[Project]_Naive_Bayes.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "[Project]: Naive Bayes",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "## Spam Filter Project"
      ],
      "metadata": {
        "id": "07Q5dneCMAjf"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this project, we're going to build a spam filter for SMS messages using the multinomial Naive Bayes algorithm. Our goal is to create a spam filter that classifies new messages with an accuracy greater than 80% — so we expect that more than 80% of the new messages will be classified correctly as spam or ham (non-spam).\n",
        "\n",
        "To classify messages as spam or non-spam, a computer:\n",
        "\n",
        "1. Learns how humans classify messages.\n",
        "2. Uses that human knowledge to estimate probabilities for new messages — probabilities for spam and non-spam.\n",
        "3. Classifies a new message based on these probability values — if the probability for spam is greater, then it classifies the message as spam. Otherwise, it classifies it as non-spam (if the two probability values are equal, then we may need a human to classify the message).\n",
        "\n",
        "The first task is to \"teach\" the computer how to classify messages. To do that, we'll use the multinomial Naive Bayes algorithm along with a dataset of 5,572 SMS messages that are already classified by humans.\n",
        "\n",
        "The dataset was put together by Tiago A. Almeida and José María Gómez Hidalgo, and it can be downloaded from the The UCI Machine Learning Repository. You can also download the dataset directly from this [link](https://dq-content.s3.amazonaws.com/433/SMSSpamCollection). The data collection process is described in more details on this [page](http://www.dt.fee.unicamp.br/~tiago/smsspamcollection/#composition), where you can also find some of the authors' papers.\n",
        "\n",
        "Note that due to the nature of spam messages, the dataset contains content that may be offensive to some users."
      ],
      "metadata": {
        "id": "H9BKXtbcc3Ff"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1. Exploring the Dataset"
      ],
      "metadata": {
        "id": "jCNy6OlmNFxi"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Import libraries"
      ],
      "metadata": {
        "id": "pmrApPuc7CO-"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "PX5blxBIL4U2"
      },
      "outputs": [],
      "source": [
        "import pandas as pd"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Explore data"
      ],
      "metadata": {
        "id": "rYU02Lp7AnLv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df = pd.read_csv('/content/SMSSpamCollection.csv', header = None, names = ['Label', 'SMS'])\n",
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "9ypDUtnKaylb",
        "outputId": "018d1a3d-4baa-4272-fd36-09169353d6bb"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Label                                                SMS\n",
              "0   ham  Go until jurong point, crazy.. Available only ...\n",
              "1   ham                      Ok lar... Joking wif u oni...\n",
              "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
              "3   ham  U dun say so early hor... U c already then say...\n",
              "4   ham  Nah I don't think he goes to usf, he lives aro..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ea00b494-7be5-4a1e-95b1-5dd9c28857db\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Ok lar... Joking wif u oni...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>spam</td>\n",
              "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ham</td>\n",
              "      <td>U dun say so early hor... U c already then say...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ea00b494-7be5-4a1e-95b1-5dd9c28857db')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ea00b494-7be5-4a1e-95b1-5dd9c28857db button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ea00b494-7be5-4a1e-95b1-5dd9c28857db');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "00Irqc20dObe",
        "outputId": "0f1e10d1-6df0-4666-9f01-fbce830e2893"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(5572, 2)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['Label'].value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MnSUJZKIdQNk",
        "outputId": "c407bd66-7bb5-4790-8284-730ecc61b5c5"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ham     4825\n",
              "spam     747\n",
              "Name: Label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Non-spam: {} %'.format((4825/(4825+747)) * 100))\n",
        "print('Spam: {} %'.format((747/(4825+747)) * 100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OG1HcNB7daSn",
        "outputId": "f93f27a4-feed-437b-91de-fe651e0f04a2"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Non-spam: 86.59368269921033 %\n",
            "Spam: 13.406317300789663 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2. Training and Test Set"
      ],
      "metadata": {
        "id": "nKnJ0216NIlY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Randomize the dataset to ensure that spam and ham messages are spread properly"
      ],
      "metadata": {
        "id": "Aa9jKxM7fMeg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "randomized_df = df.sample(frac=1, random_state=1)"
      ],
      "metadata": {
        "id": "p1PFgcKb1U00"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Split the randomized dataset into a training (80% of the dataset) and a test set (20% of the dataset)"
      ],
      "metadata": {
        "id": "qWvjQVvmfRBe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "split_index = int(len(randomized_df) * 0.8)\n",
        "split_index"
      ],
      "metadata": {
        "id": "Aq7o2XMWNvW4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5e4fb2d6-5335-4f65-ab89-76c64409c7fa"
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4457"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "training_data = randomized_df[:split_index].reset_index(drop=True)\n",
        "print(len(training_data))\n",
        "training_data.head(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 161
        },
        "id": "GqykxYa_5O1K",
        "outputId": "5ed7bca6-d23e-44c7-acfc-fc6b9b9ee5dd"
      },
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4457\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Label                                            SMS\n",
              "0   ham                   Yep, by the pretty sculpture\n",
              "1   ham  Yes, princess. Are you going to make me moan?\n",
              "2   ham                     Welp apparently he retired"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-82386be5-a406-47e5-8301-68d18ac031a0\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Yep, by the pretty sculpture</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Yes, princess. Are you going to make me moan?</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ham</td>\n",
              "      <td>Welp apparently he retired</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-82386be5-a406-47e5-8301-68d18ac031a0')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-82386be5-a406-47e5-8301-68d18ac031a0 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-82386be5-a406-47e5-8301-68d18ac031a0');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "testing_data = randomized_df[split_index:].reset_index(drop=True)\n",
        "print(len(testing_data))\n",
        "testing_data.head(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 161
        },
        "id": "ojCTyh0E3eQ9",
        "outputId": "92b91ce7-5fc6-4534-f93b-ca54689a2b19"
      },
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1115\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Label                                        SMS\n",
              "0   ham                   Wherre's my boytoy ? :-(\n",
              "1   ham  Later i guess. I needa do mcat study too.\n",
              "2   ham     But i haf enuff space got like 4 mb..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9bb6c951-82ea-4162-b514-4a158595051a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Wherre's my boytoy ? :-(</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Later i guess. I needa do mcat study too.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ham</td>\n",
              "      <td>But i haf enuff space got like 4 mb...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9bb6c951-82ea-4162-b514-4a158595051a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-9bb6c951-82ea-4162-b514-4a158595051a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-9bb6c951-82ea-4162-b514-4a158595051a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Find the percentage of spam and ham in both the raining and the test set"
      ],
      "metadata": {
        "id": "0mbLMfkg6XBz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "training_data['Label'].value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UX0G8owO5quW",
        "outputId": "4f9a9b63-6fee-4be2-9f16-072a4de20b3d"
      },
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ham     3857\n",
              "spam     600\n",
              "Name: Label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Non-spam (training data): {} %'.format((3857/(3857+600)) * 100))\n",
        "print('Spam (training data): {} %'.format((600/(3857+600)) * 100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VtU-uzOh59tD",
        "outputId": "8a70a37f-c90d-4860-a46e-23b6368949fb"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Non-spam (training data): 86.53803006506618 %\n",
            "Spam (training data): 13.461969934933812 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "testing_data['Label'].value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8SiYfews5wOu",
        "outputId": "098c7031-4cf2-42f8-bafc-d53ac95dd574"
      },
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ham     968\n",
              "spam    147\n",
              "Name: Label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Non-spam (test data): {} %'.format((968/(968+147)) * 100))\n",
        "print('Spam (test data): {} %'.format((147/(968+147)) * 100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7eGhC09R5iym",
        "outputId": "0fdd2078-af46-47de-bb07-730b0e6787a3"
      },
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Non-spam (test data): 86.81614349775785 %\n",
            "Spam (test data): 13.183856502242152 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:** the percentages in the two sets are similar to what we have in the full dataset"
      ],
      "metadata": {
        "id": "-WLybezX6fln"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3. Data Cleaning"
      ],
      "metadata": {
        "id": "HFwoHa1-D-qu"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Letter Case and Punctuation"
      ],
      "metadata": {
        "id": "e65atzIBN8ZC"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Remove punctuation and convert all the words to lower case"
      ],
      "metadata": {
        "id": "-vxtBIhMCKbF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "training_data['SMS'] = training_data['SMS'].str.replace('\\W', ' ').str.lower()\n",
        "training_data.sample(3)"
      ],
      "metadata": {
        "id": "N4BiHIfRPNlM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 198
        },
        "outputId": "203fec4b-f89d-4c80-af39-310785090dab"
      },
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:1: FutureWarning: The default value of regex will change from True to False in a future version.\n",
            "  \"\"\"Entry point for launching an IPython kernel.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Label                                                SMS\n",
              "1442   ham  i sent you the prices and do you mean the   lt...\n",
              "2555   ham  dun need to use dial up juz open da browser n ...\n",
              "1616   ham                                    i ll be late   "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-34426dce-556a-4dfc-bfa8-4dac44f1f039\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1442</th>\n",
              "      <td>ham</td>\n",
              "      <td>i sent you the prices and do you mean the   lt...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2555</th>\n",
              "      <td>ham</td>\n",
              "      <td>dun need to use dial up juz open da browser n ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1616</th>\n",
              "      <td>ham</td>\n",
              "      <td>i ll be late</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-34426dce-556a-4dfc-bfa8-4dac44f1f039')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-34426dce-556a-4dfc-bfa8-4dac44f1f039 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-34426dce-556a-4dfc-bfa8-4dac44f1f039');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "testing_data['SMS'] = testing_data['SMS'].str.replace('\\W', ' ').str.lower()\n",
        "testing_data.sample(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 198
        },
        "id": "SV-itV1sDXG7",
        "outputId": "dcf5d03e-1a9c-419e-b6a6-55c4ebcd7ab1"
      },
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:1: FutureWarning: The default value of regex will change from True to False in a future version.\n",
            "  \"\"\"Entry point for launching an IPython kernel.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "    Label                                                SMS\n",
              "762   ham  i meant as an apology from me for texting you ...\n",
              "741   ham            are you angry with me  what happen dear\n",
              "162   ham  lol     i knew that      i saw him in the doll..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0f22ac1b-2593-4f5f-b9bc-ec735cfee520\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>762</th>\n",
              "      <td>ham</td>\n",
              "      <td>i meant as an apology from me for texting you ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>741</th>\n",
              "      <td>ham</td>\n",
              "      <td>are you angry with me  what happen dear</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>162</th>\n",
              "      <td>ham</td>\n",
              "      <td>lol     i knew that      i saw him in the doll...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0f22ac1b-2593-4f5f-b9bc-ec735cfee520')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-0f22ac1b-2593-4f5f-b9bc-ec735cfee520 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-0f22ac1b-2593-4f5f-b9bc-ec735cfee520');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Creating The Vocabulary"
      ],
      "metadata": {
        "id": "BddOsH7kPWgC"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here, we will create a list with all of the unique words that occur in the messages of our training set"
      ],
      "metadata": {
        "id": "a8fW_O3pHBaK"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Transform each message from the SMS column into a list by splitting the string at the space character"
      ],
      "metadata": {
        "id": "5ia84ceOFdat"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "training_data['SMS'] = training_data['SMS'].str.split()\n",
        "training_data.sample(3)"
      ],
      "metadata": {
        "id": "vNLnQOucPgcA",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "4f213d47-22fa-420b-8da7-1c69d22886f2"
      },
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Label                                                SMS\n",
              "427    ham  [okey, doke, i, m, at, home, but, not, dressed...\n",
              "1496   ham              [watching, tv, now, i, got, new, job]\n",
              "120    ham  [get, me, out, of, this, dump, heap, my, mom, ..."
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d8402b1f-e1ca-4306-a464-38660b34e0e0\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>427</th>\n",
              "      <td>ham</td>\n",
              "      <td>[okey, doke, i, m, at, home, but, not, dressed...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1496</th>\n",
              "      <td>ham</td>\n",
              "      <td>[watching, tv, now, i, got, new, job]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>120</th>\n",
              "      <td>ham</td>\n",
              "      <td>[get, me, out, of, this, dump, heap, my, mom, ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d8402b1f-e1ca-4306-a464-38660b34e0e0')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-d8402b1f-e1ca-4306-a464-38660b34e0e0 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-d8402b1f-e1ca-4306-a464-38660b34e0e0');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Iterate each message in the SMS column (each message should be a list of strings) and append each word to the vocabulary list"
      ],
      "metadata": {
        "id": "X-al4JDDFUgO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "vocabulary = []\n",
        "\n",
        "for msg in training_data['SMS']:\n",
        "  for word in msg:\n",
        "    vocabulary.append(word)"
      ],
      "metadata": {
        "id": "0sWuQsrfFVRe"
      },
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Remove duplicates in the vocabulary list"
      ],
      "metadata": {
        "id": "L7kaNo6BGIuP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "vocabulary = list(set(vocabulary))\n",
        "len(vocabulary)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0KqNIqtuGMPP",
        "outputId": "e9aaf369-d33d-4fc6-e017-7eb509f87ff9"
      },
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "7782"
            ]
          },
          "metadata": {},
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### The Final Training Set"
      ],
      "metadata": {
        "id": "MCvSR4dUP-N0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this section, we will use the training set to teach the algorithm to classify new messages.\n",
        "\n",
        "Recall from the previous section that when a new message comes in, our Naive Bayes algorithm will make the classification based on the results it gets to these two equations (note that we replaced \"SpamC\" with \"Ham\" inside the second equation below):\n",
        "\n",
        "![Screenshot 2022-01-17 at 11.46.20.png](data:image/png;base64,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)\n",
        "\n",
        "Also, to calculate P(wi|Spam) and P(wi|Ham) inside the formulas above, recall that we need to use these equations:\n",
        "\n",
        "![Screenshot 2022-01-17 at 11.46.24.png](data:image/png;base64,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)\n",
        "\n",
        "Let's also summarize what the terms in the equations above mean:\n",
        "\n",
        "![Screenshot 2022-01-17 at 11.46.30.png](data:image/png;base64,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)\n"
      ],
      "metadata": {
        "id": "qP4-8b-_OAD0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "To make the calculations easier, we want bring the data to this format (the table below is a transformation of the table you see above):\n",
        "\n",
        "![Screenshot 2022-01-17 at 11.56.00.png](data:image/png;base64,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)\n",
        "\n",
        "About the transformation above, notice that:\n",
        "\n",
        "* The SMS column doesn't exist anymore.\n",
        "* Instead, the SMS column is replaced by a series of new columns, where each column represents a unique word from the vocabulary.\n",
        "* Each row describes a single message. For instance, the first row corresponds to the message \"SECRET PRIZE! CLAIM SECRET PRIZE NOW!!\", and it has the values spam, 2, 2, 1, 1, 0, 0, 0, 0, 0. These values tell us that:\n",
        "The message is spam.\n",
        "The word \"secret\" occurs two times inside the message.\n",
        "The word \"prize\" occurs two times inside the message.\n",
        "The word \"claim\" occurs one time inside the message.\n",
        "The word \"now\" occurs one time inside the message.\n",
        "The words \"coming\", \"to\", \"my\", \"party\", and \"winner\" occur zero times inside the message.\n",
        "\n",
        "* All words in the vocabulary are in lower case, so \"SECRET\" and \"secret\" come to be considered to be the same word.\n",
        "\n",
        "\n",
        "Eventually, we're going to create a new DataFrame. However, we'll first build a dictionary that we'll then convert to the DataFrame we need.\n",
        "\n",
        "For instance, to create the table we see above, we could use this dictionary and then convert it to a DataFrame:"
      ],
      "metadata": {
        "id": "7rFJ-FehQLHV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts_per_sms = {'secret': [2,1,1],\n",
        "                       'prize': [2,0,1],\n",
        "                       'claim': [1,0,1],\n",
        "                       'now': [1,0,1],\n",
        "                       'coming': [0,1,0],\n",
        "                       'to': [0,1,0],\n",
        "                       'my': [0,1,0],\n",
        "                       'party': [0,1,0],\n",
        "                       'winner': [0,0,1]\n",
        "                      }\n",
        "\n",
        "\n",
        "word_counts = pd.DataFrame(word_counts_per_sms)\n",
        "word_counts.head()"
      ],
      "metadata": {
        "id": "pzmuqdgsQW18",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "b1f89d0e-2f09-4ca6-d359-0e0aee9f4c98"
      },
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   secret  prize  claim  now  coming  to  my  party  winner\n",
              "0       2      2      1    1       0   0   0      0       0\n",
              "1       1      0      0    0       1   1   1      1       0\n",
              "2       1      1      1    1       0   0   0      0       1"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e2f296f1-0a04-4047-81e6-ff8d2638e02c\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>secret</th>\n",
              "      <th>prize</th>\n",
              "      <th>claim</th>\n",
              "      <th>now</th>\n",
              "      <th>coming</th>\n",
              "      <th>to</th>\n",
              "      <th>my</th>\n",
              "      <th>party</th>\n",
              "      <th>winner</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e2f296f1-0a04-4047-81e6-ff8d2638e02c')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-e2f296f1-0a04-4047-81e6-ff8d2638e02c button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-e2f296f1-0a04-4047-81e6-ff8d2638e02c');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "(As you may have noticed from the output above, the Label column is missing, but we'll get to that shortly)\n",
        "\n",
        "To create the dictionary we need for our training set, we can use the code below, where:\n",
        "\n",
        "* We start by initializing a dictionary named `word_counts_per_sms`, where each key is a unique word (a string) from the vocabulary, and each value is a list of the length of training set, where each element in the list is a 0.\n",
        "\n",
        "> The code `[0] * 5` outputs `[0, 0, 0, 0, 0]`. So the code `[0] * len(training_set['SMS'])` outputs a list of the length of `training_set['SMS']`, where each element in the list will be a 0.\n",
        "\n",
        "* We loop over training_set['SMS'] using at the same time the `enumerate()` function to get both the index and the SMS message (index and sms)."
