06November2019.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.3"
    },
    "colab": {
      "name": "06November2019.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/NEELANSHSETHI/Fake_News_Classification/blob/master/06November2019.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xNqu5CgBcwcQ",
        "colab_type": "text"
      },
      "source": [
        "# LIAR LIAR Dataset Clasification (Fake-News Analysis)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xVoeeiBxcwcT",
        "colab_type": "text"
      },
      "source": [
        "Let’s look at the steps that we will follow in this course.\n",
        "\n",
        "- Problem Statement\n",
        "- Hypothesis Generation\n",
        "- Getting the system ready and loading the data\n",
        "- Understanding the data\n",
        "- Exploratory Data Analysis (EDA)\n",
        "- Univariate Analysis\n",
        "- Bivariate Analysis\n",
        "- Missing value and outlier treatment\n",
        "- Evaluation Metrics for classification problems\n",
        "-    Model Building : Part I\n",
        "- Logistic Regression using stratified k-folds cross validation\n",
        "- Feature Engineering\n",
        "- Model Building : Part II\n",
        "- Logistic Regression\n",
        "- Decision tree\n",
        "- Random Forest\n",
        "- XGBoost\n",
        "Lets look at each section in detail."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yUzHWYLucwcU",
        "colab_type": "text"
      },
      "source": [
        "It is a classification problem where we have to predict whether a News Statement would be classified as Fake or Not. In a classification problem, we have to predict discrete values based on a given set of independent variable(s). Classification can be of two types:\n",
        "\n",
        "- Binary Classification : In this classification we have to predict either of the two given classes. For example: classifying the gender as male or female, predicting the result as win or loss, etc.\n",
        "- Multiclass Classification : Here we have to classify the data into three or more classes. For example: classifying a movie's genre as comedy, action or romantic, classify fruits as oranges, apples, or pears, etc."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pER5hD54cwcV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# importing packages\n",
        "import pandas as pd \n",
        "import numpy as np                     # For mathematical calculations \n",
        "import seaborn as sns                  # For data visualization \n",
        "import matplotlib.pyplot as plt        # For plotting graphs \n",
        "%matplotlib inline \n",
        "import warnings                        # To ignore any warnings warnings.filterwarnings(\"ignore\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "afKCp1_pcwcY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "##################\n",
        "import seaborn as sns\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "#from sklearn.cross_validation import train_test_split\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer, CountVectorizer\n",
        "from nltk.stem.porter import PorterStemmer\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import cross_val_score\n",
        "import nltk\n",
        "import re\n",
        "import codecs\n",
        "import seaborn as sns\n",
        "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
        "#from sklearn.cross_validation import train_test_split\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.metrics import f1_score"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DbLFVd_tcwcb",
        "colab_type": "text"
      },
      "source": [
        "# Data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xsG9kz5Kcwcc",
        "colab_type": "text"
      },
      "source": [
        "For this practice problem, we have been given three CSV files: train, test and sample submission.\n",
        "\n",
        "- Train file will be used for training the model, i.e. our model will learn from this file. It contains all the independent variables and the target variable.\n",
        "- Test file contains all the independent variables, but not the target variable. We will apply the model to predict the target variable for the test data.\n",
        "- Sample submission file contains the format in which we have to submit our predictions."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7kpjZ0Gkcwcd",
        "colab_type": "text"
      },
      "source": [
        "- Let's start with loading the data\n",
        "PS: The data was in 'tsv' format which is then converted into 'csv' format later for analysing purpose.\n",
        "The Dataset available for the research purpose containg train,test and validation sets."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UXBXqNCGcwce",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# converting train, test, and valid datasets into csv format\n",
        "#train\n",
        "tsv_file='/content/train.tsv'\n",
        "csv_table=pd.read_table(tsv_file,sep='\\t')\n",
        "csv_table.to_csv('train_converted.csv',index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pdYpc8TNcwch",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#test\n",
        "tsv_file1='/content/test.tsv'\n",
        "csv_table1=pd.read_table(tsv_file1,sep='\\t')\n",
        "csv_table1.to_csv('test_converted.csv',index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KjWJLg8Ccwck",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#valid\n",
        "tsv_file2='/content/valid.tsv'\n",
        "csv_table2=pd.read_table(tsv_file2,sep='\\t')\n",
        "csv_table2.to_csv('valid_converted.csv',index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ezS1HmZ5cwcn",
        "colab_type": "text"
      },
      "source": [
        "# Reading Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TB25_0ANcwcn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train=pd.read_csv('/content/train_converted.csv')\n",
        "test=pd.read_csv('/content/test_converted.csv')\n",
        "valid=pd.read_csv('/content/valid_converted.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CmpdqWaNcwcq",
        "colab_type": "code",
        "outputId": "b00180a8-a2bf-447b-8aca-9bb0169c4b6e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "source": [
        "train.columns"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['2635.json', 'false',\n",
              "       'Says the Annies List political group supports third-trimester abortions on demand.',\n",
              "       'abortion', 'dwayne-bohac', 'State representative', 'Texas',\n",
              "       'republican', '0', '1', '0.1', '0.2', '0.3', 'a mailer'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GVY_G6v_cwcs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Naming columns\n",
        "train.columns=[\"ID\",\"Label\",\"Statement\",\"Subject(s)\",\"Speaker\",\"Speaker's job title\",\"state info\",\"party affiliation\",\"barely true counts\",\"false counts\",\"half true counts\",\"mostly true counts\",\"pants on fire counts\",\"the context (venue/location of speech or statement)\"]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QuJjmTxIcwcv",
        "colab_type": "code",
        "outputId": "c3558e67-ffdc-4cb4-bcf9-4b22461c5440",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 684
        }
      },
      "source": [
        "train.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10540.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>When did the decline of coal start? It started...</td>\n",
              "      <td>energy,history,job-accomplishments</td>\n",
              "      <td>scott-surovell</td>\n",
              "      <td>State delegate</td>\n",
              "      <td>Virginia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>a floor speech.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>324.json</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>Hillary Clinton agrees with John McCain \"by vo...</td>\n",
              "      <td>foreign-policy</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>160.0</td>\n",
              "      <td>163.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>Denver</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1123.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Health care reform legislation is likely to ma...</td>\n",
              "      <td>health-care</td>\n",
              "      <td>blog-posting</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>none</td>\n",
              "      <td>7.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>44.0</td>\n",
              "      <td>a news release</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9028.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>The economic turnaround started at the end of ...</td>\n",
              "      <td>economy,jobs</td>\n",
              "      <td>charlie-crist</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Florida</td>\n",
              "      <td>democrat</td>\n",
              "      <td>15.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>an interview on CNN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>12465.json</td>\n",
              "      <td>true</td>\n",
              "      <td>The Chicago Bears have had more starting quart...</td>\n",
              "      <td>education</td>\n",
              "      <td>robin-vos</td>\n",
              "      <td>Wisconsin Assembly speaker</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>republican</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a an online opinion-piece</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ID  ... the context (venue/location of speech or statement)\n",
              "0  10540.json  ...                                    a floor speech. \n",
              "1    324.json  ...                                             Denver \n",
              "2   1123.json  ...                                     a news release \n",
              "3   9028.json  ...                                an interview on CNN \n",
              "4  12465.json  ...                          a an online opinion-piece \n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j0F65RQBcwcy",
        "colab_type": "code",
        "outputId": "41780100-1723-41de-e80d-8a6280302789",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10239, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8uTeAgnXcwc0",
        "colab_type": "code",
        "outputId": "6d6e1997-8d34-43ef-a988-22aa7761fa80",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1266, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uNiSbwvrcwc2",
        "colab_type": "code",
        "outputId": "d619cfb8-d8c8-4755-c5f3-e477f2f33fda",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "valid.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1283, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EpYxa2G4cwc5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Naming columns \n",
        "test.columns=[\"ID\",\"Label\",\"Statement\",\"Subject(s)\",\"Speaker\",\"Speaker's job title\",\"state info\",\"party affiliation\",\"barely true counts\",\"false counts\",\"half true counts\",\"mostly true counts\",\"pants on fire counts\",\"the context (venue/location of speech or statement)\"]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WBrFfnBFcwc7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Naming columns \n",
        "valid.columns=[\"ID\",\"Label\",\"Statement\",\"Subject(s)\",\"Speaker\",\"Speaker's job title\",\"state info\",\"party affiliation\",\"barely true counts\",\"false counts\",\"half true counts\",\"mostly true counts\",\"pants on fire counts\",\"the context (venue/location of speech or statement)\"]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GfPRQQp1cwc-",
        "colab_type": "text"
      },
      "source": [
        "Let’s make a copy of train and test data so that even if we have to make any changes in these datasets we would not lose the original datasets.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LOH6t8lvcwc-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_copy=pd.read_csv('train.csv').copy()\n",
        "test_copy=pd.read_csv('test.csv').copy()\n",
        "valid_copy=pd.read_csv('valid.csv').copy()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sRyTnT-FcwdB",
        "colab_type": "code",
        "outputId": "bac466ec-608a-4d23-c198-7a41b3b96489",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "train.head(10)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10540.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>When did the decline of coal start? It started...</td>\n",
              "      <td>energy,history,job-accomplishments</td>\n",
              "      <td>scott-surovell</td>\n",
              "      <td>State delegate</td>\n",
              "      <td>Virginia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>a floor speech.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>324.json</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>Hillary Clinton agrees with John McCain \"by vo...</td>\n",
              "      <td>foreign-policy</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>160.0</td>\n",
              "      <td>163.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>Denver</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1123.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Health care reform legislation is likely to ma...</td>\n",
              "      <td>health-care</td>\n",
              "      <td>blog-posting</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>none</td>\n",
              "      <td>7.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>44.0</td>\n",
              "      <td>a news release</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9028.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>The economic turnaround started at the end of ...</td>\n",
              "      <td>economy,jobs</td>\n",
              "      <td>charlie-crist</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Florida</td>\n",
              "      <td>democrat</td>\n",
              "      <td>15.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>an interview on CNN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>12465.json</td>\n",
              "      <td>true</td>\n",
              "      <td>The Chicago Bears have had more starting quart...</td>\n",
              "      <td>education</td>\n",
              "      <td>robin-vos</td>\n",
              "      <td>Wisconsin Assembly speaker</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>republican</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a an online opinion-piece</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2342.json</td>\n",
              "      <td>barely-true</td>\n",
              "      <td>Jim Dunnam has not lived in the district he re...</td>\n",
              "      <td>candidates-biography</td>\n",
              "      <td>republican-party-texas</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Texas</td>\n",
              "      <td>republican</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a press release.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>153.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>I'm the only person on this stage who has work...</td>\n",
              "      <td>ethics</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>160.0</td>\n",
              "      <td>163.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>a Democratic debate in Philadelphia, Pa.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>5602.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>However, it took $19.5 million in Oregon Lotte...</td>\n",
              "      <td>jobs</td>\n",
              "      <td>oregon-lottery</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>organization</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a website</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>9741.json</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>Says GOP primary opponents Glenn Grothman and ...</td>\n",
              "      <td>energy,message-machine-2014,voting-record</td>\n",
              "      <td>duey-stroebel</td>\n",
              "      <td>State representative</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>republican</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>an online video</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>7115.json</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>For the first time in history, the share of th...</td>\n",
              "      <td>elections</td>\n",
              "      <td>robert-menendez</td>\n",
              "      <td>U.S. Senator</td>\n",
              "      <td>New Jersey</td>\n",
              "      <td>democrat</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>a speech</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ID  ... the context (venue/location of speech or statement)\n",
              "0  10540.json  ...                                    a floor speech. \n",
              "1    324.json  ...                                             Denver \n",
              "2   1123.json  ...                                     a news release \n",
              "3   9028.json  ...                                an interview on CNN \n",
              "4  12465.json  ...                          a an online opinion-piece \n",
              "5   2342.json  ...                                   a press release. \n",
              "6    153.json  ...           a Democratic debate in Philadelphia, Pa. \n",
              "7   5602.json  ...                                         a website  \n",
              "8   9741.json  ...                                    an online video \n",
              "9   7115.json  ...                                           a speech \n",
              "\n",
              "[10 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E0_MXC_ScwdE",
        "colab_type": "code",
        "outputId": "6f1253e1-0acd-4dcd-8db0-312ac925826b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "test.head(10)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>11685.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Wisconsin is on pace to double the number of l...</td>\n",
              "      <td>jobs</td>\n",
              "      <td>katrina-shankland</td>\n",
              "      <td>State representative</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>democrat</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>a news conference</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>11096.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Says John McCain has done nothing to help the ...</td>\n",
              "      <td>military,veterans,voting-record</td>\n",
              "      <td>donald-trump</td>\n",
              "      <td>President-Elect</td>\n",
              "      <td>New York</td>\n",
              "      <td>republican</td>\n",
              "      <td>63</td>\n",
              "      <td>114</td>\n",
              "      <td>51</td>\n",
              "      <td>37</td>\n",
              "      <td>61</td>\n",
              "      <td>comments on ABC's This Week.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5209.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>Suzanne Bonamici supports a plan that will cut...</td>\n",
              "      <td>medicare,message-machine-2012,campaign-adverti...</td>\n",
              "      <td>rob-cornilles</td>\n",
              "      <td>consultant</td>\n",
              "      <td>Oregon</td>\n",
              "      <td>republican</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>a radio show</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9524.json</td>\n",
              "      <td>pants-fire</td>\n",
              "      <td>When asked by a reporter whether hes at the ce...</td>\n",
              "      <td>campaign-finance,legal-issues,campaign-adverti...</td>\n",
              "      <td>state-democratic-party-wisconsin</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>democrat</td>\n",
              "      <td>5</td>\n",
              "      <td>7</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>a web video</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5962.json</td>\n",
              "      <td>true</td>\n",
              "      <td>Over the past five years the federal governmen...</td>\n",
              "      <td>federal-budget,pensions,retirement</td>\n",
              "      <td>brendan-doherty</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Rhode Island</td>\n",
              "      <td>republican</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>a campaign website</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>7070.json</td>\n",
              "      <td>true</td>\n",
              "      <td>Says that Tennessee law requires that schools ...</td>\n",
              "      <td>county-budget,county-government,education,taxes</td>\n",
              "      <td>stand-children-tennessee</td>\n",
              "      <td>Child and education advocacy organization.</td>\n",
              "      <td>Tennessee</td>\n",
              "      <td>none</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>in a post on Facebook.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1046.json</td>\n",
              "      <td>barely-true</td>\n",
              "      <td>Says Vice President Joe Biden \"admits that the...</td>\n",
              "      <td>economy,stimulus</td>\n",
              "      <td>john-boehner</td>\n",
              "      <td>Speaker of the House of Representatives</td>\n",
