Machine Learning Classification on Successful StartUp Finale.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Machine Learning - Classification on Successfull Start Up Finale v.2",
      "provenance": [],
      "collapsed_sections": [
        "akYpe1zb0JBP"
      ],
      "toc_visible": true,
      "mount_file_id": "1S0W879Vix4XzeNvricjZH6lRMmjvfSp3",
      "authorship_tag": "ABX9TyMKngxWGtJsyZkayVi5ANtI",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/fitrieratna/DataScience-Portfolio/blob/master/Machine%20Learning%20Classification%20on%20Successful%20StartUp%20Finale.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "akYpe1zb0JBP"
      },
      "source": [
        "# Pre-Explore"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oGYmxGZ1weeV"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "import chardet\n"
      ],
      "execution_count": 184,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fD1_hjcC9KUV",
        "outputId": "e2bf8a6d-addb-4bed-f94c-a9cdc420cd37"
      },
      "source": [
        "# To connect Google Drive (GDrive) with Colab\n",
        "from google.colab import drive\n",
        "drive.mount(\"/content/gdrive\")"
      ],
      "execution_count": 185,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tpx7ejBQx-cm",
        "outputId": "46316045-bb02-4905-f709-934c048bf19e"
      },
      "source": [
        "f = open('/content/gdrive/My Drive/data-analysis-science/investments_VC.csv','rb')\n",
        "line = f.readline()\n",
        "print(chardet.detect(line))"
      ],
      "execution_count": 186,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'encoding': 'ascii', 'confidence': 1.0, 'language': ''}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CR0JUSVBxXOW"
      },
      "source": [
        "data = pd.read_csv('/content/gdrive/My Drive/data-analysis-science/investments_VC.csv', encoding='unicode_escape' )"
      ],
      "execution_count": 187,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HKcy8wKsx85e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "outputId": "48b8bacd-525b-46f9-b922-9be6b616dee4"
      },
      "source": [
        "data.head(10)"
      ],
      "execution_count": 188,
      "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>permalink</th>\n",
              "      <th>name</th>\n",
              "      <th>homepage_url</th>\n",
              "      <th>category_list</th>\n",
              "      <th>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>state_code</th>\n",
              "      <th>region</th>\n",
              "      <th>city</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_at</th>\n",
              "      <th>founded_month</th>\n",
              "      <th>founded_quarter</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>first_funding_at</th>\n",
              "      <th>last_funding_at</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>/organization/waywire</td>\n",
              "      <td>#waywire</td>\n",
              "      <td>http://www.waywire.com</td>\n",
              "      <td>|Entertainment|Politics|Social Media|News|</td>\n",
              "      <td>News</td>\n",
              "      <td>17,50,000</td>\n",
              "      <td>acquired</td>\n",
              "      <td>USA</td>\n",
              "      <td>NY</td>\n",
              "      <td>New York City</td>\n",
              "      <td>New York</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012-06-01</td>\n",
              "      <td>2012-06</td>\n",
              "      <td>2012-Q2</td>\n",
              "      <td>2012.0</td>\n",
              "      <td>2012-06-30</td>\n",
              "      <td>2012-06-30</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>/organization/tv-communications</td>\n",
              "      <td>&amp;TV Communications</td>\n",
              "      <td>http://enjoyandtv.com</td>\n",
              "      <td>|Games|</td>\n",
              "      <td>Games</td>\n",
              "      <td>40,00,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>CA</td>\n",
              "      <td>Los Angeles</td>\n",
              "      <td>Los Angeles</td>\n",
              "      <td>2.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2010-06-04</td>\n",
              "      <td>2010-09-23</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>/organization/rock-your-paper</td>\n",
              "      <td>'Rock' Your Paper</td>\n",
              "      <td>http://www.rockyourpaper.org</td>\n",
              "      <td>|Publishing|Education|</td>\n",
              "      <td>Publishing</td>\n",
              "      <td>40,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>EST</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Tallinn</td>\n",
              "      <td>Tallinn</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012-10-26</td>\n",
              "      <td>2012-10</td>\n",
              "      <td>2012-Q4</td>\n",
              "      <td>2012.0</td>\n",
              "      <td>2012-08-09</td>\n",
              "      <td>2012-08-09</td>\n",
              "      <td>40000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>/organization/in-touch-network</td>\n",
              "      <td>(In)Touch Network</td>\n",
              "      <td>http://www.InTouchNetwork.com</td>\n",
              "      <td>|Electronics|Guides|Coffee|Restaurants|Music|i...</td>\n",
              "      <td>Electronics</td>\n",
              "      <td>15,00,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>GBR</td>\n",
              "      <td>NaN</td>\n",
              "      <td>London</td>\n",
              "      <td>London</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2011-04-01</td>\n",
              "      <td>2011-04</td>\n",
              "      <td>2011-Q2</td>\n",
              "      <td>2011.0</td>\n",
              "      <td>2011-04-01</td>\n",
              "      <td>2011-04-01</td>\n",
              "      <td>1500000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>/organization/r-ranch-and-mine</td>\n",
              "      <td>-R- Ranch and Mine</td>\n",
              "      <td>NaN</td>\n",
              "      <td>|Tourism|Entertainment|Games|</td>\n",
              "      <td>Tourism</td>\n",
              "      <td>60,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>TX</td>\n",
              "      <td>Dallas</td>\n",
              "      <td>Fort Worth</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2014-01-01</td>\n",
              "      <td>2014-01</td>\n",
              "      <td>2014-Q1</td>\n",
              "      <td>2014.0</td>\n",
              "      <td>2014-08-17</td>\n",
              "      <td>2014-09-26</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>60000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>/organization/club-domains</td>\n",
              "      <td>.Club Domains</td>\n",
              "      <td>http://nic.club/</td>\n",
              "      <td>|Software|</td>\n",
              "      <td>Software</td>\n",
              "      <td>70,00,000</td>\n",
              "      <td>NaN</td>\n",
              "      <td>USA</td>\n",
              "      <td>FL</td>\n",
              "      <td>Ft. Lauderdale</td>\n",
              "      <td>Oakland Park</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2011-10-10</td>\n",
              "      <td>2011-10</td>\n",
              "      <td>2011-Q4</td>\n",
              "      <td>2011.0</td>\n",
              "      <td>2013-05-31</td>\n",
              "      <td>2013-05-31</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>/organization/fox-networks</td>\n",
              "      <td>.Fox Networks</td>\n",
              "      <td>http://www.dotfox.com</td>\n",
              "      <td>|Advertising|</td>\n",
              "      <td>Advertising</td>\n",
              "      <td>49,12,393</td>\n",
              "      <td>closed</td>\n",
              "      <td>ARG</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Buenos Aires</td>\n",
              "      <td>Buenos Aires</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2007-01-16</td>\n",
              "      <td>2007-01-16</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4912393.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>/organization/0-6-com</td>\n",
              "      <td>0-6.com</td>\n",
              "      <td>http://www.0-6.com</td>\n",
              "      <td>|Curated Web|</td>\n",
              "      <td>Curated Web</td>\n",
              "      <td>20,00,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2007-01-01</td>\n",
              "      <td>2007-01</td>\n",
              "      <td>2007-Q1</td>\n",
              "      <td>2007.0</td>\n",
              "      <td>2008-03-19</td>\n",
              "      <td>2008-03-19</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>/organization/004-technologies</td>\n",
              "      <td>004 Technologies</td>\n",
              "      <td>http://004gmbh.de/en/004-interact</td>\n",
              "      <td>|Software|</td>\n",
              "      <td>Software</td>\n",
              "      <td>-</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>IL</td>\n",