      ],
      "metadata": {
        "id": "XljVkYbXQYDC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts_per_sms = {unique_word: [0] * len(training_data['SMS']) for unique_word in vocabulary}\n",
        "\n",
        "for index, sms in enumerate(training_data['SMS']):\n",
        "    for word in sms:\n",
        "        word_counts_per_sms[word][index] += 1"
      ],
      "metadata": {
        "id": "aoxT-2NxQsvI"
      },
      "execution_count": 65,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Transform `word_counts_per_sms` into a DataFrame\n"
      ],
      "metadata": {
        "id": "3RZHy4bLQwSw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts_per_sms_df = pd.DataFrame(word_counts_per_sms)\n",
        "word_counts_per_sms_df.head(3)"
      ],
      "metadata": {
        "id": "_gQzKCQBQ9ir",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237
        },
        "outputId": "21a5307b-3cfb-4f4d-9ce1-3c0354fc4a3c"
      },
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   dogwood  dun  lyrics  easter  runs  neighbors  shuhui  brin  brains  \\\n",
              "0        0    0       0       0     0          0       0     0       0   \n",
              "1        0    0       0       0     0          0       0     0       0   \n",
              "2        0    0       0       0     0          0       0     0       0   \n",
              "\n",
              "   plenty  ...  toilet  mel  weaseling  wud  ques  dps  fees  clark  or  wrnog  \n",
              "0       0  ...       0    0          0    0     0    0     0      0   0      0  \n",
              "1       0  ...       0    0          0    0     0    0     0      0   0      0  \n",
              "2       0  ...       0    0          0    0     0    0     0      0   0      0  \n",
              "\n",
              "[3 rows x 7782 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e862ebca-c58e-48a0-966f-e834e99d67a4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>dogwood</th>\n",
              "      <th>dun</th>\n",
              "      <th>lyrics</th>\n",
              "      <th>easter</th>\n",
              "      <th>runs</th>\n",
              "      <th>neighbors</th>\n",
              "      <th>shuhui</th>\n",
              "      <th>brin</th>\n",
              "      <th>brains</th>\n",
              "      <th>plenty</th>\n",
              "      <th>...</th>\n",
              "      <th>toilet</th>\n",
              "      <th>mel</th>\n",
              "      <th>weaseling</th>\n",
              "      <th>wud</th>\n",
              "      <th>ques</th>\n",
              "      <th>dps</th>\n",
              "      <th>fees</th>\n",
              "      <th>clark</th>\n",
              "      <th>or</th>\n",
              "      <th>wrnog</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3 rows × 7782 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e862ebca-c58e-48a0-966f-e834e99d67a4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-e862ebca-c58e-48a0-966f-e834e99d67a4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-e862ebca-c58e-48a0-966f-e834e99d67a4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts_per_sms_df.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DjD7ssvWLsew",
        "outputId": "5e8ea09b-9690-4fdc-9a50-20c1c7dabb0b"
      },
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(4457, 7782)"
            ]
          },
          "metadata": {},
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Concatenate the DataFrame we just built above with the DataFrame containing the training set (this way, we'll also have the Label and the SMS columns). Use the `pd.concat()` function."
      ],
      "metadata": {
        "id": "0jTgtU83K9w2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "training_data_clean = pd.concat([training_data, word_counts_per_sms_df], axis=1)\n",
        "training_data_clean.head(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 393
        },
        "id": "D9Zy13syK-RN",
        "outputId": "2d032769-9064-4894-8e54-9174f0db90c0"
      },
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Label                                                SMS  dogwood  dun  \\\n",
              "0   ham                  [yep, by, the, pretty, sculpture]        0    0   \n",
              "1   ham  [yes, princess, are, you, going, to, make, me,...        0    0   \n",
              "2   ham                    [welp, apparently, he, retired]        0    0   \n",
              "\n",
              "   lyrics  easter  runs  neighbors  shuhui  brin  ...  toilet  mel  weaseling  \\\n",
              "0       0       0     0          0       0     0  ...       0    0          0   \n",
              "1       0       0     0          0       0     0  ...       0    0          0   \n",
              "2       0       0     0          0       0     0  ...       0    0          0   \n",
              "\n",
              "   wud  ques  dps  fees  clark  or  wrnog  \n",
              "0    0     0    0     0      0   0      0  \n",
              "1    0     0    0     0      0   0      0  \n",
              "2    0     0    0     0      0   0      0  \n",
              "\n",
              "[3 rows x 7784 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-190232a2-3498-4af8-aa43-65eec218bcd6\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "      <th>dogwood</th>\n",
              "      <th>dun</th>\n",
              "      <th>lyrics</th>\n",
              "      <th>easter</th>\n",
              "      <th>runs</th>\n",
              "      <th>neighbors</th>\n",
              "      <th>shuhui</th>\n",
              "      <th>brin</th>\n",
              "      <th>...</th>\n",
              "      <th>toilet</th>\n",
              "      <th>mel</th>\n",
              "      <th>weaseling</th>\n",
              "      <th>wud</th>\n",
              "      <th>ques</th>\n",
              "      <th>dps</th>\n",
              "      <th>fees</th>\n",
              "      <th>clark</th>\n",
              "      <th>or</th>\n",
              "      <th>wrnog</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>[yep, by, the, pretty, sculpture]</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>[yes, princess, are, you, going, to, make, me,...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ham</td>\n",
              "      <td>[welp, apparently, he, retired]</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3 rows × 7784 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-190232a2-3498-4af8-aa43-65eec218bcd6')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-190232a2-3498-4af8-aa43-65eec218bcd6 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-190232a2-3498-4af8-aa43-65eec218bcd6');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 4. Calculating Constants First"
      ],
      "metadata": {
        "id": "-y3u33QUXAUk"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The Naive Bayes algorithm will need to know the probability values of the two equations below to be able to classify new messages:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.26.52.png](data:image/png;base64,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)\n",
        "\n",
        "Also, to calculate P(wi|Spam) and P(wi|Ham) inside the formulas above, recall that we need to use these equations:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.26.55.png](data:image/png;base64,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)\n",
        "\n",
        "Some of the terms in the four equations above will have the same value for every new message. As a start, let's first calculate:\n",
        "\n",
        "\n",
        "* P(Spam) and P(Ham)\n",
        "* NSpam, NHam, NVocabulary\n",
        "\n",
        "Note that:\n",
        "\n",
        "* NSpam is equal to the number of words in all the spam messages — it's not equal to the number of spam messages, and it's not equal to the total number of unique words in spam messages.\n",
        "* NHam is equal to the number of words in all the non-spam messages — it's not equal to the number of non-spam messages, and it's not equal to the total number of unique words in non-spam messages.\n",
        "\n",
        "We'll also use Laplace smoothing and set `α = 1`."
      ],
      "metadata": {
        "id": "AZFhd-a4XDX3"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Calculate P(Spam) and P(Ham)"
      ],
      "metadata": {
        "id": "cYV4myhpMviR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "spam_messages = training_data_clean[training_data_clean['Label'] == 'spam']\n",
        "ham_messages = training_data_clean[training_data_clean['Label'] == 'ham']"
      ],
      "metadata": {
        "id": "lUBaD5-mX8ZK"
      },
      "execution_count": 72,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "p_spam = len(spam_messages) / len(training_data_clean)\n",
        "p_ham = len(ham_messages) / len(training_data_clean)"
      ],
      "metadata": {
        "id": "tITKd4z8MzrD"
      },