              "      <td>Ohio</td>\n",
              "      <td>republican</td>\n",
              "      <td>13</td>\n",
              "      <td>22</td>\n",
              "      <td>11</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>a press release.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>12849.json</td>\n",
              "      <td>true</td>\n",
              "      <td>Donald Trump is against marriage equality. He ...</td>\n",
              "      <td>gays-and-lesbians,marriage</td>\n",
              "      <td>sean-patrick-maloney</td>\n",
              "      <td>Congressman for NY-18</td>\n",
              "      <td>New York</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>a speech at the Democratic National Convention</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>13270.json</td>\n",
              "      <td>barely-true</td>\n",
              "      <td>We know that more than half of Hillary Clinton...</td>\n",
              "      <td>foreign-policy</td>\n",
              "      <td>mike-pence</td>\n",
              "      <td>Governor</td>\n",
              "      <td>Indiana</td>\n",
              "      <td>republican</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "      <td>12</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>comments on \"Meet the Press\"</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>6649.json</td>\n",
              "      <td>barely-true</td>\n",
              "      <td>We know there are more Democrats in Georgia th...</td>\n",
              "      <td>elections</td>\n",
              "      <td>mike-berlon</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Georgia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>an article</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ID  ... the context (venue/location of speech or statement)\n",
              "0  11685.json  ...                                  a news conference \n",
              "1  11096.json  ...                       comments on ABC's This Week. \n",
              "2   5209.json  ...                                       a radio show \n",
              "3   9524.json  ...                                        a web video \n",
              "4   5962.json  ...                                 a campaign website \n",
              "5   7070.json  ...                             in a post on Facebook. \n",
              "6   1046.json  ...                                   a press release. \n",
              "7  12849.json  ...     a speech at the Democratic National Convention \n",
              "8  13270.json  ...                       comments on \"Meet the Press\" \n",
              "9   6649.json  ...                                         an article \n",
              "\n",
              "[10 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Erhe6HVNcwdG",
        "colab_type": "code",
        "outputId": "74907d60-10dd-4f57-868f-68998d45e9fd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "valid.head(10)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>238.json</td>\n",
              "      <td>pants-fire</td>\n",
              "      <td>When Obama was sworn into office, he DID NOT u...</td>\n",
              "      <td>obama-birth-certificate,religion</td>\n",
              "      <td>chain-email</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>none</td>\n",
              "      <td>11</td>\n",
              "      <td>43</td>\n",
              "      <td>8</td>\n",
              "      <td>5</td>\n",
              "      <td>105</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>7891.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Says Having organizations parading as being so...</td>\n",
              "      <td>campaign-finance,congress,taxes</td>\n",
              "      <td>earl-blumenauer</td>\n",
              "      <td>U.S. representative</td>\n",
              "      <td>Oregon</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>a U.S. Ways and Means hearing</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>8169.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>Says nearly half of Oregons children are poor.</td>\n",
              "      <td>poverty</td>\n",
              "      <td>jim-francesconi</td>\n",
              "      <td>Member of the State Board of Higher Education</td>\n",
              "      <td>Oregon</td>\n",
              "      <td>none</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>an opinion article</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>929.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>On attacks by Republicans that various program...</td>\n",
              "      <td>economy,stimulus</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70</td>\n",
              "      <td>71</td>\n",
              "      <td>160</td>\n",
              "      <td>163</td>\n",
              "      <td>9</td>\n",
              "      <td>interview with CBS News</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>9416.json</td>\n",
              "      <td>false</td>\n",
              "      <td>Says when armed civilians stop mass shootings ...</td>\n",
              "      <td>guns</td>\n",
              "      <td>jim-rubens</td>\n",
              "      <td>Small business owner</td>\n",
              "      <td>New Hampshire</td>\n",
              "      <td>republican</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>in an interview at gun shop in Hudson, N.H.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>6861.json</td>\n",
              "      <td>true</td>\n",
              "      <td>Says Tennessee is providing millions of dollar...</td>\n",
              "      <td>education,state-budget</td>\n",
              "      <td>andy-berke</td>\n",
              "      <td>Lawyer and state senator</td>\n",
              "      <td>Tennessee</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>a letter to state Senate education committee c...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1122.json</td>\n",
              "      <td>false</td>\n",
              "      <td>The health care reform plan would set limits s...</td>\n",
              "      <td>health-care</td>\n",
              "      <td>club-growth</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>none</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>a TV ad</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>13138.json</td>\n",
              "      <td>true</td>\n",
              "      <td>Says Donald Trump started his career back in 1...</td>\n",
              "      <td>candidates-biography,diversity,housing</td>\n",
              "      <td>hillary-clinton</td>\n",
              "      <td>Presidential candidate</td>\n",
              "      <td>New York</td>\n",
              "      <td>democrat</td>\n",
              "      <td>40</td>\n",
              "      <td>29</td>\n",
              "      <td>69</td>\n",
              "      <td>76</td>\n",
              "      <td>7</td>\n",
              "      <td>the first presidential debate</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>1880.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>Bill White has a long history of trying to lim...</td>\n",
              "      <td>military</td>\n",
              "      <td>republican-party-texas</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Texas</td>\n",
              "      <td>republican</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>an e-mail</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>12803.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>John McCains chief economic adviser during the...</td>\n",
              "      <td>economy</td>\n",
              "      <td>tim-kaine</td>\n",
              "      <td>U.S. Senator</td>\n",
              "      <td>Virginia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>8</td>\n",
              "      <td>3</td>\n",
              "      <td>15</td>\n",
              "      <td>15</td>\n",
              "      <td>0</td>\n",
              "      <td>a speech at the Democratic National Convention...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ID  ... the context (venue/location of speech or statement)\n",
              "0    238.json  ...                                                NaN \n",
              "1   7891.json  ...                      a U.S. Ways and Means hearing \n",
              "2   8169.json  ...                                 an opinion article \n",
              "3    929.json  ...                            interview with CBS News \n",
              "4   9416.json  ...        in an interview at gun shop in Hudson, N.H. \n",
              "5   6861.json  ...  a letter to state Senate education committee c... \n",
              "6   1122.json  ...                                            a TV ad \n",
              "7  13138.json  ...                      the first presidential debate \n",
              "8   1880.json  ...                                          an e-mail \n",
              "9  12803.json  ...  a speech at the Democratic National Convention... \n",
              "\n",
              "[10 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ARaQTXkdcwdM",
        "colab_type": "text"
      },
      "source": [
        "Loaded the datasets. Let's move to Univariate Analysis"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U_pRBXpQ5hk9",
        "colab_type": "text"
      },
      "source": [
        "- Combining train and validation sets here. Later We will use validation techniques"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7tkyy9Bj5mOO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train = pd.concat([train, valid])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pIEMnB265mLn",
        "colab_type": "code",
        "outputId": "1961bb67-a412-4570-9275-82c1e7fafdf9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(11522, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZThj0VLn5mId",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Dropping Target Variable from test set\n",
        "Label_test = test.iloc[:,[1]]\n",
        "test=test.drop(['Label'], axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iY-cipWucwdN",
        "colab_type": "text"
      },
      "source": [
        "# Univariate Analysis"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jpnNWeXhcwdP",
        "colab_type": "code",
        "outputId": "e72ec647-9c80-4009-f935-7bc59d8ae665",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        }
      },
      "source": [
        "print(train.dtypes)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "ID                                                      object\n",
            "Label                                                   object\n",
            "Statement                                               object\n",
            "Subject(s)                                              object\n",
            "Speaker                                                 object\n",
            "Speaker's job title                                     object\n",
            "state info                                              object\n",
            "party affiliation                                       object\n",
            "barely true counts                                     float64\n",
            "false counts                                           float64\n",
            "half true counts                                       float64\n",
            "mostly true counts                                     float64\n",
            "pants on fire counts                                   float64\n",
            "the context (venue/location of speech or statement)     object\n",
            "dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TD8qLRlCcwdT",
        "colab_type": "code",
        "outputId": "6bf0d032-fab1-43eb-dce0-e5539e609b21",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "print(test.dtypes)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "ID                                                     object\n",
            "Statement                                              object\n",
            "Subject(s)                                             object\n",
            "Speaker                                                object\n",
            "Speaker's job title                                    object\n",
            "state info                                             object\n",
            "party affiliation                                      object\n",
            "barely true counts                                      int64\n",
            "false counts                                            int64\n",
            "half true counts                                        int64\n",
            "mostly true counts                                      int64\n",
            "pants on fire counts                                    int64\n",
            "the context (venue/location of speech or statement)    object\n",
            "dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C7HjlMsfcwdX",
        "colab_type": "code",
        "outputId": "ee922f5d-34e2-4007-b866-a94c7a38fd1b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# shape of datasets\n",
        "train.shape, test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((11522, 14), (1266, 13))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c-snudAFcwda",
        "colab_type": "code",
        "outputId": "485ecd73-d66c-4e9e-9be7-cea41609b16a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 141
        }
      },
      "source": [
        "train['Label'].value_counts(normalize=True)\n",
        "# Read as percentage after multiplying by 100"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "half-true      0.204999\n",
              "false          0.195886\n",
              "mostly-true    0.192067\n",
              "barely-true    0.164034\n",
              "true           0.160128\n",
              "pants-fire     0.082885\n",
              "Name: Label, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qwCQDcJEcwdd",
        "colab_type": "code",
        "outputId": "1f0b9bf7-e752-456a-8014-b4bb2e1a0a51",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 331
        }
      },
      "source": [
        "train['Label'].value_counts().plot.bar()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f22cb321b70>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEoCAYAAAC0OiEVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nDbsjPDAcwdf",
        "colab_type": "text"
      },
      "source": [
        "On Looking at the datsets, we identified that there are 3 Data Types:\n",
        "- Continuous : ID, barely true counts, false counts, half true counts, mostly true counts, pants on fire counts.\n",
        "\n",
        "- Categorical : Label, Subject(s), Speaker, Speaker's job title, State info, party affiliation, the context (venue/location of speech or statement)\n",
        "\n",
        "- Text : Statement"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BbQmrQsFcwdg",
        "colab_type": "text"
      },
      "source": [
        "Let's go for Continuous Data Type Exploration.\n",
        "We know that we use BarGraphs for Categorical variable, Histogram or ScatterPlot for continuous variables."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oEih0BBYcwdg",
        "colab_type": "text"
      },
      "source": [
        "- Continuous Variables"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iyq5H1SUcwdh",
        "colab_type": "code",
        "outputId": "d387d922-9418-4b23-964d-e7f42978498b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        }
      },
      "source": [
        "# barely true counts\n",
        "plt.figure(figsize=(20, 6))\n",
        "sns.countplot(train[\"barely true counts\"])\n",
        "plt.title('barely true counts')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABJgAAAGDCAYAAABnfapPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0hQ82hHDVsYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JL4eZlMtcwdj",
        "colab_type": "code",
        "outputId": "43981c06-970a-4440-fd44-0c908eca2d49",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        }
      },
      "source": [
        "# false counts\n",
        "plt.figure(figsize=(20, 6))\n",
        "sns.countplot(train[\"false counts\"])\n",
        "plt.title('false counts')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABJgAAAGDCAYAAABnfapPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nETABAAAAMIqACQAAAIBR/n8ZB3OrAwtK1gAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dR5zkCpdcwdm",
        "colab_type": "code",
        "outputId": "5c4a1e7e-aaf9-4a2a-94c9-fac1ab97cf88",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        }
      },
      "source": [
        "# half true counts\n",
        "plt.figure(figsize=(20, 6))\n",
        "sns.countplot(train[\"half true counts\"])\n",
        "plt.title('half true counts')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABJgAAAGDCAYAAABnfapPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I2-wC5Ykcwdo",
        "colab_type": "code",
        "outputId": "3b624d36-d479-47f1-e953-cecc38abf4c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        }
      },
      "source": [
        "# mostly true counts\n",
        "plt.figure(figsize=(20, 6))\n",
        "sns.countplot(train[\"mostly true counts\"])\n",
        "plt.title('mostly true counts')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABJgAAAGDCAYAAABnfapPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3de7xtdVkv/s8jeL8BsiUEClPU6BxF\nQ7Sy8lIKZqBGhnnBS5Emx8upTlrnnEx/nmNa8stLejRRvCSa151hSmhmligodyS3igeQm+LdQsHn\n/DHH1ulizb3XZqy15lyb9/v1mq815nd8x5jPHGNexvyscanuDgAAAABcXzeadwEAAAAAbGwCJgAA\nAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUARMAwAxV9Y9V9ZvzrgMAYNEJmAAAklTV\nc6vqTas0rydU1T+vxrwWmQAOANhKwAQAMAdVtcu8awAAWC0CJgBg4VXVhVX1+1V1VlV9q6peW1V7\nVdX7quobVfUPVbX7VP/Dq+rcqvrqsJfNT0yN+4OqumSY7oKqelBVHZrkD5P8elV9s6rOXPL4N6mq\nq6rqP0+13b6qvl1Vm5b0/Ykkr0ry08O8vjq0v76qXllVJ1XVt5I8YOkeQEv3fKqqu1XVycNjX1BV\nj9rGMtqjql5XVV+sqq9U1bunxv1WVW0Z5rO5qu4wtO9fVV1Vu071/X5NW+upqj8b5vn5qjpsGPeC\nJD+X5OXD83x5TRxXVVdU1der6uyq+k/bWb0AwE5AwAQAbBS/muSXktwlya8keV8modCmTLZpnp4k\nVXWXJG9J8sxh3ElJ/nYIie6a5Ngk9+7uWyd5SJILu/vvk/yvJG/t7lt19z2mH7i7v5PkxCSPnWp+\ndJJTuvvKJX3PT/KUJP86zGu3qdG/keQFSW6dZJuH0FXVLZOcnOSvk9w+yVFJ/rKqDpwxyRuT3CLJ\nTw79jxvm88Ak/zvJo5LsneQLw3NZqfskuSDJnklelOS1VVXd/UdJPpLk2OF5HpvkwUl+PpN1dNvh\nMb+8A48FAGxQAiYAYKN4WXdf3t2XZBJsnNrdn+ru/0jyriT3HPr9epK/6+6Tu/u7Sf4syc2T/EyS\na5PcNMmBVXXj7r6wuz+7wsc/Icmjq6qG+4/LJNTZEe/p7o929/eGurflYZmEX6/r7mu6+1NJ3pHk\n15Z2rKq9kxyW5Cnd/ZXu/m53f3gY/Zgkx3f3J7v76iTPyWTvqv1XWPMXuvs13X1tJstg7yR7zej7\n3UzCs7slqe4+v7svXeHjAAAbmIAJANgoLp8a/vdl7t9qGL5DJnvpJEm6+3tJLkqyT3dvyWTPpucm\nuaKqTtx6uNj2dPepSb6d5P5Vdbckd06yeQefw0U70PfHktxnOMzvq8Ohdo9J8iPL9N0vyVXd/ZVl\nxi1dHt/MZK+ifVZYx2VT0357GLzVch27+4NJXp7kFZks31dX1W1W+DgAwAYmYAIAdjZfzCScSZIM\nexztl+SSJOnuv+7u+w19OsmfDl17BfM+IZPD5B6X5O3b2Atp1ryWtn8rk8PatpoOjy5K8uHu3m3q\ndqvufuoy870oyR5Vtdsy45Yuj1smuV0my+NbQ/OsGrbnOs+zu1/a3T+V5MBMDpX7/R2YHwCwQQmY\nAICdzduS/PJw8u4bJ/ndJFcn+ZequmtVPbCqbprkPzLZ8+l7w3SXJ9m/qra1ffSmJI/IJGR6wzb6\nXZ5k36q6yXZqPSPJI6vqFlV15yRPnhr33iR3qarHVdWNh9u9p09YvtVwGNr7MjlH0+5D358fRr8l\nyROr6qDhef+vTA4vvHA4f9QlSR5bVbtU1ZOS3Gk7NS99nj++9c5Q332G5f6tTJbx92ZNDADsPARM\nAMBOpbsvyCQAelmSL2VyQvBfGU7UfdMkLxzaL8vkZNjPGSb9m+Hvl6vqkzPmfVGST2ay585HtlHG\nB5Ocm+SyqvrSNvodl+Q7mQQ1JyR589RjfSOTk2YflcleSJdlsrfVTWfM63GZnAPp00muyORQwHT3\nPyT5H5mcv+nSTAKko6am+61M9jL6ciYnCP+XbdS71F8kOXK4wtxLk9wmyWuSfCWTw/K+nOTFOzA/\nAGCDqu6V7A0OAECSVNXxSb7Y3f993rUAACyKXeddAADARjFcee2R+cEV6wAAiEPkAABWpKqen+Sc\nJC/u7s/Pux4AgEXiEDkAAAAARrEHEwAAAACjCJgAAAAAGGWnPMn3nnvu2fvvv/+8ywAAAADYaZx+\n+ulf6u5Ny43bKQOm/fffP6eddtq8ywAAAADYaVTVF2aNc4gcAAAAAKMImAAAAAAYRcAEAAAAwCgC\nJgAAAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUARMAAAAAowiYAAAAABhFwAQAAADA\nKAImAAAAAEYRMAEAAAAwyq7zLmC9XPnKN827hGx66mPnXQIAAADAqrMHEwAAAACjCJgAAAAAGEXA\nBAAAAMAoAiYAAAAARhEwAQAAADDKmgVMVXWzqvp4VZ1ZVedW1Z8M7XesqlOraktVvbWqbjK033S4\nv2UYv//UvJ4ztF9QVQ9Zq5oBAAAA2HFruQfT1Uke2N33SHJQkkOr6r5J/jTJcd195yRfSfLkof+T\nk3xlaD9u6JeqOjDJUUl+MsmhSf6yqnZZw7oBAAAA2AFrFjD1xDeHuzcebp3kgUnePrSfkOThw/AR\nw/0M4x9UVTW0n9jdV3f355NsSXLIWtUNAAAAwI5Z03MwVdUuVXVGkiuSnJzks0m+2t3XDF0uTrLP\nMLxPkouSZBj/tSS3m25fZprpxzqmqk6rqtOuvPLKtXg6AAAAACxjTQOm7r62uw9Ksm8mex3dbQ0f\n69XdfXB3H7xp06a1ehgAAAAAlliXq8h191eTfCjJTyfZrap2HUbtm+SSYfiSJPslyTD+tkm+PN2+\nzDQAAAAAzNlaXkVuU1XtNgzfPMkvJTk/k6DpyKHb0UneMwxvHu5nGP/B7u6h/ajhKnN3THJAko+v\nVd0AAAAA7Jhdt9/lets7yQnDFd9ulORt3f3eqjovyYlV9f8l+VSS1w79X5vkjVW1JclVmVw5Lt19\nblW9Lcl5Sa5J8rTuvnYN6wYAAABgB6xZwNTdZyW55zLtn8syV4Hr7v9I8msz5vWCJC9Y7RoBAAAA\nGG9dzsEEAAAAwM5LwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImACAAAAYBQB\nEwAAAACjCJgAAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwiYAIAAABg\nFAETAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAA\nAGAUARMAAAAAowiYAAAAABhFwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImAC\nAAAAYBQBEwAAAACjCJgAAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwi\nYAIAAABgFAETAAAAAKMImAAAAAAYZc0Cpqrar6o+VFXnVdW5VfWMof25VXVJVZ0x3B46Nc1zqmpL\nVV1QVQ+Zaj90aNtSVc9eq5oBAAAA2HG7ruG8r0nyu939yaq6dZLTq+rkYdxx3f1n052r6sAkRyX5\nySR3SPIPVXWXYfQrkvxSkouTfKKqNnf3eWtYOwAAAAArtGYBU3dfmuTSYfgbVXV+kn22MckRSU7s\n7quTfL6qtiQ5ZBi3pbs/lyRVdeLQV8AEAAAAsADW5RxMVbV/knsmOXVoOraqzqqq46tq96FtnyQX\nTU128dA2q33pYxxTVadV1WlXXnnlKj8DAAAAAGZZ84Cpqm6V5B1JntndX0/yyiR3SnJQJns4/flq\nPE53v7q7D+7ugzdt2rQaswQAAABgBdbyHEypqhtnEi69ubvfmSTdffnU+Nckee9w95Ik+01Nvu/Q\nlm20AwAAADBna3kVuUry2iTnd/dLptr3nur2iCTnDMObkxxVVTetqjsmOSDJx5N8IskBVXXHqrpJ\nJicC37xWdQMAAACwY9ZyD6afTfK4JGdX1RlD2x8meXRVHZSkk1yY5LeTpLvPraq3ZXLy7muSPK27\nr02Sqjo2yfuT7JLk+O4+dw3rBgAAAGAHrOVV5P45SS0z6qRtTPOCJC9Ypv2kbU0HAAAAwPysy1Xk\nAAAAANh5CZgAAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwiYAIAAABg\nFAETAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAA\nAGAUARMAAAAAowiYAAAAABhFwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImAC\nAAAAYBQBEwAAAACjCJgAAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwi\nYAIAAABgFAETAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABGETABAAAAMIqACQAAAIBRBEwAAAAA\njCJgAgAAAGAUARMAAAAAo6xZwFRV+1XVh6rqvKo6t6qeMbTvUVUnV9Vnhr+7D+1VVS+tqi1VdVZV\n3WtqXkcP/T9TVUevVc0AAAAA7Li13IPpmiS/290HJrlvkqdV1YFJnp3klO4+IMkpw/0kOSzJAcPt\nmCSvTCaBVJI/TnKfJIck+eOtoRQAAAAA87dmAVN3X9rdnxyGv5Hk/CT7JDkiyQlDtxOSPHwYPiLJ\nG3riY0l2q6q9kzwkycndfVV3fyXJyUkOXau6AQAAANgx63IOpqraP8k9k5yaZK/uvnQYdVmSvYbh\nfZJcNDXZxUPbrHYAAAAAFsCaB0xVdask70jyzO7++vS47u4kvUqPc0xVnVZVp1155ZWrMUsAAAAA\nVmBNA6aqunEm4dKbu/udQ/Plw6FvGf5eMbRfkmS/qcn3Hdpmtf+Q7n51dx/c3Qdv2rRpdZ8IAAAA\nADOt5VXkKslrk5zf3S+ZGrU5ydYrwR2d5D1T7Y8friZ33yRfGw6le3+SB1fV7sPJvR88tAEAAACw\nAHZdw3n/bJLHJTm7qs4Y2v4wyQuTvK2qnpzkC0keNYw7KclDk2xJ8u0kT0yS7r6qqp6f5BNDv+d1\n91VrWDcAAAAAO2DNAqbu/uckNWP0g5bp30meNmNexyc5fvWqAwAAAGC1rMtV5AAAAADYeQmYAAAA\nABhFwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImACAAAAYBQBEwAAAACjCJgA\nAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwiYAIAAABgFAETAAAAAKMI\nmAAAAAAYRcAEAAAAwCgCJgAAAABGETABAAAAMMqKAqaqOmUlbQAAAADc8Oy6rZFVdbMkt0iyZ1Xt\nnqSGUbdJss8a1wYAAADABrDNgCnJbyd5ZpI7JDk9PwiYvp7k5WtYFwAAAAAbxDYDpu7+iyR/UVX/\npbtftk41AQAAALCBbG8PpiRJd7+sqn4myf7T03T3G9aoLgAAAAA2iBUFTFX1xiR3SnJGkmuH5k4i\nYAIAAAC4gVtRwJTk4CQHdnevZTEAAAAAbDwrDZjOSfIjSS5dw1pu8K541XHzLiG3f8qz5l0CAAAA\nsMGsNGDaM8l5VfXxJFdvbezuw9ekKgAAAAA2jJUGTM9dyyIAAAAA2LhWehW5D691IQAAAABsTCu9\nitw3MrlqXJLcJMmNk3yru2+zVoUBAAAAsDGsdA+mW28drqpKckSS+65VUQAAAABsHDfa0Ql64t1J\nHrIG9QAAAACwwaz0ELlHTt29UZKDk/zHmlQEAAAAwIay0qvI/crU8DVJLszkMDkAAAAAbuBWeg6m\nJ651IQAAAABsTCs6B1NV7VtV76qqK4bbO6pq37UuDgAAAIDFt9KTfL8uyeYkdxhufzu0AQAAAHAD\nt9KAaVN3v667rxlur0+yaQ3rAgAAAGCDWGnA9OWqemxV7TLcHpvky2tZGAAAAAAbw0oDpicleVSS\ny5JcmuTIJE/Y1gRVdfxwvqZzptqeW1WXVNUZw+2hU+OeU1VbquqCqnrIVPuhQ9uWqnr2Djw3AAAA\nANbBSgOm5yU5urs3dfftMwmc/mQ707w+yaHLtB/X3QcNt5OSpKoOTHJUkp8cpvnLrXtLJXlFksOS\nHJjk0UNfAAAAABbESgOmu3f3V7be6e6rktxzWxN09z8luWqF8z8iyYndfXV3fz7JliSHDLct3f25\n7v5OkhOHvgAAAAAsiJUGTDeqqt233qmqPZLsej0f89iqOms4hG7rPPdJctFUn4uHtlnt11FVx1TV\naVV12pVXXnk9SwMAAABgR600YPrzJP9aVc+vqucn+ZckL7oej/fKJHdKclAm53L68+sxj2V196u7\n++DuPnjTJhe4AwAAAFgvK9oLqbvfUFWnJXng0PTI7j5vRx+suy/fOlxVr0ny3uHuJUn2m+q679CW\nbbQDAAAAsABWfJjbECjtcKg0rar27u5Lh7uPSLL1CnObk/x1Vb0kyR2SHJDk40kqyQFVdcdMgqWj\nkvzGmBoAAAAAWF3X9zxK21VVb0ly/yR7VtXFSf44yf2r6qAkneTCJL+dJN19blW9LZMA65okT+vu\na4f5HJvk/Ul2SXJ8d5+7VjUDAAAAsOPWLGDq7kcv0/zabfR/QZIXLNN+UpKTVrE0AAAAAFbRSk/y\nDQAAAADLEjABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUARMAAAAAowiYAAAAABhFwAQAAADA\nKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImACAAAAYBQBEwAAAACjCJgAAAAAGEXABAAA\nAMAoAiYAAAAARhEwAQAAADCKgAkAAACAUQRMAAAAAIwiYAIAAABgFAETAAAAAKMImAAAAAAYRcAE\nAAAAwCgCJgAAAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUARMAAAAAowiYAAAAABhF\nwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImACAAAAYBQBEwAAAACjCJgAAAAA\nGEXABAAAAMAoAiYAAAAARlmzgKmqjq+qK6rqnKm2Parq5Kr6zPB396G9quqlVbWlqs6qqntNTXP0\n0P8zVXX0WtULAAAAwPWzlnswvT7JoUvanp3klO4+IMkpw/0kOSzJAcPtmCSvTCaBVJI/TnKfJIck\n+eOtoRQAAAAAi2HNAqbu/qckVy1pPiLJCcPwCUkePtX+hp74WJLdqmrvJA9JcnJ3X9XdX0lycq4b\nWgEAAAAwR+t9Dqa9uvvSYfiyJHsNw/skuWiq38VD26z266iqY6rqtKo67corr1zdqgEAAACYaW4n\n+e7uTtKrOL9Xd/fB3X3wpk2bVmu2AAAAAGzHegdMlw+HvmX4e8XQfkmS/ab67Tu0zWoHAAAAYEGs\nd8C0OcnWK8EdneQ9U+2PH64md98kXxsOpXt/kgdX1e7Dyb0fPLQBAAAAsCB2XasZV9Vbktw/yZ5V\ndXEmV4N7YZK3VdWTk3whyaOG7icleWiSLUm+neSJSdLdV1XV85N8Yuj3vO5eeuJwAAAAAOZozQKm\n7n70jFEPWqZvJ3najPkcn+T4VSwNAAAAgFU0t5N8AwAAALBzEDABAAAAMMqaHSLHzunil//2vEvI\nvsf+n3mXAAAAAEyxBxMAAAAAowiYAAAAABhFwAQAAADAKAImAAAAAEYRMAEAAAAwioAJAAAAgFEE\nTAAAAACMImACAAAAYBQBEwAAAACjCJgAAAAAGEXABAAAAMAoAiYAAAAARhEwAQAAADCKgAkAAACA\nUQRMAAAAAIwiYAIAAABgFAETAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABGETABAAAAMIqACQAA\nAIBRBEwAAAAAjCJgAgAAAGAUARMAAAAAo+w67wKAxbX5+MPmXUIOf9L75l0CAAAA22EPJgAAAABG\nETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUV5GDOfnwa3553iXkF37r7+ZdAgAAADsBARM7\npbNeefi8S8jdn7p53iUAAADAunCIHAAAAACjCJgAAAAAGEXABAAAAMAoAiYAAAAARnGSbwAAAIAF\nc8XL53/V79sfu/Krn89lD6aqurCqzq6qM6rqtKFtj6o6uao+M/zdfWivqnppVW2pqrOq6l7zqBkA\nAACA5c3zELkHdPdB3X3wcP/ZSU7p7gOSnDLcT5LDkhww3I5J8sp1rxQAAACAmRbpHExHJDlhGD4h\nycOn2t/QEx9LsltV7T2PAgEAAAC4rnkFTJ3kA1V1elUdM7Tt1d2XDsOXJdlrGN4nyUVT0148tAEA\nAACwAOZ1ku/7dfclVXX7JCdX1aenR3Z3V1XvyAyHoOqYJPnRH/3R1asUAAAAgG2ayx5M3X3J8PeK\nJO9KckiSy7ce+jb8vWLofkmS/aYm33doWzrPV3f3wd198KZNm9ayfAAAAACmrPseTFV1yyQ36u5v\nDMMPTvK8JJuTHJ3khcPf9wyTbE5ybFWdmOQ+Sb42dSgdAAAAjHbeKy+fdwk58Kl7bb8TLKh5HCK3\nV5J3VdXWx//r7v77qvpEkrdV1ZOTfCHJo4b+JyV5aJItSb6d5InrXzIAAAAAs6x7wNTdn0tyj2Xa\nv5zkQcu0d5KnrUNpAAAAAFwP87qKHAAAAAA7CQETAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABG\nETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAAAGAUARMAAAAAowiYAAAAABhFwAQAAADAKAImAAAA\nAEYRMAEAAAAwioAJAAAAgFEETAAAAACMImACAAAAYJRd510AwFhvfP1D5l1CHveE98+7BADWwSPe\n8eF5l5B3/eovzLsEALgOARMAAABwg3L5Sz8y7xKy19N/bt4lrCqHyAEAAAAwij2YANbBK940/8P4\nnvZYh/Gtl8Pe/Yx5l5D3Pfwv5l0CAAA3IPZgAgAAAGAUARMAAAAAowiYAAAAABhFwAQAAADAKE7y\nDQAAsIxXvfPyeZeQpzxyr3mXALAi9mACAAAAYBQBEwAAAACjOEQOgCTJ89/6kHmXkP/x6++fdwkA\nAAvr0hddPO8Ssvd/23feJbCgBEwAALATOfIdZ867hLz9V+8x7xIAWGcCJgA2lCe+69B5l5DXPeLv\n510CAAAsFAETAAAAa+qjb7hy3iXkZx+/ad4lwE5NwAQAALBBbf6bL827hBz+a3vOuwRgAbiKHAAA\nAACjCJgAAAAAGMUhcgAAIzzsHa+fdwl5768+Yd4lAMD3XX7cWfMuIXs96+7zLuEGxx5MAAAAAIwi\nYAIAAABgFIfIAcAN0EPf9b/nXUJOesRz5l3CDcbD3v62eZeQ9x75qHmXwIJ5+rsumncJeekj9pt3\nCQA7DQETALCwfvmdfznvEvJ3j/ydeZdwg3H429877xKy+ciHzbsEANiQHCIHAAAAwCgbJmCqqkOr\n6oKq2lJVz553PQAAAABMbIiAqap2SfKKJIclOTDJo6vqwPlWBQAAAECyQQKmJIck2dLdn+vu7yQ5\nMckRc64JAAAAgGycgGmfJNOXmbh4aAMAAABgzqq7513DdlXVkUkO7e7fHO4/Lsl9uvvYqT7HJDlm\nuHvXJBeschl7JvnSKs9ztW2EGpONUacaV89GqFONq2cj1KnG1bMR6lTj6tkIdapx9WyEOtW4ejZC\nnWpcPRuhTjWuntWu88e6e9NyI3ZdxQdZS5ck2W/q/r5D2/d196uTvHqtCqiq07r74LWa/2rYCDUm\nG6NONa6ejVCnGlfPRqhTjatnI9SpxtWzEepU4+rZCHWqcfVshDrVuHo2Qp1qXD3rWedGOUTuE0kO\nqKo7VtVNkhyVZPOcawIAAAAgG2QPpu6+pqqOTfL+JLskOb67z51zWQAAAABkgwRMSdLdJyU5aY4l\nrNnhd6toI9SYbIw61bh6NkKdalw9G6FONa6ejVCnGlfPRqhTjatnI9SpxtWzEepU4+rZCHWqcfWs\nW50b4iTfAAAAACyujXIOJgAAAAAWlIBpiao6tKouqKotVfXsZcbftKreOow/tar2n0ONx1fVFVV1\nzozxVVUvHWo8q6ruNYca96uqD1XVeVV1blU9Y9HqrKqbVdXHq+rMocY/WabP3Nf3UMcuVfWpqnrv\nMuMWpcYLq+rsqjqjqk5bZvwivC53q6q3V9Wnq+r8qvrpRaqxqu46LL+tt69X1TMXqcapOp41vG/O\nqaq3VNXNloxf99flcp+NVbVHVZ1cVZ8Z/u4+Y9qjhz6fqaqj17nGXxuW5feqauYVPrb3/bQOdb54\neO+cVVXvqqrd5lnntr4Lq+p3q6qras8Z085zfT+3qi6Zep8/dMa0c13fQ/t/Gdb5uVX1onnWOWNZ\nHlRVH9v6vVNVh8yYdp7r+x5V9a81+X7826q6zYxp12s5Lrt9VlXPH97bZ1TVB6rqDjOmX69luc3t\nyEV4j9eM7ciqeu3QdlZNtjluNWP65wzr+4Kqesha1LitOqfGv7SqvrmN6de8zm0sy9dX1eenPi8P\nmjH9PNf3R6bq+2JVvXteNU491g/9bqjJhbJOHdbjW2ty0azlpluX1+SMGo8dHnvm+3rot16fQTO3\nyWvO3421A9/ZVXXI1HM4s6oeMWOeK3qNrEh3uw23TE4g/tkkP57kJknOTHLgkj6/k+RVw/BRSd46\nhzp/Psm9kpwzY/xDk7wvSSW5b5JT51Dj3knuNQzfOsm/LbMs51rn8Li3GoZvnOTUJPddtPU9PPZ/\nTfLXSd67zLhFqfHCJHtuY/wivC5PSPKbw/BNkuy2aDVO1bJLksuS/Nii1ZhknySfT3Lz4f7bkjxh\nSZ91f10u99mY5EVJnj0MPzvJny4z3R5JPjf83X0Y3n0da/yJJHdN8o9JDt7G62Gb30/rUOeDk+w6\nDP/pjGW5bnUuV+PQvl8mFwXp10gAAA6oSURBVAX5wnKfSQuwvp+b5Pe2M90irO8HJPmHJDcd7t9+\n0dZ3kg8kOWwYfmiSf1zA9f2JJL8wDD8pyfPnvByX3T5LcpupPk/P8Pk9x2U5cztyUd7jmbEduWRZ\nviTDd9CSaQ8c1vNNk9xxWP+7rNGynLm9m+TgJG9M8s0Z065LndtYlq9PcuR2pp3r+l7S5x1JHj+v\nGqce74d+N2SynXbUMPyqJE+d52tyRo33TLJ/tvF7Yr2X49Tjfn+bPAvw3Zgd+M5Ocov8YNtt7yRX\nbL2/ZJ7bfY2s9GYPph92SJIt3f257v5OkhOTHLGkzxGZ/EhNkrcneVBV1TrWmO7+pyRXbaPLEUne\n0BMfS7JbVe29PtVNdPel3f3JYfgbSc7P5EfpwtQ5PO7W/9jceLgtPSnZ3Nd3Ve2b5JeT/NWMLnOv\ncYXmur6r6raZfCC/Nkm6+zvd/dVFqnGJByX5bHd/YUn7otS4a5KbV9WumXx5fXHJ+HV/Xc74bJyu\n44QkD19m0ockObm7r+ruryQ5Ocmh61Vjd5/f3RdsZ9KVfD+tmhl1fqC7rxnufizJvvOscxvfhccl\n+W+57uf5VnNd3ys09/Wd5KlJXtjdVw99rphnnTNq7CRb9wi6ba77OZTMf33fJck/DcMnJ/nVZSZd\nz+W47PZZd399qtsts/z7Zz2X5ba2IxfiPT5rO3Lrshy+824+o84jkpzY3Vd39+eTbMnkdbDqZtVZ\nVbskeXEmy3KWdalzhdvks8x1fW8dX5O9Ex+YZLk9mNbtvbP0d8PwOnxgJttiyextoXV7TS7326a7\nP9XdF25n0nVbjktMb5PP/btxR76zu/vbU9tuN8sy76sdeI2siIDph+2T5KKp+xfnuqHI9/sMK+tr\nSW63LtWt3Eqex7qpyaEx98wk6Z829zqH3TPPyCTNPbm7Z9Y4x/X9/2fy5f+9GeMXocZk8oH1gao6\nvaqOWWb8vNf3HZNcmeR1wy65f1VVt1zSZ941TjsqyVuWaZ97jd19SZI/S/J/k1ya5Gvd/YEl3Rbl\ndblXd186DF+WZK9l+sx9ma7AotX4pEz2pFtqrnVW1RFJLunuM7fRbRGW5bE1OXzm+Fr+sM1FqPEu\nSX5u2GX+w1V172X6zLvOZyZ5cVVdlMln0nOW6TPvGs/ND35Y/Fome98sNZcal26fVdULhmX5mCT/\nc5lJ5l7nor3HZ21HVtXrMvnOuVuSl82zxm3UeWySzVPfkcuZ+7JM8oLh8/K4qrrpgtaYTH6Qn7Ik\nrF33GnPd3w23S/LVqZBh1mPPs8aVmtfn+fQ2+aJ+N86sq6ruU1XnJjk7yVOmXgtbrfQ1siICJtZU\nTY47f0eSZ874wJ2r7r62uw/K5D/xh1TVf5p3TdOq6mFJruju0+ddywrcr7vvleSwJE+rqp+fd0FL\n7JrJ7qSv7O57JvlWJodMLZzhuOfDk/zNvGtZzvCD+IhMQrs7JLllVT12vlVtX3d3Vv4fUWaoqj9K\nck2SN8+7lmlVdYskf5jlfxgvklcmuVOSgzIJaP98vuXMtGsmhyHcN8nvJ3nbAu4d+9Qkz+ru/ZI8\nK8MeqgvmSUl+p6pOz+RQr+/MuZ4ky2+fdfcfDcvyzZmED3M3XWcmnzsL9R6ftR3Z3U/M5Pvx/CS/\nPscSkyxb589nEnguF37NxYxl+ZxMQrp7Z/J59AdzLHF7vxseneX/MbhuNsLvho1Q47RltskX9btx\nZl3dfWp3/2Qm76Pn1JLzpq42AdMPuyQ//J+lfYe2ZfsMh4bcNsmX16W6lVvJ81hzVXXjTDYK3tzd\n71ymy0LUmSQ9OVTqQ7nubpbzXt8/m+Twqrowk90rH1hVb1qwGpN8f6+WrbtkvivX3a123uv74iQX\nT/236e2ZBE7T5l3jVocl+WR3X77MuEWo8ReTfL67r+zu7yZ5Z5KfWdJnIV6XSS7fegjh8He5XZkX\nYZluz0LUWFVPSPKwJI8ZArul5lnnnTIJPc8cPjP3TfLJqvqRJf3muiy7+/LhR8r3krwmyx+CsAjr\n++Ik7xwOC/l4Jv9pXnri1XnXeXQmnz/JZON/4ZZld3+6ux/c3T+VyY/Pzy7TbV1rXMH22Zuz/KF8\n865zYd/jy21Hdve1mWy7zX1ZTtW0tc4HJLlzki3DsrxFVW1ZhDqnl+VwqGQPh/28LgvyHl+6vmty\nUupDkvzdjEnWq8br/G5I8heZnE5h1+089txqXOa3zSzzeN8s3SZf1O/G7dbV3ecn+WaSpTtUfDkr\ne42siIDph30iyQHDWdRvksnucJuX9NmcyQZNkhyZ5IMzNrLnaXOSx9fEfTM5fGVbu7+uuiExfW2S\n87v7JTO6zbXOqtpUw1WQqurmSX4pyaeXqXFu67u7n9Pd+3b3/pm8Hj/Y3Uv3FJn7a7KqbllVt946\nnMnJgJde2Wmu67u7L0tyUVXddWh6UJLzFqnGKdv6L9gi1Ph/k9y3qm4xvNcflMl/aZfWuQifldN1\nHJ3kPcv0eX+SB1fV7sPeWQ8e2hbJSr6f1lRVHZrJLu2Hd/e3Z3SbW53dfXZ337679x8+My/O5CTB\nly3pOtf1XT98zrRH5LqflckCrO9MziPygCSpqrtkcqLSLy3pM+86v5jkF4bhByb5zDJ95r2+bz/8\nvVGS/57JyVOXWrflOGv7rKoOmOp2RK67PZSs47Jcrs5Fe4/P2I68oKruPPUcDs/yy3JzkqNqcsXV\nOyY5IMnHV7vGbdR5enf/yNSy/HZ333ledc7aJp/6B1Flcgjacp+X81zfW9ftkZmcrPo/Zky+LjXO\n+N3wmEzCsCOHbrO2hdZlXa/wt80s8/g8X7pNvqjfjcvWNdSw69D+Y5nsEXjh9ITD9vlKXiMr02t8\n1vWNdsvkKiT/lsl/mP5oaHteJhvUyeTkWH+TyYnPPp7kx+dQ41sy2a3+u5l8uT45yVMyOaYymVzl\n4BXDczg7M65KtMY13i+TQ1HOSnLGcHvoItWZ5O5JPjXUeE6S/7mI63uq3vvnB1daWKgaM7lKwpnD\n7dyp987CrO+hhoOSnDas83dncgWKRavxlpn8J+G2U20LVeNQx59ksmF1TiZXoLnpvF+XMz4bb5fk\nlEx+eP5Dkj2Gvgcn+aupaZ801LolyRPXucZHDMNXJ7k8yfuHvndIctLUtNf5flrnOrdkck6BrZ/p\nr5pnncvVuGT8hRmuRLNg6/uNw3v3rEw2OPde0PV9kyRvGt7jn0zywEVb35lsa5yeyXfPqUl+agHX\n9zOG5fNvSV6YpOa8HGdtn71jWNdnJfnbTE78Pc9luWydS/rM9T2eZbYjM/nn/UeH9/g5mewNdpuh\n/+FJnjc1/R8N6/uCDFdDXK86l+nzzanhda9zVo1JPji1LN+UH1zFbSHW99S4f8xkj6vp/nN570w9\n3v3zg98NP57JttiWTLbNtl5pbC6vyRk1Pj2Tz81rMvnnwV/Nezlm+W3yuX83Zse+sx+XyW+zM4b2\nh0/N56Qkd9jWa+T63LZ+yQEAAADA9eIQOQAAAABGETABAAAAMIqACQAAAIBRBEwAAAAAjCJgAgAA\nAGAUARMAwAxVtX9V/cbU/ftX1Xt3YPr7V9XPrE11a6+qHl5VB867DgBg8QmYAABm2z/Jb2yv0zbc\nP8myAVNV7Tpivuvl4UkETADAdgmYAICdwrC30aer6vVV9W9V9eaq+sWq+mhVfaaqDhn67VFV766q\ns6rqY1V196H9F6rqjOH2qaq6dZIXJvm5oe1ZU491o2Gem6bub9l6f2s9SZ6S5FnD9D831Paqqjo1\nyYuq6rlV9XtT05wzTJeqemxVfXyY9v9U1S7LPOd7V9W/VNWZQ99bV9XNqup1VXX28DweMPR9QlW9\nfGra91bV/Yfhb1bVC4b5fKyq9hr2vDo8yYuHGu5UVU+vqvOGZXfiaqw3AGDnIGACAHYmd07y50nu\nNtx+I8n9kvxekj8c+vxJkk91992HtjcM7b+X5GndfVCSn0vy70meneQj3X1Qdx+39UG6+3tJ3pTk\nMUPTLyY5s7uvnOpzYZJXJTlumP4jw6h9k/xMd//XWU+iqn4iya8n+dmhnmunHmtrn5skeWuSZ3T3\nPYYa/j3J0yYP3/85yaOTnFBVN9vOcrtlko8N8/mnJL/V3f+SZHOS3x/q/+ywPO45LLunbGeeAMAN\niIAJANiZfL67zx4CoHOTnNLdneTsTA53SyaB0xuTpLs/mOR2VXWbJB9N8pKqenqS3br7mu081vFJ\nHj8MPynJ61ZY499097Xb6fOgJD+V5BNVdcZw/8eX9Llrkku7+xPDc/n6UPP9Mgm/0t2fTvKFJHfZ\nzuN9J8nWc0udnh8sq6XOSvLmqnpsku0tHwDgBkTABADsTK6eGv7e1P3vJdnmOY+6+4VJfjPJzZN8\ntKrutp3+FyW5vKoemOSQJO9bYY3fmhq+Jj+8PbZ1T6NKcsKw59BB3X3X7n7uCuc/y6zHSpLvDkFc\nMtlbatay+uUkr0hyr0zCr41wHikAYB0ImACAG5qPZDjcbDgH0Ze6++tVdadh76c/TfKJTA6x+0aS\nW29jXn+Vyd5Cs/ZK2t70F2YS1qSq7pXkjkP7KUmOrKrbD+P2qKofWzLtBUn2rqp7D31uPQQ+08/v\nLkl+dOh7YZKDhvNF7ZdJKLY936+/qm6UZL/u/lCSP0hy2yS3WsE8AIAbAAETAHBD89wkP1VVZ2Vy\nEu+jh/ZnDifZPivJdzPZI+msJNcOJ79+1jLz2pxJyDLr8Li/TfKIrSf5Xmb8O5LsUVXnJjk2yb8l\nSXefl+S/J/nAUM/JSfaenrC7v5PJeZpeVlVnDn1uluQvk9yoqs7O5BxNT+juqzM5BPDzSc5L8tIk\nn9zGMtrqxCS/X1WfSnJAkjcN8/1Ukpd291dXMA8A4AagfrA3NAAAO6KqDs7kJN7LhUcAADcYjpsH\nALgequrZSZ6aJVd3AwC4IbIHEwAAAACjOAcTAAAAAKMImAAAAAAYRcAEAAAAwCgCJgAAAABGETAB\nAAAAMIqACQAAAIBR/h/h7zNTKT8PhAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MfVCH7Uxcwds",
        "colab_type": "code",
        "outputId": "918208d7-f7f8-4d38-f5b5-7e53fad55dd4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        }
      },
      "source": [
        "# pants on fire counts \n",
        "plt.figure(figsize=(20, 6))\n",
        "sns.countplot(train[\"pants on fire counts\"])\n",
        "plt.title('pants on fire counts')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABJgAAAGDCAYAAABnfapPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1440x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N689Ep8vcwdu",
        "colab_type": "code",
        "outputId": "148a550f-0bd4-4bee-97d1-7906aa0f2da1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 993
        }
      },
      "source": [
        "# Pairplot for cross visualisation of continuous variables\n",
        "plt.figure(figsize=(30,30))\n",
        "sns.pairplot(train, diag_kind='kde');"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/nonparametric/kde.py:447: RuntimeWarning: invalid value encountered in greater\n",
            "  X = X[np.logical_and(X > clip[0], X < clip[1])] # won't work for two columns.\n",
            "/usr/local/lib/python3.6/dist-packages/statsmodels/nonparametric/kde.py:447: RuntimeWarning: invalid value encountered in less\n",
            "  X = X[np.logical_and(X > clip[0], X < clip[1])] # won't work for two columns.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2160x2160 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3QAAAN4CAYAAACLUqrAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngiDopfwfI1ooP/AWYXIAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 900x900 with 30 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fAzfKzDEcwdw",
        "colab_type": "text"
      },
      "source": [
        "- Spread of Target Variable"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jeq2GD2icwdx",
        "colab_type": "code",
        "outputId": "d60e2e8b-8a42-46f8-bdb3-a989f2c3e49e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 464
        }
      },
      "source": [
        "cnt_srs = train['Label'].value_counts()\n",
        "plt.figure(figsize=(18,7))\n",
        "sns.barplot(cnt_srs.index, cnt_srs.values, alpha=0.8)\n",
        "plt.ylabel('Frequency', fontsize=12)\n",
        "plt.xlabel('Label', fontsize=12)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABCwAAAGsCAYAAAAWmjbvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1296x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rkdFqd7Kcwd1",
        "colab_type": "text"
      },
      "source": [
        "# Data Preprocessing"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ht1nMlsicwd1",
        "colab_type": "text"
      },
      "source": [
        "- Checking Missing Values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rskkXG2Icwd2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def missing_value(df):\n",
        "    total = df.isnull().sum().sort_values(ascending = False)\n",
        "    percent = (df.isnull().sum()/df.isnull().count()*100).sort_values(ascending=False)\n",
        "    missing_df = pd.concat([total, percent], axis=1, keys = ['Total', 'Percent'])\n",
        "    return missing_df"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eqzhYGOKcwd6",
        "colab_type": "code",
        "outputId": "3baf5307-4cda-4af0-a0ce-406360ffda45",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        }
      },
      "source": [
        "x_train = missing_value(train)\n",
        "x_train"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Total</th>\n",
              "      <th>Percent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Speaker's job title</th>\n",
              "      <td>3242</td>\n",
              "      <td>28.137476</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>state info</th>\n",
              "      <td>2487</td>\n",
              "      <td>21.584794</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "      <td>114</td>\n",
              "      <td>0.989412</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>pants on fire counts</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mostly true counts</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>half true counts</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>false counts</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>barely true counts</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>party affiliation</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Subject(s)</th>\n",
              "      <td>2</td>\n",
              "      <td>0.017358</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Statement</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Label</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ID</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    Total    Percent\n",
              "Speaker's job title                                  3242  28.137476\n",
              "state info                                           2487  21.584794\n",
              "the context (venue/location of speech or statem...    114   0.989412\n",
              "pants on fire counts                                    2   0.017358\n",
              "mostly true counts                                      2   0.017358\n",
              "half true counts                                        2   0.017358\n",
              "false counts                                            2   0.017358\n",
              "barely true counts                                      2   0.017358\n",
              "party affiliation                                       2   0.017358\n",
              "Speaker                                                 2   0.017358\n",
              "Subject(s)                                              2   0.017358\n",
              "Statement                                               0   0.000000\n",
              "Label                                                   0   0.000000\n",
              "ID                                                      0   0.000000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HQUC9xpDcwd8",
        "colab_type": "code",
        "outputId": "3317f909-901e-46f8-9086-695b4b8eb6eb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 437
        }
      },
      "source": [
        "x_test = missing_value(test)\n",
        "x_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Total</th>\n",
              "      <th>Percent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Speaker's job title</th>\n",
              "      <td>325</td>\n",
              "      <td>25.671406</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>state info</th>\n",
              "      <td>262</td>\n",
              "      <td>20.695103</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "      <td>17</td>\n",
              "      <td>1.342812</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>pants on fire counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mostly true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>half true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>false counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>barely true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>party affiliation</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Subject(s)</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Statement</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ID</th>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    Total    Percent\n",
              "Speaker's job title                                   325  25.671406\n",
              "state info                                            262  20.695103\n",
              "the context (venue/location of speech or statem...     17   1.342812\n",
              "pants on fire counts                                    0   0.000000\n",
              "mostly true counts                                      0   0.000000\n",
              "half true counts                                        0   0.000000\n",
              "false counts                                            0   0.000000\n",
              "barely true counts                                      0   0.000000\n",
              "party affiliation                                       0   0.000000\n",
              "Speaker                                                 0   0.000000\n",
              "Subject(s)                                              0   0.000000\n",
              "Statement                                               0   0.000000\n",
              "ID                                                      0   0.000000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pcF_ZxIucweA",
        "colab_type": "text"
      },
      "source": [
        "We need to treat the missing values here. It will reduce the model accuracy otherwise."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AVOfNIj5cweA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Filling Missing Values with Unknown Variable\n",
        "values = {\"Speaker's job title\":'Unknown'}\n",
        "train = train.fillna(value=values)\n",
        "test = test.fillna(value=values)\n",
        "#valid = valid.fillna(value=values)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fc0tLBs4cweC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Filling Missing Values with Unknown Variable\n",