              "      <td>Springfield, Illinois</td>\n",
              "      <td>Champaign</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2010-01-01</td>\n",
              "      <td>2010-01</td>\n",
              "      <td>2010-Q1</td>\n",
              "      <td>2010.0</td>\n",
              "      <td>2014-07-24</td>\n",
              "      <td>2014-07-24</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>/organization/01games-technology</td>\n",
              "      <td>01Games Technology</td>\n",
              "      <td>http://www.01games.hk/</td>\n",
              "      <td>|Games|</td>\n",
              "      <td>Games</td>\n",
              "      <td>41,250</td>\n",
              "      <td>operating</td>\n",
              "      <td>HKG</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Hong Kong</td>\n",
              "      <td>Hong Kong</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2014-07-01</td>\n",
              "      <td>2014-07-01</td>\n",
              "      <td>41250.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                          permalink                name  ... round_G round_H\n",
              "0             /organization/waywire            #waywire  ...     0.0     0.0\n",
              "1   /organization/tv-communications  &TV Communications  ...     0.0     0.0\n",
              "2     /organization/rock-your-paper   'Rock' Your Paper  ...     0.0     0.0\n",
              "3    /organization/in-touch-network   (In)Touch Network  ...     0.0     0.0\n",
              "4    /organization/r-ranch-and-mine  -R- Ranch and Mine  ...     0.0     0.0\n",
              "5        /organization/club-domains       .Club Domains  ...     0.0     0.0\n",
              "6        /organization/fox-networks       .Fox Networks  ...     0.0     0.0\n",
              "7             /organization/0-6-com             0-6.com  ...     0.0     0.0\n",
              "8    /organization/004-technologies    004 Technologies  ...     0.0     0.0\n",
              "9  /organization/01games-technology  01Games Technology  ...     0.0     0.0\n",
              "\n",
              "[10 rows x 39 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 188
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JRC-X1KBxyCm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0337fb51-c83c-4c15-a53d-b31ef593ec30"
      },
      "source": [
        "data.info()"
      ],
      "execution_count": 189,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 54294 entries, 0 to 54293\n",
            "Data columns (total 39 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   permalink             49438 non-null  object \n",
            " 1   name                  49437 non-null  object \n",
            " 2   homepage_url          45989 non-null  object \n",
            " 3   category_list         45477 non-null  object \n",
            " 4    market               45470 non-null  object \n",
            " 5    funding_total_usd    49438 non-null  object \n",
            " 6   status                48124 non-null  object \n",
            " 7   country_code          44165 non-null  object \n",
            " 8   state_code            30161 non-null  object \n",
            " 9   region                44165 non-null  object \n",
            " 10  city                  43322 non-null  object \n",
            " 11  funding_rounds        49438 non-null  float64\n",
            " 12  founded_at            38554 non-null  object \n",
            " 13  founded_month         38482 non-null  object \n",
            " 14  founded_quarter       38482 non-null  object \n",
            " 15  founded_year          38482 non-null  float64\n",
            " 16  first_funding_at      49438 non-null  object \n",
            " 17  last_funding_at       49438 non-null  object \n",
            " 18  seed                  49438 non-null  float64\n",
            " 19  venture               49438 non-null  float64\n",
            " 20  equity_crowdfunding   49438 non-null  float64\n",
            " 21  undisclosed           49438 non-null  float64\n",
            " 22  convertible_note      49438 non-null  float64\n",
            " 23  debt_financing        49438 non-null  float64\n",
            " 24  angel                 49438 non-null  float64\n",
            " 25  grant                 49438 non-null  float64\n",
            " 26  private_equity        49438 non-null  float64\n",
            " 27  post_ipo_equity       49438 non-null  float64\n",
            " 28  post_ipo_debt         49438 non-null  float64\n",
            " 29  secondary_market      49438 non-null  float64\n",
            " 30  product_crowdfunding  49438 non-null  float64\n",
            " 31  round_A               49438 non-null  float64\n",
            " 32  round_B               49438 non-null  float64\n",
            " 33  round_C               49438 non-null  float64\n",
            " 34  round_D               49438 non-null  float64\n",
            " 35  round_E               49438 non-null  float64\n",
            " 36  round_F               49438 non-null  float64\n",
            " 37  round_G               49438 non-null  float64\n",
            " 38  round_H               49438 non-null  float64\n",
            "dtypes: float64(23), object(16)\n",
            "memory usage: 16.2+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gJJTgrzXAnkW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "81356057-f6b0-4922-a31c-503de86e9f9b"
      },
      "source": [
        "data.columns"
      ],
      "execution_count": 190,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['permalink', 'name', 'homepage_url', 'category_list', ' market ',\n",
              "       ' funding_total_usd ', 'status', 'country_code', 'state_code', 'region',\n",
              "       'city', 'funding_rounds', 'founded_at', 'founded_month',\n",
              "       'founded_quarter', 'founded_year', 'first_funding_at',\n",
              "       'last_funding_at', 'seed', 'venture', 'equity_crowdfunding',\n",
              "       'undisclosed', 'convertible_note', 'debt_financing', 'angel', 'grant',\n",
              "       'private_equity', 'post_ipo_equity', 'post_ipo_debt',\n",
              "       'secondary_market', 'product_crowdfunding', 'round_A', 'round_B',\n",
              "       'round_C', 'round_D', 'round_E', 'round_F', 'round_G', 'round_H'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 190
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vU1hLgNxyxKC",
        "outputId": "3c1e4332-a5db-47bd-c61c-ddb42bb85c15"
      },
      "source": [
        "missing_data = data.isnull()\n",
        "\n",
        "for column in missing_data.columns.values.tolist():\n",
        " print (column)\n",
        " print (missing_data[column].value_counts())\n",
        " print(\"\")  "
      ],
      "execution_count": 191,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "permalink\n",
            "False    49438\n",
            "True      4856\n",
            "Name: permalink, dtype: int64\n",
            "\n",
            "name\n",
            "False    49437\n",
            "True      4857\n",
            "Name: name, dtype: int64\n",
            "\n",
            "homepage_url\n",
            "False    45989\n",
            "True      8305\n",
            "Name: homepage_url, dtype: int64\n",
            "\n",
            "category_list\n",
            "False    45477\n",
            "True      8817\n",
            "Name: category_list, dtype: int64\n",
            "\n",
            " market \n",
            "False    45470\n",
            "True      8824\n",
            "Name:  market , dtype: int64\n",
            "\n",
            " funding_total_usd \n",
            "False    49438\n",
            "True      4856\n",
            "Name:  funding_total_usd , dtype: int64\n",
            "\n",
            "status\n",
            "False    48124\n",
            "True      6170\n",
            "Name: status, dtype: int64\n",
            "\n",
            "country_code\n",
            "False    44165\n",
            "True     10129\n",
            "Name: country_code, dtype: int64\n",
            "\n",
            "state_code\n",
            "False    30161\n",
            "True     24133\n",
            "Name: state_code, dtype: int64\n",
            "\n",
            "region\n",
            "False    44165\n",
            "True     10129\n",
            "Name: region, dtype: int64\n",
            "\n",
            "city\n",
            "False    43322\n",
            "True     10972\n",
            "Name: city, dtype: int64\n",
            "\n",
            "funding_rounds\n",
            "False    49438\n",
            "True      4856\n",
            "Name: funding_rounds, dtype: int64\n",
            "\n",
            "founded_at\n",
            "False    38554\n",
            "True     15740\n",
            "Name: founded_at, dtype: int64\n",
            "\n",
            "founded_month\n",
            "False    38482\n",
            "True     15812\n",
            "Name: founded_month, dtype: int64\n",
            "\n",