      "execution_count": 73,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Calculate NSpam, NHam, NVocabulary"
      ],
      "metadata": {
        "id": "_6Sh2xvSMx2S"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "num_words_in_spam_message = spam_messages['SMS'].apply(len)\n",
        "n_spam = num_words_in_spam_message.sum()\n",
        "\n",
        "num_words_in_ham_message = ham_messages['SMS'].apply(len)\n",
        "n_ham = num_words_in_ham_message.sum()\n",
        "\n",
        "n_vocabulary = len(vocabulary)"
      ],
      "metadata": {
        "id": "4NXL9iYTNZpZ"
      },
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Laplace smoothing"
      ],
      "metadata": {
        "id": "6OXBa7GgNzvM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "alpha = 1"
      ],
      "metadata": {
        "id": "2AAtWxU1Nxmv"
      },
      "execution_count": 75,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 5. Calculating Parameters"
      ],
      "metadata": {
        "id": "TjKkfQYOX-zo"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "P(Spam), P(Ham), NSpam, NHam, and NVocabulary will have constant values in our equations for every new message (regardless of the message or each individual word in the message).\n",
        "\n",
        "However, `P(wi|Spam)` and `P(wi|Ham)` will vary depending on the individual words. For instance, P(\"secret\"|Spam) will have a certain probability value, while `P(\"cousin\"|Spam)` or `P(\"lovely\"|Spam)` will most likely have other values.\n",
        "\n",
        "Although both `P(wi|Spam)` and `P(wi|Ham)` vary depending on the word, the probability for each individual word is constant for every new message.\n",
        "\n",
        "For instance, let's say we receive two new messages:\n",
        "\n",
        "* \"secret code\"\n",
        "* \"secret party 2night\"\n",
        "\n",
        "We'll need to calculate `P(\"secret\"|Spam)` for both these messages, and we can use the training set to get the values we need to find a result for the equation below:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.31.00.png](data:image/png;base64,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)\n",
        "\n",
        "The steps we take to calculate `P(\"secret\"|Spam)` will be identical for both of our new messages above, or for any other new message that contains the word \"secret\". The key detail here is that calculating `P(\"secret\"|Spam)` only depends on the training set, and as long as we don't make changes to the training set, `P(\"secret\"|Spam)` stays constant. The same reasoning also applies to `P(\"secret\"|Ham)`.\n",
        "\n",
        "This means that we can use our training set to calculate the probability for each word in our vocabulary. If our vocabulary contained only the words \"lost\", \"navigate\", and \"sea\", then we'd need to calculate six probabilities:\n",
        "\n",
        "* P(\"lost\"|Spam) and P(\"lost\"|Ham)\n",
        "* P(\"navigate\"|Spam) and P(\"navigate\"|Ham)\n",
        "* P(\"sea\"|Spam) and P(\"sea\"|Ham)\n",
        "\n",
        "We have 7,782 words in our vocabulary, which means we'll need to calculate a total of 15,566 probabilities. For each word, we need to calculate both `P(wi|Spam)` and `P(wi|Ham)`.\n",
        "\n",
        "In more technical language, the probability values that `P(wi|Spam)` and `P(wi|Ham)` will take are called parameters.\n",
        "\n",
        "The fact that we calculate so many values before even beginning the classification of new messages makes the Naive Bayes algorithm very fast (especially compared to other algorithms). When a new message comes in, most of the needed computations are already done, which enables the algorithm to almost instantly classify the new message.\n",
        "\n",
        "If we didn't calculate all these values beforehand, then all these calculations would need to be done every time a new message comes in. Imagine the algorithm will be used to classify 1,000,000 new messages. Why repeat all these calculations 1,000,000 times when we could just do them once at the beginning?\n",
        "\n",
        "Let's now calculate all the parameters using the equations below:\n",
        "\n",
        "\n",
        "Recall that `P(wi|Spam)` and `P(wi|Ham)` are key parts of the equations that we need to classify the new messages:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.31.10.png](data:image/png;base64,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)\n",
        "\n"
      ],
      "metadata": {
        "id": "jkCpcLQ3YKtB"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Initialize two dictionaries, where each key-value pair is a unique word (from our vocabulary) represented as a string, and the value is 0. We'll need one dictionary to store the parameters for `P(wi|Spam)`, and the other for P(wi|Ham)"
      ],
      "metadata": {
        "id": "RS--yOYUPzaM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "spam_parameters = {unique_word:0 for unique_word in vocabulary}\n",
        "ham_parameters = {unique_word:0 for unique_word in vocabulary}"
      ],
      "metadata": {
        "id": "POETyGUOY7CD"
      },
      "execution_count": 79,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Iterate over the vocabulary, and, for each word, calculate `P(wi|Spam)` and `P(wi|Ham)` using the formulas we mentioned above.\n",
        "\n",
        "Nwi|Spam is equal to the number of times the word wi occurs in all the spam messages, while Nwi|Ham is equal to the number of times the word wi occurs in all the ham messages.\n",
        "\n",
        "Once done calculating an individual parameter, we update the probability value in the two dictionaries we created initially"
      ],
      "metadata": {
        "id": "TvkVSBNZQwmY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "for word in vocabulary:\n",
        "    word_given_spam = spam_messages[word].sum() \n",
        "    p_spam_given_word = (word_given_spam + alpha) / (n_spam + alpha * n_vocabulary)\n",
        "    spam_parameters[word] = p_spam_given_word\n",
        "    \n",
        "    word_given_ham = ham_messages[word].sum()  \n",
        "    p_ham_given_word = (word_given_ham + alpha) / (n_ham + alpha*n_vocabulary)\n",
        "    ham_parameters[word] = p_ham_given_word"
      ],
      "metadata": {
        "id": "8AhfjficQw4E"
      },
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 6. Classifying a New Message"
      ],
      "metadata": {
        "id": "_blPXL9QZl0M"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We will now start creating the spam filter. The spam filter can be understood as a function that:\n",
        "\n",
        "* Takes in as input a new message (w1, w2, ..., wn)\n",
        "* Calculates P(Spam|w1, w2, ..., wn) and P(Ham|w1, w2, ..., wn)\n",
        "* Compares the values of P(Spam|w1, w2, ..., wn) and P(Ham|w1, w2, ..., wn), and:\n",
        "\n",
        "If P(Ham|w1, w2, ..., wn) > P(Spam|w1, w2, ..., wn), then the message is classified as ham.\n",
        "\n",
        "If P(Ham|w1, w2, ..., wn) < P(Spam|w1, w2, ..., wn), then the message is classified as spam.\n",
        "\n",
        "If P(Ham|w1, w2, ..., wn) = P(Spam|w1, w2, ..., wn), then the algorithm may request human help.\n",
        "\n",
        "Note that:\n",
        "\n",
        "* The input variable message is assumed to be a string.\n",
        "* We perform a bit of data cleaning on the string message:\n",
        "\n",
        "We remove the punctuation using the `re.sub()` function.\n",
        "\n",
        "We bring all letters to lower case using the `str.lower()` method.\n",
        "\n",
        "We split the string at the space character and transform it into a Python list using the `str.split()` method.\n",
        "\n",
        "* We compare `p_spam_given_message` with `p_ham_given_message` and then print a classification label.\n",
        "\n",