        "values = {\"state info\":'Unknown1'}\n",
        "train = train.fillna(value=values)\n",
        "test = test.fillna(value=values)\n",
        "#valid = valid.fillna(value=values)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M4JLvYzlcweE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train['the context (venue/location of speech or statement)'].fillna(train['the context (venue/location of speech or statement)'].mode()[0], inplace=True)\n",
        "test['the context (venue/location of speech or statement)'].fillna(test['the context (venue/location of speech or statement)'].mode()[0], inplace=True)\n",
        "#valid['the context (venue/location of speech or statement)'].fillna(valid['the context (venue/location of speech or statement)'].mode()[0], inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JkLuKWG4cweG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train['pants on fire counts'].fillna(train['pants on fire counts'].median(), inplace=True)\n",
        "train['mostly true counts'].fillna(train['mostly true counts'].median(), inplace=True)\n",
        "train['half true counts'].fillna(train['half true counts'].median(), inplace=True)\n",
        "train['false counts'].fillna(train['false counts'].median(), inplace=True)\n",
        "train['barely true counts'].fillna(train['barely true counts'].median(), inplace=True)\n",
        "train['party affiliation'].fillna(train['party affiliation'].mode()[0], inplace=True)\n",
        "train['Speaker'].fillna(train['Speaker'].mode()[0], inplace=True)\n",
        "train['Subject(s)'].fillna(train['Subject(s)'].mode()[0], inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P7jKiqM-cweJ",
        "colab_type": "code",
        "outputId": "23c7c3ed-493e-45eb-dc22-c781645e27f4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        }
      },
      "source": [
        "# Checking again for missing values\n",
        "y_train = missing_value(train)\n",
        "y_train"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Total</th>\n",
              "      <th>Percent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>pants on fire counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mostly true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>half true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>false counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>barely true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>party affiliation</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>state info</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker's job title</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Subject(s)</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Statement</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Label</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ID</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    Total  Percent\n",
              "the context (venue/location of speech or statem...      0      0.0\n",
              "pants on fire counts                                    0      0.0\n",
              "mostly true counts                                      0      0.0\n",
              "half true counts                                        0      0.0\n",
              "false counts                                            0      0.0\n",
              "barely true counts                                      0      0.0\n",
              "party affiliation                                       0      0.0\n",
              "state info                                              0      0.0\n",
              "Speaker's job title                                     0      0.0\n",
              "Speaker                                                 0      0.0\n",
              "Subject(s)                                              0      0.0\n",
              "Statement                                               0      0.0\n",
              "Label                                                   0      0.0\n",
              "ID                                                      0      0.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CkY-oPX-cweN",
        "colab_type": "code",
        "outputId": "6c02e3e3-2ca8-45e4-93e9-ba32d0302415",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 437
        }
      },
      "source": [
        "# Checking again for missing values\n",
        "y_test = missing_value(test)\n",
        "y_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Total</th>\n",
              "      <th>Percent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>pants on fire counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mostly true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>half true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>false counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>barely true counts</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>party affiliation</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>state info</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker's job title</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Speaker</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Subject(s)</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Statement</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ID</th>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    Total  Percent\n",
              "the context (venue/location of speech or statem...      0      0.0\n",
              "pants on fire counts                                    0      0.0\n",
              "mostly true counts                                      0      0.0\n",
              "half true counts                                        0      0.0\n",
              "false counts                                            0      0.0\n",
              "barely true counts                                      0      0.0\n",
              "party affiliation                                       0      0.0\n",
              "state info                                              0      0.0\n",
              "Speaker's job title                                     0      0.0\n",
              "Speaker                                                 0      0.0\n",
              "Subject(s)                                              0      0.0\n",
              "Statement                                               0      0.0\n",
              "ID                                                      0      0.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yAqU0GWwcweP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Checking again for missing values\n",
        "#y_valid = missing_value(valid)\n",
        "#y_valid"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o5ShvHOncweR",
        "colab_type": "text"
      },
      "source": [
        "- Targeting Statement Column"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZoFKgtttk1BQ",
        "colab_type": "code",
        "outputId": "3555abb7-bfb3-49dd-dfd9-b1e88b8c6593",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(11522, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J2XsnEdKcweS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#removing non-letter symbols and converting text in 'Statement' to lowercase \n",
        "train['Statement'] = train['Statement'].apply(lambda x: \"\".join(re.sub(r\"[^A-Za-z\\s]\", '',str(x))))\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \"\".join(re.sub(r\"[^A-Za-z\\s]\",'',str(x))))\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \"\".join(re.sub(r\"[^A-Za-z\\s]\",'',str(x))))\n",
        "\n",
        "# lower casing the Text\n",
        "train['Statement'] = train['Statement'].apply(lambda x: \" \".join(x.lower() for x in x.split()))\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \" \".join(x.lower() for x in x.split()))\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \" \".join(x.lower() for x in x.split()))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BKgduhFccweW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Removing punctuations\n",
        "#adding characters list which needs to remove that is PUNCTUATION\n",
        "punc = ['.', ',', '\"', \"'\", '?', '!', ':', ';', '(', ')', '[', ']', '{', '}',\"%\",'/','<','>','br','�','^','XX','XXXX','xxxx','xx']\n",
        "#removing extra characters\n",
        "train['Statement'] = train['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in punc))\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in punc))\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in punc))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JzRXcvozcwea",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import nltk"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t3f0GKwhiSmB",
        "colab_type": "code",
        "outputId": "d06dc01e-3139-4d07-b990-d8db41ac2c82",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        }
      },
      "source": [
        "nltk.download('stopwords')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UCouVqlUcweY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#removal of stopwords\n",
        "from nltk.corpus import stopwords\n",
        "stop = stopwords.words('english')\n",
        "train['Statement'] = train['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in stop))\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in stop))\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in stop))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KBcbIVT8iWCZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#common words removal\n",
        "freq_train = pd.Series(' '.join(train['Statement']).split()).value_counts()[:10]\n",
        "freq_test = pd.Series(' '.join(test['Statement']).split()).value_counts()[:10]\n",
        "#freq_valid = pd.Series(' '.join(valid['Statement']).split()).value_counts()[:10]\n",
        "\n",
        "freq_train = list(freq_train.index)\n",
        "train['Statement'] = train['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_train))\n",
        "freq_test = list(freq_test.index)\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_test))\n",
        "#freq_valid = list(freq_valid.index)\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_valid))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5GyeiEs9i4nL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#rare words removal\n",
        "freq_train_rare = pd.Series(' '.join(train['Statement']).split()).value_counts()[-10:]\n",
        "freq_train_rare = list(freq_train_rare.index)\n",
        "train['Statement'] = train['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_train_rare))\n",
        "\n",
        "freq_test_rare = pd.Series(' '.join(test['Statement']).split()).value_counts()[-10:]\n",
        "freq_test_rare = list(freq_test_rare.index)\n",
        "test['Statement'] = test['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_test_rare))\n",
        "\n",
        "#freq_valid_rare = pd.Series(' '.join(valid['Statement']).split()).value_counts()[-10:]\n",
        "#freq_valid_rare = list(freq_valid_rare.index)\n",
        "#valid['Statement'] = valid['Statement'].apply(lambda x: \" \".join(x for x in x.split() if x not in freq_valid_rare))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KkX7-FbNjRgu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#STEMMING\n",
        "st = PorterStemmer()\n",
        "train['Statement']=train['Statement'].apply(lambda x: \" \".join([st.stem(w) for w in x.split()]))\n",
        "test['Statement']=test['Statement'].apply(lambda x: \" \".join([st.stem(w) for w in x.split()]))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1lYoMqrFrU2N",
        "colab_type": "code",
        "outputId": "b2543e4e-c650-4cd4-ada1-5a71a4155ea9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 584
        }
      },
      "source": [
        "train.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10540.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>declin coal start start natur ga took start be...</td>\n",
              "      <td>energy,history,job-accomplishments</td>\n",
              "      <td>scott-surovell</td>\n",
              "      <td>State delegate</td>\n",
              "      <td>Virginia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>a floor speech.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>324.json</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>hillari clinton agre john mccain vote give geo...</td>\n",
              "      <td>foreign-policy</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>160.0</td>\n",
              "      <td>163.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>Denver</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1123.json</td>\n",
              "      <td>false</td>\n",
              "      <td>care reform legisl like mandat free sex chang ...</td>\n",
              "      <td>health-care</td>\n",
              "      <td>blog-posting</td>\n",
              "      <td>Unknown</td>\n",
              "      <td>Unknown1</td>\n",
              "      <td>none</td>\n",
              "      <td>7.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>44.0</td>\n",
              "      <td>a news release</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9028.json</td>\n",
              "      <td>half-true</td>\n",
              "      <td>econom turnaround start end term</td>\n",
              "      <td>economy,jobs</td>\n",
              "      <td>charlie-crist</td>\n",
              "      <td>Unknown</td>\n",
              "      <td>Florida</td>\n",
              "      <td>democrat</td>\n",
              "      <td>15.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>an interview on CNN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>12465.json</td>\n",
              "      <td>true</td>\n",
              "      <td>chicago bear start quarterback last total numb...</td>\n",
              "      <td>education</td>\n",
              "      <td>robin-vos</td>\n",
              "      <td>Wisconsin Assembly speaker</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>republican</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a an online opinion-piece</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ID  ... the context (venue/location of speech or statement)\n",
              "0  10540.json  ...                                    a floor speech. \n",
              "1    324.json  ...                                             Denver \n",
              "2   1123.json  ...                                     a news release \n",
              "3   9028.json  ...                                an interview on CNN \n",
              "4  12465.json  ...                          a an online opinion-piece \n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IY96my6R7hE5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Removing .json extension\n",
        "train['ID'] = train['ID'].str.rstrip('.json')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MW_HRjfK7ppL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test['ID'] = test['ID'].str.rstrip('.json')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qhNZnbUwtQj2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#ID = train.iloc[:,[1]]\n",
        "#train=train.drop(['ID'], axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u6KYvzg7FnrS",
        "colab_type": "code",
        "outputId": "ac315dc0-91d6-40f9-ea38-2ae09ad956b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 212
        }
      },
      "source": [
        "train.columns, test.columns"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(Index(['ID', 'Label', 'Statement', 'Subject(s)', 'Speaker',\n",
              "        'Speaker's job title', 'state info', 'party affiliation',\n",
              "        'barely true counts', 'false counts', 'half true counts',\n",
              "        'mostly true counts', 'pants on fire counts',\n",
              "        'the context (venue/location of speech or statement)'],\n",
              "       dtype='object'),\n",
              " Index(['ID', 'Statement', 'Subject(s)', 'Speaker', 'Speaker's job title',\n",
              "        'state info', 'party affiliation', 'barely true counts', 'false counts',\n",
              "        'half true counts', 'mostly true counts', 'pants on fire counts',\n",
              "        'the context (venue/location of speech or statement)'],\n",
              "       dtype='object'))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5476GMxtbvYE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = train.loc[:,['Speaker',\"Speaker's job title\",'state info','party affiliation','the context (venue/location of speech or statement)']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wZm2kw71bvnX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X1 = test.loc[:,['Speaker',\"Speaker's job title\",'state info','party affiliation','the context (venue/location of speech or statement)']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h5KCF31Xbvkc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X3 = pd.get_dummies(X1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_vm1ISrfaxKo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X2 = pd.get_dummies(X)   # gives (1451, 221)  \n",
        "#test = pd.get_dummies(test)    # gives (1459, 206)  \n",
        "x, x1 = X2.align(X3, join='inner', axis=1)  # inner join"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UrEjX6uReYx8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x, x1 = X2.align(X3, join='inner', axis=1)  # inner join"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tW5Jf18Hb5ZS",
        "colab_type": "code",
        "outputId": "01dfe9e7-0799-4e9b-8162-755a2d2f8fba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x.shape, x1.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((11522, 1155), (1266, 1155))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xwMtlbWEs_2A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#x = x_data(train)  #train set"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "daLOLynD3JXp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#x1 = x_data(test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FCowIICLEXUD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#x.shape, x1.shape"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lB8zG5b6y-VB",
        "colab_type": "code",
        "outputId": "01c7d744-a561-4331-ff72-ad0f851ba871",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 584
        }
      },
      "source": [
        "train.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ID</th>\n",
              "      <th>Label</th>\n",
              "      <th>Statement</th>\n",
              "      <th>Subject(s)</th>\n",
              "      <th>Speaker</th>\n",
              "      <th>Speaker's job title</th>\n",
              "      <th>state info</th>\n",
              "      <th>party affiliation</th>\n",
              "      <th>barely true counts</th>\n",
              "      <th>false counts</th>\n",
              "      <th>half true counts</th>\n",
              "      <th>mostly true counts</th>\n",
              "      <th>pants on fire counts</th>\n",
              "      <th>the context (venue/location of speech or statement)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10540</td>\n",
              "      <td>half-true</td>\n",
              "      <td>declin coal start start natur ga took start be...</td>\n",
              "      <td>energy,history,job-accomplishments</td>\n",
              "      <td>scott-surovell</td>\n",
              "      <td>State delegate</td>\n",
              "      <td>Virginia</td>\n",
              "      <td>democrat</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>a floor speech.</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>324</td>\n",
              "      <td>mostly-true</td>\n",
              "      <td>hillari clinton agre john mccain vote give geo...</td>\n",
              "      <td>foreign-policy</td>\n",
              "      <td>barack-obama</td>\n",
              "      <td>President</td>\n",
              "      <td>Illinois</td>\n",
              "      <td>democrat</td>\n",
              "      <td>70.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>160.0</td>\n",
              "      <td>163.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>Denver</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1123</td>\n",
              "      <td>false</td>\n",
              "      <td>care reform legisl like mandat free sex chang ...</td>\n",
              "      <td>health-care</td>\n",
              "      <td>blog-posting</td>\n",
              "      <td>Unknown</td>\n",
              "      <td>Unknown1</td>\n",
              "      <td>none</td>\n",
              "      <td>7.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>44.0</td>\n",
              "      <td>a news release</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9028</td>\n",
              "      <td>half-true</td>\n",
              "      <td>econom turnaround start end term</td>\n",
              "      <td>economy,jobs</td>\n",
              "      <td>charlie-crist</td>\n",
              "      <td>Unknown</td>\n",
              "      <td>Florida</td>\n",
              "      <td>democrat</td>\n",
              "      <td>15.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>an interview on CNN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>12465</td>\n",
              "      <td>true</td>\n",
              "      <td>chicago bear start quarterback last total numb...</td>\n",
              "      <td>education</td>\n",
              "      <td>robin-vos</td>\n",