            "founded_quarter\n",
            "False    38482\n",
            "True     15812\n",
            "Name: founded_quarter, dtype: int64\n",
            "\n",
            "founded_year\n",
            "False    38482\n",
            "True     15812\n",
            "Name: founded_year, dtype: int64\n",
            "\n",
            "first_funding_at\n",
            "False    49438\n",
            "True      4856\n",
            "Name: first_funding_at, dtype: int64\n",
            "\n",
            "last_funding_at\n",
            "False    49438\n",
            "True      4856\n",
            "Name: last_funding_at, dtype: int64\n",
            "\n",
            "seed\n",
            "False    49438\n",
            "True      4856\n",
            "Name: seed, dtype: int64\n",
            "\n",
            "venture\n",
            "False    49438\n",
            "True      4856\n",
            "Name: venture, dtype: int64\n",
            "\n",
            "equity_crowdfunding\n",
            "False    49438\n",
            "True      4856\n",
            "Name: equity_crowdfunding, dtype: int64\n",
            "\n",
            "undisclosed\n",
            "False    49438\n",
            "True      4856\n",
            "Name: undisclosed, dtype: int64\n",
            "\n",
            "convertible_note\n",
            "False    49438\n",
            "True      4856\n",
            "Name: convertible_note, dtype: int64\n",
            "\n",
            "debt_financing\n",
            "False    49438\n",
            "True      4856\n",
            "Name: debt_financing, dtype: int64\n",
            "\n",
            "angel\n",
            "False    49438\n",
            "True      4856\n",
            "Name: angel, dtype: int64\n",
            "\n",
            "grant\n",
            "False    49438\n",
            "True      4856\n",
            "Name: grant, dtype: int64\n",
            "\n",
            "private_equity\n",
            "False    49438\n",
            "True      4856\n",
            "Name: private_equity, dtype: int64\n",
            "\n",
            "post_ipo_equity\n",
            "False    49438\n",
            "True      4856\n",
            "Name: post_ipo_equity, dtype: int64\n",
            "\n",
            "post_ipo_debt\n",
            "False    49438\n",
            "True      4856\n",
            "Name: post_ipo_debt, dtype: int64\n",
            "\n",
            "secondary_market\n",
            "False    49438\n",
            "True      4856\n",
            "Name: secondary_market, dtype: int64\n",
            "\n",
            "product_crowdfunding\n",
            "False    49438\n",
            "True      4856\n",
            "Name: product_crowdfunding, dtype: int64\n",
            "\n",
            "round_A\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_A, dtype: int64\n",
            "\n",
            "round_B\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_B, dtype: int64\n",
            "\n",
            "round_C\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_C, dtype: int64\n",
            "\n",
            "round_D\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_D, dtype: int64\n",
            "\n",
            "round_E\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_E, dtype: int64\n",
            "\n",
            "round_F\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_F, dtype: int64\n",
            "\n",
            "round_G\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_G, dtype: int64\n",
            "\n",
            "round_H\n",
            "False    49438\n",
            "True      4856\n",
            "Name: round_H, dtype: int64\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KGEgXMioy8dn",
        "outputId": "dddc89f2-b627-432e-c18f-5be11a4531f9"
      },
      "source": [
        "#closer look on value count on each feature  \n",
        "for n in data.columns.values.tolist():\n",
        "    print(n)\n",
        "    print (data[n].value_counts())\n",
        "    print(\"\")  "
      ],
      "execution_count": 192,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "permalink\n",
            "/organization/treasure-valley-urology-services         2\n",
            "/organization/prysm                                    2\n",
            "/organization/eagle-i-music                            1\n",
            "/organization/tesoro-enterprises                       1\n",
            "/organization/spo-medical                              1\n",
            "                                                      ..\n",
            "/organization/layer-7-technologies                     1\n",
            "/organization/igy-immune-technologies-life-sciences    1\n",
            "/organization/botanical-tans                           1\n",
            "/organization/speak-with-me                            1\n",
            "/organization/wanova                                   1\n",
            "Name: permalink, Length: 49436, dtype: int64\n",
            "\n",
            "name\n",
            "Spire                4\n",
            "Roost                4\n",
            "Compass              3\n",
            "Peach                3\n",
            "Roadmap              3\n",
            "                    ..\n",
            "Minded               1\n",
            "MenuSpring           1\n",
            "CloudEndure          1\n",
            "Swink.tv             1\n",
            "Grove Instruments    1\n",
            "Name: name, Length: 49350, dtype: int64\n",
            "\n",
            "homepage_url\n",
            "http://www.ikang.com                 2\n",
            "http://www.shotspotter.com           2\n",
            "http://contour.com                   2\n",
            "http://www.canwenetwork.com          2\n",
            "http://clevelandfoundation.org       2\n",
            "                                    ..\n",
            "http://1000corks.com                 1\n",
            "http://www.coraid.com                1\n",
            "http://www.brainrepublic.com         1\n",
            "http://www.caradvisorynetwork.com    1\n",
            "http://www.truviso.com               1\n",
            "Name: homepage_url, Length: 45850, dtype: int64\n",
            "\n",
            "category_list\n",
            "|Software|                                               3650\n",
            "|Biotechnology|                                          3597\n",
            "|E-Commerce|                                             1263\n",
            "|Mobile|                                                 1211\n",
            "|Curated Web|                                            1120\n",
            "                                                         ... \n",
            "|Consumer Electronics|Television|Hardware + Software|       1\n",
            "|MMO Games|Publishing|FreetoPlay Gaming|Game|Games|         1\n",
            "|Medical Marijuana Patients|                                1\n",
            "|Information Technology|SaaS|E-Commerce|                    1\n",
            "|Content|Internet|E-Commerce|Social Media|                  1\n",
            "Name: category_list, Length: 16675, dtype: int64\n",
            "\n",
            " market \n",
            " Software                     4620\n",
            " Biotechnology                3688\n",
            " Mobile                       1983\n",
            " E-Commerce                   1805\n",
            " Curated Web                  1655\n",
            "                              ... \n",
            " Biomass Power Generation        1\n",
            " Rural Energy                    1\n",
            " Early-Stage Technology          1\n",
            " Insurance Companies             1\n",
            " Musical Instruments             1\n",
            "Name:  market , Length: 753, dtype: int64\n",
            "\n",
            " funding_total_usd \n",
            " -                8531\n",
            " 10,00,000         925\n",
            " 5,00,000          761\n",
            " 1,00,000          749\n",
            " 40,000            680\n",
            "                  ... \n",
            " 27,38,34,120        1\n",
            " 1,12,49,996         1\n",
            " 1,58,70,000         1\n",
            " 2,51,30,716         1\n",
            " 1                   1\n",
            "Name:  funding_total_usd , Length: 14617, dtype: int64\n",
            "\n",
            "status\n",
            "operating    41829\n",
            "acquired      3692\n",
            "closed        2603\n",
            "Name: status, dtype: int64\n",
            "\n",
            "country_code\n",
            "USA    28793\n",
            "GBR     2642\n",
            "CAN     1405\n",
            "CHN     1239\n",
            "DEU      968\n",
            "       ...  \n",
            "MUS        1\n",
            "TTO        1\n",
            "ALB        1\n",
            "MOZ        1\n",
            "BRN        1\n",
            "Name: country_code, Length: 115, dtype: int64\n",
            "\n",
            "state_code\n",
