6hwGvltexYAp59V8+263mGkcpgmLkb1CSNUHHnZfrC06DDi7RkJUhipqICyMv2PenwzDiOKAo+P/omRr1/PZ3NS+G1OGcNPCe320y3ZwTwN99IVxNORf1QLI3E2O/ef/CuLt2Uwvc+LpC1ZgWXNGhS7/ofHnZqKUlyMoaQEa+XkQHdCAo7bb8c6eTKGEG8pLlqOqlojtc89cmdkYNq+HeW334LYqgCqXnn1KEF/8c52lLvMpMUU0ffUtBq3e3Y93nEoHrvLiNXkPurTeodpYjWMRigzRIOKXv0wxCSMAE53cOaMHKpczWtQVGKtoU+ijdEnbT9j+6/i/T/7M+nmCv5YHynVWkNA27uPfSOuZznfUYENwy9dcAwYgBYRgWn7du7d8wmm3VtRtlT9vrlbtcI5bhzmyZNR4uJwv/022rvvotpsKE88QcSgQSF8RaIl8oSR7YfiUdWaK/WcXbpg/e03jIsWoeXkoKSmBr+R/vTnn0Dte9Is2dmGFXsyUBT4dYt+Yv7RdQfG1r1qHJuWBtFRKiWlBrbl15z8ezhPKfi4WH0ot8ISAxV4P6CEUosOI8EepvEM0cTbKsJiWey//zGXr/7uydLsRJ5/tJhJD/gurfGEbSk14SeaBkVF4Hbr9dxnz6b8mon0Ks9HRcFGBaxejaWW0tlqVBSuXr3QzjwT09VXY2vTxnub8frr9a2ahQiRtm31mhkVLjP7S6JJiy3xud3dti2utDRM+/bheuQRTK+8EqKW+oGmweefA+Bq187nplKHmfPevYoKl2/dmNNOqqh1qFRR9KGaFSv1oZpjhRFPz0h8vP6zw6yHEYPDgaZpIR2ODYO3xIYL9jBNOMwXqS41poT7/zEXgMlTo/nkg6p0/f77kJioceapjkbtHtviFRToZYInTEDLyND/iiQlQUoKXH45keX5rKQ//x36Kq4vv8R1+uloZjOayYSrfXvs55+P/eWXUdetw7x4MdYHH8RYLYgI0RSYzdA2U/9DUeu8EUXBMXQoAIb330drAp/kG2zpUti6Fc1sxtnVd1l2bkk0FS4zZoObi47byIUDNjHxlFVcfP2RJ7l26aq/jW+uw4oaz469cfF66Kiw6hvlGZwVIR+qkZ4RgreaxhtGIsMjjADcOGQJW/OT+O+yQVw13kRyiouZP5l47jkAhTkLLCz4oYCRZ8eHuKVNlNuth41Zs6BXLzj9dP1y2TJ49VW0jz/2dqEe/pmlhCieYTL2ESN47LuBGOLi4IIL0MrLUXNzMaanYzI338qcomXp0FFh23Y9jAxtt6vG7c4ePVBnz8ZQVIT75Zcx3nln8BvpDx9+CICzWzewWn1uyivV52mlxhTz2W9t67SM2btHTR127/UO0yToAcZp03uzjQ47msOBclh7gqlFhxFvBVbpGWkwRYEnz/iR/SVRfLe+B6NGV93WIfEgW/OTePXZMgkjh3O59D9Kjz2m12j3uPtunJGxmMuKAD2AuJOTcXXpgrNjR9yZmRy0R3P8q7dwsDya64Ys59UvsvQgUkmJiMB4WPevEE1dhw4Kc+bUvrwXAJMJx6BB2ObO1beYnjTpqJM/mySXC+2TT1AAZ+/eNW4+WKaHkaSoMrDWbZNSzyTWLfnHDiOeYZqERN8wYlBduIuLCWUVSxmmAVxBmsBaFUZCv6bbn4wGjTcu/Mq7xXyU2cE7Yz5lxj/1cdGvFrYmZ1foZ2s3GX/9BQMHwrhxsHEjakQE9qFDcXbujMtowVxWRAVWtrUfQcm111Jy881UnHYa7o4dwWLhgTmnc7A8mh6t9vN/7yRhSpHCc6L5O9qGeR6OgQPRTCaMO3bg/vbb4DTMn379FWX/ftSICL2+z2G8YSS69jkitalXz4hnmCZJ74dwR1SFDy0/tDvJt+gwErIJrBFNd1+ahoowu/j48o94bPRP/PqvNziv+3r6pecwqM0unG4jb0w7EOomhp7LBY8/jjZoEPz1F2pEBOWjRlF8++1UjB5N2dixnNduBcezhAz20GHbAh7Kvgq3VvXfdMG2LD5c1R9F0Xj5jg1EdGme28YLcbiq5b1H3sFWi4zE0bev/sMjj9DsJqR5hmh69ap1c7x8TxiJrfuHN0/PyMGyKPLLIo54nNNtoNSpfwKPT9bf/Ew2E2VU3udQaDdvbdFhJFQTWONt4dUz4hFns3Pz0D/oknLQe901g/Rtv9/8OBans5n94fCnPXvgxBPh/vtRnE6c3bpRcvPNOE44wTtunFsSxU/berCM4xl1gl6e+dmFIxj132t4d0V/ckuiuOO7swG4evCfjLxVluCK8OGtwpp/9H2OHEOGoBkMGFesQH3wwSC0zE/Ky9G++AKoDCO18PSMpCTUfcO06GjIbKPvM7Ux78i9pMX2qvkgnjDiUxJewkjoVJWDD/YwTfjMGTmW83usJSmylN2HYvj2nYPHvkM4+uMPvTdk8WI0q5WyCy+kbMwYtOhon8O+/Lsnbs3AgIzdfPRDPB/8r4woq5M/92Zw63fn0vWZO9mSn0RqdDHTnjZLUTIRVjyjFrml0ZQ6jjzxWk1JoeL00wFQHnsM9auvgtG8xps5E6W4GDUuDndmZq2H5JVFAZCUWL8Pbt176EM6Gw8cOYx4qq9Gmh2YY/ThGp8qrBJGQifYwzSef+uWFEZsZjdX9tcL/DzzjMLPP8Pvv8P27aFtV1AcOgRvvIE2ciTKvn24U1Iovv56nH361Drx7tPVfQC49KQ9GGJjufzqSDZuNvDYLbvplnYIrXI9zdPjVpA8vPZPVkI0V/HxEBerf8LfVXD0yZuOQYOwDxiAomkoY8eirVsXhBY2gsMBL76of9u7d82qbpUOVg6zJCfVL4x066b/bcjOO/LWHAWe+SK2qs0GfXpGCgvr9Zz+1qLDiNQZCY7xA1egoPFHdhKjR8Pw4dC+Pfz0XRgOV2Vnwznn6DVBEhPh+utRHA59WObaa9ESax8P35SXxMq9GRgVlUuvjvJen97GyH0vtmHdngSWzing55fWcPnjx8leMSIspaXpv9e5pdFHP1BRqDjjDFxt26KUl6OdfLJeSExVg9DKenI64bLLYMECNJMJZ79+RzzUM2ckOaV+/7+7d9cvNx1lmMazSV6s1e5dwls9jCgSRkLHW2ckaGFET7stLYxkJRTwyGk/MzhzF71Sc0iO1OdDvPt8mE1qtdvhggvg++9R9u4F9F1vK045hbJLLqlRU6C6T1fry/xO6byFjOGda9yuKDDoH/GcOrE3htjYwLRfiBBL9YSRkmOEEQCTibJLLkGNj8eQmwsXX4zWsyd88UXTmdjqcsGVV8KXX6IZjZRddhlq8pEDg2fOSHKr+r01e8JI9lHCiGdZb2yEHaVy8qzPME1xcb2e09+kzghSZyQYJg77g4nD/gBg0fZ2nP32eH5ckozToWK2hEkmfuwxWL8eNSqKsksuwd26NdShaJGmVQ3RXHbGQRRrzTAiREvg2aoltyTq6AdW0qKjKb7+eqx//IH1jz9QNmyAf/4T9YEHMDz8cABbWgdbt8LkyfDVV3oQufTSWpfzVucJI63S6vfW3K2bfrnzUDzlThMR5poTYPcW6R9iUmKq3n98hmmKiur1nP4WJu8CDRPMYRqXCwoL9dTfEsNIdUMyd5IQUcahsggWfXv0vRSajb/+Qps2DYDys8/G3a5dnYIIwJJdmewoSCDaYueCCXXbClyIcOTZ/65OPSMeERHYTz6Zottvxz5sGACGRx5Bff/9ALTwGOx2WL4crrgCrUsXPYgYDJRdcgmuzkf/kOFyKxwqr+wZSa/fhl6tWkFCvIqGwuaDtf8NWbtfT3o925d6r/PpGSkpqe1uQdOiw0hVOfjAn4aCgqrv420tO4yYjBqndd4EwNcfh8G5cLngmmtQXC6c3bvj8vSZHmZ3YSy7C2NrXOdZrntOr43E9uoQ8OYK0VR5wsj++oQRj4gIKk47DfvgwQAo11yDtnixH1uHvn3DzJlw4YVobduide8OgwfDiSeiZWWhRUTAoEHwwQcobjfOTp0onTAB12F70NTmUHlVjZDk9Lp9kPFQFOje/egrav7O0cNI3z5VQ1g+c0ZCHEZkmIbgDNN4hmhirRWYjE1kPDOEzuyWzSer+/Ld/Dhe0JpfVWcfTz4JK1ag2WyUn3lmrYes2pvGmW+Nx+4yceVxfzJl5DwOlkVyyQdj2VscS2p0MffdUeYdyxWiJfIM0xw41gTWo6gYPRpDfj7mTZvQzjwT7ZJLULZtQ9u+HRQFJSEBEhIgKkovPGYw6BNfCwv1r9JSPRW1b69/gf5pMj8fbdYslF36vjmH/8ny/KxZLDg7d8Y+fDhqenqd2+0Zoom3lWOOrf+y/W7dFX5fXHutEbeqsD5X32yvz8CqXherFYqo/IAkYSR0qsrBBy+MxLfwIRqPUzpuxmJ0sS0vnr8XF9B7WHyom9Qwb7wB//43AOWnn45Wy94O+4qiueyjSymrrH749ooBfPJXH0wGlWKHlW4puXzz8lY6Xzw4qE0XoqlpVM+Ih8FA2UUXET1jBsbcXP3/KDXDw1GtWVPr1QqgRkTg7NsXZ/fuoGkodjuK04kaG4uamIgWFdWgT1feyatRpSiRdduXpjpPh2xtYWRrfiLlLjMRJiddBsZ7r/cZpiktrXG/YGrRYaR6nRFN0wK6XDJQk1eVQ4cw7diBcfduTLt3oxkM2E855ZgTpUIt2upkZPttzN7cma/fLWyeYeS//4XrrwfQ95bxlKmuptxpYuzHl7KvOJauyQd4+v6DPP56K/7YoC/xHdZuB1++W0TrE4cEtelCNEVVc0bqNoH1iGw2SseOxfrbb2g2G2pCAmqCXtlVqahAKS9HcTr12eOqCoqCZrOh2WxgMqEUF2M4dAhDQQEYjfptVivulBRcXbpUvXn4UdUmeeUo5vrvN1W1oqZmrRHPEE2P1rlY0lp5r7daIdcTRspCu01Jiw4jPhNYVbXWvQL8pV6b5JWXY163DsOhQ6hJSajJyahJSWgGgzeJm7OzMa9Zg2n37hp3N733Ho7jjqP8tNNAUTDm5GA4eBB3ZiZqypGL4gTbmd2ymb25M9/NjuCBUDemvt56C+1f/0IB7EOGUFF5rqsrd5q45ZtzWbk3g4SIMj5/ZiO9xg3n7Nvgmw8K2TB3FxMnRxDdo+bunUK0RJ4wklcahVtVMBoaPqStxcVRcYRh06booKf6aoy9Qff3rKjZkpdU49x5J6+2PYRizvBe7zNnRMJI6PhMYHW7gxNGIo/QM+JwYN64UQ8Ymzah1LF4j6YouNu0Qe3aFW3QIEwbN2L84gssK1di/vtvcDiqxjIBV48eVJx4Iqrnf30Ind4lmzs4m+VbU9i7tYz0Ds2kvPknn6Bdcw2KpmEfPJiK0aN9gsiK3em8/2d/vvi7F0V2GyaDm/fv+J2eV50C6Ieef0UcXFH/rlghwllKCiiKhlszkF8WQUp0+G0qeiTenpFYZ4Pun5UFVquG3W5iZ0E87ROryruvqQwjfbr7bsDnM0xTEdoilC06jPhMYHW7A/pcRxqmMW7dimXVKswbNqA4qn5R3K1a4e7aFUNuLoa9ezEcVpDGlZGBe+RIlHHjMI8YgSmiaia2Ons2XHUVhpwc/eeYGNSEBEw7d2Jetw7zunU4u3alYuTIek2w8re02BKOS9/Dyr0ZfDXjIDc/2gzCyI8/ol1xhR5EBgzQ98ioFkTeW9mPW749z/tzm7gCHhuzgjMeHCFVU4U4BpNJL4V+IE9hf0l0ywwjiQ17LzIaoWsXjdVrFDbmJfuEEU/PSN9+vh+4fXpGJIyETvVhGs3trN8Ep3ryhpFqy3qNW7cS/e673p/VuDicQ4agjB2L6dxzscTpn5w1TUM9cADNbvfO/jbExGCKqn1c1XDqqbBtG64ff0RLTMTYpw/G+HjUpUtRp0zBuGCBPsyTnY2zUyfsJ5yAu23bI+6XEEjndF/Pyr0Z3PNMKj3/4eakfzTh1SQLF6JddBGKy4WjVy8qzjrLJ4hoGrz423AATu+2iYlX5nPqVZmYM/4hQUSIOkpNVTiQ51lRkxvq5gSNZ1+alHruS1Ndt+4GVq/Rl/eO7qKXTygot7G7MB6AvkN93zN8wojdjuZ2h2xFX4sOIz7l4NWGjdPVVW2raUzbtgHgatsW9513Yrr0UqytWtW4r6IoKLVcf1Q2G6YLLvB9nMGDMcybh7pyJeqUKRjmzsW8eTPmzZtRo6Jwde6Ms0sX3KmpaPHxQQkn/xq8lLlbOrJwe3vOON3F51+4OeucJhhI1q9HO/tslPJynJ07U37BBTXOz4o9GWw6mEyEyckH77pJHCSrY4Sor9apCmv+buSKmmbIuy9NI8JIbZNY1+7X3zvaxBWQ1Nl3zqBPGHE40BwOlGq97MHUosNIqIdpPJNP3aefjvXWWwP6/NUZjjsOZs9GW7sW15QpGH/5BUNpKZZVq7CsWgWAZjKhJifj7NwZZ9++R91PoTGiLE4+G/sB4z67mJ82duX8C1Tef9fFmMub0K9mWRlcfDFKURGutm0pu/jiWucXfbRKX01zTq9sEvp3C3YrhQgLDarCGgbySvVei/puklddbRvm/b1fP6G90g6gxHbyOd5nzojdrs8bCVEYadEVWD1hREPB5QhyGFFVjJWbqSnDhwf0uY9E6dkT0/ffw6FDuN9/H9f55+NOT0czGlFcLow5OdgWLiTm5ZeJevNNLL/9hmHfPr/vjGkzu3l/zKdc1GsNLreBy64w8vLzjmPfMVhuuQXWrtX3nLn44qpfnGrsLiNf/N0LgCsuKkUxNaEwJUQz4rflvc2Mt85I64b3DHtW1GQfSPbuFeiZL9KnU2mN4WKfnhFAq14qPMha9F/M6kvFHeVu6rcbQP0cHkYMBw/qY3QmE4YRIwL4zMemRERgHDsWxo4FQCstxbVyJSxaBF98gfHPPzHt2YNpzx5AL/rj6tKFitNO0wv8+IHZqPLGhV8Rb6vgf8sHccskC/v2lPPo0xGhrc763nswYwaaolB20UW1FjUDmJXdhYKKCNJjijjtyoxajxFCHJt3s7xGVGFtjjzDNK3SG17DpEsXfTVSQUUEeaWRpESXeWuM9Oldc/jHaoUyInFjwIiKdvBgVdXZIJOekUqOCv9+2j9cfr7+i+AJI8bKN3Z3ejrGNm0C+tz1pURFYRoxAtO992Javhy2bsU1eTKuvn3RLBYM5eVY/vqL6OnTMW7d6rfnNRo0njlrJvefPBeAx5+N4OrLS3HV3IAyODZsQLvxRgDsI0eyzHoCv+9oy/b8eCqcvp9ePvxLH6K5dPBGLG0ljAjRUH6pwtrMlDtNlFZWaE6p5yZ51UVEQKcO+nvNUwtG+pSB7zuo5uPqe3kqlCn6h0rpGQkRy03XsoxV3MHzOOyB26BMVeFQ5Sqrw8OI2rkzpgBU8/MnpV07TE8/DYBWUYH7m29Q7roLw65dRL37LvYRI3AMHowW3fg/HooCd41cSOvoEm7//mze/jiK3P2lfPZ9FJHBXPnrmSdSWoorK4uvWl3H5W9e7nNIj1b7uf2E3xiRtY1fNuk7cl41Dlk5I0QjeMLIgRY0TJNfuZLGZHAT17pxczYeeUzhsss03lx6PMV2i7cMfOcB8TWOtVr1yxJiiKG46o0qBFp0z4iSvYGBrCCZPBz2wG1eV1QEqqq/QXl27PWEEQYMCNjzBoJis2EcMwbD+vWoF12EAtgWLiT2mWeIfv11rHPm6PNKtMadz6sG/Mn7Yz7BZnIy89coThlWRl6ePnJ07bXQrZvGTTdpVE678b/bboO//0aNiiL/vDHc8/MZAKRElWAz6UWJ1uW25l9fXsjxL0/ErRkYkLGb3ud2CVCDhGgZPMM0LalnJM9TfTWyDENU4z51jRmj8H/P6H+jPv6rH1CzDLyHJ4zkK4moNhuUh27vtBYdRir7qLBREdBhGs98kUizA5vZDU4nxsqCZIwcGbDnDaioKAyff4761lu427UDwLhvnz7h9fXXiX71VawLFzZqW+ozu23k66veI95Wzh9/RZKRrjJiBPzvf5CdrTB9ukLHDip3TlI5eNBfLwz44AP473+980T+b/VodhYkkBFbyMb5uyktcrNnzQH+c/M+EiIrKHbo/6MvP3k3hthYPzZEiJbH0zNysCwKp7tlvEV55oskRZb5ZWntLZMsTL23qohZz8xDKLX0wHvCyAmWpRTfcw9anz6Nfu6Gahn/0kdSPYzYAx9GvEM0OTkoqooaGYlx4MCAPW8wGMaPx7h9O+qWLbgefRTXkCFoJhPGAwewzZlD9GuvoXhOQAMMabuLWVfPICO2EIfTQJTZweX9VvH6BV8yOHMnFXYDzz1vYFA/u396SbKz0Tyb3514ItsT+/H8whMAeGzsKuKO64ohwkZ6rxQefDmNHTlWnrwrh5tPW8t199azFowQooakJDAa9Z7VvNJmUJXZD7zVV6PL/VZ0bOpjNu64qRyTwc05JxfXeownjDiCsHP9sbToOSPB7hmJP3zyamYmpia0cV1jGDp0wHD//XD//Wh5ebimT8cwfTqGffuIeu89Sq+++ogrUY6lW6s85l//Bst2t2FE1jairXoX5CV91vDL5k7c+cOZbNudwKknlLJgWRRJSQ18ESUl8M9/eueJ2EeO5MEvTqXcZWZYux2MfahHjfkgMTEKdz+VCoR+rx8hwoHBAK1SNPbl6CXh02Ib3rvaXHjDSIz/ShooCjz3SgSPP16B1XZcrcd4wojdFfooID0j6GHE6QjcnJHDd+z1Tl7tUfPNLRwoycmYHngAw4oVaOnpGA8dIur99xs1HpkcVcYZXTd6gwjo/9lO7byZb8e9S1pMEeu2RXH6iaUU1/4h4Og0DSZM0OeJREeTc/Zl/N/vJ/DV2l4YFJXn7t6NqVV4BEchmrrUNP3vYkspfObpAUqKb9gmeUdji7OheFLHYTxXuzUDbjW070USRoAIygM6gfXwHXs9YUQZNChgz9kkpKWhLFig74+zfz/Rb72Fdd48TJs2+XWiVFZCAV9d+R6JEWUsXxfFuaNK6l9Q98kn4fPPUQ1GnmrzAt1ef4ipv5wKwLVD/uT4q5vXRGMhmrPU1Mow0kJqjXhLwScGtsTE4apnFLsrtEM1EkbwzBkJRs9IOUpZGcbKK5STTw7YczYZHTvC3LloUVEYc3OxzZtH1AcfEPv001gXLPBbNddurfL48sr3ibbYmbc0mo9fr32JmqbBp59C375w5ZWwbx8waxbaffcBMMnwAvduuIZSp4VuKbm8cOXv/N8HrVAqf1eEEIHX0krCe4ZpGrNJXkNUDyMOd2iHaiSM4BmmCcKcEWt51XyRxESMPXsG7DmbEqVvX1i7FufkybhOOAE1KQlFVbHNnUvUu++iFBX55Xn6pe/j1uG/A/DwY4YavSOrV8PJJ8OYMfr3778Pl3Zciv3CS1E0jTe4jv9zTWRQm1388MjvrFkDt707DFv7TL+0TwhRNy1tea+3FHwj9qVpCJNJr9gK0jMSWj49I4F7muo9I94wkpWFoYETOpsjpV07zE8/jWnhQgwHDqC++CKa1Ypp+3aip0/HvHKlXzYrvGHwH8Tbytm4N44PXq1a7ztjBvTvrzF/PthMTm4/4TcuSpnLF+VnYi0vZAEjuIWXuOmE5cyf4+LMfw/D1FpWxwgRCt7CZ6Uto/CZN4y0Cm4YUZRqk1ilZySEQjBMY/BUuOvSgotjKQqGW26BlStRO3XCUF5O5LffEvPii1iWL6cx9d9jbQ5uGebpHTHjcsH338N112moqsLZ3dezdOIrPNz/Ez4tP59kDrJCGcCl5i/47w2LeGlmVyK6hGZvBiGErqWVhPcO07QOfiDwLu+VnpEQqj5M4wzSnBFPEbD09IA9X3Oh9OiB4e+/cd93H2psLIbCQiK+/57YJ58k8sMPsSxdilJYWO/H/dfgpSRGlLFlfyyTri/mkkv0IDK2/5+8d8mntC9dR9S772IoKcbdqhXtvn6av5aUcsXLJ7Wo3iohmirvZnktIIxoWrWekbTgbw1irdyyRnpGQqlaGLEHcJgmL0+/TIgsxyBhxJfVivGxx1D27cN9//2oCQkoTifmjRuJmDmTmJdewrR5c70eMsbq8M4deWlGDOXlCqd23sQLZ3yDbf48ov73PwyFhbiTknC99RbJ555MSv8svxUbEkI0TtUE1vAfpimyW3Gp+t+elIzal+AGkrVybr70jIRSZdndQNcZ2bNHf+y0mGJvz4jSxHbqDTUlMhLjo4+iHDiA+6efcF5/Pe62bVFcLiI/+ghTdna9Hu/aQUtJiiwFoH/6Ht478SXi334T27x5KKqKo1cv3J99hvXMMwPxcoQQjeAJI4UVETV2yA43nmW9UWYHkYmNLwVfX01lzkjoy66Fks9qmsA8RVERlJTok5LSog6hlOpvkLRtG5gnbOYUoxHjaadhPO00cDhQzzkHw88/E/npp5Sfdx5qQgKGwkKU8nKcnTujxcdX3VnTMG7bhqGoiNiMDF4//0vm/p3O48oDxM74A0XT0Gw27FdeieWJJzAkJobsdQohjiwuDiwWDYdDIbc0mrbx9R+ubS48QzSJkWUYgro1ua6qCmtoQ5+EESqHaQIURjyb88ZaK4hxFqAAmqJgyJTlosdksWD4/nvUiy7C8N13RH75pc/NNpMJ+9Ch2E84AcOhQ0TMmoVp+3bv7RdEzOBCTUOp0CvfOnr1Qnv4YaznnYdiaNmdgkI0ZYoCqa01du5SOFASFdZhxFt9NaoMbHFBf36rVf+wHOqS8BJG0MOIK0ATWD1hpPoQjRYVhRIX/F+6ZslsxvDVV6jXXIPy4YdokZGocXEogHHnTmwLF2JZvhylvFwPekYj7owMjHv3Yqis8upu3RrnTTdhuf122VVXiGYiNU1h567wX1Gzt0j/m5QWVxaS7UGayjBNy/54WH1pryMwvwTeMBJb7J28qkVHo0SF/8QsvzEaMbz9NhQWwu7dGLdvx7h9O+6330Zt1QpDZRBx9OqF48MPMW7eDAUFuL7+GscLL6D98Qe2Bx+UICJEM5JWuT/NnqLw/uDmeX0ZrQK4iuIomsrSXukZwRNGAvMU1cOIp2dEjY3FKMME9aZERFA9MhrHjYNLL8X10ku4IyMxX3ll1dJcsxnTeeeFpJ1CiMbzlGLalNfQbbibh92F+oekzPSG11dqjKbSMyJhhMDWGfGEkfSYoqplvTJx0n+sVkyTJ7fwX2Qhwk+3bvrlpoPJoW1IgO2pHKbJDNECy6bSM9KyP55Xn8BaEeg5I0VVc0ZSZCt6IYQ4Gm8YyQvzMFKoD9NkZoUmDDSVnhEJI0AE5QErB1/bMA2tZM8TIYQ4Gk8Y2V0YR4k9+JVJg0HTqiawZnYIzWuUnpGmwKfOSKDCiP64GbHVhmk8FX2EEELUKjERWqXou6lvPhie80bySiOxu00oaGR2Dn6NEWjmPSOvvPIKWVlZ2Gw2Bg8ezNKlS+t0v48//hhFUTj//PMb8rT+VxlGzLhw2Z1+f3iXC/bv17+vPkwjpeCFEOLYunXXp6yH61CNZ75Iq+gSbMmhWe3XVIqe1TuMfPLJJ0yaNImpU6eycuVK+vbty+jRo8nNzT3q/bZv387kyZMZMWJEgxvrd5VhBECr8H8YyckBVVUwGdykRJVW9YxIwTMhhDimbt30MJKdF57z7LzLeuOKUEK0Sad3mKa59Yw899xzXHfddUyYMIEePXrw2muvERkZyYwZM454H7fbzdixY/nPf/5Dhw4dGtVgv6oWRnD4f423Z75IanQJBpcDpXL9sKFdO78/lxBChJtwn8S6p3JZb0ZCaciqQjfLnhGHw8GKFSsYNWpU1QMYDIwaNYrFixcf8X4PP/wwrVq14pprrqnT89jtdoqKiny+AsJoxG3UJw0pTv/3jFRNXq22ksZkQpE5I0IIcUxhH0Yqe0batA5NwTNopj0jeXl5uN1uWrdu7XN969atycnJqfU+ixYt4n//+x9vvvlmnZ9n2rRpxMXFeb8yAzis4TbpvSOKy/8FZ6rCSIlP9VVDiLrjhBCiOfGEkc0HE3GrwS+VHmh7vAXP1JC1oVn2jNRXcXExV155JW+++SbJyXVPtvfeey+FhYXer127dgWsjW5z5VBNQMNIVc+IGhPjOzwkhBCiVm3bgs2m4XCb2FkQH+rm+J1nAmubEBU8g6q3I4e7GZWDT05Oxmg0st+zRKTS/v37Sa1l6GHLli1s376dc845x3udquoJ0GQykZ2dTceOHWvcz2q1YvXEtQBzW/R/CYM7cGEkPbpqWa8WFxeSzZCEEKK5MRqhS2eN1WsUNuYl0z7xUKib5Fe7Q1zwDKr3jJiAAO2LUgf16hmxWCwMGDCAOXPmeK9TVZU5c+YwdOjQGsd369aNNWvWsGrVKu/Xueeey8knn8yqVasCOvxSV6o3jPj/H2H3bv2yesEzTUrBCyFEnXXvob9NbQyzeSNuVWFfsT5k37ZTcD5816aqzkgz6hkBmDRpEuPGjWPgwIEcf/zxvPDCC5SWljJhwgQArrrqKjIyMpg2bRo2m41evXr53D8+Ph6gxvWhogahZyQtpgjDdqm+KoQQ9eWZN7LxQHiFkdySKFyqEYOikt4xNAXPoHoF1ma2Ud6YMWM4cOAADz74IDk5OfTr149Zs2Z5J7Xu3LkTQzPakVaz6mHE6PbvahpN81RfVUiXUvBCCNEg4bphnmclTVpMMZakuJC1o9n2jABMnDiRiRMn1nrbvHnzjnrft99+uyFPGTCeMGLwcxgpLISyMn1uiE/11bQ0vz6PEEKEs3DtGfHWGIkrQokK3QzWptIz0ny6MAKlMoyY/DxnxDNEE2crJ9LikuqrQgjRAF266Jf55ZEcLI0IbWP8aLen+mpiWUgXNTSVnpEWH0Y8PSMm1b9FZzxhJCO2CDTN2zOiSBgRQog6i4yEdm31VZjhNInV0zPSpnXoVrCA9Iw0HRGeMOJEU/1XeKaqxkgxSnk5SuVjK1IKXggh6iUcN8zbW1ljJDMjdAXP4LCeES0wu9fXhYQRmyeMOMDt9tvD+oQRT8Ezmw1DUnhuhS2EEIHSvTKMbDgQPhvmeQqehbqzvHrPiObH98D6avFhRKkMI2bVAYHoGYmuti9NdDRKVJTfnkMIIVqC447TL3/Z3CmUH979ao+n4Fn70A6P+PSMSBgJHSVSnxBlUh1+TYXe6quxxVXVV2NjUcxmvz2HEEK0BOeeCzarysa8FP7a1/w3GnW5FXJKooHQFjyDw+aMBGBblLqSMBJRrWckEMM01Zb1apUF34QQQtRdXJweSAA+Xd0ntI3xg33FMaiaAbPBTWr70PaWV+8ZkWGaEDJE6mHEotn9HEb0vsS06j0jMl9ECCEa5Ior9berL/7uhcvdvPf38hQ8S48twpgQuoJnUK1nxG1Cc0rPSMgYoirDCHZUl3/CiNMJubn69+nVC55VVqkVQghRP6efDkkJbvaXxLBgW/tQN6dRPMt60+OKUSJCWzul+p609orQrexp8WHEWNkzYqMCZ4V/wsiOHaBpCjaTk6TIMgkjQgjRSGYzjLlMf8v6pJkP1Xh6RtokhbbgGRweRmRpb8gYo6rCiMPun3+IjRv1y45JBzEYwFBUpF+Rnu6XxxdCiJboyiv1N+7vN3Sn1NF8FwOs3a/vUZaV7t9imw1hsVR9L2EkhAzVwoi93D89I54w0ik5HxwODAcP6lf06+eXxxdCiJZo8GDomOWi1GFh5oauoW5Og7jcCj9v6gzAqJP8uydaQxgMYDbrIaSiXMJIyPgM09j9M17mDSOJBzHm5KBoGmp0NMa+ff3y+EII0RIpClwxTq/L8crioeQUR4e4RfW3eGdbDpVHkhhRxokXNo2he++KGukZCR3P0l4bFTj8NHknO1u/7Jh0EOO+fQC4MzIwJCb65fGFEKKlGj8eIm1uVu1LZ/ArN/H+n/2aVSG0mdn6NsSn99iCtV1GiFujs1oqe0ZCOGrU4sOIpxy8jQq/zSTeuFH/h+2clIdx714AtK5dQz5RSQghmrusLFj8u8ZxnQsorIhg4jfn8c/3x5Jb0vSrW2sa3uGlc04tRzE0jbdg7/Je6RkJocowEkE5Tkfj/yFKS2H3bj10dKrWM6L179/oxxZCCAF9+ptYsi6eJ+49hM3sYs6WTox47XoWbmvaG5Gu3d+KHQUJ2ExOTh+TEOrmeFmt+nuWXQ3dpGAJIzb/DtNs3qxfJkaUkWAqwnDgAADKCSc0+rGFEELoTCaY8ngCK1ZAj3bF7C+J4bx3r+Lp+SP8uc2YX3mGaEZ23EZcz6zQNqaaB6cqvHD3TrpePixkbQjtDj1NQbUw4ixvfPU5z3yRTkkHMe7fr09ejYqSyatCCBEAPXqbWLo2hpuuLODdr+J57Nd/sLcolufO/oGmNjLuGaI5d2RBk9qnbPx4gLYhbYP0jFRWv7NRgb2s8WHEW2MkOd938mpycqMfWwghRE1RUfDOl/FMf64URdF4a8VA7p55Rq0TW1ftTeXazy/k1m/Pwa0GL63sLoxl1b50FDTOGdP057cEm/SMVB+m8WMYkcmrQggRXDfcEYU1ysE1N5h5c9nxmIwqE4f+TpHdxt6iWN5Ycjw/beriPf60zhs5u3u2354/tySKqbNHkRRZxjnd1zOozW48c1R/2KAP0RyfuZuMIc27nH0gSBjxCSONL0BTvfqq8W+ZvCqEEME04V8WnC47199sZfofQ5j+xxCf2w2KSqfkfDYeSOb1JYN9wsiWg4k8Nf9Ezuq2gXO6b6jXMM/uwljOf/dKNh/Ue8FfXjyM1Ohi2sYXsCU/kYNlem/IWYNzMERmNv6FhhkJI5VhxIiKs9TRqIfSNMjO1gCFLnH7MFTulqcMH97YVgohhKijf91kRVPt3HWXQrnDSJytglirnRM67mLKHXaiB/ahQ3eNhdvbs25/Cj1aH0BV4fovL2D5njZ8srovoztv5KkzZ9IuofCYz7ctP4Fz37mKXYXxtIkrYFj3fGb9mU5OSQw5JTHe4zon5XHF1ZajPFLLJWGkMowAuIvLGvVQeXlQUKCgoNHZuV6fvBoZiVF6RoQQIqiun2jlmutcUJiPIcKGYosBUx/vkPkF57r54msjbywdzAvnfM+Hf/Vj+Z42RJicuDQDP23qwoJX2nN29/W0SyigbXwBraNLiDQ7iDQ7capGtuYnsvVgIu//2Z+ckhg6Jh5k5vRNdL54MPYKjbmf5pC/NY/OnRW69Ikkrl0ihjhZzFAbCSPVtix0l1U06qE8QzRt4gqJzN2lP2ZGBiaZvCqEEEFnspqgVUqtt916h5EvvtZ3AL59+CIemj0KgH+fvYQLHhrE9VcVsnB1Ip+tqdsOwT1a7WfmjN20O0sfFrJFKJw5LhVI9cdLCXsSRhQFh2LFotlRSxsXRrzLepOrFTvr0kUmrwohRBMzYgT07enir7Vmzn57PHllUXRNPsAdT2YQ0cXK/FVWZn58iFVzDrBth4Htey3kF5kpt5soc+pvne0TC+iUXkq3zk7G/8tG8pABIX5VzZeEEcBhtGFx2dHK/dMz0ikpH+OOypU0MkQjhBBNjqLArZNMXHMN7C6KA+CZG9YR0WWk9/azLkvgrMtqVkrVVBVcLjDHy4dNP5E6I4DTqM8bUf00TNMlfl9V5VWZvCqEEE3SZZdBUrwbgHN7rufMu+r24VExGFAsFgkifiRhhKowojVyy0JPGOltzUZRVTSLRSavCiFEExURAS+9auCMfrt4/ikHhtjYUDepxZJhGsDlhzDidsPmzfqy3k7GrQCocXEYEprOZkhCCCF8XXaZwmWXZQJS+yOUpGcEcJkql/c6Gl70bOdOsNsVLEYXqc49AKhJSSgWWVMuhBBCHI2EEcBlrgwjroaHkU2b9Mv2CYcwFhUAoLVq1ciWCSGEEOFPwgjgMuub5Smuhu9Nk5+vX7aKKcFQUKD/kCndfkIIIcSxSBgB3JXDNI0JI6Wl+mWk2VkVRtq1a2TLhBBCiPAnYQRwWzxhpOHDNCUl+mW01VEVRjp1amTLhBBCiPAnYQRQK+eMGNyN7xmJMZahFBcDoHTpcpR7CCGEEAIkjACgWjxhpPE9IxnaLhRAM5kwdOzoh9YJIYQQ4U3CCKBaGx9GPD0jbVR9gzypMSKEEELUjYQRQKsMI0Y/9Iy0du4GQE1MRKm2I7AQQgghaidhBNAsjQ8jnp6RVs7K3Xpbt250u4QQQoiWQMIIgK0yjKiOBj+Ep2ck0a6HETIyGtsqIYQQokWQMELVMI2pEWHE0zOSUJGjfyM1RoQQQog6kTACEFEZRvwwTBNTfkD/RlbSCCGEEHUiYQRQbI3vGSkpASMuIsr0uvBKt25+aZsQQggR7iSMAEplz4hZa8wwjUYGezBoKprRKDVGhBBCiDqSMAIokfpGeWa1osGPUVICWWwHKmuMJCb6o2lCCCFE2JMwAhgi/dEzAu3YAYCWkOBdoSOEEEKIo5MwQtUwjUW1N+j+LhfY7UpVz0irViiK4q/mCSGEEGFNwghgjKoMI5odTdPqfX/PShpPz4jUGBFCCCHqTsIIVcM0Fs2ud3PUkyeMtK/sGZEaI0IIIUTdSRihqmfEih3NUf95I57qq55hGjp08FPLhBBCiPAnYQQwRVeGEc2OZq//vJHSUlBQaYO+Y6/SpYtf2yeEEEKEMwkjVO8ZqYAGhJGSEkhjHxacaIqC0rmzv5sohBBChC0JI1T1jERQgVZR/1ojpaVVQzRaXByG5GR/Nk8IIYQIaxJGAHNMVU0Qtbik3vf39IwAqLGxKJGRfmubEEIIEe4kjFDVMwLgOlRU7/uXlkIs+v20yEipMSKEEELUg4QRwBJlRkUPEM7C0nrfv3oYQXpFhBBCiHqRMAKYLQoV6L0jzoL6h5GSEoihGNB7RoQQQghRdxJGALMZytE3y3MWltX7/j49IzEx/myaEEIIEfYkjABGI96eEVdRw3pGvGEkOtqfTRNCCCHCnoSRSnalMowUl9f7vqWlVcM00jMihBBC1I+EkUqeMOIuqX8Y8ekZiY31Z7OEEEKIsCdhpJLDE0ZKG1b0TMKIEEII0TASRio5DHoYURsYRrzDNBJGhBBCiHqRMFLJYawMI+X1DyM+wzTx8X5slRBCCBH+JIxUcnp6Rsoatmuvp2dESUjwa7uEEEKIcCdhpJLT0zNS0ZBde7WqnhEJI0IIIUS9NCiMvPLKK2RlZWGz2Rg8eDBLly494rFffvklAwcOJD4+nqioKPr168d7773X4AYHitNUuT+N3VHv+zpKHNjQQ4ySlOTPZgkhhBBhr95h5JNPPmHSpElMnTqVlStX0rdvX0aPHk1ubm6txycmJnL//fezePFiVq9ezYQJE5gwYQI//fRToxvvT67KnhFc7nrfVykprvpeekaEEEKIeql3GHnuuee47rrrmDBhAj169OC1114jMjKSGTNm1Hr8SSedxAUXXED37t3p2LEjt912G3369GHRokWNbrw/uTw9I05Xve6nqmAq04doVJMZoqL83TQhhBAirNUrjDgcDlasWMGoUaOqHsBgYNSoUSxevPiY99c0jTlz5pCdnc2JJ554xOPsdjtFRUU+X4HmCSOKy1mv+5WXV9skz2pFsVr93jYhhBAinNUrjOTl5eF2u2ndurXP9a1btyYnJ+eI9yssLCQ6OhqLxcJZZ53FSy+9xKmnnnrE46dNm0ZcXJz3KzMzsz7NbBCnRe/RMDjrN2fEp8aI1SJhRAghhKinoKymiYmJYdWqVSxbtozHHnuMSZMmMW/evCMef++991JYWOj92rVrV8DbWGGNA8DkrF85eJ/qq1arvuueEEIIIerMVJ+Dk5OTMRqN7N+/3+f6/fv3k5qaesT7GQwGOnXqBEC/fv1Yv34906ZN46STTqr1eKvVijXIPQx2mx5GzI6yet2vesEzzWpFURS/t00IIYQIZ/XqGbFYLAwYMIA5c+Z4r1NVlTlz5jB06NA6P46qqtjt9a/nEUgOm17G3eysXxipPkyjRUT4vV1CCCFEuKtXzwjApEmTGDduHAMHDuT444/nhRdeoLS0lAkTJgBw1VVXkZGRwbRp0wB9/sfAgQPp2LEjdrudmTNn8t577zF9+nT/vpJGckbqPSOWeoYRn1LwspJGCCGEqLd6h5ExY8Zw4MABHnzwQXJycujXrx+zZs3yTmrduXMnBkNVh0tpaSk33XQTu3fvJiIigm7duvH+++8zZswY/70KP3BE6GHE6ipD07Q6D7f4zBmJjg5U84QQQoiwVe8wAjBx4kQmTpxY622HT0x99NFHefTRRxvyNEHlivKEkVJwOPTJqHVQUlJtmEbCiBBCCFFvsjdNJc8wjdVVhlZe9xU10jMihBBCNI6EkUpqjB5GjLjRDh2q8/18wkhMTCCaJoQQQoQ1CSOVDNGRuNBrhGh5eXW+X/VhGmJjA9E0IYQQIqxJGKkUHaNQiN47Up8w4tMzImFECCGEqDcJI5Wio/GGEfLz63w/n56R+Hj/N0wIIYQIcxJGKlUPI1o9wkj1nhElLi4gbRNCCCHCmYSRStXDiFKPCaw+5eATEgLSNiGEECKcSRip5DNMU1BQ5/uVlmjeYRpFwogQQghRbxJGKvmEkcLCOt/PWVyBGRcASmJiIJomhBBChDUJI5Wio6GIytUw9egZoajI+62EESGEEKL+JIxU8pnAWo8wopToQzROsw1Fdu0VQggh6k3CSCWfYZpqvR3HYijRh3RUs63O+9kIIYQQooqEkUrVw4haVFrn+5nK9OCiWq0oFktA2iaEEEKEMwkjlSIjq4WRcked72cqryx4ZrWiGI2BaJoQQggR1iSMVDIYwG6tnMBaYa/TfRwOiFI9YUR6RYQQQoiGkDBSjSOisuiZ3Y6macc8vnrBM4PNHNC2CSGEEOFKwkg17mg9jBgd5eB0HvP40tKqfWmUSFtA2yaEEEKEKwkj1XjCiMlRjlZRcczjfXbsjY4OZNOEEEKIsCVhpLq4eAAMqrtOtUaqD9MQGxu4dgkhhBBhTMJINYbYaNyVp0TLyzvm8dWHaYiPD2DLhBBCiPAlYaSa6BjFWxK+LmHEp2ckJiaQTRNCCCHCloSRanz2p8nPP+bxPj0jMkwjhBBCNIiEkWp8SsIfPHjM4316RmSYRgghhGgQCSPV+ISROkxg9VlNExcXuIYJIYQQYUzCSDU+YeTQoWMe71NnRMKIEEII0SASRqqJiqoWRgoLj3m8zzBNYmIAWyaEEEKELwkj1fj0jBQXH/P4kmKtagJrQkIAWyaEEEKELwkj1fiEkaKiYx5fnleKAX0PGyUpKZBNE0IIIcKWhJFq6tsz4sjTA4uqGGTOiBBCCNFAEkaq8QkjJSXHPN59SA8jTlMEik02yhNCCCEaQsJINdXDiFJaeszj3QV674nbYkOxWALaNiGEECJcSRipxqdnpKzsmMdrlStuVIsNJIwIIYQQDSJhpBqfnpHycjRNO+rxSnHlJFerGcUgp1IIIYRoCHkHrcYnjNjt4HId8Vi3G4zl+rwSxSa9IkIIIURDSRippvpGeUqFHc1uP+KxRUVVBc+MEeagtE8IIYQIRxJGqrFYoMxUGUbcLu+ckNoUFFSFEUNURDCaJ4QQQoQlCSOH0aJiqr7PyzvicYcOVe1LQ2xsoJslhBBChC0JI4eJiDZShB5IjhZGCgogiYP6D7IvjRBCCNFgEkYOEx1TbRLrwYNHPK6gAFLJ0X9ITQ1Cy4QQQojwJGHkMNHRijeMaPn5RzzOJ4ykpQWhZUIIIUR4kjByGJ/CZwUFRzyuoADS2Kf/kJER8HYJIYQQ4UrCyGHqHEbyVVqzHwClbdsgtEwIIYQITxJGDlM9jGhHCSP2nENYcAISRoQQQojGkDByGJ+ekaKiIx6n7Nfni5SZYzAkJQWjaUIIIURYkjByGJ+S8EcJI8Y8PYyU2+JRoqOD0jYhhBAiHEkYOYxPz0hx8RGPs+brYcQREYdiMgWjaUIIIURYkjByGJ8wUlJyxOMiCvUw4q5WsVUIIYQQ9Sdh5DA+m+UdJYzElOrLerXoqKC0SwghhAhXEkYOExUFBcTrP5SVHfG4uHJ9Wa8xXuaLCCGEEI0hYeQw0dGQSysADCUlaJpW4xinE5Ld+jCNpU1KUNsnhBBChBsJI4eJjoZ96OXdleJitFp6R6qXgo/oIKXghRBCiMaQMHKY6GjIQd/4TnG70XburHFM9TBi6pgVxNYJIYQQ4UfCyGGio8GJhYPohczUrVtrHFN4wEEKeQAonToFtX1CCCFEuJEwchhP/bK9Srr+zY4dNY4p254LgBMTxg4dgtU0IYQQIixJGDmMJ4zs0SrDyO7dNY5x7NSHaA4Zk1Hi44PUMiGEECI8SRg5jLdnhMowsndvjWNcu/UwUmhKQomICFbThBBCiLAkYeQwkZH6pWdFDTk5NQ+qvK7YkohikFMohBBCNIa8kx7GYICoKK2qZ+TAgRrHGA9UbZInhBBCiMaRMFKL6CitqtbIwYM1bjdXbpJnj4gLaruEEEKIcCRhpBbRMYq3Z8RQWFijCmtEgR5GnFGxQW+bEEIIEW4kjNQiKuroVVijS/QwokXLvjRCCCFEY0kYqUV0tOJbhXXXLp/b48r1MGJMkmEaIYQQorEkjNQiOhocWCm3xACgbtvmc3uCo3KTvMzWQW+bEEIIEW4kjNTCM/pSYtVLwvtUYS0uJkorBSCiU3qQWyaEEEKEHwkjtfCEkSKLHkZ8Nsvz1BghmvhOrYLdNCGEECLsSBiphSeMHDKlAKBUq8Jq36GHkRxSScqSOSNCCCFEYzUojLzyyitkZWVhs9kYPHgwS5cuPeKxb775JiNGjCAhIYGEhARGjRp11OObAk8YOWiq7PnYv997W9m2qjASmxYT7KYJIYQQYafeYeSTTz5h0qRJTJ06lZUrV9K3b19Gjx5Nbm5urcfPmzePyy67jF9//ZXFixeTmZnJaaedxp49exrd+EDxhJEDhsoJqtVem6OyZ+SgIQVjpC3YTRNCCCHCTr3DyHPPPcd1113HhAkT6NGjB6+99hqRkZHMmDGj1uM/+OADbrrpJvr160e3bt3473//i6qqzJkzp9GNDxRPGNlvqFzem5/vvc25Sw8jBaZkFJuEESGEEKKx6hVGHA4HK1asYNSoUVUPYDAwatQoFi9eXKfHKCsrw+l0kpiYeMRj7HY7RUVFPl/B5N25V9MLn1WvwqpVTmAtsSSAxRLUdgkhhBDhqF5hJC8vD7fbTevWvvU1WrduTU5tu9vWYsqUKaSnp/sEmsNNmzaNuLg471dmZmZ9mtlonjCyW20DgFJUhFZeDoAhV3+dpbZEFEUJaruEEEKIcBTU1TRPPPEEH3/8MV999RW2owxx3HvvvRQWFnq/dh1WATXQPGFkhzsDqKzCuns3aBrRe7IBKI+ID2qbhBBCiHBlqs/BycnJGI1G9ldbXQKwf/9+UlNTj3rfZ555hieeeIJffvmFPn36HPVYq9WK1WqtT9P8ql07/XJ9fjpqRASG8nLULVswVlQQd2ALFVjZk9AzZO0TQgghwkm9ekYsFgsDBgzwmXzqmYw6dOjQI97vqaee4pFHHmHWrFkMHDiw4a0Nkm7dwGTSKLLbcERW1hLZuRM++wyAHzmDiFSpMSKEEEL4Q72HaSZNmsSbb77JO++8w/r167nxxhspLS1lwoQJAFx11VXce++93uOffPJJHnjgAWbMmEFWVhY5OTnk5ORQUlLiv1fhZxYLdOmsT1gtNCcD6JvlVYaRz7iY+PhQtU4IIYQIL/UOI2PGjOGZZ57hwQcfpF+/fqxatYpZs2Z5J7Xu3LmTffv2eY+fPn06DoeDf/7zn6SlpXm/nnnmGf+9igDo1Vs/NTmKPvxk+P13yM7GoVj4nrOJj9dC2TwhhBAibNRrzojHxIkTmThxYq23zZs3z+fn7du3N+QpQq53b/j0U9juzqQvYFywAIBlESMoLoslPl5W0gghhBD+IHvTHEGvXvrlhor2gL6iBuA70/kAJCRKGBFCCCH8QcLIEfTurV+uKunivU4zGnm9aCwAfftJGBFCCCH8QcLIEbRvDxE2lZ2Vhc8A9qb2pYAEOiQeJGtwRghbJ4QQQoQPCSNHYDBAz56wl3TvdXMizgJgRJc9GGQ5jRBCCOEXEkaOoncfA3tJp8Qch2az8UbR5QCcNMwR4pYJIYQQ4UPCyFH06gUOrNzZ7iP2XDmR33O7AnDyGREhbpkQQggRPhq0tLel8ExinXvoeEYWF6Kh0CX5AJnHB3fjPiGEECKcSc/IUXiW927NT+CXTZ0AGNF1L4bY2BC2SgghhAgvEkaOIjUVEhNUVM3Ap6v1zf1OGu4McauEEEKI8CJh5CgUBXr30euJlDotAJx0RmQomySEEEKEHQkjx9CrV1Vxs+4puaQPkPkiQgghhD9JGDkGzyRWgBHd9mGIiQldY4QQQogwJGHkGDyTWAFOOkHmiwghhBD+Jkt7j6FXL7BaVDS3yklnRYe6OUIIIUTYkTByDHFx8MO3bhyrVpE6qG+omyOEEEKEHQkjdXDKaDOMHhTqZgghhBBhSeaMCCGEECKkJIwIIYQQIqQkjAghhBAipCSMCCGEECKkJIwIIYQQIqQkjAghhBAipCSMCCGEECKkJIwIIYQQIqQkjAghhBAipCSMCCGEECKkJIwIIYQQIqQkjAghhBAipCSMCCGEECKkmsWuvZqmAVBUVBTilgghhBCirjzv25738SNpFmGkuLgYgMzMzBC3RAghhBD1VVxcTFxc3BFvV7RjxZUmQFVV9u7dS0xMDIqi+O1xi4qKyMzMZNeuXcTGxvrtcZsTOQdyDkDOAcg5ADkHIOfA369f0zSKi4tJT0/HYDjyzJBm0TNiMBho06ZNwB4/Nja2Rf7SVSfnQM4ByDkAOQcg5wDkHPjz9R+tR8RDJrAKIYQQIqQkjAghhBAipFp0GLFarUydOhWr1RrqpoSMnAM5ByDnAOQcgJwDkHMQqtffLCawCiGEECJ8teieESGEEEKEnoQRIYQQQoSUhBEhhBBChJSEESGEEEKEVIsOI6+88gpZWVnYbDYGDx7M0qVLQ92kgJg2bRqDBg0iJiaGVq1acf7555Odne1zTEVFBTfffDNJSUlER0dz0UUXsX///hC1OPCeeOIJFEXh9ttv917XEs7Bnj17uOKKK0hKSiIiIoLevXuzfPly7+2apvHggw+SlpZGREQEo0aNYtOmTSFssX+53W4eeOAB2rdvT0REBB07duSRRx7x2Tcj3M7BggULOOecc0hPT0dRFL7++muf2+vyevPz8xk7diyxsbHEx8dzzTXXUFJSEsRX0ThHOwdOp5MpU6bQu3dvoqKiSE9P56qrrmLv3r0+jxHO5+BwN9xwA4qi8MILL/hcH8hz0GLDyCeffMKkSZOYOnUqK1eupG/fvowePZrc3NxQN83v5s+fz80338wff/zB7NmzcTqdnHbaaZSWlnqPueOOO/juu+/47LPPmD9/Pnv37uXCCy8MYasDZ9myZbz++uv06dPH5/pwPweHDh1i+PDhmM1mfvzxR9atW8ezzz5LQkKC95innnqKF198kddee40lS5YQFRXF6NGjqaioCGHL/efJJ59k+vTpvPzyy6xfv54nn3ySp556ipdeesl7TLidg9LSUvr27csrr7xS6+11eb1jx45l7dq1zJ49m++//54FCxbwr3/9K1gvodGOdg7KyspYuXIlDzzwACtXruTLL78kOzubc8891+e4cD4H1X311Vf88ccfpKen17gtoOdAa6GOP/547eabb/b+7Ha7tfT0dG3atGkhbFVw5ObmaoA2f/58TdM0raCgQDObzdpnn33mPWb9+vUaoC1evDhUzQyI4uJirXPnztrs2bO1kSNHarfddpumaS3jHEyZMkU74YQTjni7qqpaamqq9vTTT3uvKygo0KxWq/bRRx8Fo4kBd9ZZZ2lXX321z3UXXnihNnbsWE3Twv8cANpXX33l/bkur3fdunUaoC1btsx7zI8//qgpiqLt2bMnaG33l8PPQW2WLl2qAdqOHTs0TWs552D37t1aRkaG9vfff2vt2rXTnn/+ee9tgT4HLbJnxOFwsGLFCkaNGuW9zmAwMGrUKBYvXhzClgVHYWEhAImJiQCsWLECp9Ppcz66detG27Ztw+583HzzzZx11lk+rxVaxjn49ttvGThwIBdffDGtWrWif//+vPnmm97bt23bRk5Ojs85iIuLY/DgwWFzDoYNG8acOXPYuHEjAH/99ReLFi3ijDPOAFrGOaiuLq938eLFxMfHM3DgQO8xo0aNwmAwsGTJkqC3ORgKCwtRFIX4+HigZZwDVVW58sorueuuu+jZs2eN2wN9DprFRnn+lpeXh9vtpnXr1j7Xt27dmg0bNoSoVcGhqiq33347w4cPp1evXgDk5ORgsVi8//E8WrduTU5OTghaGRgff/wxK1euZNmyZTVuawnnYOvWrUyfPp1JkyZx3333sWzZMm699VYsFgvjxo3zvs7a/l+Eyzm45557KCoqolu3bhiNRtxuN4899hhjx44FaBHnoLq6vN6cnBxatWrlc7vJZCIxMTEsz0lFRQVTpkzhsssu824U1xLOwZNPPonJZOLWW2+t9fZAn4MWGUZasptvvpm///6bRYsWhbopQbVr1y5uu+02Zs+ejc1mC3VzQkJVVQYOHMjjjz8OQP/+/fn777957bXXGDduXIhbFxyffvopH3zwAR9++CE9e/Zk1apV3H777aSnp7eYcyCOzOl0cskll6BpGtOnTw91c4JmxYoV/N///R8rV65EUZSQtKFFDtMkJydjNBprrJTYv38/qampIWpV4E2cOJHvv/+eX3/9lTZt2nivT01NxeFwUFBQ4HN8OJ2PFStWkJuby3HHHYfJZMJkMjF//nxefPFFTCYTrVu3DvtzkJaWRo8ePXyu6969Ozt37gTwvs5w/n9x1113cc8993DppZfSu3dvrrzySu644w6mTZsGtIxzUF1dXm9qamqNif0ul4v8/PywOieeILJjxw5mz57t7RWB8D8HCxcuJDc3l7Zt23r/Pu7YsYM777yTrKwsIPDnoEWGEYvFwoABA5gzZ473OlVVmTNnDkOHDg1hywJD0zQmTpzIV199xdy5c2nfvr3P7QMGDMBsNvucj+zsbHbu3Bk25+OUU05hzZo1rFq1yvs1cOBAxo4d6/0+3M/B8OHDayzp3rhxI+3atQOgffv2pKam+pyDoqIilixZEjbnoKysDIPB98+e0WhEVVWgZZyD6uryeocOHUpBQQErVqzwHjN37lxUVWXw4MFBb3MgeILIpk2b