              "      <td>Wisconsin Assembly speaker</td>\n",
              "      <td>Wisconsin</td>\n",
              "      <td>republican</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>a an online opinion-piece</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      ID  ... the context (venue/location of speech or statement)\n",
              "0  10540  ...                                    a floor speech. \n",
              "1    324  ...                                             Denver \n",
              "2   1123  ...                                     a news release \n",
              "3   9028  ...                                an interview on CNN \n",
              "4  12465  ...                          a an online opinion-piece \n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RwdyWK6c17rj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "rest_df_train = train.loc[:,['ID','barely true counts','false counts',\"half true counts\",'mostly true counts','pants on fire counts']]\n",
        "rest_df_test = test.loc[:,['ID','barely true counts','false counts',\"half true counts\",'mostly true counts','pants on fire counts']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "833ElA6f2qzC",
        "colab_type": "code",
        "outputId": "7d281f78-d14e-4d9c-a498-bdcbb0e05971",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        }
      },
      "source": [
        "rest_df_train.head(), rest_df_test.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(      ID  barely true counts  ...  mostly true counts  pants on fire counts\n",
              " 0  10540                 0.0  ...                 1.0                   0.0\n",
              " 1    324                70.0  ...               163.0                   9.0\n",
              " 2   1123                 7.0  ...                 5.0                  44.0\n",
              " 3   9028                15.0  ...                19.0                   2.0\n",
              " 4  12465                 0.0  ...                 5.0                   1.0\n",
              " \n",
              " [5 rows x 6 columns],\n",
              "       ID  barely true counts  ...  mostly true counts  pants on fire counts\n",
              " 0  11685                   2  ...                   0                     0\n",
              " 1  11096                  63  ...                  37                    61\n",
              " 2   5209                   1  ...                   1                     1\n",
              " 3   9524                   5  ...                   2                     7\n",
              " 4   5962                   1  ...                   1                     0\n",
              " \n",
              " [5 rows x 6 columns])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "70e3EqojoN0v",
        "colab_type": "code",
        "outputId": "1393d615-4c2c-43b1-c9a4-fe988d8a9e6e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 107
        }
      },
      "source": [
        "#NOw We will convert the target variable into LabelEncoder\n",
        "y_train = train.loc[:,['Label']]\n",
        "labelencoder1 = LabelEncoder()\n",
        "labelencoder1.fit(y_train.values)\n",
        "y_train=labelencoder1.transform(y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/label.py:219: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/label.py:252: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rfZc3BfpoR25",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#pd.DataFrame(y_train).nunique()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kgxvggtCsLm4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Converting the Text into list\n",
        "stat_train = train['Statement'].tolist()\n",
        "stat_test = test['Statement'].tolist()\n",
        "\n",
        "# Adding train and test data\n",
        "lis_stat = stat_train + stat_test "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z2__ULMPsLkd",
        "colab_type": "code",
        "outputId": "206fa2ec-f407-4f04-d713-b2d071369d3f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 90
        }
      },
      "source": [
        "#Fitting TFIDF on summary\n",
        "tfv_sum = TfidfVectorizer(max_features=5000, lowercase=True, analyzer='word',\n",
        " stop_words= 'english',ngram_range=(1,1))\n",
        "# Fitting TF-IDF to both training and test sets (semi-supervised learning)\n",
        "tfv_sum.fit(lis_stat)\n",
        "print(tfv_sum.vocabulary_)\n",
        "print(tfv_sum.idf_)\n",
        "xtrain_tfv_sum =  tfv_sum.transform(stat_train).toarray()\n",
        "xtest_tfv_sum = tfv_sum.transform(stat_test).toarray()\n",
        "type(xtrain_tfv_sum)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'declin': 1111, 'coal': 840, 'start': 4254, 'natur': 2927, 'ga': 1805, 'took': 4556, 'begin': 393, 'georg': 1846, 'bush': 596, 'administr': 64, 'hillari': 2067, 'clinton': 826, 'agre': 99, 'john': 2360, 'mccain': 2708, 'vote': 4812, 'benefit': 410, 'doubt': 1313, 'iran': 2298, 'care': 650, 'reform': 3683, 'legisl': 2500, 'like': 2542, 'mandat': 2648, 'free': 1769, 'sex': 4027, 'chang': 728, 'surgeri': 4383, 'econom': 1383, 'turnaround': 4637, 'end': 1442, 'term': 4475, 'chicago': 756, 'bear': 380, 'quarterback': 3574, 'total': 4563, 'number': 3025, 'tenur': 4474, 'uw': 4748, 'faculti': 1575, 'decad': 1105, 'jim': 2353, 'live': 2561, 'district': 1269, 'repres': 3743, 'im': 2171, 'person': 3294, 'stage': 4241, 'work': 4947, 'activ': 51, 'pass': 3217, 'russ': 3884, 'feingold': 1625, 'toughest': 4568, 'ethic': 1497, 'sinc': 4088, 'howev': 2129, 'million': 2797, 'oregon': 3108, 'lotteri': 2597, 'fund': 1795, 'port': 3401, 'newport': 2962, 'eventu': 1504, 'land': 2460, 'new': 2958, 'noaa': 2984, 'marin': 2671, 'oper': 3093, 'gop': 1892, 'primari': 3474, 'oppon': 3096, 'glenn': 1873, 'grothman': 1925, 'joe': 2359, 'cast': 674, 'compromis': 903, 'cost': 1005, 'higher': 2054, 'electr': 1406, 'time': 4538, 'histori': 2074, 'share': 4034, 'nation': 2923, 'popular': 3395, 'margin': 2667, 'smaller': 4127, 'latino': 2475, 'nearli': 2937, 'american': 165, 'slip': 4119, 'middl': 2780, 'class': 805, 'poverti': 3421, 'mitt': 2830, 'romney': 3850, 'governor': 1900, 'massachusett': 2688, 'didnt': 1205, 'slow': 4121, 'rate': 3619, 'growth': 1930, 'govern': 1898, 'actual': 54, 'cut': 1062, 'economi': 1384, 'billion': 437, 'shutdown': 4068, 'afford': 84, 'act': 49, 'alreadi': 149, 'sens': 4003, 'waiv': 4821, 'otherwis': 3119, 'suspend': 4399, 'elect': 1404, 'novemb': 3019, 'chose': 771, 'young': 4986, 'lowincom': 2610, 'worker': 4948, 'oppos': 3098, 'requir': 3752, 'buy': 603, 'motorcycl': 2884, 'said': 3898, 'provis': 3544, 'quot': 3580, 'disgrac': 1246, 'rep': 3732, 'ron': 3853, 'kind': 2424, 'dwi': 1362, 'fellow': 1627, 'democrat': 1153, 'went': 4881, 'spend': 4207, 'spree': 4226, 'credit': 1034, 'card': 649, 'max': 2703, 'water': 4850, 'philippin': 3314, 'rais': 3600, 'world': 4953, 'bank': 343, 'becam': 384, 'partial': 3203, 'owner': 3165, 'left': 2497, 'puerto': 3554, 'rico': 3808, 'women': 4942, 'men': 2747, 'make': 2638, 'adjust': 61, 'inflat': 2228, 'kitzhab': 2430, 'unit': 4705, 'state': 4256, 'highest': 2055, 'corpor': 998, 'best': 420, 'auto': 308, 'industri': 2221, 'america': 164, 'scott': 3956, 'walker': 4825, 'favor': 1608, 'famili': 1588, 'children': 764, 'want': 4831, 'rid': 3809, 'plan': 3349, 'parenthood': 3197, 'dont': 1306, 'know': 2437, 'jonathan': 2368, 'hate': 2004, 'crime': 1036, 'muslim': 2908, 'mosqu': 2875, 'tripl': 4619, 'pari': 3198, 'san': 3908, 'bernardino': 417, 'rick': 3807, 'perri': 3290, 'lost': 2595, 'remain': 3715, 'texa': 4485, 'governorship': 1901, 'landslid': 2462, 'isi': 2307, 'support': 4372, 'tweet': 4642, 'shoot': 4052, 'began': 391, 'chattanooga': 742, 'youth': 4988, 'unemploy': 4691, 'minor': 2809, 'commun': 882, 'paul': 3236, 'ryan': 3889, 'endors': 1445, 'trump': 4626, 'look': 2586, 'right': 3817, 'constantli': 951, 'budget': 569, 'deficit': 1129, 'good': 1890, 'busi': 599, 'climat': 822, 'come': 868, 'sever': 4024, 'month': 2859, 'abl': 11, 'burn': 592, 'exist': 1530, 'plant': 3352, 'tax': 4441, 'today': 4545, 'pay': 3241, 'mark': 2672, 'pryor': 3547, 'feder': 1619, 'think': 4507, 'author': 306, 'regul': 3695, 'toilet': 4548, 'seat': 3971, 'austin': 303, 'citi': 795, 'basic': 368, 'doubl': 1311, 'size': 4106, 'everi': 1505, 'nuclear': 3023, 'test': 4482, 'conduct': 921, 'earth': 1371, 'th': 4490, 'perman': 3285, 'member': 2742, 'secur': 3986, 'council': 1009, 'case': 669, 'catastroph': 679, 'event': 1503, 'offic': 3061, 'center': 701, 'diseas': 1244, 'control': 975, 'prevent': 3468, 'ad': 55, 'trillion': 4616, 'debt': 1103, 'weve': 4886, 'suprem': 4376, 'court': 1019, 'justic': 2386, 'suggest': 4354, 'decid': 1108, 'base': 364, 'south': 4181, 'african': 89, 'law': 2479, 'marco': 2665, 'rubio': 3872, 'skip': 4109, 'defens': 1126, 'includ': 2200, 'arm': 243, 'fight': 1650, 'david': 1090, 'cicillin': 786, 'respons': 3770, 'loan': 2568, 'guarante': 1932, 'solyndra': 4161, 'fail': 1577, 'solar': 4151, 'compani': 886, 'ohio': 3066, 'item': 2317, 'later': 2471, 'sign': 4075, 'gov': 1897, 'kasich': 2395, 'seek': 3987, 'abort': 15, 'undergo': 4681, 'mandatori': 2649, 'probe': 3494, 'director': 1225, 'homeland': 2094, 'use': 4742, 'terrorist': 4481, 'attack': 286, 'instead': 2260, 'substitut': 4344, 'manmad': 2655, 'disast': 1233, 'africanamerican': 90, 'hispan': 2072, 'obamacar': 3038, 'provid': 3543, 'insur': 2266, 'nonu': 2999, 'resid': 3758, 'illeg': 2168, 'barack': 348, 'exactli': 1514, 'reduct': 3674, 'commiss': 875, 'idea': 2160, 'rule': 3875, 'engag': 1451, 'probabl': 3493, 'life': 2534, 'cosponsor': 1004, 'jail': 2324, 'flag': 1681, 'paperback': 3192, 'edit': 1387, 'book': 486, 'delet': 1142, 'line': 2549, 'individu': 2220, 'model': 2839, 'countri': 1015, 'maggi': 2627, 'hassan': 2002, 'tuition': 4632, 'terri': 4477, 'mcauliff': 2707, 'invent': 2287, 'scheme': 3943, 'rent': 3730, 'lincoln': 2547, 'creat': 1031, 'job': 2356, 'half': 1959, 'york': 4982, 'privat': 3487, 'sector': 3985, 'town': 4574, 'west': 4884, 'truck': 4623, 'georgia': 1847, 'earn': 1369, 'cent': 700, 'dollar': 1295, 'verifi': 4774, 'employe': 1434, 'inaccur': 2190, 'anywher': 204, 'way': 4856, 'excess': 1521, 'build': 574, 'wall': 4826, 'ago': 98, 'long': 2581, 'environment': 1475, 'impact': 2178, 'statement': 4259, 'sen': 3999, 'obama': 3037, 'campaign': 621, 'announc': 185, 'choos': 770, 'cabinet': 607, 'regist': 3691, 'turn': 4636, 'public': 3548, 'high': 2053, 'school': 3948, 'offer': 3060, 'republican': 3747, 'candid': 632, 'militari': 2791, 'recruit': 3663, 'harvard': 1997, 'elena': 1410, 'kagan': 2390, 'posit': 3410, 'unanim': 4669, 'wrong': 4969, 'teacher': 4449, 'mayor': 2706, 'sanctuari': 3910, 'transgend': 4590, 'chanc': 727, 'murder': 2901, 'bain': 334, 'capit': 641, 'help': 2036, 'earli': 1365, 'childhood': 762, 'learn': 2491, 'bright': 541, 'ladi': 2455, 'michel': 2775, 'prais': 3426, 'man': 2646, 'rush': 3881, 'social': 4144, 'media': 2728, 'post': 3413, 'phil': 3311, 'gingrey': 1862, 'gave': 1829, 'power': 3424, 'electron': 1407, 'spi': 4210, 'warrant': 4839, 'alex': 127, 'sink': 4094, 'funnel': 1799, 'quarter': 3573, 'nobid': 2985, 'contract': 969, 'popul': 3394, 'virginia': 4796, 'net': 2950, 'zero': 4994, 'creation': 1032, 'bob': 473, 'mcdonnel': 2714, 'home': 2093, 'valu': 4758, 'florida': 1696, 'collaps': 856, 'construct': 955, 'foreclosur': 1719, 'arizona': 239, 'immigr': 2175, 'polic': 3377, 'stop': 4287, 'anyon': 201, 'appear': 212, 'reason': 3637, 'suspici': 4400, 'undocu': 4689, 'war': 4832, 'ii': 2166, 'factor': 1572, 'drive': 1339, 'recess': 3648, 'consum': 958, 'nativ': 2925, 'begun': 394, 'save': 3925, 'nomine': 2989, 'confirm': 927, 'hear': 2022, 'homeless': 2095, 'veteran': 4780, 'charli': 737, 'crist': 1041, 'medicar': 2732, 'advantag': 71, 'program': 3508, 'unaccept': 4667, 'chines': 767, 'larger': 2468, 'wind': 4916, 'panel': 3186, 'manufactur': 2658, 'europ': 1500, 'togeth': 4547, 'forget': 1727, 'brown': 557, 'let': 2511, 'employ': 1433, 'deni': 1156, 'coverag': 1022, 'birth': 447, 'point': 3370, 'past': 3222, 'egypt': 1394, 'great': 1914, 'stride': 4308, 'polit': 3380, 'attent': 289, 'paid': 3176, 'week': 4872, 'chri': 773, 'christi': 775, 'benghazi': 411, 'signific': 4078, 'steve': 4276, 'southerland': 4183, 'import': 2184, 'nurs': 3029, 'crimin': 1038, 'transport': 4596, 'built': 575, 'prison': 3484, 'day': 1091, 'mass': 2687, 'incarcer': 2195, 'explos': 1548, 'nonviol': 3001, 'offend': 3058, 'virtual': 4798, 'noth': 3014, 'burden': 584, 'fastestgrow': 1602, 'increas': 2207, 'major': 2637, 'phoenix': 3317, 'kidnap': 2419, 'belong': 401, 'aflcio': 86, 'commission': 876, 'squat': 4232, 'poll': 3383, 'beat': 381, 'ronald': 3854, 'reagan': 3632, 'sent': 4004, 'troop': 4620, 'conflict': 929, 'circumst': 792, 'peacekeep': 3251, 'set': 4020, 'jolli': 2366, 'lowest': 2609, 'depress': 1169, 'averag': 313, 'student': 4321, 'pocket': 3369, 'univers': 4706, 'die': 1206, 'gunfir': 1944, 'effect': 1390, 'mean': 2718, 'face': 1566, 'modifi': 2843, 'mortgag': 2874, 'short': 4056, 'sale': 3902, 'lender': 2506, 'depart': 1163, 'agricultur': 101, 'websit': 4868, 'permit': 3287, 'carri': 661, 'load': 2567, 'hidden': 2051, 'gun': 1943, 'leav': 2493, 'hous': 2123, 'atlanta': 281, 'streetcar': 4300, 'expect': 1536, 'minimum': 2805, 'wage': 4819, 'hour': 2121, 'lift': 2538, 'secretari': 3982, 'kerri': 2413, 'senat': 4000, 'flew': 1687, 'met': 2761, 'communist': 883, 'dictat': 1204, 'daniel': 1078, 'accus': 40, 'terror': 4480, 'nato': 2926, 'open': 3091, 'divis': 1277, 'sure': 4379, 'larg': 2467, 'say': 3927, 'critic': 1043, 'alway': 153, 'prochoic': 3499, 'gina': 1861, 'raimondo': 3597, 'invest': 2288, 'fee': 1621, 'pension': 3269, 'portfolio': 3402, 'gone': 1888, 'annual': 186, 'grow': 1928, 'longer': 2582, 'lay': 2484, 'firefight': 1668, 'old': 3072, 'twothird': 4653, 'wisconsinit': 4928, 'receiv': 3646, 'check': 748, 'search': 3969, 'current': 1059, 'exempt': 1527, 'healthcar': 2017, 'premium': 3450, 'ir': 2297, 'investig': 2289, 'liber': 2520, 'group': 1927, 'progress': 3509, 'everybodi': 1507, 'promis': 3515, 'fastest': 1601, 'pace': 3170, 'meet': 2735, 'prime': 3476, 'minist': 2807, 'benjamin': 412, 'netanyahu': 2951, 'marri': 2676, 'love': 2605, 'peter': 3301, 'defazio': 1123, 'earl': 1364, 'blumenau': 468, 'socialist': 4145, 'openli': 3092, 'serv': 4014, 'congress': 934, 'rudi': 3873, 'giuliani': 1867, 'presid': 3460, 'britain': 543, 'european': 1501, 'union': 4703, 'ks': 2444, 'medicaid': 2731, 'second': 3975, 'duti': 1361, 'colleg': 859, 'columbu': 864, 'contribut': 973, 'cfo': 713, 'shut': 4067, 'scam': 3936, 'refund': 3685, 'victim': 4786, 'comput': 905, 'scienc': 3952, 'degre': 1135, 'plu': 3364, 'air': 107, 'forc': 1715, 'taxpay': 4443, 'footbal': 1712, 'leagu': 2489, 'hussein': 2147, 'accept': 29, 'islam': 2309, 'sharia': 4035, 'domest': 1299, 'disput': 1258, 'commit': 877, 'legal': 2499, 'purchas': 3559, 'neighborhood': 2948, 'close': 829, 'told': 4549, 'gay': 1830, 'organ': 3111, 'stronger': 4314, 'advoc': 77, 'special': 4199, 'ted': 4457, 'kennedi': 2407, 'speaker': 4197, 'consist': 948, 'halfway': 1961, 'goal': 1878, 'seven': 4022, 'bennet': 413, 'guantanamo': 1931, 'bay': 377, 'bring': 542, 'colorado': 861, 'oregonian': 3109, 'amaz': 157, 'donat': 1303, 'syrian': 4419, 'iraqi': 2301, 'refuge': 3684, 'research': 3754, 'vast': 4764, 'arriv': 249, 'believ': 399, 'sourc': 4180, 'prosper': 3531, 'newt': 2966, 'gingrich': 1863, 'amnesti': 167, 'mr': 2892, 'caprio': 644, 'career': 651, 'politician': 3381, 'rhode': 3796, 'island': 2311, 'temporari': 4467, 'disabl': 1227, 'jersey': 2345, 'exceed': 1519, 'revenu': 3786, 'money': 2850, 'street': 4299, 'anybodi': 199, 'donald': 1301, 'ok': 3069, 'taken': 4426, 'lobbyist': 2570, 'expressli': 1553, 'ban': 342, 'racial': 3584, 'profil': 3506, 'beg': 390, 'send': 4001, 'coast': 843, 'guard': 1933, 'patrol': 3232, 'lake': 2458, 'usmexico': 4744, 'border': 493, 'va': 4750, 'wait': 4820, 'rob': 3835, 'cornil': 995, 'dan': 1074, 'branch': 520, 'tri': 4612, 'expand': 1534, 'improv': 2189, 'michigan': 2776, 'bottl': 500, 'deposit': 1168, 'profit': 3507, 'admit': 67, 'gener': 1838, 'goe': 1881, 'educ': 1388, 'fought': 1742, 'develop': 1194, 'drain': 1327, 'johnson': 2362, 'ponzi': 3387, 'type': 4656, 'polici': 3379, 'refer': 3677, 'north': 3006, 'carolina': 657, 'directli': 1224, 'relat': 3706, 'spent': 4208, 'car': 646, 'finland': 1665, 'windmil': 4918, 'china': 766, 'sarah': 3918, 'wasilla': 4844, 'highli': 2058, 'respect': 3767, 'librarian': 2525, 'refus': 3686, 'consid': 946, 'remov': 3721, 'librari': 2524, 'milwauke': 2801, 'approv': 225, 'reduc': 3673, 'maximum': 2704, 'forfeitur': 1725, 'marijuana': 2669, 'possess': 3411, 'differ': 1207, 'bicycl': 427, 'violat': 4793, 'wrote': 4972, 'letter': 2513, 'hillarycar': 2068, 'patent': 3226, 'trademark': 4580, 'complaint': 894, 'washington': 4843, 'redskin': 3672, 'propos': 3526, 'caus': 688, 'decreas': 1113, 'hundr': 2137, 'lose': 2593, 'mexican': 2768, 'citizen': 796, 'statist': 4265, 'veto': 4781, 'packag': 3173, 'illinoi': 2169, 'gubernatori': 1934, 'bradi': 515, 'prioriti': 3483, 'win': 4915, 'sponsor': 4219, 'shelter': 4044, 'dog': 1291, 'cat': 678, 'chamber': 723, 'worsen': 4958, 'action': 50, 'peak': 3252, 'trail': 4586, 'rest': 3771, 'consecut': 942, 'thoma': 4515, 'eric': 1486, 'duncan': 1359, 'came': 618, 'elev': 1411, 'temperatur': 4466, 'afghanistan': 85, 'foreign': 1720, 'price': 3471, 'webster': 4869, 'form': 1728, 'marriag': 2677, 'trap': 4598, 'abus': 25, 'relationship': 3707, 'judg': 2374, 'better': 422, 'lower': 2607, 'els': 1418, 'lot': 2596, 'casino': 672, 'bankrupt': 345, 'small': 4125, 'senior': 4002, 'kid': 2418, 'arrest': 248, 'leader': 2487, 'area': 234, 'rank': 3611, 'oil': 3068, 'rig': 3815, 'combin': 867, 'payday': 3244, 'poor': 3389, 'charg': 732, 'trust': 4627, 'sound': 4178, 'anyth': 202, 'function': 1794, 'voter': 4813, 'incumb': 2211, 'break': 527, 'pledg': 3362, 'strickland': 4305, 'incom': 2201, 'hike': 2065, 'defeat': 1124, 'soundli': 4179, 'tuesday': 4631, 'alert': 126, 'explod': 1545, 'greater': 1915, 'percentag': 3275, 'bar': 347, 'gen': 1836, 'petraeu': 3304, 'armi': 244, 'damag': 1073, 'method': 2763, 'stab': 4234, 'club': 835, 'enemi': 1448, 'order': 3104, 'minim': 2804, 'carbon': 647, 'output': 3136, 'singl': 4090, 'block': 460, 'costofliv': 1007, 'januari': 2330, 'male': 2642, 'babi': 321, 'mani': 2652, 'strike': 4309, 'syria': 4418, 'assad': 262, 'chemic': 751, 'weapon': 4863, 'cash': 671, 'clunker': 836, 'got': 1895, 'local': 2572, 'youv': 4990, 'gut': 1946, 'recogn': 3653, 'real': 3633, 'dewhurst': 1198, 'record': 3657, 'promot': 3516, 'beach': 378, 'mind': 2802, 'conclud': 913, 'shouldnt': 4060, 'kill': 2420, 'lead': 2486, 'pack': 3172, 'access': 30, 'hb': 2010, 'needl': 2943, 'core': 991, 'accord': 35, 'agenc': 94, 'away': 318, 'tom': 4552, 'kean': 2403, 'largest': 2469, 'biggest': 434, 'readi': 3631, 'command': 870, 'chief': 758, 'median': 2729, 'seen': 3988, 'barrett': 359, 'held': 2031, 'properti': 3523, 'unlik': 4711, 'assembl': 265, 'bipartisan': 444, 'josh': 2371, 'mandel': 2650, 'lie': 2530, 'leas': 2492, 'drill': 1337, 'hide': 2052, 'possibl': 3412, 'turnpik': 4640, 'surplu': 4386, 'project': 3512, 'nomin': 2988, 'trip': 4618, 'east': 1377, 'counti': 1013, 'settl': 4021, 'lawsuit': 2481, 'decis': 1109, 'sheriff': 4045, 'clark': 804, 'jr': 2372, 'untrain': 4723, 'correct': 999, 'deputi': 1171, 'fbi': 1610, 'jame': 2326, 'comey': 869, 'discuss': 1243, 'consider': 947, 'privatesector': 3488, 'biden': 430, 'gulf': 1942, 'surg': 4381, 'partit': 3209, 'iraq': 2300, 'limbaugh': 2544, 'clear': 814, 'hed': 2028, 'succeed': 4346, 'wasnt': 4845, 'perdu': 3278, 'abolish': 14, 'elimin': 1413, 'campu': 624, 'twice': 4648, 'graduat': 1906, 'theyv': 4505, 'imprison': 2188, 'grown': 1929, 'score': 3955, 'flat': 1682, 'affect': 80, 'thousand': 4518, 'parti': 3202, 'cornyn': 996, 'greg': 1921, 'abbott': 5, 'gdp': 1832, 'nobodi': 2986, 'mile': 2789, 'closer': 831, 'eastern': 1378, 'need': 2941, 'photo': 3320, 'id': 2158, 'twenti': 4644, 'portland': 3404, 'pearl': 3254, 'allow': 145, 'payrol': 3247, 'holiday': 2088, 'expir': 1542, 'anoth': 188, 'estim': 1495, 'nafta': 2912, 'trade': 4579, 'pennsylvania': 3268, 'numer': 3027, 'studi': 4323, 'shown': 4063, 'socal': 4142, 'righttowork': 3818, 'gross': 1924, 'product': 3501, 'histor': 2073, 'affair': 79, 'scandal': 3938, 'satisfi': 3921, 'driven': 1340, 'initi': 2237, 'heard': 2023, 'christma': 778, 'underwear': 4688, 'bomber': 481, 'isol': 2313, 'extremist': 1562, 'sherrod': 4046, 'stimulu': 4280, 'specif': 4201, 'claim': 802, 'rich': 3802, 'charleston': 736, 'given': 1869, 'background': 324, 'fast': 1599, 'okay': 3070, 'soldier': 4153, 'twelv': 4643, 'thrown': 4525, 'challeng': 722, 'reject': 3704, 'notion': 3017, 'unconstitut': 4675, 'warren': 4840, 'buffett': 573, 'lemieux': 2504, 'precinct': 3432, 'allen': 140, 'murphi': 2902, 'recount': 3659, 'kathleen': 2398, 'sebeliu': 3973, 'determin': 1190, 'yearold': 4979, 'patient': 3229, 'klein': 2433, 'repeatedli': 3739, 'enforc': 1450, 'data': 1083, 'treasur': 4603, 'saw': 3926, 'houston': 2125, 'drop': 1343, 'level': 2514, 'columbia': 862, 'barri': 360, 'richest': 3805, 'millionair': 2798, 'solut': 4158, 'crisi': 1040, 'vacat': 4753, 'hypocrisi': 2154, 'foundat': 1743, 'staff': 4239, 'femal': 1632, 'violenc': 4794, 'adopt': 68, 'fallen': 1586, 'question': 3575, 'miranda': 2813, 'warn': 4837, 'minut': 2810, 'pregnanc': 3443, 'unintend': 4702, 'borrow': 495, 'saudi': 3923, 'result': 3776, 'depend': 1164, 'welcom': 4877, 'complain': 893, 'gain': 1807, 'entir': 1467, 'previou': 3469, 'thing': 4506, 'head': 2011, 'brought': 554, 'dairi': 1070, 'workforc': 4949, 'nationwid': 2924, 'someth': 4164, 'portman': 3406, 'irrespons': 2305, 'dekalb': 1136, 'palin': 3181, 'rightw': 3819, 'pat': 3224, 'buchanan': 