            "CA    9917\n",
            "NY    2914\n",
            "MA    1969\n",
            "TX    1466\n",
            "WA     974\n",
            "      ... \n",
            "MB      13\n",
            "AK      12\n",
            "NB       8\n",
            "SK       4\n",
            "PE       2\n",
            "Name: state_code, Length: 61, dtype: int64\n",
            "\n",
            "region\n",
            "SF Bay Area              6804\n",
            "New York City            2577\n",
            "Boston                   1837\n",
            "London                   1588\n",
            "Los Angeles              1389\n",
            "                         ... \n",
            "Shropshire                  1\n",
            "San Sebastian               1\n",
            "Tarragona                   1\n",
            "Scottish Borders, The       1\n",
            "Dki Jakarta                 1\n",
            "Name: region, Length: 1089, dtype: int64\n",
            "\n",
            "city\n",
            "San Francisco    2615\n",
            "New York         2334\n",
            "London           1257\n",
            "Palo Alto         597\n",
            "Austin            583\n",
            "                 ... \n",
            "East Amherst        1\n",
            "Nuevo León          1\n",
            "Douglas             1\n",
            "Dar Es Salaam       1\n",
            "Targu Jiu           1\n",
            "Name: city, Length: 4188, dtype: int64\n",
            "\n",
            "funding_rounds\n",
            "1.0     32039\n",
            "2.0      9219\n",
            "3.0      4026\n",
            "4.0      1997\n",
            "5.0      1001\n",
            "6.0       560\n",
            "7.0       252\n",
            "8.0       152\n",
            "9.0        84\n",
            "10.0       43\n",
            "11.0       35\n",
            "12.0       12\n",
            "13.0        8\n",
            "15.0        4\n",
            "14.0        4\n",
            "16.0        1\n",
            "18.0        1\n",
            "Name: funding_rounds, dtype: int64\n",
            "\n",
            "founded_at\n",
            "2012-01-01    2181\n",
            "2011-01-01    2161\n",
            "2010-01-01    1855\n",
            "2009-01-01    1603\n",
            "2013-01-01    1575\n",
            "              ... \n",
            "2010-05-27       1\n",
            "1962-01-01       1\n",
            "1998-11-20       1\n",
            "2011-02-06       1\n",
            "2006-09-09       1\n",
            "Name: founded_at, Length: 3369, dtype: int64\n",
            "\n",
            "founded_month\n",
            "2012-01    2327\n",
            "2011-01    2286\n",
            "2010-01    1952\n",
            "2013-01    1722\n",
            "2009-01    1655\n",
            "           ... \n",
            "1929-01       1\n",
            "1985-06       1\n",
            "1970-06       1\n",
            "1924-12       1\n",
            "1960-06       1\n",
            "Name: founded_month, Length: 420, dtype: int64\n",
            "\n",
            "founded_quarter\n",
            "2012-Q1    2904\n",
            "2011-Q1    2768\n",
            "2010-Q1    2259\n",
            "2013-Q1    2206\n",
            "2009-Q1    1852\n",
            "           ... \n",
            "1986-Q3       1\n",
            "1908-Q1       1\n",
            "1969-Q3       1\n",
            "1973-Q2       1\n",
            "1924-Q4       1\n",
            "Name: founded_quarter, Length: 218, dtype: int64\n",
            "\n",
            "founded_year\n",
            "2012.0    5211\n",
            "2011.0    4905\n",
            "2013.0    4044\n",
            "2010.0    3773\n",
            "2009.0    2979\n",
            "          ... \n",
            "1907.0       1\n",
            "1905.0       1\n",
            "1903.0       1\n",
            "1918.0       1\n",
            "1920.0       1\n",
            "Name: founded_year, Length: 103, dtype: int64\n",
            "\n",
            "first_funding_at\n",
            "2012-01-01    468\n",
            "2013-01-01    463\n",
            "2008-01-01    422\n",
            "2011-01-01    392\n",
            "2007-01-01    342\n",
            "             ... \n",
            "2010-01-24      1\n",
            "0029-09-14      1\n",
            "2007-02-25      1\n",
            "2008-03-22      1\n",
            "2004-07-26      1\n",
            "Name: first_funding_at, Length: 3914, dtype: int64\n",
            "\n",
            "last_funding_at\n",
            "2013-01-01    387\n",
            "2014-01-01    364\n",
            "2012-01-01    348\n",
            "2008-01-01    302\n",
            "2011-01-01    272\n",
            "             ... \n",
            "2012-03-18      1\n",
            "1983-06-01      1\n",
            "2001-12-10      1\n",
            "2007-03-04      1\n",
            "1992-06-30      1\n",
            "Name: last_funding_at, Length: 3657, dtype: int64\n",
            "\n",
            "seed\n",
            "0.0          35598\n",
            "40000.0        654\n",
            "1000000.0      570\n",
            "500000.0       543\n",
            "100000.0       538\n",
            "             ...  \n",
            "315317.0         1\n",
            "51701.0          1\n",
            "190417.0         1\n",
            "78820.0          1\n",
            "3008970.0        1\n",
            "Name: seed, Length: 3337, dtype: int64\n",
            "\n",
            "venture\n",
            "0.0            26161\n",
            "10000000.0       493\n",
            "5000000.0        487\n",
            "1000000.0        470\n",
            "2000000.0        422\n",
            "               ...  \n",
            "697065.0           1\n",
            "178450000.0        1\n",
            "566000.0           1\n",
            "12201999.0         1\n",
            "7598390.0          1\n",
            "Name: venture, Length: 9300, dtype: int64\n",
            "\n",
            "equity_crowdfunding\n",
            "0.0           48916\n",
            "100000.0         22\n",
            "50000.0          19\n",
            "1000.0           16\n",
            "10000.0          14\n",
            "              ...  \n",
            "10020000.0        1\n",
            "300.0             1\n",
            "1250000.0         1\n",
            "290000.0          1\n",
            "540.0             1\n",
            "Name: equity_crowdfunding, Length: 252, dtype: int64\n",
            "\n",
            "undisclosed\n",
            "0.0          48486\n",
            "500000.0        20\n",
            "1000000.0       18\n",
            "100000.0        18\n",
            "2000000.0       16\n",
            "             ...  \n",
            "5034690.0        1\n",
            "445588.0         1\n",
            "239173.0         1\n",
            "190016.0         1\n",
            "750565.0         1\n",
            "Name: undisclosed, Length: 687, dtype: int64\n",
            "\n",
            "convertible_note\n",
            "0.0          48881\n",
            "500000.0        20\n",
            "100000.0        20\n",
            "50000.0         15\n",
            "1000000.0       15\n",
            "             ...  \n",
            "444334.0         1\n",
            "1120500.0        1\n",
            "35000.0          1\n",
            "1643000.0        1\n",
            "2027669.0        1\n",
            "Name: convertible_note, Length: 299, dtype: int64\n",
            "\n",
            "debt_financing\n",
            "0.0          45213\n",
            "1000000.0      117\n",
            "500000.0        99\n",
            "100000.0        92\n",
            "2000000.0       86\n",
            "             ...  \n",
            "823069.0         1\n",
            "3641000.0        1\n",
            "1253000.0        1\n",
            "561000.0         1\n",
            "3537825.0        1\n",
            "Name: debt_financing, Length: 1872, dtype: int64\n",
            "\n",
            "angel\n",
            "0.0          46309\n",
            "500000.0       195\n",
            "1000000.0      189\n",
            "100000.0       123\n",
            "250000.0       119\n",
            "             ...  \n",
            "47619.0          1\n",
            "174000.0         1\n",
            "1785000.0        1\n",
            "1391700.0        1\n",
            "870000.0         1\n",
            "Name: angel, Length: 999, dtype: int64\n",
            "\n",
            "grant\n",
            "0.0          48296\n",
            "25000.0         37\n",
            "50000.0         36\n",
            "40000.0         33\n",
            "100000.0        33\n",
            "             ...  \n",
            "264000.0         1\n",
            "410172.0         1\n",
            "38999.0          1\n",
            "7090000.0        1\n",
            "9214.0           1\n",
            "Name: grant, Length: 532, dtype: int64\n",
            "\n",
            "private_equity\n",
            "0.0            48065\n",