+OWXX0hKSvK5PdzPwZVXXsnq1at9/j6mp6dz11138dNPPwFBOAeNngLbTH388cea1WrV3n77bW3dunXav/71Ly0+Pl7LyckJddP87sb/b+f+QZKJ4ziO+4CkHNIfEByMA4XAoaVFuLnFKWl0Otqqpc0lHAOnFreWllwFsVkbHCyCizZpuPEmlwRFBz/PdmT4RM9D9nuw9wtu8sfx/X3gzs/gz5MTbWxs6O7uTkEQhNdoNArXHB8fy7ZttdttPT4+ynEcOY5jcOrle3uaRlr9DB4eHhSNRnVxcaGXlxfV63VZlqWbm5twTbVa1ebmpprNpp6fn1UsFpXJZDQejw1O/nVc11U6ndbt7a1831ej0VAymVS5XA7XrFoGw+FQnufJ8zxFIhFdXl7K87zwpMhn9lsoFLS3t6f7+3t1u13t7OyoVCqZ2tJf+yiD6XSqg4MDbW9v6+npae4dOZlMwnuscgaLvD9NIy03gx9bRiSpVqvJtm2tra0pn8+r1+uZHmkpIpHIwuv6+jpcMx6PdXp6qq2tLVmWpcPDQwVBYG7ob/C+jPyEDFqtlnZ3dxWLxZTL5XR1dTX3+Ww2U6VSUSqVUiwW0/7+vvr9vqFpv97r66vOzs5k27bi8biy2azOz8/nvnRWLYNOp7Pw+XddV9Ln9jsYDFQqlZRIJLS+vq6joyMNh0MDu/k3H2Xg+/4f35GdTie8xypnsMiiMrLMDH5Jb/56EAAA4Jv9yN+MAACA/wdlBAAAGEUZAQAARlFGAACAUZQRAABgFGUEAAAYRRkBAABGUUYAAIBRlBEAAGAUZQQAABhFGQEAAEZRRgAAgFG/AbS8EUEiEjp4AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "encoded_data = autoencoder.encoder(normal_test_data).numpy()\n",
    "decoded_data = autoencoder.decoder(encoded_data).numpy()\n",
    "\n",
    "plt.plot(normal_test_data[0], 'b')\n",
    "plt.plot(decoded_data[0], 'r')\n",
    "plt.fill_between(np.arange(140), decoded_data[0], normal_test_data[0], color='lightcoral')\n",
    "plt.legend(labels=[\"Input\", \"Reconstruction\", \"Error\"])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ocA_q9ufB_aF"
   },
   "source": [
    "Create a similar plot, this time for an anomalous test example."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:54.469916Z",
     "iopub.status.busy": "2022-12-14T06:14:54.469323Z",
     "iopub.status.idle": "2022-12-14T06:14:54.605389Z",
     "shell.execute_reply": "2022-12-14T06:14:54.604490Z"
    },
    "id": "vNFTuPhLwTBn"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "encoded_data = autoencoder.encoder(anomalous_test_data).numpy()\n",
    "decoded_data = autoencoder.decoder(encoded_data).numpy()\n",
    "\n",
    "plt.plot(anomalous_test_data[0], 'b')\n",
    "plt.plot(decoded_data[0], 'r')\n",
    "plt.fill_between(np.arange(140), decoded_data[0], anomalous_test_data[0], color='lightcoral')\n",
    "plt.legend(labels=[\"Input\", \"Reconstruction\", \"Error\"])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ocimg3MBswdS"
   },
   "source": [
    "### Detect anomalies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Xnh8wmkDsypN"
   },
   "source": [
    "Detect anomalies by calculating whether the reconstruction loss is greater than a fixed threshold. In this tutorial, you will calculate the mean average error for normal examples from the training set, then classify future examples as anomalous if the reconstruction error is higher than one standard deviation from the training set.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "TeuT8uTA5Y_w"
   },
   "source": [
    "Plot the reconstruction error on normal ECGs from the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:54.612266Z",
     "iopub.status.busy": "2022-12-14T06:14:54.612003Z",
     "iopub.status.idle": "2022-12-14T06:14:55.026507Z",
     "shell.execute_reply": "2022-12-14T06:14:55.025717Z"
    },
    "id": "N7FltOnHu4-l"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "74/74 [==============================] - 1s 4ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "reconstructions = autoencoder.predict(normal_train_data)\n",
    "train_loss = tf.keras.losses.mae(reconstructions, normal_train_data)\n",
    "\n",
    "plt.hist(train_loss[None,:], bins=50)\n",
    "plt.xlabel(\"Train loss\")\n",
    "plt.ylabel(\"No of examples\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mh-3ChEF5hog"
   },
   "source": [
    "Choose a threshold value that is one standard deviations above the mean."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:55.033843Z",
     "iopub.status.busy": "2022-12-14T06:14:55.033178Z",
     "iopub.status.idle": "2022-12-14T06:14:55.037649Z",
     "shell.execute_reply": "2022-12-14T06:14:55.037000Z"
    },
    "id": "82hkl0Chs3P_"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold:  0.031506725\n"
     ]
    }
   ],
   "source": [
    "threshold = np.mean(train_loss) + np.std(train_loss)\n",
    "print(\"Threshold: \", threshold)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uEGlA1Be50Nj"