563, 'energi': 1449, 'jeann': 2336, 'shaheen': 4031, 'measur': 2724, 'dc': 1092, 'strictest': 4307, 'skyrocket': 4111, 'toughen': 4566, 'penalti': 3263, 'workplac': 4951, 'gender': 1837, 'discrimin': 1242, 'cook': 982, 'clean': 811, 'infrastructur': 2234, 'proper': 3521, 'collect': 858, 'predatori': 3438, 'target': 4436, 'decemb': 1106, 'period': 3283, 'difficult': 1208, 'hire': 2071, 'cbo': 692, 'congression': 935, 'thompson': 4516, 'cap': 637, 'fiscal': 1672, 'reserv': 3756, 'strateg': 4295, 'petroleum': 3305, 'fertil': 1637, 'uninsur': 4701, 'firearm': 1667, 'homicid': 2098, 'upward': 4734, 'ultrasound': 4665, 'harbor': 1983, 'youd': 4984, 'common': 879, 'pill': 3329, 'jason': 2333, 'kander': 2392, 'suspect': 4398, 'soil': 4150, 'shelbi': 4041, 'sorri': 4174, 'memphi': 2746, 'intellig': 2269, 'offici': 3062, 'seriou': 4013, 'threat': 4519, 'suburb': 4345, 'absolut': 23, 'juli': 2378, 'locat': 2573, 'miss': 2820, 'teenag': 4460, 'daughter': 1087, 'partner': 3210, 'blame': 456, 'mike': 2788, 'nanci': 2914, 'pelosi': 3260, 'straight': 4293, 'beneficiari': 409, 'account': 36, 'transfer': 4589, 'mechan': 2726, 'place': 3346, 'internet': 2276, 'defund': 1133, 'report': 3742, 'insist': 2251, 'fals': 1587, 'barackobama': 349, 'play': 3356, 'clip': 827, 'samesex': 3907, 'conni': 941, 'mack': 2621, 'iv': 2318, 'eisenhow': 1400, 'personnel': 3296, 'wisconsin': 4927, 'counterpart': 1012, 'airplan': 111, 'list': 2555, 'flow': 1698, 'wont': 4944, 'default': 1122, 'homeown': 2096, 'lend': 2505, 'practic': 3425, 'wipe': 4925, 'doddfrank': 1287, 'brand': 521, 'cup': 1054, 'atmospher': 283, 'condit': 918, 'push': 3564, 'pressur': 3464, 'approxim': 226, 'pound': 3419, 'squar': 4230, 'inch': 2198, 'unlimit': 4712, 'demand': 1151, 'moment': 2848, 'partialbirth': 3204, 'obtain': 3046, 'scholarship': 3947, 'attend': 288, 'star': 4251, 'hook': 2104, 'falcon': 1583, 'stadium': 4237, 'repaid': 3733, 'outsid': 3139, 'missouri': 2824, 'lone': 2580, 'refin': 3680, 'resolut': 3763, 'disapprov': 1231, 'condemn': 917, 'behavior': 396, 'younger': 4987, 'competitor': 891, 'recent': 3647, 'intern': 2275, 'near': 2935, 'academ': 26, 'achiev': 41, 'fulltim': 1792, 'doctor': 1284, 'theyr': 4504, 'sick': 4070, 'feed': 1622, 'owe': 3163, 'mistak': 2826, 'situat': 4099, 'arkansa': 241, 'highspe': 2061, 'train': 4587, 'fulli': 1791, 'extend': 1554, 'add': 57, 'ceil': 695, 'cover': 1021, 'despit': 1182, 'recordbreak': 3658, 'scrutini': 3962, 'conserv': 945, 'mahmoud': 2630, 'abba': 4, 'ali': 131, 'establish': 1493, 'vladimir': 4805, 'putin': 3565, 'board': 470, 'led': 2494, 'appointe': 219, 'jefferson': 2339, 'textbook': 4489, 'standard': 4248, 'figur': 1652, 'inspir': 2254, 'salari': 3901, 'perk': 3284, 'berni': 418, 'pull': 3555, 'truth': 4628, 'matter': 2701, 'physic': 3322, 'logist': 2576, 'wealth': 4860, 'penni': 3267, 'foster': 1741, 'ineffect': 2222, 'routin': 3868, 'cell': 697, 'phone': 3318, 'negoti': 2946, 'clackama': 801, 'renegoti': 3727, 'agreement': 100, 'trimet': 4617, 'portlandmilwauki': 3405, 'light': 2539, 'rail': 3594, 'feloni': 1629, 'dramat': 1329, 'registr': 3692, 'impos': 2185, 'camp': 620, 'deal': 1096, 'born': 494, 'anytim': 203, 'defend': 1125, 'civil': 799, 'postal': 3415, 'servic': 4016, 'financi': 1659, 'loss': 2594, 'delay': 1140, 'opposit': 3099, 'amend': 162, 'confeder': 922, 'battl': 373, 'gift': 1858, 'shop': 4054, 'hurt': 2145, 'sheila': 4040, 'jackson': 2322, 'lee': 2495, 'hey': 2049, 'spot': 4221, 'planet': 3351, 'googl': 1891, 'iron': 2303, 'ore': 3107, 'northern': 3009, 'pristin': 3485, 'stream': 4298, 'wast': 4847, 'highesttax': 2057, 'black': 453, 'wealthiest': 4862, 'sold': 4152, 'seller': 3996, 'onlin': 3090, 'amanda': 155, 'fritz': 1783, 'manag': 2647, 'child': 759, 'gotten': 1896, 'wors': 4957, 'floridian': 1697, 'repeal': 3737, 'congressman': 936, 'pascrel': 3215, 'option': 3101, 'lawmak': 2480, 'effort': 1392, 'draw': 1331, 'surrog': 4390, 'rape': 3613, 'incest': 2197, 'confront': 932, 'issu': 2316, 'offshor': 3064, 'medic': 2730, 'forev': 1723, 'microchip': 2778, 'implant': 2181, 'opt': 3100, 'becom': 386, 'enact': 1439, 'predat': 3437, 'conceal': 906, 'havent': 2006, 'oversight': 3156, 'charter': 740, 'appli': 216, 'elorza': 1417, 'municip': 2900, 'overrul': 3151, 'roughli': 3865, 'ri': 3797, 'nathan': 2922, 'hope': 2108, 'liter': 2557, 'georgian': 1848, 'technic': 4454, 'badli': 329, 'ive': 2319, 'teach': 4448, 'reach': 3628, 'explicitli': 1544, 'disclosur': 1237, 'moroccan': 2870, 'potteri': 3418, 'huge': 2135, 'influenc': 2231, 'arabia': 230, 'alqaida': 148, 'williamson': 4913, 'attorney': 291, 'prosecut': 3528, 'adult': 69, 'fundrais': 1797, 'unmatch': 4713, 'reflect': 3682, 'shrink': 4065, 'labor': 2452, 'convict': 981, 'overse': 3154, 'raid': 3593, 'retir': 3779, 'balanc': 336, 'battlefield': 374, 'green': 1918, 'packer': 3174, 'player': 3357, 'peopl': 3271, 'republ': 3746, 'brag': 517, 'shot': 4059, 'appar': 210, 'civilian': 800, 'fall': 1585, 'faster': 1600, 'hospit': 2114, 'mostli': 2877, 'restaur': 3772, 'highpay': 2059, 'proport': 3525, 'bought': 501, 'vehicl': 4769, 'airport': 112, 'store': 4289, 'sell': 3995, 'watt': 4853, 'bulb': 576, 'row': 3869, 'reelect': 3676, 'carter': 667, 'associ': 271, 'press': 3463, 'patrick': 3230, 'problem': 3495, 'suicid': 4355, 'congressmen': 937, 'addit': 59, 'agent': 96, 'master': 2692, 'lock': 2574, 'run': 3876, 'capac': 639, 'reed': 3675, 'overhaul': 3147, 'origin': 3113, 'frank': 1759, 'pallon': 3182, 'fingerprint': 1663, 'inde': 2212, 'independ': 2214, 'cours': 1018, 'attitud': 290, 'celebr': 696, 'miamidad': 2771, 'omnibu': 3080, 'earmark': 1368, 'talk': 4429, 'tough': 4565, 'kim': 2423, 'myer': 2910, 'reloc': 3714, 'headquart': 2014, 'california': 616, 'prop': 3519, 'film': 1656, 'harass': 1982, 'toll': 4551, 'privileg': 3489, 'cross': 1045, 'bridg': 538, 'isakson': 2306, 'chambliss': 724, 'estat': 1494, 'legislatur': 2501, 'kay': 2402, 'bailey': 332, 'hutchison': 2149, 'mica': 2772, 'mari': 2668, 'burk': 591, 'tabl': 4422, 'partisan': 3208, 'pick': 3324, 'multipl': 2897, 'regular': 3696, 'mid': 2779, 'rage': 3592, 'ran': 3605, 'constitut': 953, 'convent': 978, 'rise': 3823, 'flint': 1690, 'mich': 2773, 'poison': 3372, 'complic': 899, 'compli': 897, 'code': 849, 'prayer': 3429, 'room': 3857, 'rug': 3874, 'pray': 3428, 'allah': 136, 'bulk': 577, 'drug': 1347, 'gang': 1814, 'motor': 2883, 'ask': 259, 'england': 1453, 'grant': 1912, 'institut': 2261, 'miami': 2770, 'carlo': 655, 'curbelo': 1056, 'abil': 9, 'pollut': 3385, 'combat': 866, 'disclos': 1236, 'donor': 1305, 'suspicion': 4401, 'usa': 4738, 'continu': 967, 'wonder': 4943, 'rand': 3607, 'billionair': 438, 'super': 4365, 'pac': 3169, 'sept': 4007, 'intend': 2270, 'recov': 3660, 'newli': 2961, 'hold': 2084, 'payment': 3245, 'advisori': 76, 'mitchel': 2828, 'park': 3199, 'dome': 1298, 'millionplu': 2800, 'littl': 2559, 'evid': 1511, 'red': 3666, 'camera': 619, 'accid': 31, 'intersect': 2278, 'camden': 617, 'delinqu': 1145, 'pastor': 3223, 'priest': 3473, 'imam': 2173, 'moral': 2865, 'tradit': 4581, 'sexual': 4028, 'guilti': 1940, 'jobless': 2358, 'colleagu': 857, 'crack': 1024, 'attract': 292, 'affirm': 82, 'procedur': 3497, 'known': 2439, 'global': 1874, 'aid': 104, 'instabl': 2255, 'recipi': 3650, 'thank': 4491, 'nowher': 3020, 'award': 317, 'date': 1085, 'pupil': 3558, 'compar': 887, 'path': 3227, 'citizenship': 797, 'welfar': 4878, 'tammi': 4431, 'baldwin': 338, 'introduc': 2284, 'lesbian': 2509, 'bisexu': 450, 'ident': 2161, 'acorn': 45, 'slaveri': 4117, 'girl': 1865, 'survey': 4394, 'transit': 4591, 'ridership': 3812, 'age': 93, 'render': 3725, 'complet': 895, 'june': 2381, 'hard': 1984, 'jacob': 2323, 'someon': 4163, 'region': 3690, 'fair': 1579, 'fact': 1570, 'reid': 3699, 'rebecca': 3640, 'bradley': 516, 'diploma': 1220, 'implement': 2182, 'exchang': 1522, 'tape': 4435, 'featur': 1616, 'systemat': 4420, 'minnesota': 2808, 'cori': 992, 'booker': 487, 'safeti': 3897, 'modern': 2841, 'releas': 3708, 'return': 3781, 'alon': 147, 'eu': 1499, 'export': 1549, 'korea': 2441, 'market': 2673, 'farmer': 1595, 'suffer': 4351, 'luxuri': 2615, 'tim': 4537, 'kain': 2391, 'ambit': 160, 'dnc': 1281, 'chairman': 719, 'flipflop': 1692, 'easi': 1374, 'jet': 2349, 'zack': 4992, 'space': 4191, 'formal': 1729, 'sanction': 3909, 'subcommitte': 4330, 'hampshir': 1971, 'thirdhighest': 4510, 'rack': 3586, 'carol': 656, 'sheaport': 4038, 'creator': 1033, 'sixti': 4103, 'somewher': 4166, 'stay': 4269, 'interven': 2280, 'libya': 2526, 'cleveland': 818, 'perpupil': 3289, 'distanc': 1262, 'interview': 2282, 'sander': 3912, 'indefinit': 2213, 'detent': 1189, 'stand': 4247, 'minutemen': 2811, 'hunt': 2140, 'incent': 2196, 'fouryear': 1748, 'gallup': 1809, 'egyptian': 1395, 'violent': 4795, 'maddox': 2622, 'avail': 312, 'prepar': 3453, 'exxon': 1563, 'aarp': 2, 'resel': 3755, 'supplement': 4370, 'monthli': 2860, 'incorpor': 2206, 'purpos': 3561, 'read': 3630, 'grade': 1904, 'cuyahoga': 1063, 'clinic': 825, 'emerg': 1426, 'zika': 4995, 'troubl': 4622, 'pet': 3299, 'kilmartin': 2422, 'statut': 4267, 'occupi': 3051, 'protest': 3537, 'forward': 1739, 'parol': 3200, 'fisher': 1674, 'watch': 4849, 'limit': 2545, 'contracept': 968, 'cheaper': 746, 'easier': 1375, 'udal': 4659, 'defin': 1130, 'del': 1137, 'closest': 832, 'groceri': 1923, 'flood': 1693, 'pork': 3396, 'protect': 3535, 'warner': 4838, 'prohibit': 3511, 'prescript': 3456, 'fcc': 1612, 'neutral': 2955, 'essenti': 1492, 'massiv': 2691, 'typic': 4657, 'white': 4893, 'treati': 4606, 'certainli': 708, 'registri': 3693, 'confisc': 928, 'remind': 3719, 'math': 2698, 'apprehend': 221, 'pakistan': 3179, 'stood': 4286, 'big': 432, 'fli': 1688, 'perform': 3280, 'angel': 176, 'tavera': 4440, 'kendrick': 2406, 'meek': 2734, 'advanc': 70, 'inmat': 2241, 'dead': 1093, 'soda': 4148, 'philadelphia': 3312, 'teamster': 4451, 'countless': 1014, 'beverag': 424, 'tobacco': 4544, 'corn': 993, 'spike': 4211, 'sugar': 4352, 'smoke': 4134, 'nineti': 2976, 'texan': 4486, 'uniqu': 4704, 'staffer': 4240, 'committe': 878, 'race': 3582, 'particip': 3205, 'sought': 4177, 'hurrican': 2144, 'katrina': 2401, 'god': 1880, 'punish': 3557, 'cathol': 683, 'church': 784, 'antichrist': 195, 'filer': 1654, 'stock': 4281, 'rival': 3828, 'middleclass': 2781, 'founder': 1744, 'nelson': 2949, 'hardearn': 1986, 'secondari': 3976, 'produc': 3500, 'dropout': 1344, 'embryon': 1425, 'stem': 4272, 'exampl': 1517, 'success': 4347, 'treatment': 4607, 'dane': 1076, 'hand': 1972, 'count': 1011, 'presidenti': 3462, 'ambassador': 159, 'christoph': 779, 'steven': 4277, 'request': 3751, 'household': 2124, 'friend': 1781, 'switch': 4413, 'usag': 4739, 'instanc': 2257, 'harm': 1993, 'death': 1099, 'risen': 3824, 'dismiss': 1249, 'incompet': 2203, 'longest': 2583, 'inform': 2232, 'accumul': 37, 'africa': 88, 'thirdbiggest': 4508, 'precipit': 3434, 'centuri': 704, 'ye': 4976, 'stupid': 4328, 'folk': 1705, 'submit': 4334, 'earliest': 1367, 'oblig': 3042, 'updat': 4728, 'address': 60, 'commonli': 880, 'spoken': 4217, 'languag': 2465, 'opinion': 3094, 'referendum': 3678, 'thug': 4526, 'youll': 4985, 'southern': 4184, 'trigger': 4615, 'vaccin': 4754, 'brazilian': 525, 'inject': 2238, 'pregnant': 3444, 'execut': 1526, 'carli': 654, 'fiorina': 1666, 'assault': 264, 'bond': 483, 'multipli': 2898, 'platform': 3355, 'cop': 988, 'eightyear': 1398, 'hybrid': 2151, 'honor': 2102, 'repeat': 3738, 'failur': 1578, 'variou': 4763, 'buck': 564, 'mccollum': 2710, 'nra': 3021, 'leadership': 2488, 'pretti': 3466, 'intact': 2267, 'turkey': 4635, 'arab': 229, 'bid': 428, 'hiv': 2077, 'leffingwel': 2496, 'osama': 3116, 'bin': 440, 'laden': 2454, 'assassin': 263, 'drone': 1342, 'final': 1657, 'classroom': 808, 'suppos': 4373, 'design': 1179, 'aclu': 44, 'newark': 2959, 'immedi': 2174, 'monitor': 2854, 'hpv': 2131, 'highway': 2063, 'el': 1401, 'paso': 3216, 'bullet': 580, 'competit': 890, 'elector': 1405, 'sic': 4069, 'wed': 4870, 'nikita': 2975, 'extra': 1559, 'shoe': 4051, 'session': 4018, 'desk': 1181, 'renacci': 3722, 'lieberman': 2531, 'daili': 1069, 'expens': 1538, 'ceo': 705, 'franklin': 1763, 'delano': 1138, 'roosevelt': 3858, 'fasttrack': 1603, 'deport': 1167, 'dreamer': 1335, 'vet': 4779, 'piec': 3326, 'ray': 3626, 'champion': 725, 'storm': 4291, 'mahal': 2629, 'franc': 1755, 'surveil': 4393, 'invas': 2286, 'onehalf': 3083, 'blue': 465, 'chip': 768, 'indic': 2218, 'forecast': 1718, 'predict': 3440, 'fourth': 1745, 'ix': 2320, 'oversea': 3155, 'followup': 1707, 'backlog': 325, 'view': 4791, 'pornographi': 3400, 'molest': 2846, 'mccrori': 2712, 'advis': 74, 'romneycar': 3851, 'blueprint': 467, 'ballot': 340, 'especi': 1490, 'nytim': 3035, 'apolog': 208, 'bad': 328, 'hole': 2086, 'argument': 238, 'capita': 642, 'switzerland': 4414, 'raini': 3599, 'infant': 2225, 'follow': 1706, 'attempt': 287, 'spous': 4223, 'everyon': 1508, 'everywher': 1510, 'paycheck': 3243, 'pump': 3556, 'pose': 3409, 'risk': 3825, 'latina': 2474, 'craft': 1026, 'breweri': 534, 'threshold': 4522, 'surround': 4391, 'offens': 3059, 'dalla': 1071, 'lo': 2566, 'detroit': 1192, 'carcieri': 648, 'host': 2115, 'tree': 4608, 'secondmost': 3980, 'heavili': 2026, 'deploy': 1166, 'superintend': 4367, 'taylor': 4445, 'travi': 4601, 'ignor': 2165, 'laugh': 2476, 'ton': 4554, 'narcot': 2916, 'potenti': 3417, 'deregul': 1173, 'telecommun': 4462, 'french': 1774, 'loretta': 2592, 'lynch': 2616, 'team': 4450, 'whitewat': 4895, 'corrupt': 1001, 'manhattan': 2651, 'ralli': 3602, 'acceler': 28, 'preced': 3431, 'slower': 4122, 'pathway': 3228, 'deliv': 1146, 'jack': 2321, 'robust': 3841, 'outpac': 3134, 'gogeb': 1882, 'driver': 1341, 'licens': 2528, 'mexico': 2769, 'central': 703, 'gasolin': 1826, 'fuel': 1788, 'richard': 3803, 'burr': 593, 'insid': 2250, 'brave': 523, 'christian': 776, 'vermont': 4775, 'sentenc': 4005, 'coupl': 1016, 'unarm': 4670, 'email': 1420, 'moni': 2852, 'tarp': 4437, 'treasuri': 4604, 'treat': 4605, 'game': 1813, 'client': 820, 'haiti': 1957, 'doc': 1282, 'viabl': 4782, 'gold': 1884, 'medal': 2727, 'nonpartisan': 2996, 'choic': 769, 'ohioan': 3067, 'ride': 3810, 'passeng': 3219, 'reveal': 3784, 'instat': 2259, 'healthcaregov': 2018, 'user': 4743, 'privaci': 3486, 'expert': 1541, 'al': 117, 'qaeda': 3568, 'kelli': 2404, 'ayott': 320, 'lobbi': 2569, 'network': 2953, 'financ': 1658, 'cranston': 1028, 'grew': 1922, 'aircraft': 109, 'carrier': 662, 'submarin': 4333, 'complianc': 898, 'plumber': 3365, 'sheet': 4039, 'metal': 2762, 'effici': 1391, 'loophol': 2588, 'decept': 1107, 'secondhighest': 3977, 'toy': 4576, 'describ': 1176, 'boston': 498, 'roy': 3870, 'cooper': 987, 'qatar': 3569, 'restrict': 3774, 'goldman': 1885, 'sach': 3890, 'boss': 497, 'unilater': 4699, 'disarm': 1232, 'moratorium': 2866, 'news': 2963, 'stori': 4290, 'railroad': 3596, 'michael': 2774, 'william': 4912, 'tea': 4447, 'gather': 1828, 'harri': 1994, 'alien': 132, 'equal': 1479, 'citat': 793, 'kasim': 2396, 'respond': 3769, 'worst': 4960, 'redistrict': 3671, 'mention': 2750, 'mate': 2695, 'inflationari': 2230, 'soar': 4141, 'relief': 3711, 'lamar': 2459, 'smith': 4132, 'bailout': 333, 'assur': 273, 'happen': 1979, 'direct': 1223, 'parent': 3196, 'ninetyseven': 2979, 'array': 247, 'baltimor': 341, 'distress': 1267, 'india': 2216, 'payasyougo': 3242, 'entitl': 1470, 'increasingli': 2208, 'opportun': 3097, 'ocean': 3053, 'process': 3498, 'facil': 1568, 'greenhous': 1919, 'northeastern': 3008, 'length': 2507, 'elig': 1412, 'realli': 3636, 'realiz': 3635, 'dian': 1202, 'sandi': 3913, 'destruct': 1186, 'atlant': 280, 'gwinnett': 1949, 'henri': 2039, 'debat': 1100, 'commanderinchief': 871, 'ganley': 1815, 'toomey': 4558, 'jeb': 2337, 'aliv': 135, 'suit': 4356, 'vodka': 4806, 'netherland': 2952, 'tie': 4532, 'subsidi': 4340, 'hotel': 2117, 'jump': 2379, 'closeddoor': 830, 'walmart': 4827, 'compact': 885, 'fluoresc': 1702, 'cfl': 712, 'pop': 3392, 'tell': 4464, 'lag': 2456, 'howel': 2128, 'moder': 2840, 'lever': 2515, 'langevin': 2464, 'transpacif': 4594, 'partnership': 3211, 'tini': 4540, 'doesnt': 1290, 'barbara': 350, 'boxer': 507, 'worri': 4956, 'weather': 4865, 'overwhelmingli': 3162, 'caught': 687, 'merrick': 2758, 'garland': 1824, 'islamist': 2310, 'fatal': 1605, 'nondefens': 2993, 'disenfranchis': 1245, 'felon': 1628, 'onequart': 3084, 'threaten': 4520, 'bomb': 480, 'consult': 957, 'jerri': 2344, 'ed': 1386, 'gillespi': 1859, 'recommend': 3654, 'jerusalem': 2347, 'israel': 2315, 'draft': 1325, 'yr': 4991, 'staf': 4238, 'ratio': 3622, 'seattl': 3972, 'spiral': 4213, 'turnout': 4639, 'experi': 1539, 'kenya': 2411, 'resign': 3761, 'visit': 4802, 'nashvil': 2921, 'redistribut': 3670, 'characterist': 731, 'tampa': 4432, 'strip': 4311, 'lt': 2612, 'casey': 670, 'marquett': 2675, 'constitu': 952, 'cuba': 1050, 'venezuela': 4772, 'wildfir': 4909, 'link': 2551, 'far': 1592, 'earlier': 1366, 'crowd': 1046, 'sprawl': 4224, 'capitol': 643, 'monument': 2861, 'supervisor': 4369, 'noncitizen': 2991, 'roll': 3849, 'richardson': 3804, 'withdraw': 4933, 'cantor': 636, 'appoint': 218, 'stamp': 4246, 'farm': 1594, 'turbin': 4634, 'particularli': 3207, 'seizur': 3992, 'march': 2663, 'hijack': 2064, 'revolut': 3791, 'brutal': 560, 'theocraci': 4497, 'spanish': 4194, 'tv': 4641, 'object': 3041, 'criteria': 1042, 'boundari': 503, 'resourc': 3766, 'bodi': 476, 'armor': 245, 'helicopt': 2032, 'ammunit': 166, 'video': 4789, 'alltim': 146, 'franchis': 1756, 'dealer': 1097, 'autism': 307, 'prior': 3481, 'rifl': 3814, 'round': 3867, 'residenti': 3759, 'handgun': 1974, 'balloon': 339, 'pole': 3376, 'basketbal': 369, 'coach': 839, 'drastic': 1330, 'alcohol': 125, 'underag': 4677, 'bureaucrat': 588, 'scan': 3937, 'keyston': 2415, 'xl': 4974, 'pipelin': 3340, 'speech': 4205, 'guest': 1936, 'takeov': 4427, 'quit': 3578, 'present': 3458, 'percentof': 3277, 'flore': 1695, 'laid': 2457, 'govt': 1902, 'permiss': 3286, 'medicin': 2733, 'shortfal': 4058, 'assumpt': 272, 'siphon': 4095, 'certain': 707, 'apart': 205, 'weekend': 4873, 'flight': 1689, 'bail': 331, 'costli': 1006, 'knew': 2436, 'sixth': 4101, 'fine': 1661, 'asthma': 274, 'uk': 4661, 'studio': 4324, 'ordinari': 3106, 'soccer': 4143, 'pend': 3265, 'fewest': 1642, 'victor': 4787, 'nunn': 3028, 'handpick': 1977, 'sununu': 4364, 'agenda': 95, 'gonna': 1889, 'write': 4967, 'lack': 2453, 'defer': 1127, 'museum': 2904, 'isnt': 2312, 'sequestr': 4011, 'fort': 1732, 'hood': 2103, 'advisor': 75, 'secondlowest': 3979, 'tip': 4541, 'physician': 3323, 'distribut': 1268, 'hezbollah': 2050, 'latin': 2473, 'immin': 2176, 'food': 1708, 'magazin': 2626, 'beckham': 385, 'cargo': 652, 'cruis': 1047, 'experienc': 1540, 'burger': 589, 'king': 2426, 'deer': 1121, 'bidder': 429, 'alli': 141, 'sean': 3968, 'duffi': 1355, 'ribbl': 3798, 'hall': 1962, 'controversi': 976, 'consent': 943, 'pride': 3472, 'night': 2972, 'iowa': 2293, 'caucus': 686, 'focus': 1704, 'contain': 961, 'phoni': 3319, 'hors': 2113, 'key': 2414, 'smart': 4130, 'launch': 2477, 'recycl': 3665, 'theu': 4502, 'fewer': 1641, 'detaine': 1188, 'hn': 2080, 'rwi': 3888, 'birther': 448, 'equival': 1481, 'low': 2606, 'barrow': 362, 'dga': 1200, 'ownership': 3166, 'schoolchildren': 3949, 'justifi': 2387, 'governmentrun': 1899, 'automobil': 311, 'antibusi': 193, 'litig': 2558, 'expans': 1535, 'pasco': 3214, 'injuri': 2240, 'assist': 270, 'nutrit': 3031, 'outcom': 3124, 'misconduct': 2816, 'grandfath': 1910, 'fidel': 1643, 'castro': 676, 'prefer': 3442, 'significantli': 4079, 'ninetyfour': 2978, 'scale': 3934, 'clock': 828, 'sayspresid': 3930, 'unelect': 4690, 'antiabort': 192, 'altogeth': 152, 'similar': 4082, 'indiana': 2217, 'religi': 3712, 'freedom': 1771, 'restor': 3773, 'prenat': 3451, 'exce': 1518, 'nose': 3011, 'dive': 1272, 'declar': 1110, 'qualifi': 3571, 'healthi': 2019, 'billiondollar': 439, 'tallahasse': 4430, 'milk': 2792, 'elementari': 1409, 'tear': 4452, 'squarefoot': 4231, 'jeff': 2338, 'bp': 512, 'plummet': 3366, 'thom': 4514, 'tilli': 4536, 'wealthi': 4861, 'bibl': 426, 'princip': 3478, 'sewer': 4026, 'asid': 258, 'exclus': 1525, 'aspirin': 261, 'secretli': 3983, 'tortur': 4562, 'cartel': 666, 'rebat': 3639, 'smuggl': 4136, 'danni': 1079, 'missil': 2821, 'shield': 4047, 'russia': 3885, 'voucher': 4815, 'fortysix': 1737, 'lineitem': 2550, 'profession': 3503, 'huckabe': 2132, 'statu': 4266, 'thirdgrad': 4509, 'caucu': 685, 'instructor': 2263, 'traffic': 4582, 'tower': 4573, 'appl': 214, 'conflictfre': 930, 'crimea': 1037, 'ukrain': 4662, 'circul': 791, 'coin': 854, 'catherin': 682, 'cortez': 1002, 'masto': 2693, 'sweetheart': 4407, 'worth': 4961, 'firm': 1671, 'pled': 3361, 'fraud': 1764, 'salt': 3905, 'revolutionari': 3792, 'guy': 1947, 'footag': 1711, 'broad': 545, 'secret': 3981, 'bobbi': 474, 'fish': 1673, 'kathi': 2397, 'entireti': 1468, 'canada': 626, 'accuraci': 39, 'barn': 355, 'anymor': 200, 'uncompens': 4674, 'track': 4578, 'travel': 4600, 'deborah': 1102, 'ross': 3863, 'ransom': 3612, 'hostag': 2116, 'heroin': 2045, 'addict': 58, 'opioid': 3095, 'chapter': 730, 'fetus': 1640, 'baucu': 376, 'deduct': 1116, 'st': 4233, 'van': 4760, 'wanggaard': 4830, 'drawn': 1333, 'falk': 1584, 'identif': 2162, 'statewid': 4263, 'alison': 134, 'handel': 1973, 'chairwoman': 721, 'fulton': 1793, 'wendi': 4880, 'davi': 1089, 'iranian': 2299, 'rouhani': 3866, 'fighter': 1651, 'dozen': 1323, 'zip': 4996, 'subject': 4332, 'patriot': 3231, 'sit': 4097, 'conced': 907, 'tourism': 4570, 'advertis': 72, 'overal': 3144, 'redirect': 3669, 'mode': 2838, 'twitter': 4650, 'garden': 1821, 'express': 1552, 'strong': 4313, 'pasch': 3213, 'childcar': 761, 'onetenth': 3085, 'frankli': 1762, 'sue': 4350, 'portion': 3403, 'compens': 888, 'fortyf': 1735, 'bruce': 558, 'braley': 519, 'diowa': 1217, 'nascar': 2920, 'abraham': 16, 'danc': 1075, 'ebay': 1380, 'whitman': 4896, 'separ': 4006, 'porn': 3399, 'philli': 3315, 'handl': 1975, 'loom': 2587, 'slash': 4113, 'predecessor': 3439, 'miller': 2796, 'judici': 2376, 'barrier': 361, 'innov': 2245, 'innerc': 2243, 'whatev': 4887, 