            "100000000.0       39\n",
            "5000000.0         30\n",
            "10000000.0        22\n",
            "20000000.0        18\n",
            "               ...  \n",
            "79003.0            1\n",
            "255000000.0        1\n",
            "4530000.0          1\n",
            "1025000.0          1\n",
            "226785400.0        1\n",
            "Name: private_equity, Length: 847, dtype: int64\n",
            "\n",
            "post_ipo_equity\n",
            "0.0            49122\n",
            "20000000.0         8\n",
            "10000000.0         7\n",
            "5000000.0          7\n",
            "12000000.0         6\n",
            "               ...  \n",
            "48700000.0         1\n",
            "89000000.0         1\n",
            "24345292.0         1\n",
            "422921270.0        1\n",
            "1900000.0          1\n",
            "Name: post_ipo_equity, Length: 239, dtype: int64\n",
            "\n",
            "post_ipo_debt\n",
            "0.000000e+00    49363\n",
            "1.000000e+07        4\n",
            "5.000000e+06        4\n",
            "1.500000e+08        3\n",
            "3.000000e+07        3\n",
            "2.000000e+07        3\n",
            "3.500000e+07        3\n",
            "2.500000e+07        3\n",
            "5.000000e+07        2\n",
            "5.000000e+08        2\n",
            "2.000000e+06        2\n",
            "1.500000e+07        1\n",
            "7.498349e+06        1\n",
            "6.000000e+07        1\n",
            "3.200000e+07        1\n",
            "2.223000e+08        1\n",
            "2.390000e+08        1\n",
            "1.500000e+06        1\n",
            "8.000000e+05        1\n",
            "3.000000e+05        1\n",
            "8.000000e+08        1\n",
            "1.900000e+09        1\n",
            "7.000000e+07        1\n",
            "7.500000e+07        1\n",
            "2.200000e+07        1\n",
            "4.300000e+05        1\n",
            "9.075000e+04        1\n",
            "3.500000e+09        1\n",
            "5.800000e+09        1\n",
            "5.900000e+05        1\n",
            "1.200000e+07        1\n",
            "1.150000e+07        1\n",
            "3.500000e+06        1\n",
            "9.200000e+08        1\n",
            "1.200000e+08        1\n",
            "2.700000e+07        1\n",
            "2.000000e+09        1\n",
            "7.250000e+08        1\n",
            "3.100000e+07        1\n",
            "2.400000e+08        1\n",
            "1.055750e+09        1\n",
            "4.500000e+07        1\n",
            "6.000000e+08        1\n",
            "5.100000e+06        1\n",
            "5.300000e+07        1\n",
            "1.100000e+06        1\n",
            "3.900000e+08        1\n",
            "2.103847e+08        1\n",
            "3.000000e+08        1\n",
            "9.663821e+07        1\n",
            "8.820564e+07        1\n",
            "6.500000e+07        1\n",
            "5.500000e+06        1\n",
            "1.990000e+07        1\n",
            "7.500000e+06        1\n",
            "2.400000e+07        1\n",
            "1.750000e+08        1\n",
            "Name: post_ipo_debt, dtype: int64\n",
            "\n",
            "secondary_market\n",
            "0.0            49419\n",
            "200000000.0        1\n",
            "7343789.0          1\n",
            "20000000.0         1\n",
            "12500000.0         1\n",
            "126700000.0        1\n",
            "680611554.0        1\n",
            "400000000.0        1\n",
            "63750000.0         1\n",
            "19300000.0         1\n",
            "4700000.0          1\n",
            "78800355.0         1\n",
            "210000000.0        1\n",
            "2500000.0          1\n",
            "156929.0           1\n",
            "700000.0           1\n",
            "7718867.0          1\n",
            "6192396.0          1\n",
            "210000.0           1\n",
            "60000000.0         1\n",
            "Name: secondary_market, dtype: int64\n",
            "\n",
            "product_crowdfunding\n",
            "0.0          49225\n",
            "1000000.0        7\n",
            "500000.0         6\n",
            "100000.0         5\n",
            "600000.0         4\n",
            "             ...  \n",
            "17693.0          1\n",
            "174253.0         1\n",
            "95238.0          1\n",
            "16000.0          1\n",
            "557000.0         1\n",
            "Name: product_crowdfunding, Length: 176, dtype: int64\n",
            "\n",
            "round_A\n",
            "0.0           40435\n",
            "5000000.0       525\n",
            "3000000.0       378\n",
            "2000000.0       372\n",
            "10000000.0      324\n",
            "              ...  \n",
            "2498400.0         1\n",
            "755700.0          1\n",
            "6570000.0         1\n",
            "1904761.0         1\n",
            "4969429.0         1\n",
            "Name: round_A, Length: 2035, dtype: int64\n",
            "\n",
            "round_B\n",
            "0.0           43991\n",
            "10000000.0      391\n",
            "15000000.0      206\n",
            "8000000.0       201\n",
            "5000000.0       184\n",
            "              ...  \n",
            "4086441.0         1\n",
            "825000.0          1\n",
            "12150000.0        1\n",
            "955943.0          1\n",
            "5501029.0         1\n",
            "Name: round_B, Length: 1269, dtype: int64\n",
            "\n",
            "round_C\n",
            "0.0           46601\n",
            "10000000.0      164\n",
            "15000000.0      159\n",
            "20000000.0      129\n",
            "25000000.0       97\n",
            "              ...  \n",
            "2200000.0         1\n",
            "35632090.0        1\n",
            "22720000.0        1\n",
            "48218802.0        1\n",
            "2620442.0         1\n",
            "Name: round_C, Length: 740, dtype: int64\n",
            "\n",
            "round_D\n",
            "0.0            48150\n",
            "10000000.0        67\n",
            "20000000.0        64\n",
            "15000000.0        59\n",
            "12000000.0        44\n",
            "               ...  \n",
            "4700000.0          1\n",
            "40800000.0         1\n",
            "10100000.0         1\n",
            "106000000.0        1\n",
            "43950544.0         1\n",
            "Name: round_D, Length: 458, dtype: int64\n",
            "\n",
            "round_E\n",
            "0.0            48922\n",
            "20000000.0        24\n",
            "25000000.0        20\n",
            "10000000.0        17\n",
            "30000000.0        17\n",
            "               ...  \n",
            "9150000.0          1\n",
            "10024997.0         1\n",
            "40099568.0         1\n",
            "20029000.0         1\n",
            "175445382.0        1\n",
            "Name: round_E, Length: 225, dtype: int64\n",
            "\n",
            "round_F\n",
            "0.0            49266\n",
            "20000000.0        12\n",
            "50000000.0         9\n",
            "40000000.0         7\n",
            "25000000.0         6\n",
            "               ...  \n",
            "3800204.0          1\n",
            "210000000.0        1\n",
            "286000000.0        1\n",
            "15500000.0         1\n",
            "109000000.0        1\n",
            "Name: round_F, Length: 110, dtype: int64\n",
            "\n",
            "round_G\n",
            "0.000000e+00    49404\n",
            "1.000000e+07        3\n",
            "2.500000e+07        2\n",
            "6.998544e+07        1\n",
            "6.300000e+07        1\n",
            "6.000000e+07        1\n",
            "1.900000e+07        1\n",
            "2.950000e+07        1\n",
            "3.700000e+06        1\n",
            "5.000000e+07        1\n",
            "6.937919e+06        1\n",
            "1.720000e+08        1\n",
            "3.000000e+07        1\n",
            "6.800000e+06        1\n",
            "4.000000e+08        1\n",
            "1.608457e+07        1\n",
            "1.700000e+06        1\n",
            "1.250000e+07        1\n",
            "2.170000e+07        1\n",
            "5.600000e+07        1\n",
            "2.600000e+07        1\n",
            "3.500000e+08        1\n",
            "1.000000e+08        1\n",
            "8.184270e+05        1\n",
            "8.199999e+06        1\n",
            "4.089011e+07        1\n",
            "1.500000e+07        1\n",
            "7.366771e+07        1\n",
            "9.900000e+07        1\n",
            "3.638297e+06        1\n",
            "1.000000e+09        1\n",
            "3.500000e+07        1\n",
            "Name: round_G, dtype: int64\n",
            "\n",
            "round_H\n",
            "0.0            49434\n",
            "600000000.0        1\n",