        "To write the code we need for calculating `p_spam_given_message` and `p_ham_given_message`, we need to use these two equations:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.38.09.png](data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAfwAAACXCAYAAADwFuk+AAAYVGlDQ1BJQ0MgUHJvZmlsZQAAWIWVeQVUVF8X77mTzDAM3d0l3SAxdHeDwNAdQ4NBiYSKIKCUCioIIliEiIUgooigAgYiIUqpoIIiIO8S+v2//7fWe+udtc69v9lnnx1nn9pzAeBMIUdGhiLoAAgLj6HYGunyObu48mGnAAFgARFAAEn2iY4kWVubA7j8ef93WR6C+eDyXGpT1v+2/18Lva9ftA8AkDWMvX2jfcJgfB0AVIZPJCUGAIwqTBeMj4ncxO4wZqLABsI4chMHbOOMTey9jU9u8djb6sG4HgAqAplMCQCA2ArT+eJ8AmA5xJdwG0O4b1A4zDoPYy2fQLIvAJy7YJ5dYWERm9gZxmLe/5AT8F8yvf/KJJMD/uJtX7YKlX5QdGQoOfH/czj+3yUsNPaPDhG4EgIpxrabPsPj9jIkwmwTE2A8H+5taQVjBhj/DPLd4ocxAh8Ya+ywzY/g8onWg8cMsMBY1pesbwZjLhgbhodamu/Qvf2DDE1gDM8QREJQjIn9Tt8sv2gDux2Z5ZQIW6s/2J+iR9rp20CmbOnd5O+MDXEg7ch/Gehn8kf+96RAe6dtm5H4uCBHSxgTYcwSHWJnts2DFEoK1LP8w0OJtd20XwjG6n7hRrrb8pEe/hRD2x1+Slj0H3+RWYFBJpY7uDQm0N54R069D3nLfjYYt/qFkxz+yPGLdjb/44uvn77Btu/Ifr9whx1/kWORMbq2O32/RoZa7/Cj8H6hRpt0ARhzRcfZ7fRFacXAE3JbPsoyMsbafttOlHcw2dR62x5UAjAHekAf8IFYuHqDCBAMgvrmW+bhX9sthoAMKCAA+AGpHcqfHk5bLeHw0w4kgU8w8gPRf/vpbrX6gTiYvv6Xuv2UAv5brXFbPULABxiHATMQCv+O3eoV/lebI3gPU4L+R7sPbGsoXDfb/pdGginmO5TYP3L5aP9wYgww+hhjjCFGHMWB0kJpoMzhpw5c5VGqKLU/1v6HH/0BPYCeQA+ix9CvPIPSKP/yhw9YgDFYg+GOz97/9BklAktVQumiNGH5sGwUC4oDSKEUYU0klDasWwmm6u1Yvun9v2X/lw//GPUdPpwsDoFjxengxP7dkyhBVPorZXNM/zlC27Z6/x1Xvb8t/9av94+R9oXfZv/mRGYhryG7kfeQPch2ZAvgQ95BtiJ7kbc28d9Z9H5rFv3RZrtlTwgsJ+h/9JF3dG6OZLTsRdkZ2bXtthi/hJjNBaYXEZlICQoIjOEjwTu/H59JuI/0Lj55WXk5ADbPke1t6pvt1vkAsTz9D81vGoDd8NrA9f+HFnwcgLouAFhz/kMTcQOAHd6XrzzziaXEbdNQmw80wANaeEWxAx4gCMRgf+SBMtAAOsAAmAIrYA9cgAc8yoHwfKaAeLAXpIJMkAuOgSJQCk6Ds+ACuASughbQDu6BB+Ax6AeD4A08e6bAHFgAy2AVgiAsRAMxQuwQLyQMSULykCqkBRlA5pAt5AJ5QQFQOBQL7YXSoVyoACqFKqFa6Ap0A7oH9UAD0CtoHJqBvkK/EEgEAcGE4EaIIGQQqggSwgxhj9iDCEBEIZIQGYijiJOIKkQ9ohlxD/EYMYgYQ8whlpAASY1kQfIjpZCqSD2kFdIV6Y+kIPcjc5DFyCpkA7INjvNz5BhyHrmCwqAYUXwoKXgGG6McUD6oKNR+1GFUKeoCqhnViXqOGkctoH6jadBcaEm0OtoE7YwOQMejM9HF6Gp0E7oLXk1T6GUMBsOCEcWowKvRBROMScYcxlRgGjF3MQOYScwSFotlx0piNbFWWDI2BpuJLcHWY+9gn2GnsD+pqKl4qeSpDKlcqcKp0qiKqeqoblM9o/pItYqjwwnj1HFWOF9cIi4Pdw7XhnuKm8Kt4unxonhNvD0+GJ+KP4lvwHfhR/DfqKmpBajVqG2og6hTqE9SX6Z+SD1OvUJgIEgQ9AjuhFjCUUIN4S7hFeEbDQ2NCI0OjStNDM1Rmlqa+zSjND+JjERpognRl3iAWEZsJj4jfqbF0QrTkmg9aJNoi2mv0T6lnafD0YnQ6dGR6fbTldHdoBumW6JnpJejt6IPoz9MX0ffQz/NgGUQYTBg8GXIYDjLcJ9hkhHJKMiox+jDmM54jrGLcYoJwyTKZMIUzJTLdImpj2mBmYFZkdmROYG5jPkW8xgLkkWExYQllCWP5SrLEMsvVm5WEqsfazZrA+sz1h9snGw6bH5sOWyNbINsv9j52A3YQ9jz2VvY33KgOCQ4bDjiOU5xdHHMczJxanD6cOZwXuV8zYXgkuCy5UrmOsvVy7XEzcNtxB3JXcJ9n3ueh4VHhyeYp5DnNs8MLyOvFm8QbyHvHd5ZPmY+El8o30m+Tr4Ffi5+Y/5Y/kr+Pv5VAVEBB4E0gUaBt4J4QVVBf8FCwQ7BBSFeIQuhvUIXhV4L44RVhQOFTwh3C/8QERVxEjkk0iIyLcomaiKaJHpRdESMRkxbLEqsSuyFOEZcVTxEvEK8XwIhoSQRKFEm8VQSIaksGSRZITmwC71LbVf4rqpdw1IEKZJUnNRFqXFpFmlz6TTpFunPMkIyrjL5Mt0yv2WVZENlz8m+kWOQM5VLk2uT+yovIe8jXyb/QoFGwVDhgEKrwqKipKKf4inFl0qMShZKh5Q6lNaVVZQpyg3KMypCKl4q5SrDqkyq1qqHVR+qodV01Q6otautqCurx6hfVf+iIaURolGnMb1bdLff7nO7JzUFNMmalZpjWnxaXlpntMa0+bXJ2lXaEzqCOr461TofSeKkYFI96bOurC5Ft0n3h5663j69u/pIfSP9HP0+AwYDB4NSg1FDAcMAw4uGC0ZKRslGd43RxmbG+cbDJtwmPia1JgumKqb7TDvNCGZ2ZqVmE+YS5hTzNguEhanFcYsRS2HLcMsWK2BlYnXc6q21qHWU9U0bjI21TZnNB1s527223XaMdp52dXbL9rr2efZvHMQcYh06HGkd3R1rHX846TsVOI05yzjvc37swuES5NLqinV1dK12XXIzcCtym3JXcs90H9ojuidhT48Hh0eoxy1PWk+y5zUvtJeTV53XGtmKXEVe8jbxLvde8NHzOeEz56vjW+g746fpV+D30V/Tv8B/OkAz4HjATKB2YHHgfJBeUGnQYrBx8OngHyFWITUhG6FOoY1hVGFeYTfCGcJDwjsjeCISIgYiJSMzI8ei1KOKohYoZpTqaCh6T3RrDBN8Ye+NFYs9GDsepxVXFvcz3jH+WgJ9QnhCb6JEYnbixyTDpPPJqGSf5I69/HtT947vI+2r3A/t997fcUDwQMaBqRSjlAup+NSQ1CdpsmkFad/TndLbMrgzUjImDxodvJhJzKRkDh/SOHQ6C5UVlNWXrZBdkv07xzfnUa5sbnHu2mGfw4+OyB05eWTjqP/RvjzlvFPHMMfCjw3la+dfKKAvSCqYPG5xvLmQrzCn8HuRZ1FPsWLx6RP4E7Enxk6an2wtESo5VrJWGlg6WKZb1ljOVZ5d/qPCt+LZKZ1TDae5T+ee/nUm6MzLSqPK5iqRquKzmLNxZz+cczzXfV71fG01R3Vu9XpNeM3YBdsLnbUqtbV1XHV5FxEXYy/O1LvX91/Sv9TaINVQ2cjSmHsZXI69PHvF68rQVbOrHddUrzVcF75e3sTYlNMMNSc2L7QEtoy1urQO3DC90dGm0dZ0U/pmTTt/e9kt5lt5t/G3M25v3Em6s3Q38u78vYB7kx2eHW/uO99/0WnT2ddl1vXwgeGD+92k7jsPNR+296j33Hik+qjlsfLj5l6l3qYnSk+a+pT7mp+qPG3tV+tvG9g9cPuZ9rN7z/WfP3hh8uLxoOXgwJDD0Mth9+Gxl74vp1+Fvlp8Hfd69U3KCHok5y3d2+JRrtGqd+LvGseUx26N64/3TthNvJn0mZx7H/1+bSrjA82H4o+8H2un5afbZwxn+mfdZqfmIudW5zM/0X8q/yz2+foXnS+9C84LU4uUxY2vh7+xf6v5rvi9Y8l6aXQ5bHn1R85P9p8XVlRXun85/fq4Gr+GXTu5Lr7e9tvs98hG2MZGJJlC3roKIOGK8PcH4GsNADQuADD2A4B3287zdgoSvnwg4LcjZIAgIVVRbGg8hgorS+WCS8ffIWBoyMQWOjx9KMMjJiXmclbAFsLex6nMdYx7jleHL49/QBAvpCbsIhIiGibmLq4rwS2xKPlgV4lUiLSmDI3MO9lGuRR5GwV+hU+KN5QOKtuocKlMqTaoJaiTNPAaz3eXa/pq7dL6qt2is5ekq0vQfad3W7/OoMIw32i/MdlE25TNdNGs17zBosKy0qrdetIWbcduz+FA54h0XHNadQGuODeiO80e1J4ljwnPfq+75Gve1T4lvjl+if4BAfaBukGKwRIh/KHsYbThyPDvEROR/VE3Keeij8YciM2Ma0pAJfol3d0L9onsVz9gkuKWGpt2NL0oI/mg4sHJzLxD1lnC2dQ5IBdxmP6I2FGtPMtjTvmuBa7HnQsdi+yLbU5YnjQrMSrVLdMqV6tQOCV1WuKMbKVZVfrZsfMm1fU1c7X0dcIX5eo1Luk3WDQ6Xfa8Eng18lr89f1Nac0HW7Jac2/ktRXdLG+vvnX9dted4btj94Y6Gu/7d7J1PuwqfhDf7f9wT4/TI5vHZr1GT4z77J9G9Z8ZePWc+oXMoN6QybDBS9VXwq+Jr1feTI+8fHtv9Oy79LGAcYcJy0mL91ZTVh9MP6pNs06PzeTMKs6OzV2YT/pk/Jnqc+0Xoy+TC2cXE756fLP6brEUvNzx89CvlnX9jY2d+MshUcgZ1Bh6ErNAhcQp4wOpywljRAnaeLoHDOyMiUwvWORZ09jecihxZnL183DwOvPl87cLjAguCS0Lz4o8ET0rRhHXkqCSeCF5elewlJLUb+kHMkdlneR45T7KNyjEKWoqQUpdyjkqVqqMqkNqJepuGtwaI/AscNdi1xrWPqHjRhIhreoO6l3RP2zgZ7jbiN7og3G7SZFpnJmfubdFoGWEVZi1t42VrYadhD2nA9ER4bjs9NF5yOW+a4NbmXvOniSPIE9nL32yjDebD+Qz6zvo1+nfFFAdWByUERwR4hKqEyYaTgPPhPHI0ajv0fwxnrElcffiXyZMJs4nreyl3sezX+wAXwom5V1qU1peOiXD46BDpvOhoKz07IqcS7lNh5uPXD96Je/Ssdr88wVnjpcVFhXlFWefSDuZWBJRGlAWVJ5Scee0+JkLVaJnC849P79SQ7zAUStYJwHPA5VLWg36jRaXXa6EXs28dvb67aaB5tGW6dZvbcibrO2StzRu69xRuct/D3FvoqP7flNnTVfZg2PdBx8m9VAexTzO7m3vY3m6r//tM47n2i/sB/2HUobPv3z66vsbhhGpt+ajke9OjN0cfzYxOjnxfu4DGo5+6szAHP287CelzyJfaL/8XPiwOPz10bcb3yuXDiw7/hD9sfyzfSXpl8YqYU1/fWYn/tLQHKIC6YESR2PRi5gZ7CzVBG6RGk8QpiERXWlT6erpBxg2mISZDViCWQ+ynWa/ztHF+ZDrAfdNnkreBD5dvl/85wTMBOYEs4REhTqEPYRXRApFZUUfiQWIY8VrJIwlPkpm7hLb1SXlIw2kK2R2y7yUjYVvN43y5vLTCumKPIqtSrZK88oHVXhVWuBby7TaAXUW9YsaJI1nu312f9ZM1sJqlWkrag/pJJF4SK26Vrqv9AL1NvSrDKwNcYb3jfYaKxrPmlSZupuxmQ2ZF1nYWdJa9lilW2tYf7dptA2xE7V7b1/psMeR3fGFU56zsfOGS5NrqJuQ21v34j2We5Y9Cj2FPa97kbxekxO8BbxfwvtIoJ+Rv0qAWqBJEDk4LIQcqh1GFzYSfj4iLFIpci3qPiUn2jqGOeZN7Ok433iR+A8JpxINEkeSQpOZkp/vvbnv9v7OA/dTbqTWphWnp2dEHHTLNDgkkYXOepFdkuOaK5S7enjsyJOjN/LOHNuf71agfpzj+ErhUNHV4hMnjpwsKKksvVb2oPxlxeyp1TM0lXxVCmeNz7mfj6jeX5N94XBtSh35oko9sf7rpU8NK5cJV3iuyl+zvp7cdL35Z6vajci2kpuX21tv3bzdc2fpnlHHjU67rqXu4h6FRy96j/R59Zs8I73QHQp9RRyZm+ibXfq+shn/7f/7NgtGGYDjqXCGmgmAgzYA+Z1wnjkI5514AKxpALBXAwgRf4Ag9AJIffzv+QHBpw0GUAN6wAZ4gSiQBepwlmwFXIE/iIazyzxwCjSA2+ApGAff4cyRC5KDjCBPKB7Kh+qhh9AHBAYhhjBHRCMq4DxvA87r4pA3kL9RRqjjqAm0AjoL/Q6jjinBrMIZ1iMqFaoaHCcuH0+Nz6bGUx8jcBBqaBRp2omaxDZaVdqbdMZ0b+hjGOgYLjHqMw4w2TMNMFsxP2PxZPnJWsKmyTbKvo+Dk6ON04MLx9XOHcejyPON9yofhV+Jf02gW7BYKFB4twhRZEz0mliWuLcESVJkF3HXqtRn6fcyg7JNcsnycvKjClmKSopflFqVC1QSVX3VzNVlNVh3EzWltcp0JElHdHv0vhhQGTIbsRtzmQiZKppZmkdZnLTstPpqI2jrZHfUvtsR5aTvnOnS68bi7r2nzuO9F4ZM743xXvKZ8h3xmw2gDTQLKgr+GLo7rDD8c6RpVF00ISYq9nW8YUJrklRy9T6+/WUpLKn56fiM1INLh4Kz5nJyD4cdbcqnP85R+Km49qRnKUtZf8WR00ZnlqryzjGdz6pevhBS+/XisUsGjfSXF69+uD7dPNf6sW2yffEO6z29+x5dXt12PdqPZZ6IP1UeCH/+cxj1Gjdy+h3j+O0p4vTeOdKnxi+rX5W/Gy7jfxz5+Whl+tfU6qu16+vHfntvyG7tH5vxxwICYADsgB9IAAWgCYyBPfACYSAZZIMSUAtugMfgLViA0BAHJLsV/USoELoM9UGfELQIBYQrIh1xFTGF5EV6Is8h51HKqAzUIFocnYoegWNfhgXYQOwglQFVK04GV4cXx9dTK1LfIVgTJmkSiDhiES0/7WU4f31DH8/AwtDC6Mj4iWkfM575JIsUyyPWCDZWtrvsQRxMHHc5I7iEuEa4S3icedl4X/FV8PsKyAoCwRdCF4UzRNxFFeFcbla8V+IafIrlSaVL75WJkfWR05EnyPcp5CiaKbEqLSq/UulWbVarUj+skbQ7TjNbq1X7B0lB11cvV7/aoNnwptFN41smPabj5ggLCUtHq4PWLTbzdkL2ng4VjqPOAi7Brs3u2D1OHqWeXV4D5A7vWp8s3yA/W3/jAJfAtKC7ITSh3mHtERyRSVFvo3VjauNo4yMTHifxJ8ft7d+vdOBcKmdaYQb+YHLmfBY5eyI36YhsHuLY24IrhXHFiie+llwpi61QP/XrTHWV/NmKcx+rRWsCL1yuY71Yfkmz4dPlkqtq1/qayM2rrVVtNu3gVu0d87uLHac7vR+oP+R/hHr85EncU0x/zjPC86pBz2GLV6Fvat5+HOOdsH6f+uH2DOvcsc8iC0++FS4fXjFZlV87tf7+9+JO/FEAB+jg1c8PJIEy0APWwAOO/T545VeC6+AhGIXXPQESgXSgPVAyVAbdgsYRODjqZEQRoh/JjPRD3kJxoVJQs2gX9BOMHuYWVhN7j8qc6i0uGk+Lv0ztSEASWmiiiHLEn7RddCX0sQwujCZMpsw2LKasKmzi7EocnpyJXDHc3jz2vJZ8FvwWAuaCFkK2wp4i0aJHxOrEH0rM7KKRUpH2lymVHZLnUPBVbFRaVbFWfaKevdtFC619TGdN10wvHY5gi2G70W3jPpNVMzPzZktpq3obadtmez2HIacwF7xrvbujB70Xtbenr5vf+wCNwNygDyG2ob3hFhHPotwo0zHJcTzxo4kPku/uqzjgkPIrrTLDMZP30EL2rdzDR/zzjPLZCx4X+hctn0gvoS+tKleueHLavxKqKj+nen6wJraWs+5h/YEGo8syVw2vH2iuas1rc2lnvTV8p+yey31s5/kHit03ewweDfcm9Mn0IwcWnk8PDgznvxJ9XfHm91uD0Zx3j8dpJxwmz7yf+SD3MWT6zMzD2dl59Ceuz7Jf9BecFslffb9Zfxf4vrR0ZJlrue6H2o/SHys/nX42r7CsUFaaV1Z/6fzK+NWzSly1Wz2x2r9GtaazlrB2ZW1mnX/dZb1g/dH6+m+5376/T/x+/Pv3htyG38bJjd7N+Ef7K8hvHR8QQRcA9OjGxjcRALAFAKznb2ysVm1srJ+Fk40RAO6Gbn9D2jpr6AAoX9hEj1u/vvj3t5z/A8V42gGgG5FCAAABnWlUWHRYTUw6Y29tLmFkb2JlLnhtcAAAAAAAPHg6eG1wbWV0YSB4bWxuczp4PSJhZG9iZTpuczptZXRhLyIgeDp4bXB0az0iWE1QIENvcmUgNS40LjAiPgogICA8cmRmOlJERiB4bWxuczpyZGY9Imh0dHA6Ly93d3cudzMub3JnLzE5OTkvMDIvMjItcmRmLXN5bnRheC1ucyMiPgogICAgICA8cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0iIgogICAgICAgICAgICB4bWxuczpleGlmPSJodHRwOi8vbnMuYWRvYmUuY29tL2V4aWYvMS4wLyI+CiAgICAgICAgIDxleGlmOlBpeGVsWERpbWVuc2lvbj41MDg8L2V4aWY6UGl4ZWxYRGltZW5zaW9uPgogICAgICAgICA8ZXhpZjpQaXhlbFlEaW1lbnNpb24+MTUxPC9leGlmOlBpeGVsWURpbWVuc2lvbj4KICAgICAgPC9yZGY6RGVzY3JpcHRpb24+CiAgIDwvcmRmOlJERj4KPC94OnhtcG1ldGE+ClMqlGgAAEAASURBVHgB7V0FvFRFFz+kgJS0goB0dyOliNIhIF0fIN3SSHcIiIi0tJQgXdKNdEl3SglKCLzv/Gd3lru7d3ffe+x7b/e9c/jx9t65c+fO/CfOnDNnzkQKYCIhQUAQEAQEAUFAEAjXCEQO16WTwgkCgoAgIAgIAoKAQkAYvjQEQUAQEAQEAUEgAiAgDD8CVLIUURAQBAQBQUAQEIYvbUAQEAQEAUFAEIgACAjDjwCVLEUUBASB8IHAixcv6NiJM6owL/77j06cOktXr90MH4WTUoQ4AlFD/AvyAUFAEBAEBAGvIPDtwO/pypUblCjRe/Tq5SvKny87zZ63nAZ8246KFs7rlW9IIuEXAWH44bdupWSCgCAQjhC4dOU6JUmcgD5IlpiWr9xEi+aNpwTvxaOduw/SiZPnhOGHo7oOqaKISj+kkJV0BQFBQBDwIgJg9K2a16F9B47RV9XLKWYPFf+BP45T+vSpvPglSSq8IiAMP7zWrJRLEBAEwhUC0aNHJ7Vuf/Is5c6VRZVt957D6jdPrmyylh+uajtkCiMq/ZDBVVIVBAQBQcDrCBw7/qdKM3vWDOp305Y99HmZYrRz1x/0338v6cMU73v9m5Jg+EFAJPzwU5dSEkFAEAjnCJw5e4lKFMtPMWK8o0qaPm1KOv3nedrPav1KFT4J56WX4r0tApHEl/7bQijvCwKCgCAQOgg8e/acokSJQtGivVHOPvr7CcWJ/S5FjhwpdDIhX/FbBITh+23VScYFAUFAEBAEBIHAIyAq/cBjJTEFAUFAEBAEBAG/RUAYvt9WnWRcEBAEBAFBQBAIPALC8AOPlcQUBAQBQUAQEAT8FoE3lh9+WwTJuCAgCAgC/o/A69evqVqtNiFWkLkzR9O7sWKGWPqSsO8jIAzf9+tIcigICAIRAIFIkSJRjWpf0MVL12ntum30z79PbaVOk+ZDql+3su3e1cXiJevYze5Zu8dlPy9OmTJ8RNGjRbMLl5uIh4Aw/IhX51JiQUAQ8EEEwPDr1qqkctaoflWqWK2FLZcfJk9GVSqUtt27ujhw4Lgdwx83phcVL5rfVXQJj2AIyBp+BKtwKa4gIAj4PgIpmMFnz57xrTKaIkUyYfZvhWD4e9nvGf7r1wHUf/APNHvuMo+1A9eTRnr16hXhfV8ldopEdRp29tXsSb68jADWcGvW68Cnn/3h5ZTDLrmTp85RlZqt6J9//nWbCZQd/dFI8BvvqzRo2I80edovvpc91hKEBQVlHA5u/s6eu0SDh08K1OvSnsxhCnOVPipmzLgZ9PyFfedGu40SOTJFYu9RKdk/dJnPiqnToRyLMW3mIjpx6ix179LM8ZG6B9PcuHkXTZqygC5cuEoJEsSjgvlzUpuW9ahL9+H0ba/WvL6VxvTdsA5E3k+dPh/W2ZDvhxICkbm9d+vclJq26E3zZo2mzBnThtKXXX9m7/4jtGnzbuKmaCP0zUiRIhMcu8WMFUP1J/QpR/rrrwfUttNA6tqpGb37bizHx+r+7l/3aeJP82nZbxvUffr0qalZkxr034uX6rujh3c3fS+sA1s2r00167anhAni05dVPw/r7Hjl+78sWk3nLlyxS0uPw1zhlCB+XCpVqhClS+N8Mp+ncdgu0WDePOFJ4+07f7l9W9qTW3gozCX8V69eU2I+4/nBw79p8dK16n+ihPHp/aSJmTnHp3v3HtHw0VOpEq9nHTh43K40kIQm/jSPRgz5ht55J7rdM9xg1tlv0ATqP3AC1eO1sfWrplOv7i3p5q27VL5Kc8VM036U0uk9CRAEwgqBvLmz0ddNv6I2HQYQ3KiGNUVjQ68kiRPSmrVbVd+8ceM2JUuaiJImSUCx2OJ7w6Zd1KJNX+racyT3t9e27EJa/6bnCPq4SD76/LOPbeHGi9NnLlDVGq0Jg/Qvc76jXxf9QJ+WLERde4ykXn2/o+zZLAfEGN/xlWsw+qEDOxEkfZQjPFDs2LEoTpxYtnEYZUJdJ0r0HmHG9/OcZVSjdnuaM/83u+J6GoftIofgjbSnQIALX/q+QAsWrgrIVaByQPvOg5yyM2vuMvWsco2WAS9fvlTPeXAJqFG3fUCf/uOc4uuAhYtXq/e4Qeog9Xv+whUV3qJtX7twX7tBWYGJUMRC4NHfj1W9o937AvGkPKBoyVoqT2wBbpclnpTYnrFlue3Z5i17VPzLV67bwowXz58/D/jkiwYBjZp1t/Vp/bxb71Hq3UNHTuogn/1F/lmLESL5q/+/rgoHV+Oi2Ud79Rtre6dCta/NorgNO3P2ou19PdbqF1D3yAv+X7t+SwUHZhzW77/t7x+HTrjEWtpT4NANcwlfz0kOHj6pLvPny66DbL8ZeUsJ6PLlG/T48T/qmgcDOssnR9WqUU7dO/7h4tOkqQuUCr9IoTx2j9N89CElTZqQCuR1/pZdRLkRBMIAgbhxYqt2/RMvQ718ab+uHQbZ4W1iV21bxHRf1PmAZg1qeNDxk2fUL/7M/WWFOtUt5Ycf2MKMF2vXb6f79x9R9Spl1GEwxmfFi+ZTt1kypzMG++R13a8q0vYdB4gZpU/mL6iZOnLMcvwuTuTDIT1GSpfujSofYy/I0zisIoXCH2lPgQPZJxg+mDPOcwblzpnFKef6DGg4jYgT5131fPHSdcqJRCYX65x/3XugBhQMKmaq0UQJE1C+vNmcviUBgoAvIFCoQC7FZGGfEtZ0+MgplQUzJoC+pZ/rs9gvXb5GB/i41sIFc7vM+olTFtuUO3fvO8WJFy82FS2S1y/2jWsBZefug07l8McAvWyqy2Usg3F/f/LkSdUjT+Ow8f23vzYYkjgkJu3JARAXt2FutId8Xbp83aUE8fz5C1q2YpPKfgs2lNGzThgTgWG7OhISUpKmuo27UP3alShvnmykB6UZk4fajpiE9eeosdPpBRsOdmrXSJ01/f2Pc9Ss/avq5ahhvar8nTdzI3x74eI1dOnKdSpSMBeVL1eKrl29RaU/LaI++ffjJzRkxE+EdDu3b0w5c2Sm6TMX0979RylylMj0eemiar8tZseTpy0kVo9RhnSpqXOHJvTB+0l0tk1/Ye08gfO2e99hKlP6Y2rZrDbb07D1FNOqNVvoyZN/6SvWety4eUetq5bhPDXg/HuDUO5xE2ZRzJjvUI+uX9uMd7AGO2vOr1SXMU7Gthczfl5CazZsp4lj+1rW/7zxcR9JA+vUi5aspfmLVhImm726tVBHkyJ7x9nhybr126hD20Yqt2pdmyez343ooe6D8idLFot0u4/bTM7smTy+evvOPZo6YxGfi36UMqZPw9ux8tIX7HBF9xesqS/5dR2tYckak+Hhg7pQYCVoVqWq75sxgfUbd6hnMIbF90BHjlqkRHdr8EkS87ow0/iJs+ne/YdUvFg+ypEtk+p70MhhwgOCMDBl+kJasXozlSpegFp9XZcWczl+WbxarS8D/9SpUqi4+IO+hzXmXbsPqXbatHEN+osnFQUK5KTEiRKoeBt+36mMeCFcdGjTUBn9Lvl1PZ05d5GyZclATdloMFmSRPQTawi37TxA70SPTpUrfEo1q5e19TX9wXjx4lCqVB/Qrj2HqHGDL3WwX/4C6x2srQDlyZXVqQy/Lt+owjDxS8/jFcjdOBya44W0J1UdHv+84WIeo4ZchENWdb5RgsAAtWnLbmrcvAddu3aL/teoOjPJiioTsP6F5J4u7RsVk2PuoGqsxJ0UBOt8bN2r9GVLqvhlC/rj0HEbs0cj795nNFUq/wkb892hBv/rpiyLa9csT/37tKPxP8ymsd/PVOngDwYTGCnl5D2yGMgvXLxGtet3onETZ9nizJy1lD5IlpigzmzdfgB9ye4y/3v5koYM6EhZWU056rvp6psjx0wjONgYOrAz/b5lD40YPcWWBi44a07UrfdoesmGjjDumsKThfMXr6g4Dx4+ot79xtLvW/eoe2CKGflSq/WzTgiD69nzl/VtoH+hZencdZiaYBw8dJK/vcj2LrZEzpm/QhlDIvCXJavVcsuho5ZlGmB8+OgpwkCLyY0/0+x5y2nVuq1UiyeC65h54r+mAUN+UDi84kkBDENRp1u27lVGaTpOYH/BnKDROnfeUr/u3kOdNmzalZ4+fUY1q5WlP89eoN79x1HnbsNU2IsXL6hj16GqLUNdPm5UL/oo9Rsm6S5t1N227ftVFCMTuHX7Lu+umU59B36vGN6kCQNIT7JP/2mR3lO7+UYJZt6a5i5YQV+37kuFS3yl+ikMefVEBe15GzMhLN3NnvebinOd29D40b0pXry4VLVmGxu+93giU7dRF9qybR8b57agWtyH0VeBxcGDlkkL4ozmyX3rFnXVBKh6nbY0dORPVKF8SRo5pKuaNGOi1qBpN4oePRqNHdlTjTPDRk2m03+aG+dhgnXKqrHQZfLHX6PglSF9alsRMJHt0mMErVj1u1qmGdC3vXrmbhx+m/ECiT9kI262tbLlARdoi65I2pMrZOzDI9vfhs2dViNt5YHl41K11f98Rb6kLt2G81p7Ipozc6TaRqclWTZqUhmNa1Xvu8p1nx6t1BYfSB+aMHlo32mwbZA4wfuEIdl/UaYYPXn8r4o29cfBSh2JrUZwXrGKLZRBm7ftUYMFJGtIzWDodWpVUM8gxYMw6M5ght+gbhV6wdoJUIWypVhz0FjFz5PbsmRxmKX7H8Z9q7Y0ZcuSXg3ux068WQNVLzr8QSc6zIy8Q5sGdIsZCihqVIuSRttA5MphkQY/Yy1C5kxpedCy7F4As23Zrh+VLttIMSGHpD3eYlsjJl3x48dRcaPxYKgJUhAoo3V7Y9fOli2S+DZ2SnTrNUoxjdVrtynvYdj+448E9TUmWdDaPLPWbdToFvzVRIrXNXPlzKxU0XCcoieoug6CWub3uN3ef/DQ42tzF/ymrOEH9evA7bEiLZo7nooUyk3oTz3Z2v17lqLB8GbPGEmNG35JsGGJGTOGx3QRwcgE4BMC/TN3wSpUtlIzJf2CcS6cM5bSGybf9x/8rdKOGeMdl9/A1i7060KsITMStuf9OHmeLWg5a/fq16lsmwCA8Xfr0lxNWD4rVVjF27vviNrD37bTICUcDO7fUW1p/LRkYdUHEClz5rQq7tJl65VmLHFCi4YBfX7S9wOUgxpIrZDWsT5d4YsS1LJ5HaUR1H3qggMDUgnyn7hxYysNpaMfAf1c/165eoPw31dJC17IX4nS9Wx1Xb/xN7xb6gENHdSJxvJkUU/s3I3DwR0vNDYjxkxVWld97+k3IrYnT5iYPbeMVmZPQikMDEFLEJCYc+bMRJH4H1TfsWLGZIYWxSkn/1p9TMcxqO2dInEA3oUKEP+hfoe0BZU0fFRDSsV2oXeYcXVp30TNJhH+CW8LwmANwozyAWsSEP7w0WNWpc9V4c3/V1P94g8kfFB+Xi4APWWm0OOb5hQ79rtKzYfJBvbsajp92iIlQH3/XnzLRATSEr6RNWt6HU39Os5oI/GywhDudGA2UCHCv3bqlMlVXKyZgvLktqji4DcbKtAbNy0TA0h0PVkN36xVbye1pHrRw598XL6aPNhOsg7GpRknEAYCGFOC0cWyMhHtyjMtM5bDLOXv4vXNhfPGqeWK/rxNEtISlj+wtcmf6F+ezLVpWVep2CHNg4rxtjMQJmKgAgaj01IlC7IKfRvFY4YQHIoXN45tYurufWgSMLnVFC1aVBoxtCvVqt/RNrmbOL5voKV6nQ5+D1rV+ajfMSMse+IjR46i+parg1iePPlHTWD1BN2YnvE6a+b09OP4fgRcMZn9gbfYHmOjsfWbdlLbVvVV1EI86S5aOA9rEsar+6qVP7Ml8djqzOcca6zQt+GzApNxPfkAA0YY+qA2HkzJ/SVjhtTK0A4JdeLJG7aegbAchrYMpl+fJ+yaTlp9YXxgXbfW4foXW9lAGCNctWmMP9iCCNq4egYltE44VICP/Nn3xzGVk8YNqim//RiHMebEYG2p2bZnd+NwcMcLDUW71g3Yd4P9QT9uBHz1WkRqTxqnoP6GuYRvtADGbB9MMD47eMAs0ozZo4BohCBHhqgC+Q+sZjEAGAmMsVH9alSNrYJBj3mtD4RZPdRBhw5bDJMK5MuhwvHn6rWbihFjYDt/4bJaGoB1f9o0KW1xtGSdO5dFwk/wXjxe6ytn25tbmNcjtXoSL2EdH5THYJy4e89hFQapzB1BEwBmutlatopsO6Bp3wFLutmzvXHHeZFtIwoWtJQHM2DYL2BfdXAIkiE6vl7HK2zN6xEeqEEF89vjBpwwcdLSD7AEaZx4u5a696c/qNtaNSuoySGWiVBfeoA/ZDVsy5vHMuFCucA8SpWwTIygVtcaHHiQgzGexsQVBsZ97a7i6O9gsDMS2uyAb9vZgjJlTGO7DsrFvgMWJlDY2jfRPzGBccXskXYUk0m6/iaYMJbXjH0XE0Vo00YM7qKiYZKtCRqLWOzcB1I8vmlcxjt89LSKlor79gJe0wcV/9gyAcP1GbahAaEPasIkH2v++60+PfIY6uvYCUtbhrGhnqygDjZbl8mycv8zJa1pdsOR0Hd8mSB46fV79G09DqOuzZg9yuJuHA7ueIF00S/+4z4Sh4UmTxRR25MnXFw9D3OGryUIqLpjuFEBGgsQnw1lQI9ZkjAjrFvDiMmMtLRl3GKCeLYBwKpyV2HWwa5kiYJ03rqWWjD/GxUkGtt+ZuBagjCq6w5ZJaN8hvRgcId1dUgQxhn+al4TBpXi78AA7tHflsmICjT5s5ule1Ax6/alRyxZgAHB97aWsuFGGJqTwgWcJxHGwdYkeZdBWFODJgJSlB4EtGSbO1cW23uwWC5bxmLAhYnGgV1L1ICOCNqTlyODsr3sBxdam1LKquVAlrV0lC3rmwnXth37qcTH+VWJvuW17qFsyNmu8yDq2GUI7ee2xXu43brRRfuOy1K+J8K2uFt3/nKKpo3U8AD2IkElMDvb7hlD/XpKB8wC7cSsnYEJQ22vGaoxrdjvWgZ4Rx/yut0VYkakjWfBoLawZgME7YNeQ89pXdJC+CHW4oHy5c1K6CMw6ANhqyOEAvTbVLwsp0mrtPMaDNZgWAt7IRjmRo0S1WldGe8+Zs0AyJU3QTyDMeuKpZOU8y9j38ezwNBrN5OJwLzvKQ5sgVBnoOxZM3iKrp57God1vQVlvFi3YYfSpMIuCtpDTxRR25MnXFw9D3OGv4dn7iAz15yuMg2pEfS3CWPUTPj6DWfjMDBSWNdjpg6LYE0Y2PZYmaiW3jGg/PrbRiVVwJJXSzRpP3pj7ARDNHQS5B0TjDoNOtu8jekJRC7DQHnUKg0btytBnQ8GgvV2SODtOw+mo1bJRefP8VdPZj788H316LZ1sNeqTASuZKtm7GKAFXFgCBhgbR3eDs0GaqSBdWpQWl5K0HT9xh11mdKaF0iyc+Yvp88+Laqj2DQcmLnPnrtcLS3oCQMinWPtycSf5vJ6sWV5xPai4WL33kM0/efFplssEQ31Bevt31ZuMrxlf4n6xw4CrZWxf2q5C0xeMCkDpU5lWU7BNdZ9odXQE64LF6+qeoVlO9S5SdibZCG2FMeEbyCvtUPbhLX0EyfP4XVTgqSrNQimEayBBfkbmrnpeFCTw/gME2nkaw0ftwor/aAQVOVBZQJIP8F7cdVnkAdH0vmEUZYjzeO9+6A6rEUxkmbEuXm5TxMmIsgb3PBiueqdGBYJOvkHlu1iaMObt1m0fLl4l8MP3L7WsA0JSHvGM0ryCN9l1bRhAqFp7fod6vJzbs87dh8gdvSlH9l+MZkAeRJYoPEyTsJsCfjAhTZqDIrg5W4cRpGCOl6gD8N/A3ZJoG4fPHJuI45QReT25IhFYO7DjOHfuXuPTrE1L9YfQVCX3rrtLKWYFQLqfgxiZtbmCENjgUXvtp372SDPYjgHJtmn/1hKxNuBJnzXx247n56J4lu4BvObOmOhGpz79mmjtpblsEoOZ6ySPrbcYUAFgQFiO1ZZNvSBBIJJh5kE8YfVUti4/x/b50AleVlhPc9u/336lIoUdpbKVSTrnxzZLTPwI9aJwULeJgYCw8RAh0nEhElzqE2retY3HH5MpAUwIqyt43AKPTFxeIsypP9IBWGbFsoIRqbr7yobQwK3mbN/ZYkri9OWL6yPduazC7DWXOPLsnZJj/1+Fm+9WuRSCsW3WrXrz8Znc2gt42xGUJEPZjensBpHvsxoOUuW2Ab2TY/hLic1nvKCdLUBmNZOrbYadd6+fU9JkWh/Y8bPpI7tLFs8sWOjFRuAQT2ObZ5o62iXmOilT2++0wT9A+lgV4cnKvdFSfp57q+2HRBwAtPxmyGqLX3HVubfjeipJqxwAwuDOOCj+4VZ2mDU2E0Bq2wQNFLYyufpABydVjrrktfFS84TOD0RxrZV7IoBoc0Cwx8nz1f2L8UMank832e1Tzl67Iytffcb/L1y+NO0SU1EUUaL+MWkEksmMFTUmhi46F69Zisv3Vm0LdpvQH6D4y2j9k25kkViTHoSim2Sk6cuVFt0LU/e/MU4VsCwpPXmie9fYeKF9rCWt9GCUiR/X9VLYA4ucjcOI62gjhfYYj2ejQLZS6OyT4J20BNJe/KEkP3zqPa3oXOHwcR41jO+qk9B2rt9oc2y3F1uivFaNvx7g8loNR/iozPDmK1U8YLKGh9hYLDo/AiHRS5sBIwEAz4QBrbGzXooaQF+wieM7cMGQ3nVM0jf3dk6GExx956DKo0+3VvRRm6cM35eqt6ZO3O0inv/wSP1i727RtUlmCI0BcaBJjurgKGSxWAHFSO2NxnX/FVCDn/q1a5MGPyat+qjJj5JeM8wDPIwiJat3JTXv15SdzYcdNVh9JKjMVmsjaL8WHeGtGq29xvSZpeOTdS2wjIVmtDzZy/UPbZBfd36W5X/fHmy87ebGZNWAzC2PrZsXovrpZDangc89aFFn/BSBiQ2GCKaEfCAShW4u1pHhWYG+YehJ85hMKP8bJ8B/Et/UtSuXoxxPeUFcYvymnBF3saJOlu6fL3CeyBvVfru+5lUjq3XQfXrVqYqlUqr6+jRo6uJACZV7XmHBUjbbeTJlU0xKeBhJO1wR+/qMD5zvIZ26NueralWvY4Umw3IMPGAfcHEcf2ULUzcTLHp52nDCZbP8PuA/1N+HEQwrDIjqP/1YTZ4jjaB/tqWJ5BNGlY3e8UuLJ/VDuY47zqB3Ykm9FUsgUEq38LLTeUqN1e7YLBjAG2uXev6yv5Fx8cv3tHaN2hfKlVvqSzx1ZY79kEB41RQWz4MC5PwytVbqTquVrUMGy5+Q8NGTlZ1gvhQq4N0PzTuELjJk2RQ5fKfql/9p0rF0mrswO6WerUr8nLWG80V4mA7LHb+YFuvP1KtBh1Ve9F5h4YM/2dMGUq5rDuP9DOzX1fjMOIGZ7yA1gB9qkGdKmrSqQ2ozb4t7ckMFQ9hPLv2S9qyfZ/y6Xz02Gm7/OP+7t37KoylpIAduw4E/LpiQwDCuYHYxdU3nboOVWnxTD2AJZkAXot38u+t48JnM6tr7Z7DXzivD+oo6vf6jdt297jhLXtO8XQklkwCkLaReLugypcxTF+jLKwRsfm0RjjeZw1FALsf1tGcfuFfmx2KOIXrAJYCFV763uwX5cB38KsJWLMBpr61/cIPe48+owNadxgQsGjJmgCcmcB7pQMc8eF968q3uu3FMLwIbF5YOlJtgTUQKrf4RT3yEotT7tEO4YNcY4YzIOAznp0lBSz7baNTfNQDzo5w1WadXuAAnuip/MAXvytiS/IAVtUHPH/xwlUUr4TjnAvUuZHwTbYrUUHw085GjKr88MHvKs9/nrmgcEMfRX+4ePlaAHB3RegTaIua0BfwjpEQZvY9s3rDe7zrwLRO8Qx5R726ehdxgktGX/qB9df/tr70g5pXV+OwMZ2gjBesXVR4oh85lnn/H8fswqQ9GVEO3HWYSPge5iCBelyMXW9CYp+/cBWfqvXGUMp4jfUyT2tmUBdrC2CsgUOadLduCmnN0WmJ3vJjzLiZxzys8cWgd4zRbNfuZrK2SIYLaA6SJkloCCGlGcG6sBlhr/a6jez4hqWRWXzq1Q42rPu2Zys7LQDUvNt5GaRhvTdbkszSQjkcv+MK66079qr1Y6SjDcBw7Zh3bN3T2/nwPCwpsHmBfYTRRgKaJlf1eIYlWDiW0uu86dOmpCUsycSMEUPVg7G8ULPClqI3n+xo1BAZ45hdY1eLY9t0jAejVfwPacKJf/CjAbW+zhOkca2uRz+D9G/UAJjlSdtb5GctHbYb6m2oZnER5tiucAIc/hvJ8V4/S8KnApoRjPFcGeTNZkdcn7MPD1fvmqUXnsJcjcPGMgZlvMDYCy0cDFzhWdUdSXtyh475M79l+Bhc4Qa3TYeBBPeZjgzIvLj2oVjvw3oR1krhYGcnq4wL5stps0C3j+3fd3X5eGD8d0c4bAjW866Ylrt3XT2DCv/Q3mWuHqtwGMGNGTdDqZndRgyFhyGVl9psiIalGE3Y512J1cXYeuTI1GfNW6Yms+XLltTR/e4X9Y7lF7jFHTKgU7DyDzuYlau3qHdhcAmVvXbpGqwEvfwSjoVVRphWz3NeTt4uORi0+SJ5Yxw2lguC0toV01jwihxo51DG991dS3siCjOjPXcVE9hnWF+HJIFzt1ltFNjXbPGw9ryVt05hjT8RO8KYy+5h4RY1otKDB4+pJ/smD22CFAgDQ6xFhzWFVF4g5UBCNRIkbcezILAbYeGiNcqlq9YGGN/xl2uUa9SwbsrX+ireMRJUwha6KTMXKX/4sGPAbp7jJ84GNZkQi4+JYbeeo2jU8G42DUaIfczHE37bcdixeNDABNYTpOO7ru6lPVmQsR+BXKHlw+FfN62l9tjCohgOb4JCcNmJ/75KGDSNhkUhnc++vVqH9CdM04ezI6PDI9NIoRQYlnmBFAeDurGje9kOeQqlYofIZ+DBDp70+rF3xdK8rc24FdPTB2EBPmvqcE/Rwuw5zlSAx0247xUiFryCPw4HFj8sn8F7Z3BI2pMFtUhY6g8OgPKOICAICAKCQMghgAN84GoYhF0XP4zr6/FjOEALp2aCsEy5YskkdS1/BAEg4NcqfalCQUAQEATCIwJskW5j9igftqzivAF3hG2LRpfiMNDd9Psud6/IswiGgEj4EazCpbiCgCDgmwhA2Tpt5mI2TrxM6zfucMokrNeL8u6kPOy98ys+yErTQj5H4CjbN8DVMAyQHQkHgsE+BofiePLx4fiu3IcvBPx+DT98VYeURhAQBCIqAlhchSExvNzBGZcZsU8B9np4347h40Cuq+zMzNVpfniG/+ogopjOp4+afUfCwicCIuGHz3qVUgkCgoAgIAgIAnYIyBq+HRxyIwgIAoKAICAIhE8EhOGHz3qVUgkCgoCfIwAjPXeHHHkqHhwV6XM9PMWV5xEDAWH4EaOepZSCgCDgRwjAAVSD/3XjE0UvBDnXfFYG1azXgWrW7WA77S/IicgL4RIBYfjhslqlUIKAIODPCKROlZwWzhtnemqlp3LBWdcbJ1qRPEWX5xEIAWH4EaiypaiCgCDg+whAjY+jhVN9+H6wMoujrRMlTBCsd+Wl8I2AbMsL3/UrpRMEBAE/QuD58xfERwHTSz7FM2uWdNSuVQO1t56PNHZbish8ema2rBmcDmIiEkeqboGLYA+F4UewCpfiCgKCgO8isG7DdoKjnCPsUvf69Tsqo8+ePadTp8+bZjoSWVT2zO8pU8a0Tgc0mb4kgREWAWH4EbbqpeCCgCDgawiUYmaPQ5QGDfuRxozsrrKHM+Jr1SgfrKzKUSnBgi3cviQMP9xWrRRMEBAE/A2BOLHfpd9WbiK40c2fJwfdvnOPJwCvaPHSdW6LEolF/OZ8VHj0aNHs4snRaHZwRPgbYfgRvgkIAIKAIOBLCKzftIuqVCpNy1dupI9Sp6D8eXNQ1cqfecyiI7PHCyLhe4QtQkUQhh+hqlsKKwgIAr6OQMoPk9Hho6fp4aPHVLdWJZXdFMmTBTrbvfuP4zX/cyp+996jCFv8pk0aEuj3JWL4RUB86YffupWSCQKCgB8igDX8+/cfUqJE7/lh7iXLvoyAMHxfrh3JmyAgCAgCgoAg4CUExPGOl4CUZAQBQUAQEAQEAV9GQBi+L9eO5E0QEAQEAUFAEPASAsLwvQSkJCMICAKCgCAgCPgyAsLwfbl2JG+CgCAgCAgCgoCXEBCG7yUgJRlBQBAQBAQBQcCXERCG78u1I3kTBAQBQUAQEAS8hIAwfC8BKckIAoKAICAICAK+jIAwfF+uHcmbICAICAKCgCDgJQSE4XsJSElGEBAEBAFBQBDwZQSE4fty7UjeBAFBQBAQBAQBLyEgDN9LQEoygoAgIAgIAoKALyMgDN+Xa0fyJggIAoKAICAIeAkBYfheAlKSEQQEAUFAEBAEfBkBYfi+XDuSN0FAEBAEBAFBwEsICMP3EpCSjCAgCAgCgoAg4MsICMP35dqRvAkCgoAgIAgIAl5CQBi+l4CUZAQBQUAQEAQEAV9GQBi+L9eO5E0QEAQEAUFAEPASAsLwvQSkJCMICAKCgCAgCPgyAsLwfbl2JG+CgCAgCAgCgoCXEBCG7yUgJRlBQBAQBAQBQcCXERCG78u1I3kTBAQBQUAQEAS8hIAwfC8BKckIAoKAICAICAK+jIAwfF+uHcmbICAICAKCgCDgJQSE4XsJSElGEBAEBAFBQBDwZQSE4fty7UjeBAFBQBAQBAQBLyEgDN9LQEoygoAgIAgIAoKALyMgDN+Xa0fyJggIAoKAICAIeAkBYfheAlKSEQQEAUEgpBA4duIMvfjvP3r9OoAuXLxKJ0+do4CAgJD6nKQbThGIGk7LJcUSBAQBQSBcIDBn/m+0dfs+uvvXfcqQ7iNK81EK2rX3MOXPk43atqofLsoohQgdBETCDx2c5SuCgCAgCAQZAUj0u/ccogZ1qtDlyzeo9lflqUWz2lSsSF46fPRUkNOTFyI2AsLwI3b9S+kFAUHAhxGIHDkSDRnYmQ4dPUl5cmeh3DmzqNwePnKKsmRK68M5l6z5IgLC8H2xViRPgoAgIAhYEYgXNzbt2XuECubPqULuP3hEu1jqL1G8oFrPF6AEgcAiIAw/sEhJPEFAEBAEwgCBx0/+oVOnz7OEn1V9fcfOA5QgQTxKmiQh/TxnWRjkSD7prwgIw/fXmpN8CwKCQIRAAFb578aKSdmzZlDlTZnyA/U7ZMRP9E3H/0UIDKSQ3kEgEm/tkL0d3sFSUhEEBAFBwOsIvH79mp4+fUbvvhvLlvY//z6laFGjUPTo0W1hciEIeEJAGL4nhOS5ICAICAKCgCAQDhAQlX44qEQpgiAgCAgCgoAg4AkBYfieEJLngoAgIAgIAoJAOEBAGH44qEQpgiAgCAgCgoAg4AkBca3rCSF5LggIAoJACCKwaMkamvvLihD5QtVKn1HDelVDJG1J1P8QEIbvf3UmORYEBIFwhEA23m5X5tOidIi95x3447hdyZo1qUEfJE9qF+Z4c/fOfZr40zy74MzshS9f3myUK2dmu3C5idgIiJV+xK5/Kb0gIAj4EAJDRkyiRUvW2nI0d+YoypI5ne3e7OLSletUtUZr2yMw+Z9+GEDRo0WzhcmFIAAEZA1f2oEgIAgIAj6CwKelCr91TmrVKC/M/q1RDJ8JhHuGj3Ojq9RsRf/8869f1+CSX9fRqjVb/LoMkvnAI4C67tBlML148SLwL/lwTJz61n/wDzR7rv+6goUk/WXttnT9xu1QQzpSpEih9q3gfCg0xte+A78nYO/P5Cv92efX8PfuP0KbNu8moz9A9IFIkSITHyRFMWPFUIdK6IMljI3ir78eUNtOA6lrp2Z2XqoQ57//XtKYcdNV9ChRonB6pL7xOuA1tWxeh65evUkr12ymqPwMz+GQ8NWrV5QqZXKqXu0L42dC5frho78p+lPxqhUqYPvAR8p+Xpzb3xbqN2gCDerXkXBqWlgSvL2NGTeDnr/4zy4b6DdRIkemSJy/lCnepzKfFaME78Wzi4ObaTMX0YlTZ6l7l2ZOz7bvOEDbdx1Q/SxqFIsM8oq/lzN7Zvr8s49p8rRf6PHjJ+o5GCD64Uv+X79OZXo/WRKn9EIqIDX3/bJlilPrDv1p1rQRFDdO7JD6VKilG1LjqzcLcOv2XZcC2+x5y+na9dvcNiKpdgg+gTG8YvlP6IP3k9IUbjuqjaoxPhK9fPlStaN2rRtwnwo9eddX+nPolTiYLSAar0MlSZyQ1qzdSouXrqUbPLtOljQRHxyRgGKxf+kNm3ZRizZ9qWvPkYRBSRMGhW96jqCPi+RTg4YO17+Im5TTuXP3Ps1dsILmzF/BbiqjUbIkiShmjBgUmQeeRAnfo/WbdtLM2b/ygPQHxY0bh5IkTaiTkF9BIMQQwGA0uH9H2rZ9P6/prgmx7wQ24VevXlPixAnowcO/VT9EX0yUMD69nzQxH+QSn+7de0TDR0+lStVa0IGD9oZnO3f/oYzKRgz5ht55x3nS+k6M6KqPr1j5u+qHx0+eo4ScZhLu46C4cd6lJ/88UwfFoC9isp44UQKnSXxgy/I28Zo0/FLlbcCQiW+TjM+8G1Lja2gVECcJvhM9Ks3/ZZVqOw8fPabECRNQ/HhxmB+8ooTcRs+cu6zaDmwjwDMwSQxNZg8sfKY/w5e+rxMPNgFFS9YKyFWgcsCJk2ftsvvs2XPbs7Xrttmebd6yR8W/fOW6LczsYubspSpeq3b9zB4HVK7RUj1nlYzp89AKnDJ9YcCsuctC63PyHR9BYOJPc1X7Zt/pPpGjBQtXqf7QvvMgp/ygfaKPos+wJKWe88Q6oEbd9gF9+o9zim8MYCau3sX7h4+eMj5S1xs371LPP6/YxOlZaAfs2XdY5eXYiTNe/7ROGzjgP6vMPX7j4uVrNuzwztr12z2+Y4wQkuOr8TvBvW7euk/A8ZOusT567LSt/M9fvHD6zPDRk9XzAUN+cHoW2gFh3Z99XsLH7OjipauEwyJAGTN8pH71H0gM6dOnVrfcKHSw2tdaolh+Svmh5WQp2wOHi0OHT6qQ/PmyOzwhlmYe0eXLN1S4PprSKZIECAIhiACWj9D2If36Ah10019030Sfefz4H5XdQ0dO0tmzl6hWjXJus3/m7EXb8yyZnK3SD1u/W6xoflu8sLookC8HpUr1Ac34eUlYZcGr3w3J8dWrGXWR2JFjf6onRQrlNjVW3Lv/qHpeIK/zGO8iyRALDuv+7BcM/zDvTwWBgWM93Ugs4ZN+/iGvIYIuXb6m9rMWLpjbGNXpGmr/rawyBeW1njVtjMSShrpNkSKZWkYwPgvta56JhvYn5Xs+gABU15jQbtthaadhmSW0wZ28tAXKnTOLU1aOHbcMvDjKNQ6r4UGLl65TR7tmypjWKb4xABMDEAbtaNGiGh+p6117D6vffCb91ClyCAfAjqB40Xz0+5Y99OI/e5uGEP50iCSvx09vj6/eyqxxqdYszQN/HFPBZkLb/QeP6MKFq+p5rlzObdYsvZAMC+v+7NyzQrK0wUz7j0Mn1JtmFbp+4w71LEGCePQFGzqBjhy1DDzZs2VQ967+nDl3yfbIbEDaf8DSkAoXzGWLZ7z4mw2J5sz/jXbtPkQxY75DTRvXoL/YJqBAgZxqjRFxN/y+kyZNWaAGyA5tGirDpSW/rud1pYuULUsGasqONWA38NPUBbRt5wFej4pOlSt8SjWrl2VjE/eGWjC4GTdhlvp2j65fU7o0qVT27v51n2bN+ZXq1q7EE5XEShJZs2E7TRzblxIles9YhGBdowNiPWz+opUE3Hp1a0FxYlsG+OMnz9K69duoQ9tGKm1lW8GM4rsRPYL1LV9+6eq1mzTyu2kE49DGDavRZ58UVdkFE5g1Zxmh/cGYdNeegzRizFTq3b0V5cuTLchFyssD1YJFqwmT2xgx3vH4/m8rN9Ga9dvpEa9nflw0rzI0+yh1Ctt7OF8da+FHj5+mHNky0YBv29meubu4dPm6S03b8+cvaNmKTer1Fs1r2ybmaKNwAOPJ6PDAQUsfL5A/h1MW7t1/aBu0c+c2H7TxnYWL1yhr7iLcX8uXK0XXrt6i0p8WUemhr+L8+LPc5zu3b0w5c2Sm6TMXE6Q/2Ot8Xroo1a1ViZ3fnGQjwYVsCHaLMqRLTZ07NGHjL2fDwOzZM6p0j584Q3lyZXXKsz8FeHN8xW6oCT/Ood37DlOZ0h9Ty2a1beMYLNWfPPmXvmJtz42bd5TdVRmunwZv4QkQQpt2VmQmtOmJJIS2pEnM7a88tR1on4aPnqKOIu7GRqcwUkX/gUAIe4DaNctR0cJ5aemy9bSM+97ff3O/4/s2LeuZ9teg9mdvtiWfZ/iQKmC4BDJ2LFhuzmNju9nzflPqtZFDu9msZk//eV7FT20Y5FSAw59DhyxSBSQSDAaOtG3HPhWUN7fzIH3v3gNq1LwHM9sY1L9PW9WAYTwIGjawM31ephgbMj2g0WOnU9fOzahz12G0gy2RMWB34gHn/aQ1qGbdDsz4L6l3ShUvQGNH9qQp0xfRsFGTKQcPKPCWZSMHCR/SFNIcyt9q13kQW6MuouGDu6jo2PoEI8RPeE8vGP4vS1bT7dv36NDRkzamZEs3GBewjN3E0k2t6uWUoRakLr1zgdfJlAq3LVvB3rlzT0lB+AQmIZjdhhcCA2nUrDu1/rouLVi8mvoPnEClSxVRg9tungD+MGkud/i6iuFv2LRbLQ1t+n1XsBh+2jQpFWzYDqavXeE4Y9YSNQltxIPopSs3uF0sVP8njO2jBqXdew9Rq3b9qWiRvNSTJ4laK+YqPWO4Xv4ySoIYcLds30fTZiyma9du0f8aVWfGWVG9honQ/fuPKF1ay0TUmJbxGmnst6pdd+05xBo6yzKajoO2A8KgDQNeR8KkG/2sY9uG1L5NAxo+agrVrt9JxdcMf+aspfRBssSEiUnr9gPYYDehYkhDBnSk+QtX0ajvptOxE2fpCmOGNHD2fP3G36gdAWNH9XL8JOkJ1IWL1+zGJaeIPh7g7fG1W+/RzAQTs8Y0m2p3ZXgiBUEEy6O9+40lTOjA8NGW2B6LjTH/eSuGj/FTL/f+whO+pcs32iGOnSEgV0Kbp7YDfGCg2bB+FdXH6zXsQjAy/fp/X1HjBtWoZ9/vqFvPUWpSG52FtT7dW6qyDR05md5/P7GaRNpliG90Hw5Mf3Z8923vfZ7hG6WKOg07K/WgrmCso7VuUZca1K2sZl8ajPsP/laXMT1IQ/ut1sSFWI2YJnVy/br6vcvMGoMVKHcue/eUGKDadhqkBriF88ZReh7QMrOkCwZ96vR5ypzZwqgx48MsNzFb+4OePP6XZk4ZblseQP6xvomBSs9yc+XIRCtW/c4SzRV7hq9SePMHWgMMrvHjx1GB0XiHgSZoCkAZM6RRv3rCgQZppH0HjlD+vDlsM3DjM1fXkDLBRH4Y31c1bMSLylayIEhiKA88fcHLV4rkydTgj10Q+tuYBB0+dppi8UQJa6GOSzQqIT/4g7rNkP4jKl+2JA0cOpEt1ePZcNzLuILgMhUE96jLfttAxjpSDwL5R2tPHrHk4I6gvgTjXTDnO4U94mLAxUS0TYeBNGxQFxrIEzJMzqCVCSpp63ssg31cqrZ6XffFkiUKUq/uLShr5vS2ZHV+YWXvjoyD9seF89hFxYDrbtDevG2PYvboZ7oP1alVgbUqh5QUj8SePn1GM5jhb143i3rxAA2qULaUksBwnYe1BmyMyEuDJ2n+rDH0XnzLtkIIAmyYhyhOFCe2ZUveA8bcHcGPwkmMCTw+mO1QcPduaDzz5vgKIQS2FsNXTlPSO/IfNaplbNC2HxjfQJ9xu8S4gO2VRkIasAXR44Xxmdm1Ftqw7JWR/xsJbQf9DmQmtAWm7UD7AY1dqeKFqOu/I1VaI4Z+w0tPlnaalcf8Y1YbgmGDOitrfF3mc+euqPiOfwLbnx3f88a9pTa8kVIIpXHQqs4HExkzorv6SuTIUbghRVHM3+yzT578o565U4kryWTrXvV6o/pVWb3+ZqBC4KYtu9lJyHJTqWILvwfGjkEGzB6E9BCGgV8bCqbkfbsZM6Qm7DMGQbLHlkIQVFuXWZIB069ft4oKwx8MDiBH/9kBKvTNH6iGa/JMedLkeSqwdMlC6hcMFekCLzBVUHGroVPajz6kh7ytaiK/gzVhSP0Hdi0JEtP9lwdPSK45s7MqmJkHqBhvfQRpw6oCBgPIUiULsnp5G2H7DFRgkHzzsIp6MTsSSspLGeNH9/bKMoPKQCj+gSqvc4fGtGvvQfXVctblJNyA2YCyZ7OofaEShlo7I6uIg0Ox48RSr2FLnDs6yGpxTDYx0dKEJYVve7ZWddW99yjF3Mz2wuv4rn7hOEdr2rA8kzNnJorE/6AOjxUzpuqPju/+azW0jeNhv7ru48U+zmfXF3R6S10M2sjThB/nqmjN/1dTRydI3aD81uWTpzxJ7fFNc4rNy06oG/TRlrzsoOn06QvqEup7zezV3m/Of9as9uOCfif2uzHVJSRXdwTpEKrshvWq2Ja53MV3fAamFZKksffG+BqJVd1DBnWiZ6xFAc5p0nxI8F0A0mp3bfwMgQBazRs379qKd/HSNWrwv240c+owNb7oB+7W8PcdOKqiVeSJt3EcRSAEkO++/1k9dxTaAtt2oJFt37o+nT5jaSPYT6+ZPRKGVgjUiyV7vdXvz7OWuClSJFXPHP8Etj87vueNe5832ttnWEdHZ8R/MA/Mvl1RFJ4MeCJdgYiXySoJG9/5Q0v/BZzX76HCBRXnAUqTVs0Xts78EA6nIalTpSCtSciTJ6uOzg3lT3UNw0I9MUHD3rx1jwrP6jABsb1ovWjM+4Fj8A6FX60qrMKspQAdsRpOFTSshWKtGSpMMAJIGRV5fbNp4zcDpDXJQP3AqUqtmhXoPGsgYAwDIyvsmQbh8A9QXkM5MfkoVcIyGQGzz8p+weHY6Oepw9UEae6C39Q7/vYHdQtV5dp1O1TWMfkDYZBxnHBhcDl16jzlZab/VsTpuKPbd//iNXnLJMMYr0ql0lTIaoeC58HRqhgtuZEW+mH8+HHVMhom32YEBgDyxLS0FbXZGqzGE+mAKRnp0JETqg2ibWs1KZ5raVIP8mizNXn5Sfd59FEjBvr7eQyGiLv3HFafQvs2I82H9a9ZHITFiW2ZrLljWq7eDY1wb46vEJogXGy2ClIYZzRpxqwnwQi/yDYhBQu+sdlInSo5QWMKYSIwBCFLG107tg28bzS6dlwKCmzbQZmwPn/IKngWKPAmv7BXwLIEhDbjcuW27RYBDxpMt+ShP7t9N5gPfZrho5PYrIKDYGGJwQiqRncDDaQhEDq02YClpTQzq2AM3qCcVvUUrrVqKV/erMpYCmu8oJcvXykJH1JFKsMWQb0emtdg8AMDEywjwOAoapSoiqmqRFz8AdNFOcFstLpQS9m5DXjt3H1QGW4hGdgcoNPF4TXKtyE9Yy9l1SwgrX1Wa9lsWd8wHWgSSnycX30KnUKrSDH5ACYw8vNXAiNH+8TkMysbYIKOH7eogI0GprApSZnyA2VPgThgYtpV6GPWRh3hJQ6o410RloJAsT2oxlPwLpXbbDfhSJhQasMzGBudu3DZMYrHey0JQv0dGMNBJAjnJyCU0RXp/oHn0Pw4km7PZkZX589bVKYF87+ZlIMJwB5Aa9quXL1hS1IP2vkMhn/GQTuhdekNL6xet1W9V4qXKmBD8OhvS3/WiT2xuurWuxF0uONvx3aNaP7sMcGS7pFWAP8LKQqp8XW3VcNVjHcygGA8CuEAho5a6wjnSdAYFS5gmVBh6QMGkKk+fD/QxT31p0WSxgtmRtd6jCpkIrQFte1ooS2XYTJy1CpcFTakDw3omnXbVPvDmIAdY2jjRgpsfza+461rn2b4585fthlkZLeuhwam4Anei6uiAXxXtGffEfXIODDruOjgkNJAUF06Eow2QMk/sKhsMLHYvM2yPIAG8cNPc9kz4DYVxyZVGCR5PNhllSCMM9O16y3S4ud8VOaO3QeInZWoNFz9AeMApWXVmabrN+6oy5TWjoOONGf+cvqM0wwsgSGwgwjVWF29ow2pMCvXhPV7SFu6U8MaHJ1OYwwXsVMmDlbRsbSAyY2jBIXlD7hSdXX2AZgslgNgie6KoBpEGu58nsN4bfrPi5Xlu1k6GAx/Yct4eJRzNXF89vy5ap8ZM35ks0K/dfsvlZxWZeIGhkHlviihwtEmu7IHyP993ZPgI/zbAeNpL7dFeKiDkZsZ6cljPCsDNYuDsJw8kUO5HPO7dsM2ZUEMdSSoS/fhpNNUAYH4o/uLmQtrV6/HZU0c6G8HZmmMf+r0Odut6aBt1bSZfVdr+dJ+lMKWBqQ6TIIRH5OfOg062zxw2gZtw8TCNmgbtvBCnY92C5scGDW27zyYjh49bfsGLvQkJgFrOdxR9OjRlQZRq3vdxQ3tZyE1vupJ54fWMej2HUuf0MufKOfK1ZvVMhfaNAwpO34zlCawBnAS71YKLOnteBhDzLZyoi+A8uZxnkgGpe3oSakS2qxLFEj3D6vQaNx5s91qYI6dVg8ePKSqNdvYeBjeAem+56k/W2J7969PMnwMitgWA+M1EFQmf/HatCsm4AhJOqtVMwZ+I2EgRGOEEZCW4KHu0w0UBmlgEpqZoFFgJqorSKcFd70gqMph0PH9xNm2NSq4GV29ZiuVKG6RavUe1/wGpw9GqcK4TQ6zQVCWLOlo8tSFvO5XVd27+gOjMRAMSyDZQGrE3mDQVbaYBtOCRJeLVZWejthUL1n/jP1+ltotMHLMNGOw3bU2TNSS32p2fQyCXQDwwqA7ZvxMgoSjJUJszdLbs37g87tRr9WqfG5LF2XAjoMfJ8+nhS7cyaLuBg/7UTFLLSXbErBejGaf70gD/80I34Gl+vcT59Bati8wI6jqsFti8PBJvH3NsvziGA8TG5ThIO/2AANA+1m4dI2KhrYBwqBz8PAJqmE9f2H9ph3UoE4VZb+ACdvoYd2oScPqCi/2mKbecfxzlie+oDQedp1AxZ4vT3aaOmORag/ID9zy9ug9Rq0xYsIF7REms415hwm2I7nTLOCbd+7eo1OsodDtCv1FT2rw3B3B1zwmgDr/xrjoA6qP88APgtEVGAPaDQgTSrynJ9JJ2K0vGDHatKYcVg3bGaukjy132AYKwiQY20PL8kQLzBZ1jsmko6bNNmgblluwZQxUkteY12/YQf8+fUpFClskUfWA/+i93RlMlgN1HF/9DanxVZc3R3aLtuuIdZK00HrcL8Y3jMGoxwmT5lCbVvXUK+t4y/AnrCmETc/16xbsdVpmv7CbQN9fv3Gnepw8eVK6eeuOGqtRz0gf/U4LbTCSc5xMB6XtaKdQWArSy6/4sJ4EG/f3X7H2+4K8NRv9sMaXX6glaGM5dH/w1J+N73jrOqq3EvJmOmA02roS6aLiKrIE1JYbCAZHT5TPunYCFZHRGA8WqdW+amP3OqQs0KG9y5TEh21xmjD4VPqyJdWrXZENtP6ng6kt76/E4FK5eiulzq1WtQzBcnMYb8UoV6kZr3GXt6lvwXgxcdBrqEjkJjdIUOXyn6pf/adKxdJq4lC6bCP1TVhYuyOsnXfp2ERtKSpToQk9f/ZC3W/Zto++bv2tGtzAALp/08xlMmZrkJ+wGhOqaj0pMnu5KDd+HFABprp0+XrV2Qb2bc9GMjMVBninPu+ewPqxI2FCBYY6bdIQu86AdVWkiYne6T8vOr6m7rFWCyYLQzH4cTcjDNRQBRfmTmdG+A4Y327eH+/KVgJbyfAdtD1oLlytK/bq1pI6dhlCxT+tq/D+pGRhwj5w7KLA4TdYQoFhol5y+ZSfv2ajR39zAAAJ2ElEQVS/9DDwnPLjIMWMjp2wSI/GdWhjvjGhg9FfdJYWPdGgfh2UJXqJ0pbBFPFhsFa9qmVi1b93GzYeTMV+H35R1vtQ0aMezAgMGf3OSJgAgfZuXxio/MAzHs7BAKM2SrlDeevcKiuzR3rAGN/q3qW52vlQpnwTBNtIT+B+X/ezzbAO0jfiY2KGusSEpw/7OtjIk94ZPy9V26fmzhyt0tATG0hexkFb90/jhDw7L0lhAoJvYoIwacIAuzV/JHjcaoOTKWMaWx5D4sKsf77td0JqfNX5qle7Mh09doaat+qjJnxJmJFjCyi2Ppet3FSNFd25TWq/IVgWhOZuEE/kx4zsrpNx+Vu9djvbDipEgk8Q/P+Fd6dgEo3JvJGwQwUGhEvmW8Z6PAtK27l+87ZKrkxpi08HnfYlFigxrmsbJoRj4rJg0SrVt9BnRw93Lk9Q+rP+ltd+ecYVLgn+u1t3GBCiZWNJJ4AlINs32J1oAPxaGwlhvD3JGKSuWZpxCkMA7zAIMHvGjnlc+tLnbUcBvJ4fgF9NyBcbWulbp1+cOwC/2y9e/Of0DAE3b90N+OSLBqbPjIGsmg9g1X0Az6xVMH5ZujUtAyLwQSoBLdr2DeA10ABmAgFjv59pTE5dI2+9+o11Cg+LAB6EAn5dscHtp+EHnteKA+4/eGiLx1qOAJYMbD7lbQ/4gv0xKOx1fY0eOy2gydc9AoAlD1jGqKoeUE8bNu60C/d0wxoxlSft094xPsLRVo3t1zGON+55j74qK/ydhxQ9f/5ctUFjWXGGBurASKy9M96qa9SBYzwdCe0YaTsSfM/jvIBhI39yfPTW946+9APjr/9tfekHJ9Oexlf0bYyPPGm0JQ8sMU5hTHSk5Ss2qjMj8AzvGQl9w50vfWPcoF4Hpu2gf5v1E/QxY5vT30YY2o4eE3U4fjGuBqc/G9N4m+vIXps5+FhCXzf9Skmpjmp9b2YTnpuM1pmx2SLXuHaLbyEMqk1HcrQa1c/h8MPVMx3H8Rcq8zS85U6rzvEc+cIOAU/kyihoFxv66e187tLAOhSckGjpDb8wyDMrA6Tarj1GqnV7SHc/TplPDx/ZG0PhW1vZ0K+wYZeBu++H5DPYP2zfuZ/ymTheMn4XRp+QGGAsqglqxPTpUjtJhniOGb7R+G0VS8DlPi/B2yUZD4etdwvZaRKkTOxzDwpB6kCejNboxvcRjrZqbL/G5966LsYOfiBdwblNSBE0H2iDxrJia6ze76y/qw0X9T1+0Wcc4+nnaMdmWpXdvBUTmp8aX5bVUSPcr6fxFVoUjI/azgkAAUuMUxgTHWk9n3oKbeDylRvZWNmyhOUYJyTuA9N20L/N+gn6mLHN6fwhDG1Hj4k6HL/B7c/GNN7mOtwyfDhKgLqWT5l7G3zC3btYX8f67XcTZqqy1azbns85n25XTqyf4uxzuPf1JvXuP9ayts9pwxMVHPh8xGpzI+EoVax5B5XBGdPw1jUMiOBIBp3Xm/QnL1forYpIF+30F15rz8QORzAgakI9LFy0hjq1a2y6k0TH8+VfDHqd2I4Dlssw4vR3whrxVHZuVKXSZ3Z1FVLlYmkupJJ+q3S9Pb6m/DAZb6M7zUtdF1ggsDi1easM+uDLvtCffXIN3xt1BeOwUWwQVbdRZ7VWiDVbIZwtHptmTB7qFgpoRWBQY+fa1+0bgXtoXENz9QacpsBdMDQdYU0PHjymnsHwSOcp30PZI5dR64M1fkj2WIPWBMbS89sxVKniJ2pNW4f74y/2MUMi/IZ3J8yZPlLZNfhjOZDnn3j3Bwx5g+O8yF/LbJZvb4+vXTo05XX5h37phMsMH8cwX+nP4VbCB+Dwavfj+H58VO5KtfXDsRL86R5lSZYsUahkGQ4jatUoHyrfcvxI/TqVTU9ic4wXGvd9e7W2M8jx1jeNzF6naWT2CIM3xwTvxWfpuImO4te/XzetRTieVO+88cfCwBJ8F5+TAI+f2ggzxMvhoxI+yu3N8RUTCOOOJSOuOMTIrM8Y4/j6ta/050gwAPB1sCR/goAgIAhEBASwVVIfwoXyOrqZNcMAW9Sq1mhte6QP77IFyIUgYEUgXEv4UsuCgCAgCPgLAnDwggNljDTj5yVqy5oxzHgNeW0OnxhqpJ/5tEw4sxESBBwREAnfERG5FwQEAUEgFBHACXHr2MEPzgDAThZHglFnbvb4WbVyGZuhIOxseCubclUNx0+OhJ0R8MFRoWwJ2yFOjnHkPuIhEG6N9iJeVUqJBQFBwB8RuHLlhu0cCjj8cSR489zLh4jBO5zexcH7wvmkRssBP2bvIA2czQHf9MYDaxzTlvuIhYBI+BGrvqW0goAgIAgIAhEUAVnDj6AVL8UWBAQBQUAQiFgICMOPWPUtpRUEBAFBQBCIoAgIw4+gFS/FFgQEAd9DAA6Y2N98sDOGw2P0aWzBTkReDLcICMMPt1UrBRMEBAF/Q2DEmKk0aqy9q+vAlGHD7zupZbt+hJM24eRFSBAwQ0CM9sxQkTBBQBAQBMIAAT4pjt1Kx3R5oI+rLJ3jA2fe4cNpmrXqzUchf0FNG9dwFVXCIzACsi0vAle+FF0QEAR8BwE+UlVlJk7soLvQ1mfLR4sWzXcKJDnxOQSE4ftclUiGBAFBIKIhAMc7v2/Zo470Xr54IiVM+B7hdMHHT/5xC0WypInVMbTGSOIt3YiGXBsREIZvREOuBQFBQBAIZQRevw7gA75WUJ8erWj9xh304NHfiuFfvXqTbt6+65SbSBTJLgznzgsJAoFBQBh+YFCSOIKAICAIhBACOClu/KhetHDJGoJLXK2eL1G8QPC+KOehBQ+3CPCWMPwIUMlSREFAEPBtBOLGjU1Ll6+nBnWq0LXrtyhF8mS0bOVGunrFsq7vKvf58majwgVz2z2W40/t4JAbAwLC8A1gyKUgIAgIAmGBwIlT5+j27XtUvFh+wta88aN7U7Ei+ehp7mdus4OJgpAgEFgEZFteYJGSeIKAICAIhBACN27eoZp12lNOPiCnRfPalD1rhiB9ae6C32jdxp107Nif9G6smJQmbUr6tmcr2/JAkBKTyOEWAWH44bZqpWCCgCDgTwg8efIvRYkSmWLGjOFP2Za8+hECwvD9qLIkq4KAICAICAKCQHARENe6wUVO3hMEBAFBQBAQBPwIAWH4flRZklVBQBAQBAQBQSC4CAjDDy5y8p4gIAgIAoKAIOBHCAjD96PKkqwKAoKAICAICALBRUAYfnCRk/cEAUFAEBAEBAE/QkAYvh9VlmRVEBAEBAFBQBAILgLC8IOLnLwnCAgCgoAgIAj4EQLC8P2osiSrgoAgIAgIAoJAcBEQhh9c5OQ9QUAQEAQEAUHAjxD4P4amaRRlmYryAAAAAElFTkSuQmCC)\n",
        "\n",
        "Note that some new messages will contain words that are not part of the vocabulary. Recall from the previous session that we simply ignore these words when we're calculating the probabilities.\n",
        "\n",
        "* Initiate `p_spam_given_message` and `p_ham_given_message` with `p_spam_given_message = p_spam` and `p_ham_given_message = p_ham (p_spam and p_ham are P(Spam)` and `P(Ham)`.\n",
        "\n",
        "* Iterate over each word in message (the input of the `classify()` function), which should be a list of strings by the time you start this loop. For each word:\n",
        "\n",
        "If the word is present in the dictionary containing the spam parameters, then update the value of `p_spam_given_message` by multiplying with the parameter value specific to that word. We'll need to code something similar to `p_spam_given_message *= parameters_spam[word]`.\n",
        "\n",
        "If the word is present in the dictionary containing the ham parameters, then update the value of `p_ham_given_message` by multiplying with the parameter value specific to that word. We'll need to do something like `p_ham_given_message *= parameters_spam[word]`.\n",
        "\n",
        "If the word is not present in any of the two dictionaries, then we don't do anything. We ignore words that are not part of the vocabulary."
      ],
      "metadata": {
        "id": "7V7fxT-wZw1n"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import re\n",
        "\n",
        "def classify(message):\n",
        "\n",
        "    # clean the message\n",
        "    message = re.sub('\\W', ' ', message)\n",
        "    message = message.lower()\n",
        "    message = message.split()\n",
        "    \n",
        "    p_spam_given_message = p_spam\n",
        "    p_ham_given_message = p_ham\n",
        "\n",
        "    for word in message:\n",
        "      if word in spam_parameters:\n",
        "        p_spam_given_message *= spam_parameters[word]\n",
        "\n",
        "      if word in ham_parameters:\n",
        "        p_ham_given_message *= ham_parameters[word]\n",
        "\n",
        "    print('P(Spam|message):', p_spam_given_message)\n",
        "    print('P(Ham|message):', p_ham_given_message)\n",
        "\n",
        "    if p_ham_given_message > p_spam_given_message:\n",
        "        print('Label: Ham')\n",
        "    elif p_ham_given_message < p_spam_given_message:\n",
        "        print('Label: Spam')\n",
        "    else:\n",
        "        print('Equal proabilities, have a human classify this!')"
      ],
      "metadata": {
        "id": "N_EGiqcJbrWJ"
      },
      "execution_count": 82,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Use the `classify()` function to classify two new messages. You can use any messages you want, but we suggest that one message is obviously spam, and the other is obviously ham. For instance, you can use these two messages:\n",
        "\n",
        "'WINNER!! This is the secret code to unlock the money: C3421.'\n",
        "\n",
        "'Sounds good, Tom, then see u there'"
      ],