'cheat': 747, 'freed': 1770, 'slave': 4116, 'ku': 2445, 'klux': 2435, 'klan': 2432, 'paint': 3178, 'uber': 4658, 'rideshar': 3813, 'eye': 1565, 'perfect': 3279, 'convers': 979, 'substanc': 4341, 'nj': 2983, 'dishonest': 1247, 'passag': 3218, 'futur': 1803, 'everyth': 1509, 'factual': 1574, 'accur': 38, 'husband': 2146, 'charit': 733, 'la': 2449, 'vega': 4766, 'boehner': 478, 'wreck': 4965, 'onethird': 3087, 'commerci': 874, 'debbi': 1101, 'wasserman': 4846, 'schultz': 3951, 'termin': 4476, 'ill': 2167, 'antichoic': 194, 'deadli': 1094, 'analysi': 170, 'anniversari': 184, 'jobkil': 2357, 'sunday': 4361, 'morn': 2869, 'usual': 4745, 'delta': 1149, 'anim': 181, 'neg': 2944, 'holder': 2085, 'involv': 2292, 'panther': 3189, 'intimid': 2283, 'richmond': 3806, 'chees': 750, 'glad': 1870, 'word': 4946, 'clearli': 816, 'exoner': 1532, 'retire': 3780, 'contributor': 974, 'machin': 2620, 'field': 1644, 'text': 4488, 'wouldv': 4963, 'bonu': 484, 'grab': 1903, 'unaccount': 4668, 'accommod': 33, 'shooter': 4053, 'enrolle': 1461, 'match': 2694, 'simpli': 4084, 'river': 3829, 'downstream': 1319, 'rice': 3801, 'photograph': 3321, 'acr': 47, 'recreat': 3662, 'enter': 1463, 'kindergarten': 2425, 'dental': 1161, 'ken': 2405, 'wish': 4931, 'recal': 3644, 'petit': 3303, 'signatur': 4076, 'rebuild': 3642, 'prove': 3541, 'chronic': 780, 'traumat': 4599, 'wive': 4938, 'televis': 4463, 'plane': 3350, 'late': 2470, 'pursu': 3562, 'dream': 1334, 'prank': 3427, 'blogger': 461, 'koch': 2440, 'publicli': 3550, 'wouldnt': 4962, 'onechild': 3081, 'steril': 4275, 'roland': 3847, 'burri': 594, 'octob': 3055, 'silent': 4081, 'feel': 1623, 'farrakhan': 1597, 'spoke': 4216, 'audienc': 296, 'edward': 1389, 'professor': 3504, 'extens': 1555, 'mittromney': 2831, 'fox': 1750, 'json': 2373, 'barelytru': 353, 'mostlytru': 2878, 'prek': 3446, 'rd': 3627, 'cite': 794, 'true': 4624, 'barrel': 358, 'comment': 872, 'halftru': 1960, 'pb': 3249, 'dmv': 1279, 'allianc': 142, 'newsmak': 2964, 'pantsfir': 3190, 'radio': 3590, 'lawyer': 2482, 'froze': 1784, 'channel': 729, 'jon': 2367, 'runyan': 3879, 'redefin': 3667, 'cruz': 1049, 'mathemat': 2699, 'eighti': 1397, 'fed': 1618, 'livestock': 2562, 'zone': 4998, 'expenditur': 1537, 'undermin': 4684, 'purg': 3560, 'russian': 3886, 'crop': 1044, 'ethanol': 1496, 'mail': 2632, 'tonight': 4555, 'unoppos': 4715, 'santorum': 3917, 'interst': 2279, 'road': 3832, 'cotton': 1008, 'navi': 2929, 'sail': 3899, 'ship': 4049, 'tour': 4569, 'join': 2363, 'hip': 2070, 'fashion': 1598, 'grand': 1908, 'mahon': 2631, 'valley': 4757, 'fy': 1804, 'chesterfield': 753, 'duke': 1357, 'joni': 2370, 'ernst': 1487, 'copi': 990, 'certif': 709, 'canadian': 627, 'sand': 3911, 'mcdonald': 2713, 'pittsburgh': 3344, 'snow': 4138, 'safe': 3894, 'marathon': 2662, 'fifth': 1645, 'examin': 1516, 'felt': 1630, 'crash': 1029, 'kenosha': 2408, 'tribe': 4614, 'howard': 2127, 'robert': 3838, 'wastewat': 4848, 'qualiti': 3572, 'pentagon': 3270, 'martial': 2681, 'faith': 1581, 'cobb': 845, 'shock': 4050, 'older': 3073, 'bigger': 433, 'kept': 2412, 'ultim': 4664, 'golf': 1887, 'undercov': 4678, 'memori': 2745, 'ceremoni': 706, 'inflationadjust': 2229, 'drawdown': 1332, 'danger': 1077, 'atf': 277, 'trace': 4577, 'broward': 556, 'firework': 1670, 'safer': 3895, 'woman': 4941, 'northeast': 3007, 'quick': 3576, 'identifi': 2163, 'desir': 1180, 'broke': 548, 'abandon': 3, 'ensur': 1462, 'quickli': 3577, 'warmer': 4836, 'mar': 2661, 'polar': 3375, 'vortex': 4811, 'calib': 614, 'mormon': 2868, 'obey': 3040, 'bruun': 561, 'commerc': 873, 'decor': 1112, 'rare': 3616, 'jan': 2327, 'answer': 189, 'craig': 1027, 'robinson': 3840, 'affili': 81, 'display': 1255, 'mirror': 2814, 'fifti': 1647, 'strongli': 4316, 'disagre': 1229, 'bargain': 354, 'coauthor': 844, 'principl': 3479, 'equip': 1480, 'rebel': 3641, 'destroy': 1185, 'mitch': 2827, 'mcconnel': 2711, 'tommi': 4553, 'human': 2136, 'granit': 1911, 'stater': 4261, 'renew': 3728, 'gardner': 1822, 'outlaw': 3126, 'birthright': 449, 'brother': 552, 'chai': 716, 'subsaharan': 4337, 'noshow': 3012, 'steadili': 4270, 'connecticut': 940, 'solv': 4159, 'scientist': 3954, 'fare': 1593, 'shore': 4055, 'particular': 3206, 'insecur': 2248, 'gi': 1854, 'earner': 1370, 'traffick': 4583, 'calendar': 613, 'dime': 1211, 'task': 4438, 'domin': 1300, 'youtub': 4989, 'paper': 3191, 'rigel': 3816, 'weaken': 4859, 'regulatori': 3698, 'campaignfin': 622, 'imposs': 2186, 'plain': 3347, 'simpl': 4083, 'cancer': 631, 'screen': 3960, 'altern': 151, 'regard': 3687, 'flaw': 1684, 'outer': 3125, 'continent': 966, 'deep': 1118, 'royalti': 3871, 'oh': 3065, 'wife': 4906, 'ebola': 1381, 'sequest': 4010, 'randi': 3608, 'hopper': 2109, 'rfond': 3795, 'du': 1351, 'kati': 2400, 'wing': 4922, 'bureau': 586, 'attribut': 293, 'sharp': 4036, 'workrel': 4952, 'understand': 4686, 'english': 1454, 'obes': 3039, 'audit': 297, 'ratifi': 3621, 'pain': 3177, 'religion': 3713, 'replac': 3740, 'territori': 4479, 'integr': 2268, 'clerk': 817, 'emiss': 1429, 'freshman': 1777, 'waterboard': 4851, 'main': 2633, 'mother': 2879, 'bunch': 582, 'korean': 2442, 'bag': 330, 'hell': 2033, 'summit': 4359, 'structur': 4318, 'paytoplay': 3248, 'wentworth': 4882, 'sonogram': 4170, 'stephen': 4274, 'da': 1067, 'clayton': 810, 'millag': 2794, 'metro': 2765, 'oversaw': 3153, 'bonus': 485, 'rememb': 3718, 'contractor': 970, 'tap': 4434, 'devil': 1196, 'vanderbeek': 4761, 'arbitr': 231, 'tennesse': 4471, 'file': 1653, 'subpoena': 4335, 'courag': 1017, 'recoveri': 3661, 'donut': 1307, 'demograph': 1154, 'singlepay': 4092, 'merit': 2755, 'reward': 3793, 'corzin': 1003, 'taxpayerfund': 4444, 'bracket': 513, 'bulli': 581, 'color': 860, 'unpreced': 4718, 'yearlong': 4978, 'vacanc': 4751, 'plot': 3363, 'loi': 2578, 'nonprofit': 2997, 'uniform': 4698, 'winter': 4924, 'olymp': 3077, 'stroke': 4312, 'pen': 3261, 'schrader': 3950, 'explain': 1543, 'suppli': 4371, 'cnn': 838, 'casualti': 677, 'fitzgerald': 1677, 'necessari': 2939, 'defraud': 1132, 'nofli': 2987, 'written': 4968, 'drag': 1326, 'inherit': 2236, 'repair': 3734, 'lane': 2463, 'dedic': 1115, 'analyst': 171, 'overdos': 3146, 'accident': 32, 'killer': 2421, 'commonwealth': 881, 'toledo': 4550, 'bell': 400, 'panhandl': 3187, 'bed': 387, 'rushhour': 3882, 'commut': 884, 'downtown': 1320, 'sunset': 4362, 'freeway': 1772, 'vice': 4784, 'kirk': 2428, 'moin': 2845, 'aug': 299, 'depriv': 1170, 'protein': 3536, 'lunch': 2613, 'thwart': 4528, 'incid': 2199, 'superior': 4368, 'meat': 2725, 'festiv': 1638, 'horrend': 2110, 'prosecutor': 3529, 'consumpt': 959, 'earthquak': 1372, 'step': 4273, 'walk': 4824, 'nixon': 2982, 'duck': 1353, 'garbag': 1819, 'allan': 137, 'fung': 1798, 'web': 4866, 'site': 4098, 'pour': 3420, 'atwat': 295, 'lavish': 2478, 'mental': 2749, 'twentyf': 4645, 'habit': 1951, 'recklessli': 3651, 'susan': 4397, 'uranium': 4735, 'enrich': 1459, 'capabl': 638, 'pot': 3416, 'element': 1408, 'door': 1308, 'mcchrystal': 2709, 'fix': 1678, 'espionag': 1491, 'snowden': 4139, 'whistleblow': 4892, 'cannabi': 634, 'music': 2906, 'art': 251, 'instruct': 2262, 'wit': 4932, 'reauthor': 3638, 'exportimport': 1550, 'prekindergarten': 3447, 'frame': 1754, 'salmonella': 3904, 'fda': 1613, 'peanut': 3253, 'quonset': 3579, 'antonio': 198, 'cervic': 711, 'hedg': 2029, 'littleknown': 2560, 'preciou': 3433, 'fresh': 1776, 'ahead': 102, 'magic': 2628, 'marketplac': 2674, 'inadequ': 2192, 'seced': 3974, 'patdown': 3225, 'hewlettpackard': 2048, 'nebraska': 2938, 'pant': 3188, 'moon': 2862, 'lax': 2483, 'eras': 1484, 'crush': 1048, 'mammogram': 2645, 'ice': 2157, 'hondura': 2100, 'accomplish': 34, 'basi': 367, 'hydraul': 2153, 'frack': 1751, 'midterm': 2784, 'review': 3788, 'await': 316, 'occas': 3048, 'boon': 489, 'forest': 1722, 'dallasfort': 1072, 'bowl': 505, 'unknown': 4707, 'begich': 392, 'drunk': 1349, 'input': 2247, 'cool': 984, 'reaction': 3629, 'somebodi': 4162, 'odd': 3056, 'monopoli': 2857, 'instant': 2258, 'downgrad': 1317, 'devot': 1197, 'lent': 2508, 'stateown': 4260, 'brazil': 524, 'explor': 1547, 'automat': 310, 'plastic': 3353, 'hegar': 2030, 'sometim': 4165, 'shift': 4048, 'abrahamlincoln': 17, 'twentyfour': 4646, 'cold': 855, 'blood': 462, 'milliondollar': 2799, 'repay': 3736, 'anglo': 178, 'millenni': 2795, 'provision': 3545, 'disqualifi': 1259, 'adam': 56, 'segment': 3989, 'beer': 389, 'aggress': 97, 'healthier': 2020, 'rural': 3880, 'illiter': 2170, 'morocco': 2871, 'fenc': 1633, 'septic': 4009, 'inspect': 2252, 'tank': 4433, 'multnomah': 2899, 'busiest': 600, 'centerpiec': 702, 'slot': 4120, 'outsourc': 3140, 'connect': 939, 'stringent': 4310, 'holperin': 2091, 'outofst': 3133, 'whenev': 4890, 'rene': 3726, 'ellmer': 1416, 'czar': 1065, 'reopen': 3731, 'tremend': 4609, 'proud': 3539, 'surviv': 4395, 'dri': 1336, 'testimoni': 4484, 'sotomayor': 4176, 'darryl': 1082, 'forti': 1733, 'fortun': 1734, 'emir': 1428, 'formula': 1730, 'rocket': 3843, 'gaza': 1831, 'militar': 2790, 'log': 2575, 'spring': 4227, 'perriello': 3291, 'riski': 3826, 'solid': 4156, 'oklahoma': 3071, 'louisiana': 2603, 'kansa': 2393, 'utah': 4746, 'transpar': 4595, 'visa': 4800, 'overstay': 3157, 'contrast': 972, 'sing': 4089, 'hawaii': 2007, 'alaska': 122, 'chevrolet': 754, 'volt': 4809, 'coop': 986, 'enterpris': 1464, 'freddi': 1767, 'mac': 2619, 'indict': 2219, 'brain': 518, 'facebook': 1567, 'page': 3175, 'levi': 2516, 'werent': 4883, 'abel': 7, 'finish': 1664, 'arent': 236, 'sb': 3933, 'depict': 1165, 'homosexu': 2099, 'tactic': 4423, 'schedul': 3942, 'revers': 3787, 'underground': 4683, 'ayer': 319, 'comprehens': 901, 'publish': 3552, 'trayvon': 4602, 'martin': 2682, 'tattoo': 4439, 'solvent': 4160, 'whitehous': 4894, 'easili': 1376, 'classifi': 806, 'miner': 2803, 'aim': 106, 'cia': 785, 'interrog': 2277, 'brendan': 531, 'doherti': 1293, 'occup': 3050, 'sam': 3906, 'tent': 4472, 'pitch': 3343, 'bankruptci': 346, 'unlaw': 4708, 'japan': 2331, 'embassi': 1422, 'tragedi': 4584, 'alabama': 119, 'mississippi': 2823, 'wyom': 4973, 'hud': 2133, 'soon': 4171, 'strengthen': 4302, 'cuccinelli': 1052, 'halt': 1965, 'epa': 1477, 'gase': 1825, 'measl': 2723, 'pictur': 3325, 'subhuman': 4331, 'mongrel': 2851, 'hit': 2075, 'enlist': 1457, 'specialist': 4200, 'split': 4215, 'thirdlargest': 4511, 'surpass': 4385, 'latest': 2472, 'repatri': 3735, 'usborn': 4740, 'occur': 3052, 'twoandahalf': 4651, 'investor': 2290, 'bust': 601, 'gm': 1876, 'weight': 4875, 'warm': 4835, 'mayb': 2705, 'haitian': 1958, 'drink': 1338, 'ground': 1926, 'gibb': 1857, 'poorer': 3390, 'western': 4885, 'corridor': 1000, 'preexist': 3441, 'winner': 4923, 'superdeleg': 4366, 'forbidden': 1714, 'remedi': 3717, 'jeopard': 2341, 'cleaner': 812, 'custom': 1061, 'statehous': 4257, 'staterun': 4262, 'dump': 1358, 'unregul': 4719, 'guevara': 1937, 'descent': 1175, 'middleincom': 2782, 'hitler': 2076, 'soviet': 4188, 'hadnt': 1953, 'ward': 4833, 'fouryearold': 1749, 'preschool': 3454, 'betti': 423, 'sutton': 4402, 'pete': 3300, 'mission': 2822, 'cocain': 847, 'japanes': 2332, 'embargo': 1421, 'madison': 2624, 'fraudul': 1765, 'runoff': 3878, 'previous': 3470, 'genet': 1839, 'makeup': 2639, 'ant': 190, 'elsewher': 1419, 'bushera': 598, 'kurt': 2447, 'enrol': 1460, 'august': 300, 'orient': 3112, 'hw': 2150, 'encount': 1440, 'detain': 1187, 'buyer': 605, 'disrupt': 1261, 'discov': 1239, 'hurd': 2143, 'acrosstheboard': 48, 'fri': 1779, 'empow': 1436, 'matern': 2697, 'chile': 765, 'bucket': 565, 'overhead': 3148, 'liquor': 2554, 'republicanl': 3749, 'pool': 3388, 'strategi': 4296, 'apprais': 220, 'categori': 681, 'blind': 458, 'worldwid': 4955, 'lowerincom': 2608, 'mainli': 2634, 'gordon': 1893, 'cofound': 852, 'invad': 2285, 'saddam': 3892, 'acquir': 46, 'captur': 645, 'offset': 3063, 'tune': 4633, 'francisco': 1758, 'flee': 1685, 'emin': 1427, 'domain': 1297, 'incub': 2210, 'rapid': 3614, 'herbert': 2040, 'hoover': 2107, 'quadrupl': 3570, 'bare': 352, 'budg': 568, 'rio': 3820, 'ongo': 3089, 'concentr': 908, 'outrag': 3137, 'kathryn': 2399, 'starkey': 4252, 'divers': 1273, 'bike': 436, 'radiat': 3588, 'tsa': 4629, 'feet': 1624, 'todd': 4546, 'akin': 116, 'academi': 27, 'pediatr': 3255, 'hoax': 2081, 'blackston': 455, 'profici': 3505, 'assess': 267, 'capandtrad': 640, 'chuck': 783, 'hagel': 1955, 'filibust': 1655, 'stabil': 4235, 'ethnic': 1498, 'ought': 3121, 'bench': 406, 'hardwork': 1991, 'gunfre': 1945, 'alaskan': 123, 'section': 3984, 'claus': 809, 'outstand': 3142, 'frequent': 1775, 'midwest': 2785, 'unfairli': 4695, 'everifi': 1506, 'materi': 2696, 'server': 4015, 'vietnam': 4790, 'gallon': 1808, 'franken': 1761, 'factori': 1573, 'faulti': 1607, 'stack': 4236, 'lyndon': 2617, 'ram': 3604, 'artist': 255, 'justin': 2388, 'jay': 2334, 'cancel': 630, 'stagger': 4242, 'technolog': 4456, 'prescrib': 3455, 'databas': 1084, 'pharmaci': 3307, 'prostat': 3533, 'breast': 529, 'jd': 2335, 'hayworth': 2008, 'tech': 4453, 'timelin': 4539, 'lorann': 2590, 'ausley': 302, 'concess': 912, 'passport': 3221, 'holli': 2089, 'turner': 4638, 'discretionari': 1241, 'dwight': 1363, 'mercuri': 2752, 'outlin': 3128, 'forcibl': 1716, 'sayshillari': 3929, 'poorest': 3391, 'speak': 4196, 'deliber': 1143, 'nonexist': 2995, 'download': 1318, 'jennif': 2340, 'brunner': 559, 'mismanag': 2819, 'alter': 150, 'viru': 4799, 'shell': 4043, 'roger': 3846, 'pois': 3371, 'seventi': 4023, 'segreg': 3990, 'inact': 2191, 'stole': 4283, 'stretch': 4304, 'tourist': 4571, 'outbreak': 3123, 'loudli': 2599, 'dirti': 1226, 'fossil': 1740, 'error': 1488, 'uptick': 4733, 'reconcili': 3655, 'curb': 1055, 'spread': 4225, 'botch': 499, 'logo': 2577, 'fanci': 1590, 'joann': 2355, 'kloppenburg': 2434, 'societi': 4146, 'normandi': 3005, 'carpool': 660, 'februari': 1617, 'minimumwag': 2806, 'arlen': 242, 'specter': 4203, 'interestingli': 2272, 'kentucki': 2409, 'print': 3480, 'lindsey': 2548, 'graham': 1907, 'shortag': 4057, 'sonia': 4169, 'propaganda': 3520, 'father': 1606, 'thensen': 4496, 'eat': 1379, 'hottest': 2120, 'sheldon': 4042, 'upheld': 4730, 'surpris': 4388, 'subprim': 4336, 'mortal': 2873, 'unfund': 4697, 'waiver': 4822, 'hmo': 2079, 'taint': 4425, 'beef': 388, 'cafeteria': 609, 'lewi': 2517, 'hero': 2044, 'alleg': 138, 'remark': 3716, 'entiti': 1469, 'overwhelm': 3161, 'missourian': 2825, 'thirti': 4512, 'menendez': 2748, 'pursuit': 3563, 'proven': 3542, 'theft': 4495, 'petersburg': 3302, 'dr': 1324, 'chu': 782, 'reli': 3709, 'insuffici': 2264, 'amtrak': 169, 'instal': 2256, 'derail': 1172, 'underfund': 4680, 'realiti': 3634, 'signer': 4077, 'administ': 63, 'studentathlet': 4322, 'ipad': 2294, 'applic': 217, 'nfl': 2967, 'buse': 595, 'limousin': 2546, 'fleet': 1686, 'newspap': 2965, 'victori': 4788, 'nevadan': 2957, 'doubledigit': 1312, 'nevada': 2956, 'meal': 2717, 'chariti': 734, 'harrisburg': 1995, 'pawlenti': 3240, 'horribl': 2111, 'dad': 1068, 'prolif': 3513, 'util': 4747, 'cheap': 745, 'barton': 363, 'deadlin': 1095, 'crippl': 1039, 'viagra': 4783, 'rapist': 3615, 'mao': 2659, 'vern': 4776, 'genocid': 1842, 'bosnia': 496, 'boat': 472, 'ration': 3623, 'evalu': 1502, 'statemand': 4258, 'exam': 1515, 'overlook': 3149, 'southwest': 4186, 'gelber': 1835, 'needi': 2942, 'jewish': 2351, 'saturday': 3922, 'disturb': 1270, 'trend': 4610, 'massacr': 2689, 'wound': 4964, 'geithner': 1834, 'currenc': 1058, 'secondlargest': 3978, 'fcat': 1611, 'economist': 1385, 'boost': 490, 'unabl': 4666, 'hasner': 2001, 'halliburton': 1963, 'su': 4329, 'eighth': 1396, 'hamilton': 1969, 'buddi': 567, 'upstat': 4732, 'hog': 2083, 'cass': 673, 'sunstein': 4363, 'ferguson': 1635, 'mo': 2833, 'regim': 3689, 'behalf': 395, 'charl': 735, 'incred': 2209, 'boe': 477, 'wineri': 4921, 'vatican': 4765, 'consul': 956, 'tool': 4557, 'sworn': 4415, 'ginsburg': 1864, 'lien': 2532, 'unpaid': 4716, 'powel': 3423, 'butt': 602, 'reservoir': 3757, 'chavez': 743, 'struggl': 4319, 'newborn': 2960, 'shade': 4029, 'receipt': 3645, 'version': 4777, 'pharmaceut': 3306, 'che': 744, 'nickel': 2971, 'censu': 699, 'cheerlead': 749, 'cowboy': 1023, 'entertain': 1465, 'pizza': 3345, 'invit': 2291, 'fanni': 1591, 'aig': 105, 'disappear': 1230, 'mountain': 2887, 'revok': 3790, 'argu': 237, 'terribl': 4478, 'algebra': 129, 'admiss': 66, 'palestinian': 3180, 'isra': 2314, 'geneva': 1840, 'normal': 3004, 'semiautomat': 3998, 'misdemeanor': 2817, 'norcross': 3002, 'exact': 1513, 'visitor': 4803, 'scientif': 3953, 'springsteen': 4228, 'orang': 3103, 'gym': 1950, 'heart': 2024, 'counsel': 1010, 'sooner': 4172, 'coffman': 851, 'preserv': 3459, 'roadsid': 3833, 'frederica': 1768, 'wilson': 4914, 'testifi': 4483, 'bermuda': 415, 'fulfil': 1790, 'seri': 4012, 'condom': 920, 'label': 2451, 'soft': 4149, 'disadvantag': 1228, 'outspent': 3141, 'guess': 1935, 'blow': 464, 'confidenti': 925, 'sharron': 4037, 'angl': 177, 'spill': 4212, 'slush': 4124, 'rosen': 3862, 'monkey': 2855, 'inaugur': 2193, 'snowstorm': 4140, 'attach': 285, 'spare': 4195, 'vegan': 4767, 'starr': 4253, 'throw': 4524, 'innoc': 2244, 'rachel': 3583, 'zenzing': 4993, 'junket': 2383, 'porkbarrel': 3397, 'fundament': 1796, 'warhead': 4834, 'meant': 2721, 'approach': 223, 'tend': 4469, 'greec': 1917, 'metropolitan': 2766, 'lieuten': 2533, 'cigarett': 787, 'budgetrepair': 570, 'ala': 118, 'pinprick': 3336, 'intervent': 2281, 'residu': 3760, 'footprint': 1713, 'admir': 65, 'highvalu': 2062, 'deeper': 1119, 'gambl': 1811, 'buono': 583, 'diabet': 1201, 'cure': 1057, 'endang': 1443, 'swine': 4410, 'cattl': 684, 'unveil': 4725, 'honest': 2101, 'floor': 1694, 'gerald': 1849, 'daugherti': 1086, 'nondiscrimin': 2994, 'brookhaven': 551, 'unincorpor': 4700, 'mosquito': 2876, 'premis': 3449, 'candidaci': 633, 'manipul': 2653, 'koster': 2443, 'volunt': 4810, 'station': 4264, 'ninetyf': 2977, 'southwestern': 4187, 'desert': 1177, 'enjoy': 1456, 'demonstr': 1155, 'penc': 3264, 'divert': 1274, 'therapi': 4499, 'bureaucraci': 587, 'dave': 1088, 'brat': 522, 'philosoph': 3316, 'teddi': 4458, 'dovilla': 1316, 'brennan': 532, 'geograph': 1844, 'roe': 3845, 'wade': 4818, 'tick': 4529, 'cochair': 848, 'gore': 1894, 'suzann': 4404, 'bonamici': 482, 'epidem': 1478, 'sacrif': 3891, 'till': 4535, 'fraction': 1752, 'dock': 1283, 'mobil': 2836, 'twin': 4649, 'zuckerberg': 4999, 'struck': 4317, 'subsid': 4339, 'wildli': 4910, 'unpopular': 4717, 'resolv': 3764, 'doorstep': 1309, 'lifetim': 2537, 'loui': 2601, 'toothpast': 4559, 'fluorid': 1703, 'maryland': 2686, 'ralph': 3603, 'lifesav': 2536, 'reput': 3750, 'activist': 52, 'strangul': 4294, 'norm': 3003, 'wildlif': 4911, 'arctic': 233, 'dole': 1294, 'atheist': 278, 'denver': 1162, 'sidelin': 4071, 'frankel': 1760, 'shower': 4062, 'bathroom': 372, 'pell': 3259, 'absorb': 24, 'hama': 1968, 'thought': 4517, 'hillsborough': 2069, 'wear': 4864, 'sock': 4147, 'pleas': 3360, 'pipe': 3339, 'reus': 3782, 'ann': 182, 'mclane': 2716, 'kuster': 2448, 'boccieri': 475, 'stress': 4303, 'syndrom': 4417, 'seamen': 3967, 'ablebodi': 12, 'cohen': 853, 'policemen': 3378, 'adler': 62, 'trenton': 4611, 'blank': 457, 'encourag': 1441, 'rose': 3860, 'carolinian': 658, 'inconsist': 2204, 'interior': 2274, 'gmo': 1877, 'miscarriag': 2815, 'whatsoev': 4888, 'april': 227, 'cycl': 1064, 'aboard': 13, 'meanstest': 2720, 'oshaughnessi': 3117, 'jimmi': 2354, 'dimora': 1213, 'silberkrau': 4080, 'nake': 2913, 'degrad': 1134, 'stalk': 4245, 'apartheid': 206, 'scream': 3959, 'unfair': 4694, 'cliff': 821, 'politifact': 3382, 'valid': 4756, 'document': 1285, 'pave': 3239, 'guidelin': 1939, 'sixtytwo': 4105, 'enorm': 1458, 'languish': 2466, 'ireland': 2302, 'poland': 3374, 'raw': 3625, 'unless': 4710, 'poizner': 3373, 'corp': 997, 'hammer': 1970, 'broken': 549, 'teen': 4459, 'belief': 398, 'contrari': 971, 'saysberni': 3928, 'immun': 2177, 'surplus': 4387, 'proce': 3496, 'implic': 2183, 'truman': 4625, 'eas': 1373, 'septemb': 4008, 'cdc': 693, 'transmit': 4593, 'czech': 1066, 'fool': 1709, 'naval': 2928, 'appel': 213, 'proposit': 3527, 'discharg': 1235, 'yard': 4975, 'stolen': 4284, 'liquid': 2553, 'wellpaid': 4879, 'align': 133, 'halloween': 1964, 'escap': 1489, 'heavi': 2025, 'anonym': 187, 'wolf': 4939, 'impeach': 2179, 'asset': 268, 'deem': 1117, 'healthinsur': 2021, 'feasibl': 1615, 'radic': 3589, 'airlin': 110, 'ticket': 4530, 'retain': 3777, 'paulson': 3238, 'bernank': 416, 'orlando': 3114, 'fl': 1679, 'rock': 3842, 'sandwich': 3914, 'racist': 3585, 'goat': 1879, 'yesterday': 4981, 'hemispher': 2037, 'concret': 915, 'loughlin': 2600, 'gari': 1823, 'remodel': 3720, 'loung': 2604, 'dismantl': 1248, 'servicemen': 4017, 'divorc': 1278, 'robberi': 3837, 'annex': 183, 'overcrowd': 3145, 'drove': 1345, 'note': 3013, 'observ': 3044, 'span': 4193, 'gap': 1817, 'viciou': 4785, 'jerseyan': 2346, 'boozman': 492, 'exot': 1533, 'ninth': 2981, 'jone': 2369, 'oldest': 3074, 'sport': 4220, 'steel': 4271, 'jeopardi': 2342, 'wrongdo': 4970, 'breath': 530, 'stanford': 4250, 'tuberculosi': 4630, 'rel': 3705, 'spokesman': 4218, 'democraci': 1152, 'subsequ': 4338, 'andi': 174, 'liabil': 2519, 'hagan': 1954, 'musharraf': 2905, 'dip': 1219, 'wheel': 4889, 'garcia': 1820, 'wild': 4908, 'pharmacist': 3308, 'happi': 1980, 'worship': 4959, 'campus': 625, 'whop': 4899, 'marxist': 2685, 'sudden': 4348, 'dividend': 1276, 'anight': 180, 'broker': 550, 'entri': 1473, 'janitor': 2329, 'dine': 1214, 'weigh': 4874, 'distinguish': 1264, 'headtohead': 2015, 'malaysia': 2641, 'gate': 1827, 'catch': 680, 'maverick': 2702, 'summer': 4358, 'stockpil': 4282, 'non': 2990, 'swap': 4405, 'bow': 504, 'bergdahl': 414, 'automak': 309, 'luther': 2614, 'outperform': 3135, 'princ': 3477, 'personhood': 3295, 'conform': 931, 'sat': 3919, 'reinvest': 3703, 'overturn': 3159, 'vs': 4817, 'concept': 909, 'sierra': 4073, 'sestak': 4019, 'appropri': 224, 'rescu': 3753, 'marshal': 2678, 'employersponsor': 1435, 'mansion': 2656, 'bribe': 537, 'wide': 4903, 'nick': 2970, 'digit': 1209, 'contact': 960, 'propon': 3524, 'suddenli': 4349, 'imag': 2172, 'anchor': 172, 'lobster': 2571, 'ghetto': 1853, 'emori': 1432, 'sole': 4154, 'certifi': 