            "50000000.0         1\n",
            "4600000.0          1\n",
            "49000000.0         1\n",
            "Name: round_H, dtype: int64\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "KoO38e44JaGo",
        "outputId": "1da8ea31-ff9b-4f09-827d-5f4147cdfcb8"
      },
      "source": [
        "data.tail(5)"
      ],
      "execution_count": 193,
      "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>permalink</th>\n",
              "      <th>name</th>\n",
              "      <th>homepage_url</th>\n",
              "      <th>category_list</th>\n",
              "      <th>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>state_code</th>\n",
              "      <th>region</th>\n",
              "      <th>city</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_at</th>\n",
              "      <th>founded_month</th>\n",
              "      <th>founded_quarter</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>first_funding_at</th>\n",
              "      <th>last_funding_at</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>54289</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>54290</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>54291</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>54292</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>54293</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      permalink name homepage_url  ... round_F round_G round_H\n",
              "54289       NaN  NaN          NaN  ...     NaN     NaN     NaN\n",
              "54290       NaN  NaN          NaN  ...     NaN     NaN     NaN\n",
              "54291       NaN  NaN          NaN  ...     NaN     NaN     NaN\n",
              "54292       NaN  NaN          NaN  ...     NaN     NaN     NaN\n",
              "54293       NaN  NaN          NaN  ...     NaN     NaN     NaN\n",
              "\n",
              "[5 rows x 39 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 193
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "htv3uHpLJlPk",
        "outputId": "e6fee132-aae7-401a-d1d0-a3ff1ddb32f8"
      },
      "source": [
        "data.duplicated().sum()"
      ],
      "execution_count": 194,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4855"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 194
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Do1pp8YY0pQU"
      },
      "source": [
        "Seeing this we can say that the data has duplicates, missing values, and there's spacing in column 'market' and 'funding_total_usd'. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XkNnJaTm0Ua5"
      },
      "source": [
        "# Data Preprocessing\n",
        "\n",
        "\n",
        "> 1. Fixing space in columns\n",
        "\n",
        "> 2. Remove Duplicates\n",
        "\n",
        "> 3. Change uncommon data format\n",
        "\n",
        "> 4. Handling other missing values\n",
        "\n",
        "> 5. Add target variable\n",
        "\n",
        "> 6. Change Object data type into numeric\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jhQy_--71D74"
      },
      "source": [
        "## 1. Fixing space in column\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v91LBmqnRpx_"
      },
      "source": [
        "data = data[~data.name.isna()] #remove companies with no name"
      ],
      "execution_count": 195,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nfGDskEMT7a3"
      },
      "source": [
        "data1 = data.copy()"
      ],
      "execution_count": 196,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o1IZO11_0Zdo"
      },
      "source": [
        "data1.rename(columns= {' market ':'market', ' funding_total_usd ':'funding_total_usd' }, inplace= True)"
      ],
      "execution_count": 197,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y_uml5as1pGU",
        "outputId": "c5645aba-70c5-4e44-d98d-6de84441d56f"
      },
      "source": [
        "data1.info()"
      ],
      "execution_count": 198,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 49437 entries, 0 to 49437\n",
            "Data columns (total 39 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   permalink             49437 non-null  object \n",
            " 1   name                  49437 non-null  object \n",
            " 2   homepage_url          45988 non-null  object \n",
            " 3   category_list         45476 non-null  object \n",
            " 4   market                45469 non-null  object \n",
            " 5   funding_total_usd     49437 non-null  object \n",
            " 6   status                48123 non-null  object \n",
            " 7   country_code          44165 non-null  object \n",
            " 8   state_code            30161 non-null  object \n",
            " 9   region                44165 non-null  object \n",
            " 10  city                  43322 non-null  object \n",
            " 11  funding_rounds        49437 non-null  float64\n",
            " 12  founded_at            38553 non-null  object \n",
            " 13  founded_month         38481 non-null  object \n",
            " 14  founded_quarter       38481 non-null  object \n",
            " 15  founded_year          38481 non-null  float64\n",
            " 16  first_funding_at      49437 non-null  object \n",
            " 17  last_funding_at       49437 non-null  object \n",
            " 18  seed                  49437 non-null  float64\n",
            " 19  venture               49437 non-null  float64\n",
            " 20  equity_crowdfunding   49437 non-null  float64\n",
            " 21  undisclosed           49437 non-null  float64\n",
            " 22  convertible_note      49437 non-null  float64\n",
            " 23  debt_financing        49437 non-null  float64\n",
            " 24  angel                 49437 non-null  float64\n",
            " 25  grant                 49437 non-null  float64\n",
            " 26  private_equity        49437 non-null  float64\n",
            " 27  post_ipo_equity       49437 non-null  float64\n",
            " 28  post_ipo_debt         49437 non-null  float64\n",
            " 29  secondary_market      49437 non-null  float64\n",
            " 30  product_crowdfunding  49437 non-null  float64\n",
            " 31  round_A               49437 non-null  float64\n",
            " 32  round_B               49437 non-null  float64\n",
            " 33  round_C               49437 non-null  float64\n",
            " 34  round_D               49437 non-null  float64\n",
            " 35  round_E               49437 non-null  float64\n",
            " 36  round_F               49437 non-null  float64\n",
            " 37  round_G               49437 non-null  float64\n",
            " 38  round_H               49437 non-null  float64\n",
            "dtypes: float64(23), object(16)\n",
            "memory usage: 15.1+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tmkU1v0hSEf2",
        "outputId": "052eb86f-74a0-440b-8d65-a26e3a46c0d2"
      },
      "source": [
        "data1.name.isnull().sum()"
      ],
      "execution_count": 199,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 199
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CcRWdXeO10t_"
      },
      "source": [
        "## 2. Remove Duplicate (All NaN in whole columns) \n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EosAgVck2Wk8",
        "outputId": "49f5dfef-ac9e-47cc-8b72-7e40e19fcbaf"
      },
      "source": [
        "data1.duplicated().sum()"
      ],
      "execution_count": 200,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 200
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "id": "OkGIP3Bv--D-",
        "outputId": "3232d0e4-58b4-4dcb-9228-14cab4e47a19"
      },
      "source": [
        "data1.tail()"
      ],
      "execution_count": 201,
      "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>permalink</th>\n",
              "      <th>name</th>\n",
              "      <th>homepage_url</th>\n",
              "      <th>category_list</th>\n",
              "      <th>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>state_code</th>\n",
              "      <th>region</th>\n",
              "      <th>city</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_at</th>\n",
              "      <th>founded_month</th>\n",
              "      <th>founded_quarter</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>first_funding_at</th>\n",
              "      <th>last_funding_at</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>49433</th>\n",
              "      <td>/organization/zzish</td>\n",
              "      <td>Zzish</td>\n",
              "      <td>http://www.zzish.com</td>\n",
              "      <td>|Analytics|Gamification|Developer APIs|iOS|And...</td>\n",
              "      <td>Education</td>\n",
              "      <td>3,20,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>GBR</td>\n",
              "      <td>NaN</td>\n",
              "      <td>London</td>\n",
              "      <td>London</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2013-01-28</td>\n",
              "      <td>2013-01</td>\n",
              "      <td>2013-Q1</td>\n",