   },
   "source": [
    "Note: There are other strategies you could use to select a threshold value above which test examples should be classified as anomalous, the correct approach will depend on your dataset. You can learn more with the links at the end of this tutorial. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zpLSDAeb51D_"
   },
   "source": [
    "If you examine the reconstruction error for the anomalous examples in the test set, you'll notice most have greater reconstruction error than the threshold. By varing the threshold, you can adjust the [precision](https://developers.google.com/machine-learning/glossary#precision) and [recall](https://developers.google.com/machine-learning/glossary#recall) of your classifier. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:55.044418Z",
     "iopub.status.busy": "2022-12-14T06:14:55.043946Z",
     "iopub.status.idle": "2022-12-14T06:14:55.337397Z",
     "shell.execute_reply": "2022-12-14T06:14:55.336544Z"
    },
    "id": "sKVwjQK955Wy"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14/14 [==============================] - 0s 3ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "reconstructions = autoencoder.predict(anomalous_test_data)\n",
    "test_loss = tf.keras.losses.mae(reconstructions, anomalous_test_data)\n",
    "\n",
    "plt.hist(test_loss[None, :], bins=50)\n",
    "plt.xlabel(\"Test loss\")\n",
    "plt.ylabel(\"No of examples\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PFVk_XGE6AX2"
   },
   "source": [
    "Classify an ECG as an anomaly if the reconstruction error is greater than the threshold."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:55.341193Z",
     "iopub.status.busy": "2022-12-14T06:14:55.340648Z",
     "iopub.status.idle": "2022-12-14T06:14:55.345005Z",
     "shell.execute_reply": "2022-12-14T06:14:55.344424Z"
    },
    "id": "mkgJZfhh6CHr"
   },
   "outputs": [],
   "source": [
    "def predict(model, data, threshold):\n",
    "  reconstructions = model(data)\n",
    "  loss = tf.keras.losses.mae(reconstructions, data)\n",
    "  return tf.math.less(loss, threshold)\n",
    "\n",
    "def print_stats(predictions, labels):\n",
    "  print(\"Accuracy = {}\".format(accuracy_score(labels, predictions)))\n",
    "  print(\"Precision = {}\".format(precision_score(labels, predictions)))\n",
    "  print(\"Recall = {}\".format(recall_score(labels, predictions)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T06:14:55.348433Z",
     "iopub.status.busy": "2022-12-14T06:14:55.347978Z",
     "iopub.status.idle": "2022-12-14T06:14:55.362492Z",
     "shell.execute_reply": "2022-12-14T06:14:55.361886Z"
    },
    "id": "sOcfXfXq6FBd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.94\n",
      "Precision = 0.9940711462450593\n",
      "Recall = 0.8982142857142857\n"
     ]
    }
   ],
   "source": [
    "preds = predict(autoencoder, test_data, threshold)\n",
    "print_stats(preds, test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HrJRef8Ln945"
   },
   "source": [
    "## Next steps\n",
    "\n",
    "To learn more about anomaly detection with autoencoders, check out this excellent [interactive example](https://anomagram.fastforwardlabs.com/#/) built with TensorFlow.js by Victor Dibia. For a real-world use case, you can learn how [Airbus Detects Anomalies in ISS Telemetry Data](https://blog.tensorflow.org/2020/04/how-airbus-detects-anomalies-iss-telemetry-data-tfx.html) using TensorFlow. To learn more about the basics, consider reading this [blog post](https://blog.keras.io/building-autoencoders-in-keras.html) by François Chollet. For more details, check out chapter 14 from [Deep Learning](https://www.deeplearningbook.org/) by Ian Goodfellow, Yoshua Bengio, and Aaron Courville.\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Acknowledgments\n",
    "\n",
    "Thanks to TensorFlow Core for creating the open-source course [autoencoder](https://www.tensorflow.org/tutorials/generative/autoencoder). It inspires the majority of the content in this chapter."
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "autoencoder.ipynb",
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}