710, 'motiv': 2882, 'drown': 1346, 'regardless': 3688, 'donna': 1304, 'campbel': 623, 'wic': 4902, 'aspect': 260, 'climb': 824, 'cocacola': 846, 'compet': 889, 'pillowtex': 3330, 'devast': 1193, 'substanti': 4343, 'pilot': 3331, 'simpsonbowl': 4086, 'urg': 4737, 'aaron': 1, 'bean': 379, 'drugrel': 1348, 'surfac': 4380, 'hunger': 2139, 'thanksgiv': 4492, 'heck': 2027, 'layoff': 2485, 'shale': 4032, 'shred': 4064, 'understood': 4687, 'grandchildren': 1909, 'freez': 1773, 'walton': 4828, 'deleg': 1141, 'fla': 1680, 'obrien': 3043, 'dispensari': 1253, 'gao': 1816, 'crackdown': 1025, 'dean': 1098, 'cannon': 635, 'haridopolo': 1992, 'castl': 675, 'coon': 985, 'biggovern': 435, 'proof': 3518, 'speed': 4206, 'vehicular': 4770, 'endur': 1447, 'transvagin': 4597, 'prep': 3452, 'strau': 4297, 'ricci': 3800, 'putnam': 3566, 'skill': 4107, 'sealevel': 3966, 'inequ': 2224, 'wash': 4842, 'flu': 1701, 'simpson': 4085, 'replic': 3741, 'diplomat': 1221, 'glass': 1871, 'strongest': 4315, 'simultan': 4087, 'congresswoman': 938, 'closur': 833, 'aka': 115, 'primarili': 3475, 'melt': 2740, 'attrit': 294, 'hp': 2130, 'saint': 3900, 'scrap': 3958, 'tighten': 4534, 'uncontest': 4676, 'parttim': 3212, 'hemp': 2038, 'seal': 3965, 'ipod': 2296, 'dolphin': 1296, 'plale': 3348, 'tag': 4424, 'avoid': 315, 'scholar': 3946, 'eightyf': 1399, 'rebuilt': 3643, 'boy': 508, 'jane': 2328, 'norton': 3010, 'freshmen': 1778, 'doyl': 1322, 'guinta': 1941, 'role': 3848, 'pattern': 3233, 'maintain': 2636, 'circl': 789, 'confer': 923, 'landrieu': 2461, 'pacif': 3171, 'appalachian': 209, 'revel': 3785, 'root': 3859, 'forg': 1726, 'ventur': 4773, 'urban': 4736, 'massag': 2690, 'savannah': 3924, 'prevail': 3467, 'huey': 2134, 'environ': 1474, 'recus': 3664, 'solicitor': 4155, 'rang': 3609, 'margaret': 2666, 'touch': 4564, 'prostitut': 3534, 'intent': 2271, 'uwmadison': 4749, 'lab': 2450, 'surtax': 4392, 'engin': 1452, 'ordin': 3105, 'restroom': 3775, 'reclaim': 3652, 'whove': 4900, 'uncollect': 4673, 'rev': 3783, 'wright': 4966, 'confin': 926, 'twentythre': 4647, 'startup': 4255, 'narrow': 2918, 'compris': 902, 'multimillionair': 2896, 'sticker': 4278, 'helmet': 2035, 'harder': 1987, 'arena': 235, 'dick': 1203, 'cheney': 752, 'exploit': 1546, 'duggar': 1356, 'unsustain': 4722, 'libyan': 2527, 'versu': 4778, 'hardi': 1989, 'columbiahca': 863, 'coffer': 850, 'mankind': 2654, 'safest': 3896, 'ben': 404, 'carson': 664, 'consequ': 944, 'globe': 1875, 'precis': 3435, 'transact': 4588, 'arsen': 250, 'seiz': 3991, 'elderli': 1403, 'underemploy': 4679, 'wake': 4823, 'devic': 1195, 'unsuccess': 4721, 'congest': 933, 'slate': 4114, 'infus': 2235, 'cayman': 691, 'supposedli': 4374, 'endow': 1446, 'hardli': 1990, 'forfeit': 1724, 'hyde': 2152, 'championship': 726, 'payout': 3246, 'stopandfrisk': 4288, 'widget': 4905, 'joint': 2364, 'infract': 2233, 'ncaa': 2933, 'jesu': 2348, 'christ': 774, 'leak': 2490, 'prospect': 3530, 'legitim': 2502, 'widespread': 4904, 'visibl': 4801, 'waukesha': 4854, 'monday': 2849, 'longterm': 2585, 'marilinda': 2670, 'twoyear': 4655, 'basebal': 365, 'pink': 3335, 'poni': 3386, 'contest': 963, 'hank': 1978, 'boycott': 510, 'martinez': 2683, 'stuck': 4320, 'chicken': 757, 'outstrip': 3143, 'tide': 4531, 'eclips': 1382, 'proabort': 3492, 'era': 1483, 'naral': 2915, 'reproduct': 3745, 'hill': 2066, 'binat': 441, 'concern': 910, 'listen': 2556, 'exclud': 1524, 'fetu': 1639, 'leg': 2498, 'kick': 2416, 'harvest': 1998, 'entrepreneurship': 1472, 'overnight': 3150, 'hair': 1956, 'beaten': 382, 'jurisdict': 2385, 'vagina': 4755, 'madigan': 2623, 'appeal': 211, 'merri': 2757, 'marwig': 2684, 'architect': 232, 'federalbudget': 1620, 'bu': 562, 'selfdefens': 3994, 'lynn': 2618, 'republicancontrol': 3748, 'peer': 3258, 'hovd': 2126, 'infect': 2226, 'acid': 42, 'perhap': 3281, 'whod': 4897, 'properli': 3522, 'hobbi': 2082, 'stuff': 4326, 'decrimin': 1114, 'motorist': 2885, 'bass': 370, 'chain': 717, 'broadband': 546, 'telecom': 4461, 'spain': 4192, 'patterson': 3234, 'mob': 2835, 'greenlight': 1920, 'pinella': 3334, 'fred': 1766, 'injur': 2239, 'rarer': 3617, 'er': 1482, 'amash': 156, 'upcom': 4727, 'biennium': 431, 'balancedbudget': 337, 'chairmanship': 720, 'ozon': 3167, 'disneyland': 1251, 'saysther': 3932, 'unexpect': 4693, 'chair': 718, 'prematur': 3448, 'afscm': 91, 'forum': 1738, 'prompt': 3517, 'anticip': 196, 'supremacist': 4377, 'haslam': 2000, 'slaughter': 4115, 'extent': 1556, 'southeast': 4182, 'facilit': 1569, 'rake': 3601, 'hack': 1952, 'enabl': 1438, 'knowledg': 2438, 'shadi': 4030, 'obviou': 3047, 'oct': 3054, 'sullivan': 4357, 'survivor': 4396, 'lilli': 2543, 'weak': 4858, 'afterward': 92, 'perstud': 3297, 'ranch': 3606, 'tort': 4561, 'morgan': 2867, 'carrol': 663, 'alan': 120, 'nye': 3034, 'malaria': 2640, 'dorm': 1310, 'notifi': 3016, 'midnight': 2783, 'map': 2660, 'gerrymand': 1852, 'outreach': 3138, 'lid': 2529, 'pa': 3168, 'german': 1850, 'nazi': 2931, 'resist': 3762, 'amendmentban': 163, 'neighbor': 2947, 'abid': 8, 'caution': 689, 'puddl': 3553, 'squander': 4229, 'texasmexico': 4487, 'apprehens': 222, 'ford': 1717, 'wave': 4855, 'surrend': 4389, 'goldplat': 1886, 'reinstat': 3702, 'elabor': 1402, 'hollywood': 2090, 'miracl': 2812, 'marci': 2664, 'kaptur': 2394, 'denni': 1159, 'kucinich': 2446, 'smog': 4133, 'barbecu': 351, 'gambler': 1812, 'bend': 407, 'brad': 514, 'schimel': 3944, 'index': 2215, 'ellen': 1415, 'destabil': 1183, 'threequart': 4521, 'boyfriend': 511, 'tenth': 4473, 'megan': 2737, 'perceiv': 3274, 'alexi': 128, 'reimburs': 3700, 'tb': 4446, 'vacant': 4752, 'lloyd': 2565, 'doggett': 1292, 'merger': 2754, 'perimet': 3282, 'tout': 4572, 'downturn': 1321, 'ar': 228, 'brian': 536, 'rasmussen': 3618, 'australia': 305, 'thirtyfour': 4513, 'favorit': 1609, 'yearli': 4977, 'outofpocket': 3132, 'riot': 3821, 'represent': 3744, 'select': 3993, 'embrac': 1423, 'ahmadinejad': 103, 'bachelor': 322, 'thereaft': 4500, 'nearbi': 2936, 'angri': 179, 'extrem': 1561, 'referr': 3679, 'notori': 3018, 'therel': 4501, 'cuban': 1051, 'ampl': 168, 'meltdown': 2741, 'disrespect': 1260, 'outag': 3122, 'wisconsinmilwauke': 4929, 'guid': 1938, 'worldclass': 4954, 'defi': 1128, 'standdown': 4249, 'son': 4167, 'boot': 491, 'dark': 1080, 'dewin': 1199, 'sewag': 4025, 'eri': 1485, 'header': 2012, 'foot': 1710, 'calcul': 612, 'sort': 4175, 'webb': 4867, 'reliabl': 3710, 'ms': 2893, 'mopac': 2864, 'spew': 4209, 'discret': 1240, 'context': 964, 'buyback': 604, 'surgic': 4384, 'pm': 3368, 'bird': 445, 'presenc': 3457, 'occident': 3049, 'meanwhil': 2722, 'athlet': 279, 'dire': 1222, 'bet': 421, 'kicker': 2417, 'asianamerican': 257, 'romneyryan': 3852, 'carl': 653, 'dodg': 1288, 'auditor': 298, 'oreilli': 3110, 'nd': 2934, 'renam': 3724, 'beltlin': 403, 'marta': 2679, 'olsen': 3076, 'outmigr': 3130, 'grayson': 1913, 'retent': 3778, 'cnbc': 837, 'denmark': 1158, 'renaiss': 3723, 'overtim': 3158, 'passiv': 3220, 'advocaci': 78, 'dodd': 1286, 'conspiraci': 949, 'abramoff': 18, 'upgrad': 4729, 'duplic': 1360, 'cloth': 834, 'doug': 1314, 'deserv': 1178, 'rauner': 3624, 'foreignborn': 1721, 'frustrat': 1786, 'carpet': 659, 'dildo': 1210, 'forprofit': 1731, 'extraordinari': 1560, 'flip': 1691, 'hartsfieldjackson': 1996, 'censor': 698, 'whine': 4891, 'tire': 4542, 'rein': 3701, 'window': 4919, 'msnbc': 2894, 'movement': 2889, 'friday': 1780, 'exit': 1531, 'spiritu': 4214, 'atm': 282, 'surcharg': 4378, 'expos': 1551, 'usda': 4741, 'furiou': 1800, 'zombi': 4997, 'bit': 452, 'actor': 53, 'crct': 1030, 'sanger': 3916, 'vow': 4816, 'emit': 1430, 'rutger': 3887, 'hewent': 2047, 'pioneer': 3337, 'lowwag': 2611, 'kentuckian': 2410, 'bashar': 366, 'salem': 3903, 'membership': 2743, 'malfeas': 2643, 'movi': 2890, 'endoflif': 1444, 'blunt': 469, 'insert': 2249, 'giant': 1856, 'philip': 3313, 'morri': 2872, 'girlfriend': 1866, 'mock': 2837, 'heller': 2034, 'inelig': 2223, 'acknowledg': 43, 'shun': 4066, 'matt': 2700, 'meg': 2736, 'memo': 2744, 'byrn': 606, 'recidiv': 3649, 'regularli': 3697, 'fourthlowest': 1747, 'chosen': 772, 'buffalo': 572, 'andrew': 175, 'unleash': 4709, 'extort': 1558, 'wood': 4945, 'joke': 2365, 'trial': 4613, 'harvey': 1999, 'oswald': 3118, 'hotlin': 2119, 'abc': 6, 'corner': 994, 'aronberg': 246, 'numberon': 3026, 'fear': 1614, 'avid': 314, 'piraci': 3341, 'mix': 2832, 'unwant': 4726, 'bachmann': 323, 'coalit': 842, 'toxic': 4575, 'climatechang': 823, 'chevron': 755, 'wiretap': 4926, 'brief': 539, 'besid': 419, 'stimul': 4279, 'germani': 1851, 'dem': 1150, 'louie': 2602, 'gohmert': 1883, 'theater': 4494, 'satellit': 3920, 'environmentalist': 1476, 'wolv': 4940, 'onetim': 3088, 'wasgovernor': 4841, 'votingag': 4814, 'cincinnati': 788, 'sellwood': 3997, 'wi': 4901, 'obsolet': 3045, 'mainstream': 2635, 'dinner': 1215, 'idaho': 2159, 'christin': 777, 'allig': 143, 'viewer': 4792, 'bat': 371, 'mess': 2759, 'pardon': 3195, 'slam': 4112, 'hivaid': 2078, 'greatest': 1916, 'unequivoc': 4692, 'judgment': 2375, 'robin': 3839, 'manual': 2657, 'gear': 1833, 'wedlock': 4871, 'stone': 4285, 'exercis': 1528, 'deriv': 1174, 'delici': 1144, 'theyll': 4503, 'overweight': 3160, 'jew': 2350, 'rooftop': 3856, 'donaldtrump': 1302, 'fifthhighest': 1646, 'condo': 919, 'allegedli': 139, 'perpetr': 3288, 'priorit': 3482, 'kkk': 2431, 'johnni': 2361, 'translat': 4592, 'shape': 4033, 'hotelmotel': 2118, 'prize': 3490, 'battleground': 375, 'london': 2579, 'mere': 2753, 'factcheck': 1571, 'roadway': 3834, 'pine': 3333, 'conclus': 914, 'unsaf': 4720, 'egg': 1393, 'disastr': 1234, 'mutual': 2909, 'firemen': 1669, 'soro': 4173, 'bioweapon': 443, 'nc': 2932, 'marti': 2680, 'stuf': 4325, 'birmingham': 446, 'rider': 3811, 'hunter': 2141, 'herd': 2041, 'hoosier': 2105, 'exxonmobil': 1564, 'fit': 1676, 'bankrol': 344, 'alberta': 124, 'darl': 1081, 'entranc': 1471, 'rip': 3822, 'abject': 10, 'bloomberg': 463, 'presidentelect': 3461, 'temporarili': 4468, 'galveston': 1810, 'nh': 2968, 'tougher': 4567, 'inner': 2242, 'hazard': 2009, 'assert': 266, 'playoff': 3359, 'strength': 4301, 'motion': 2881, 'surgeon': 4382, 'hypocrit': 2155, 'afraid': 87, 'kit': 2429, 'pope': 3393, 'franci': 1757, 'kingston': 2427, 'budweis': 571, 'leticia': 2512, 'putt': 3567, 'swim': 4409, 'fairli': 1580, 'breakfast': 528, 'doughnut': 1315, 'percapita': 3273, 'buri': 590, 'staunch': 4268, 'furnitur': 1802, 'publicsector': 3551, 'gillnet': 1860, 'voic': 4807, 'iphon': 2295, 'dubiou': 1352, 'debunk': 1104, 'nusbaum': 3030, 'asia': 256, 'bloat': 459, 'homestead': 2097, 'yorker': 4983, 'teller': 4465, 'peac': 3250, 'juri': 2384, 'specul': 4204, 'holton': 2092, 'yemen': 4980, 'fincher': 1660, 'genit': 1841, 'haven': 2005, 'coalfir': 841, 'fortyseven': 1736, 'wholli': 4898, 'columnist': 865, 'absent': 21, 'sea': 3964, 'ratepay': 3620, 'distract': 1266, 'copay': 989, 'solitari': 4157, 'infest': 2227, 'fernandez': 1636, 'migrant': 2786, 'rothstein': 3864, 'brotherhood': 553, 'taxat': 4442, 'evolut': 1512, 'dioxid': 1218, 'contin': 965, 'redesign': 3668, 'backroom': 326, 'longstand': 2584, 'drainag': 1328, 'ditch': 1271, 'navig': 2930, 'canal': 629, 'oath': 3036, 'outliv': 3129, 'concert': 911, 'contend': 962, 'broader': 547, 'irregular': 2304, 'reviv': 3789, 'saysth': 3931, 'metric': 2764, 'unfortun': 4696, 'rivera': 3830, 'percentincreas': 3276, 'heterosexu': 2046, 'strict': 4306, 'jumpstart': 2380, 'ny': 3032, 'compil': 892, 'stalin': 4244, 'hardcor': 1985, 'wikipedia': 4907, 'chafe': 715, 'pam': 3184, 'portug': 3408, 'smallest': 4128, 'workman': 4950, 'ibm': 2156, 'aplu': 207, 'fourthgrad': 1746, 'fiftyseven': 1648, 'complex': 896, 'heritag': 2042, 'porkfil': 3398, 'virginian': 4797, 'swiss': 4412, 'disproportion': 1257, 'pediatrician': 3256, 'proxim': 3546, 'thc': 4493, 'mcginti': 2715, 'alarm': 121, 'misinform': 2818, 'imperson': 2180, 'liberti': 2523, 'dinosaur': 1216, 'flatout': 1683, 'constant': 950, 'suv': 4403, 'citru': 798, 'delawar': 1139, 'algeria': 130, 'oman': 3078, 'moor': 2863, 'withdrawn': 4934, 'deliveri': 1147, 'juic': 2377, 'rita': 3827, 'penal': 3262, 'libertarian': 2522, 'brew': 533, 'screwdriv': 3961, 'mubarak': 2895, 'headlin': 2013, 'progun': 3510, 'scenario': 3941, 'chart': 739, 'barr': 357, 'titl': 4543, 'orlean': 3115, 'patti': 3235, 'murray': 2903, 'paula': 3237, 'creflo': 1035, 'windfal': 4917, 'circuit': 790, 'tier': 4533, 'pow': 3422, 'symbol': 4416, 'lessen': 2510, 'convert': 980, 'empti': 1437, 'mph': 2891, 'austinit': 304, 'behead': 397, 'bing': 442, 'moammar': 2834, 'gadhafi': 1806, 'refinanc': 3681, 'uphold': 4731, 'impress': 2187, 'swift': 4408, 'burdensom': 585, 'glasssteagal': 1872, 'ak': 114, 'hust': 2148, 'fat': 1604, 'jam': 2325, 'belt': 402, 'aaa': 0, 'deepwat': 1120, 'excis': 1523, 'persist': 3293, 'valuabl': 4759, 'ukrainian': 4663, 'absente': 22, 'sixteen': 4100, 'postag': 3414, 'lord': 2591, 'penitentiari': 3266, 'dispos': 1256, 'neglect': 2945, 'underperform': 4685, 'wise': 4930, 'discourag': 1238, 'doe': 1289, 'drunken': 1350, 'baker': 335, 'incometax': 2202, 'proudli': 3540, 'compon': 900, 'lightn': 2540, 'inperson': 2246, 'liberia': 2521, 'mall': 2644, 'sanford': 3915, 'feral': 1634, 'modest': 2842, 'oneterm': 3086, 'abroad': 19, 'publicemploye': 3549, 'farmland': 1596, 'prototyp': 3538, 'garag': 1818, 'wrongli': 4971, 'withheld': 4936, 'mickey': 2777, 'mous': 2888, 'sidewalk': 4072, 'definit': 1131, 'extinct': 1557, 'tenet': 4470, 'notic': 3015, 'anderson': 173, 'taliban': 4428, 'loos': 2589, 'season': 3970, 'smoker': 4135, 'elizabeth': 1414, 'renov': 3729, 'blueberri': 466, 'cousin': 1020, 'happili': 1981, 'custodi': 1060, 'cumberland': 1053, 'pedophil': 3257, 'veget': 4768, 'scout': 3957, 'junk': 2382, 'lgbt': 2518, 'fudg': 1787, 'runaway': 3877, 'sneez': 4137, 'bishop': 451, 'affluent': 83, 'absenc': 20, 'olen': 3075, 'plate': 3354, 'unusu': 4724, 'sky': 4110, 'airstrik': 113, 'raila': 3595, 'odinga': 3057, 'rocki': 3844, 'sixtyon': 4104, 'merkley': 2756, 'sadli': 3893, 'hooter': 2106, 'tab': 4421, 'nsa': 3022, 'furlough': 1801, 'fema': 1631, 'specifi': 4202, 'oral': 3102, 'antonin': 197, 'scalia': 3935, 'ranger': 3610, 'beyonc': 425, 'schip': 3945, 'singlepar': 4091, 'cairo': 611, 'mom': 2847, 'ugli': 4660, 'incandesc': 2194, 'ninetysix': 2980, 'dna': 1280, 'brewpub': 535, 'paperwork': 3193, 'mount': 2886, 'rushmor': 3883, 'hardest': 1988, 'highrisk': 2060, 'inconveni': 2205, 'regress': 3694, 'cart': 665, 'megyn': 2738, 'monica': 2853, 'rezko': 3794, 'phd': 3310, 'stake': 4243, 'barnett': 356, 'horrif': 2112, 'unauthor': 4671, 'comptrol': 904, 'alloc': 144, 'roof': 3855, 'neumann': 2954, 'distort': 1265, 'vital': 4804, 'distinct': 1263, 'astronaut': 275, 'dispar': 1252, 'loud': 2598, 'pile': 3328, 'wine': 4920, 'click': 819, 'pro': 3491, 'precondit': 3436, 'anthoni': 191, 'highestpaid': 2056, 'sox': 4189, 'concuss': 916, 'handout': 1976, 'disney': 1250, 'boomer': 488, 'swing': 4411, 'preach': 3430, 'palm': 3183, 'pip': 3338, 'jihadist': 2352, 'melania': 2739, 'geopolit': 1845, 'beneath': 408, 'finest': 1662, 'gentrifi': 1843, 'omit': 3079, 'holi': 2087, 'unnecessari': 4714, 'chase': 741, 'pig': 3327, 'weinstein': 4876, 'parrish': 3201, 'inspector': 2253, 'threw': 4523, 'hat': 2003, 'onefourth': 3082, 'interim': 2273, 'articl': 253, 'prosser': 3532, 'prejudg': 3445, 'rosemari': 3861, 'lehmberg': 2503, 'pluto': 3367, 'astronom': 276, 'sun': 4360, 'frozen': 1785, 'parad': 3194, 'grader': 1905, 'phase': 3309, 'duckworth': 1354, 'confid': 924, 'swat': 4406, 'tragic': 4585, 'rnc': 3831, 'stove': 4292, 'flower': 1699, 'smear': 4131, 'necessarili': 2940, 'sixthhighest': 4102, 'mitig': 2829, 'cleanup': 813, 'pamphlet': 3185, 'canadianstyl': 628, 'uncl': 4672, 'mae': 2625, 'classmat': 807, 'artifici': 254, 'aunt': 301, 'conven': 977, 'sister': 4096, 'advic': 73, 'raes': 3591, 'lightrail': 2541, 'nonunion': 3000, 'bushcheney': 597, 'cafe': 608, 'fahrenheit': 1576, 'thurmond': 4527, 'bull': 578, 'ambiti': 161, 'nightmar': 2973, 'moham': 2844, 'nonprolifer': 2998, 'fiftythre': 1649, 'british': 544, 'delphi': 1148, 'childless': 763, 'linn': 2552, 'slowest': 4123, 'undergradu': 4682, 'denounc': 1160, 'ideolog': 2164, 'ribboncut': 3799, 'messag': 2760, 'halv': 1966, 'giveaway': 1868, 'watersh': 4852, 'withhold': 4937, 'pistol': 3342, 'herman': 2043, 'cain': 610, 'metrorail': 2767, 'suppress': 4375, 'techniqu': 4455, 'backward': 327, 'ham': 1967, 'ceas': 694, 'trooper': 4621, 'shovel': 4061, 'volcano': 4808, 'embroil': 1424, 'outnumb': 3131, 'boardwalk': 471, 'nasa': 2919, 'portray': 3407, 'respiratori': 3768, 'overrun': 3152, 'amazon': 158, 'noncollegeeduc': 2992, 'displac': 1254, 'song': 4168, 'outlet': 3127, 'theorist': 4498, 'soy': 4190, 'chrysler': 781, 'atom': 284, 'spotti': 4222, 'destin': 1184, 'smallbusi': 4126, 'persuad': 3298, 'skin': 4108, 'chad': 714, 'milken': 2793, 'topperform': 4560, 'nuke': 3024, 'pollster': 3384, 'livingwag': 2563, 'scare': 3940, 'blackberri': 454, 'huntsman': 2142, 'rabid': 3581, 'scuttl': 3963, 'speci': 4198, 'meaning': 2719, 'promin': 3514, 'fell': 1626, 'wand': 4829, 'ottumwa': 3120, 'wayn': 4857, 'smallpox': 4129, 'exhaust': 1529, 'montana': 2858, 'isil': 2308, 'insult': 2265, 'vari': 4762, 'frisk': 1782, 'bound': 502, 'buckybal': 566, 'applaud': 215, 'radar': 3587, 'bogu': 479, 'slice': 4118, 'bread': 526, 'playground': 3358, 'lizbeth': 2564, 'benacquisto': 405, 'box': 506, 'juvenil': 2389, 'emmi': 1431, 'cooki': 983, 'sugari': 4353, 'lifelong': 2535, 'enthusiast': 1466, 'arizonastyl': 240, 'vendor': 4771, 'diminish': 1212, 'motherfriendli': 2880, 'childbirth': 760, 'persian': 3292, 'naacp': 2911, 'beaverton': 383, 'flown': 1700, 'robber': 3836, 'jeremiah': 2343, 'brigad': 540, 'denier': 1157, 'gwen': 1948, 'narragansett': 2917, 'charlott': 738, 'hung': 2138, 'mercedesbenz': 2751, 'nyc': 3033, 'sight': 4074, 'pretend': 3465, 'constitution': 954, 'deterr': 1191, 'broun': 555, 'cascad': 668, 'owen': 3164, 'twopar': 4652, 'monogam': 2856, 'hourli': 2122, 'twotoon': 4654, 'scandalplagu': 3939, 'niagara': 2969, 'profess': 3502, 'fugit': 1789, 'giannoulia': 1855, 'fan': 1589, 'pepper': 3272, 'southsid': 4185, 'boyc': 509, 'fake': 1582, 'health': 2016, 'singleus': 4093, 'divest': 1275, 'fishermen': 1675, 'resort': 3765, 'enhanc': 1455, 'clap': 803, 'bulldoz': 579, 'caviti': 690, 'withdrew': 4935, 'assign': 269, 'arthur': 252, 'substant': 4342, 'musician': 2907, 'rainforest': 3598, 'stupak': 4327, 'nike': 2974, 'calibr': 615, 'exception': 1520, 'alzheim': 154, 'migrat': 2787, 'airbu': 108, 'pin': 3332, 'reconstruct': 3656, 'clearanc': 815, 'fractur': 1753}\n",
            "[8.84690279 9.35772842 8.84690279 ... 9.35772842 7.97143406 9.35772842]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "numpy.ndarray"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k5PzG5rSEHEH",
        "colab_type": "code",
        "outputId": "478930a4-7890-4aca-f457-3fd982429389",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "xtrain_tfv_sum.shape, xtest_tfv_sum.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((11522, 5000), (1266, 5000))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LaTCj_onsLiH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train = np.concatenate((x, xtrain_tfv_sum, rest_df_train),axis = 1)\n",
        "x_test = np.concatenate((x1, xtest_tfv_sum, rest_df_test),axis = 1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RLZqbgHtDqB-",
        "colab_type": "code",
        "outputId": "ce223dfd-326b-4687-94d1-470f34c4a799",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x_train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(11522, 6161)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JE1yYVOnDtr3",
        "colab_type": "code",
        "outputId": "cf369a9c-5dd1-4c75-99a4-00a458fcb372",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x_test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1266, 6161)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WVK2_NfKDtxB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Fitting Naive Bayes to the Training set\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "classifier1 = GaussianNB()\n",
        "classifier1.fit(x_train, y_train)\n",
        "y_pred1 = classifier1.predict(x_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KzUVVx7sTgdy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.naive_bayes import MultinomialNB\n",
        "clf = MultinomialNB()\n",
        "clf.fit(x_train, y_train)\n",
        "y_pred2 = clf.predict(x_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dsmCMuVvsLaY",
        "colab_type": "code",
        "outputId": "41fcd0a0-c3a7-4948-bf19-95710aa13139",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 72
        }
      },
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "classifier=LogisticRegression(solver='lbfgs',multi_class='multinomial',random_state=0,C=10)\n",
        "classifier.fit(x_train,y_train)\n",
        "y_pred = classifier.predict(x_test)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/linear_model/logistic.py:947: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
            "  \"of iterations.\", ConvergenceWarning)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WgC3mKhaeuLs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ans = labelencoder1.inverse_transform(y_pred)\n",
        "type(ans)\n",
        "ans = pd.DataFrame(ans)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5ZT18RC4e7MA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "id1=test.loc[:,['ID']]\n",
        "final_ans = [id1, ans]\n",
        "final_ans = pd.concat(final_ans, axis=1)\n",
        "final_ans.columns = ['ID', 'Label']\n",
        "final_ans.to_csv('NLPv1.csv',index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u-JaHYWhfJPJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "######################################################################################################"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kM-YH_1l-u8d",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}