              "      <td>2013.0</td>\n",
              "      <td>2014-03-24</td>\n",
              "      <td>2014-03-24</td>\n",
              "      <td>320000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49434</th>\n",
              "      <td>/organization/zznode-science-and-technology-co...</td>\n",
              "      <td>ZZNode Science and Technology</td>\n",
              "      <td>http://www.zznode.com</td>\n",
              "      <td>|Enterprise Software|</td>\n",
              "      <td>Enterprise Software</td>\n",
              "      <td>15,87,301</td>\n",
              "      <td>operating</td>\n",
              "      <td>CHN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Beijing</td>\n",
              "      <td>Beijing</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2012-04-01</td>\n",
              "      <td>2012-04-01</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1587301.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1587301.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49435</th>\n",
              "      <td>/organization/zzzzapp-com</td>\n",
              "      <td>Zzzzapp Wireless ltd.</td>\n",
              "      <td>http://www.zzzzapp.com</td>\n",
              "      <td>|Web Development|Advertising|Wireless|Mobile|</td>\n",
              "      <td>Web Development</td>\n",
              "      <td>97,398</td>\n",
              "      <td>operating</td>\n",
              "      <td>HRV</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Split</td>\n",
              "      <td>Split</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2012-05-13</td>\n",
              "      <td>2012-05</td>\n",
              "      <td>2012-Q2</td>\n",
              "      <td>2012.0</td>\n",
              "      <td>2011-11-01</td>\n",
              "      <td>2014-09-10</td>\n",
              "      <td>71525.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>25873.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49436</th>\n",
              "      <td>/organization/a-list-games</td>\n",
              "      <td>[a]list games</td>\n",
              "      <td>http://www.alistgames.com</td>\n",
              "      <td>|Games|</td>\n",
              "      <td>Games</td>\n",
              "      <td>93,00,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2011-11-21</td>\n",
              "      <td>2011-11-21</td>\n",
              "      <td>9300000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49437</th>\n",
              "      <td>/organization/x</td>\n",
              "      <td>[x+1]</td>\n",
              "      <td>http://www.xplusone.com/</td>\n",
              "      <td>|Enterprise Software|</td>\n",
              "      <td>Enterprise Software</td>\n",
              "      <td>4,50,00,000</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>NY</td>\n",
              "      <td>New York City</td>\n",
              "      <td>New York</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1999-01-01</td>\n",
              "      <td>1999-01</td>\n",
              "      <td>1999-Q1</td>\n",
              "      <td>1999.0</td>\n",
              "      <td>2008-06-01</td>\n",
              "      <td>2013-04-04</td>\n",
              "      <td>0.0</td>\n",
              "      <td>28000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>17000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>16000000.0</td>\n",
              "      <td>10000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               permalink  ... round_H\n",
              "49433                                /organization/zzish  ...     0.0\n",
              "49434  /organization/zznode-science-and-technology-co...  ...     0.0\n",
              "49435                          /organization/zzzzapp-com  ...     0.0\n",
              "49436                         /organization/a-list-games  ...     0.0\n",
              "49437                                    /organization/x  ...     0.0\n",
              "\n",
              "[5 rows x 39 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 201
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WwMsCyQx15ya",
        "outputId": "e2f4491d-4e69-4a9f-9f52-99e1759e84f8"
      },
      "source": [
        "print('Duplicated entries:',data1.duplicated().sum())\n",
        "data1[data1.duplicated()].isna().sum(axis=0)"
      ],
      "execution_count": 202,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Duplicated entries: 0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "permalink               0\n",
              "name                    0\n",
              "homepage_url            0\n",
              "category_list           0\n",
              "market                  0\n",
              "funding_total_usd       0\n",
              "status                  0\n",
              "country_code            0\n",
              "state_code              0\n",
              "region                  0\n",
              "city                    0\n",
              "funding_rounds          0\n",
              "founded_at              0\n",
              "founded_month           0\n",
              "founded_quarter         0\n",
              "founded_year            0\n",
              "first_funding_at        0\n",
              "last_funding_at         0\n",
              "seed                    0\n",
              "venture                 0\n",
              "equity_crowdfunding     0\n",
              "undisclosed             0\n",
              "convertible_note        0\n",
              "debt_financing          0\n",
              "angel                   0\n",
              "grant                   0\n",
              "private_equity          0\n",
              "post_ipo_equity         0\n",
              "post_ipo_debt           0\n",
              "secondary_market        0\n",
              "product_crowdfunding    0\n",
              "round_A                 0\n",
              "round_B                 0\n",
              "round_C                 0\n",
              "round_D                 0\n",
              "round_E                 0\n",
              "round_F                 0\n",
              "round_G                 0\n",
              "round_H                 0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 202
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n8dZndj22343"
      },
      "source": [
        "#data1.dropna(how='all', inplace=True)"
      ],
      "execution_count": 203,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pHWAYPJPByLY"
      },
      "source": [
        "data2 = data1.copy()"
      ],
      "execution_count": 204,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y0v5hQLz39QL",
        "outputId": "e429b44e-eb3f-44fe-81fe-d8cce2bd5696"
      },
      "source": [
        "data2.info()"
      ],
      "execution_count": 205,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 49437 entries, 0 to 49437\n",
            "Data columns (total 39 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   permalink             49437 non-null  object \n",
            " 1   name                  49437 non-null  object \n",
            " 2   homepage_url          45988 non-null  object \n",
            " 3   category_list         45476 non-null  object \n",
            " 4   market                45469 non-null  object \n",
            " 5   funding_total_usd     49437 non-null  object \n",
            " 6   status                48123 non-null  object \n",
            " 7   country_code          44165 non-null  object \n",
            " 8   state_code            30161 non-null  object \n",
            " 9   region                44165 non-null  object \n",
            " 10  city                  43322 non-null  object \n",
            " 11  funding_rounds        49437 non-null  float64\n",
            " 12  founded_at            38553 non-null  object \n",
            " 13  founded_month         38481 non-null  object \n",
            " 14  founded_quarter       38481 non-null  object \n",
            " 15  founded_year          38481 non-null  float64\n",
            " 16  first_funding_at      49437 non-null  object \n",
            " 17  last_funding_at       49437 non-null  object \n",
            " 18  seed                  49437 non-null  float64\n",
            " 19  venture               49437 non-null  float64\n",
            " 20  equity_crowdfunding   49437 non-null  float64\n",
            " 21  undisclosed           49437 non-null  float64\n",
            " 22  convertible_note      49437 non-null  float64\n",
            " 23  debt_financing        49437 non-null  float64\n",
            " 24  angel                 49437 non-null  float64\n",
            " 25  grant                 49437 non-null  float64\n",
            " 26  private_equity        49437 non-null  float64\n",
            " 27  post_ipo_equity       49437 non-null  float64\n",
            " 28  post_ipo_debt         49437 non-null  float64\n",
            " 29  secondary_market      49437 non-null  float64\n",
            " 30  product_crowdfunding  49437 non-null  float64\n",
            " 31  round_A               49437 non-null  float64\n",