      "metadata": {
        "id": "Y5pLBXpmXHu1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "classify('WINNER!! This is the secret code to unlock the money: C3421.')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8BU4wqa_XRAy",
        "outputId": "260d944d-0e00-4172-c2d5-d765efc9c9c5"
      },
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "P(Spam|message): 1.3489598779101096e-25\n",
            "P(Ham|message): 1.9380782419077522e-27\n",
            "Label: Spam\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classify('Sounds good, Tom, then see u there')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y_R9HOr3XY7F",
        "outputId": "490ff16c-7ccd-4f53-8558-4d487f677fa4"
      },
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "P(Spam|message): 2.4385273359614485e-25\n",
            "P(Ham|message): 3.6893872875947e-21\n",
            "Label: Ham\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 7. Measuring the Spam Filter's Accuracy"
      ],
      "metadata": {
        "id": "fP8ZfXMDa_47"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We'll now try to determine how well the spam filter does on our test set of 1,114 messages.\n",
        "\n",
        "The algorithm will output a classification label for every message in our test set, which we'll be able to compare with the actual label (given by a human). Note that, in training, our algorithm didn't see these 1,114 messages, so every message in the test set is practically new from the perspective of the algorithm.\n",
        "\n",
        "First off, we'll change the `classify()` function that we wrote previously to return the labels instead of printing them. Below, note that we now have return statements instead of `print()` functions:\n",
        "\n",
        "**Tasks**\n",
        "\n",
        "1. Measure the accuracy of the spam filter.\n",
        "\n",
        "* Initialize a variable named correct with a value of 0.\n",
        "* Initialize a variable named total with the number of messages in the test set.\n",
        "* Iterate over the test set DataFrame (you can use the `DataFrame.iterrows()` method). For each row:\n",
        "* If the actual label is the same as the predicted label, then increment correct by 1.\n",
        "* Use `correct` and `total` in combination with the above formula to calculate the accuracy of the spam filter.\n",
        "\n",
        "2. What do you think about the accuracy value? Is it better or worse than you expected?\n"
      ],
      "metadata": {
        "id": "OWFtOyrEb0CH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def classify_test_set(message):\n",
        "\n",
        "    message = re.sub('\\W', ' ', message)\n",
        "    message = message.lower()\n",
        "    message = message.split()\n",
        "\n",
        "    p_spam_given_message = p_spam\n",
        "    p_ham_given_message = p_ham\n",
        "\n",
        "    for word in message:\n",
        "        if word in spam_parameters:\n",
        "            p_spam_given_message *= spam_parameters[word]\n",
        "\n",
        "        if word in ham_parameters:\n",
        "            p_ham_given_message *= ham_parameters[word]\n",
        "\n",
        "    if p_ham_given_message > p_spam_given_message:\n",
        "        return 'ham'\n",
        "    elif p_spam_given_message > p_ham_given_message:\n",
        "        return 'spam'\n",
        "    else:\n",
        "        return 'needs human classification'"
      ],
      "metadata": {
        "id": "biagLyECcx9K"
      },
      "execution_count": 85,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "Now that we have a function that returns labels instead of printing them, we can use it to create a new column in our test set.\n",
        "\n",
        "```\n",
        "test_set['predicted'] = test_set['SMS'].apply(classify_test_set)\n",
        "test_set.head()\n",
        "```\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.48.40.png](data:image/png;base64,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)\n",
        "\n",
        "Now we can compare the predicted values with the actual values to measure how good our spam filter is with classifying new messages. To make the measurement, we'll use accuracy as a metric:\n",
        "\n",
        "![Screenshot 2022-01-17 at 12.48.44.png](data:image/png;base64,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)\n"
      ],
      "metadata": {
        "id": "yhVxxmBhZGEU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "testing_data['predicted'] = testing_data['SMS'].apply(classify_test_set)\n",
        "testing_data.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "cTmmL4-KYozg",
        "outputId": "8436ffb8-f8ec-4407-a7d3-1d7d29cab37d"
      },
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Label                                                SMS predicted\n",
              "0   ham                           wherre s my boytoy             ham\n",
              "1   ham          later i guess  i needa do mcat study too        ham\n",
              "2   ham             but i haf enuff space got like 4 mb          ham\n",
              "3  spam  had your mobile 10 mths  update to latest oran...      spam\n",
              "4   ham  all sounds good  fingers   makes it difficult ...       ham"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-31749fcc-363c-4e2a-88e8-c86fbbc7b287\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Label</th>\n",
              "      <th>SMS</th>\n",
              "      <th>predicted</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>wherre s my boytoy</td>\n",
              "      <td>ham</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>later i guess  i needa do mcat study too</td>\n",
              "      <td>ham</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ham</td>\n",
              "      <td>but i haf enuff space got like 4 mb</td>\n",
              "      <td>ham</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>spam</td>\n",
              "      <td>had your mobile 10 mths  update to latest oran...</td>\n",
              "      <td>spam</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>all sounds good  fingers   makes it difficult ...</td>\n",
              "      <td>ham</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-31749fcc-363c-4e2a-88e8-c86fbbc7b287')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-31749fcc-363c-4e2a-88e8-c86fbbc7b287 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-31749fcc-363c-4e2a-88e8-c86fbbc7b287');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Accuracy"
      ],
      "metadata": {
        "id": "bs-zn7LIZ1rn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "classified_messages = testing_data.shape[0]\n",
        "classified_messages"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c2ef_YijZoRM",
        "outputId": "cafa4526-37e5-4763-867f-fe838bff729e"
      },
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1115"
            ]
          },
          "metadata": {},
          "execution_count": 87
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "correctly_classified = 0\n",
        "incorrectly_classified = 0\n",
        "\n",
        "for row in testing_data.iterrows():\n",
        "  row = row[1]\n",
        "  if row['Label'] == row['predicted']:\n",
        "    correctly_classified += 1\n",
        "  else:\n",
        "    incorrectly_classified +=1   "
      ],
      "metadata": {
        "id": "hmtlRPr3aD37"
      },
      "execution_count": 98,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print('Correctly classified: {}'.format(correctly_classified))\n",
        "print('Incorrectly classified: {} '.format(incorrectly_classified))\n",
        "print('Accuracy: {}%'.format((correctly_classified / classified_messages) * 100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1pOkbcZcacTw",
        "outputId": "59db7fe6-9791-4368-9912-6b2c05e17707"
      },
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Correctly classified: 1101\n",
            "Incorrectly classified: 14 \n",
            "Accuracy: 98.7443946188341%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 8. Next Steps"
      ],
      "metadata": {
        "id": "TJeuKALfdUQZ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this project, we managed to build a spam filter for SMS messages using the multinomial Naive Bayes algorithm. The filter had an accuracy of 98.74% on the test set, which is an excellent result. We initially aimed for an accuracy of over 80%, but we managed to do way better than that.\n",
        "\n",
        "If you want to keep working on this project, here's a few next steps you can take:\n",
        "\n",
        "* Isolate the 14 messages that were classified incorrectly and try to figure out why the algorithm reached the wrong conclusions.\n",
        "* Make the filtering process more complex by making the algorithm sensitive to letter case."
      ],
      "metadata": {
        "id": "PVEy8CWldWvp"
      }
    }
  ]
}