            " 32  round_B               49437 non-null  float64\n",
            " 33  round_C               49437 non-null  float64\n",
            " 34  round_D               49437 non-null  float64\n",
            " 35  round_E               49437 non-null  float64\n",
            " 36  round_F               49437 non-null  float64\n",
            " 37  round_G               49437 non-null  float64\n",
            " 38  round_H               49437 non-null  float64\n",
            "dtypes: float64(23), object(16)\n",
            "memory usage: 15.1+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DUP-vYOoaKyW"
      },
      "source": [
        "data3= data2.copy()"
      ],
      "execution_count": 206,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SGbUm7vzYlPf"
      },
      "source": [
        "## 3. Change Uncommon Format\n",
        "\n",
        "---\n",
        "\n",
        "  Then we tackle this by eliminate the comma and change the type into float, as process following "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yH258NcaWKvf",
        "outputId": "b64ba17b-9fdb-46a5-8099-5e5faacdaa34"
      },
      "source": [
        "data3.columns"
      ],
      "execution_count": 207,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['permalink', 'name', 'homepage_url', 'category_list', 'market',\n",
              "       'funding_total_usd', 'status', 'country_code', 'state_code', 'region',\n",
              "       'city', 'funding_rounds', 'founded_at', 'founded_month',\n",
              "       'founded_quarter', 'founded_year', 'first_funding_at',\n",
              "       'last_funding_at', 'seed', 'venture', 'equity_crowdfunding',\n",
              "       'undisclosed', 'convertible_note', 'debt_financing', 'angel', 'grant',\n",
              "       'private_equity', 'post_ipo_equity', 'post_ipo_debt',\n",
              "       'secondary_market', 'product_crowdfunding', 'round_A', 'round_B',\n",
              "       'round_C', 'round_D', 'round_E', 'round_F', 'round_G', 'round_H'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 207
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "agIqAoyAYjp9"
      },
      "source": [
        "data3['funding_total_usd'] = data3['funding_total_usd'].str.replace(',', '')\n",
        "data3['funding_total_usd'] = data3['funding_total_usd'].str.replace('-', '')\n",
        "data3['funding_total_usd'] = data3['funding_total_usd'].str.replace(' ', '')\n",
        "\n",
        "data3['funding_total_usd'] = pd.to_numeric(data3['funding_total_usd'], errors='coerce')"
      ],
      "execution_count": 208,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xt9BfqUFMv3g",
        "outputId": "b2220d12-396b-4451-92f3-a3e069cd5a46"
      },
      "source": [
        "data3.dtypes"
      ],
      "execution_count": 209,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "permalink                object\n",
              "name                     object\n",
              "homepage_url             object\n",
              "category_list            object\n",
              "market                   object\n",
              "funding_total_usd       float64\n",
              "status                   object\n",
              "country_code             object\n",
              "state_code               object\n",
              "region                   object\n",
              "city                     object\n",
              "funding_rounds          float64\n",
              "founded_at               object\n",
              "founded_month            object\n",
              "founded_quarter          object\n",
              "founded_year            float64\n",
              "first_funding_at         object\n",
              "last_funding_at          object\n",
              "seed                    float64\n",
              "venture                 float64\n",
              "equity_crowdfunding     float64\n",
              "undisclosed             float64\n",
              "convertible_note        float64\n",
              "debt_financing          float64\n",
              "angel                   float64\n",
              "grant                   float64\n",
              "private_equity          float64\n",
              "post_ipo_equity         float64\n",
              "post_ipo_debt           float64\n",
              "secondary_market        float64\n",
              "product_crowdfunding    float64\n",
              "round_A                 float64\n",
              "round_B                 float64\n",
              "round_C                 float64\n",
              "round_D                 float64\n",
              "round_E                 float64\n",
              "round_F                 float64\n",
              "round_G                 float64\n",
              "round_H                 float64\n",
              "dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 209
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DWKD8-OBuNrr"
      },
      "source": [
        "data4=data3.drop(columns=['permalink', 'name','homepage_url', 'category_list','state_code', 'city', 'founded_at', 'founded_month',\n",
        "       'founded_quarter','first_funding_at','last_funding_at'])"
      ],
      "execution_count": 210,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UzgvXZPjJkAG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "outputId": "e8197653-4abf-4a7c-8fdc-79edd8a706a1"
      },
      "source": [
        "data4.head(3)"
      ],
      "execution_count": 211,
      "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>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>region</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>News</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>acquired</td>\n",
              "      <td>USA</td>\n",
              "      <td>New York City</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.0</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Games</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>Los Angeles</td>\n",
              "      <td>2.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Publishing</td>\n",
              "      <td>40000.0</td>\n",
              "      <td>operating</td>\n",
              "      <td>EST</td>\n",
              "      <td>Tallinn</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.0</td>\n",
              "      <td>40000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         market  funding_total_usd     status  ... round_F round_G  round_H\n",
              "0         News           1750000.0   acquired  ...     0.0     0.0      0.0\n",
              "1        Games           4000000.0  operating  ...     0.0     0.0      0.0\n",
              "2   Publishing             40000.0  operating  ...     0.0     0.0      0.0\n",
              "\n",
              "[3 rows x 28 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 211
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "e18yyXgvu2dz",
        "outputId": "5918abe6-e9f7-4925-c165-29f2256919bc"
      },
      "source": [
        "missing_data = data4.isnull()\n",
        "\n",
        "for column in missing_data.columns.values.tolist():\n",
        " print (column)\n",
        " print (missing_data[column].value_counts())\n",
        " print(\"\")  "
      ],
      "execution_count": 212,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "market\n",
            "False    45469\n",
            "True      3968\n",
            "Name: market, dtype: int64\n",
            "\n",
            "funding_total_usd\n",
            "False    40906\n",
            "True      8531\n",
            "Name: funding_total_usd, dtype: int64\n",
            "\n",
            "status\n",
            "False    48123\n",
            "True      1314\n",
            "Name: status, dtype: int64\n",
            "\n",
            "country_code\n",
            "False    44165\n",
            "True      5272\n",
            "Name: country_code, dtype: int64\n",
            "\n",
            "region\n",
            "False    44165\n",
            "True      5272\n",
            "Name: region, dtype: int64\n",
            "\n",
            "funding_rounds\n",
            "False    49437\n",
            "Name: funding_rounds, dtype: int64\n",
            "\n",
            "founded_year\n",
            "False    38481\n",
            "True     10956\n",
            "Name: founded_year, dtype: int64\n",
            "\n",
            "seed\n",
            "False    49437\n",
            "Name: seed, dtype: int64\n",
            "\n",
            "venture\n",
            "False    49437\n",
            "Name: venture, dtype: int64\n",
            "\n",
            "equity_crowdfunding\n",
            "False    49437\n",
            "Name: equity_crowdfunding, dtype: int64\n",
            "\n",
            "undisclosed\n",
            "False    49437\n",
            "Name: undisclosed, dtype: int64\n",
            "\n",
            "convertible_note\n",
            "False    49437\n",
            "Name: convertible_note, dtype: int64\n",
            "\n",
            "debt_financing\n",
            "False    49437\n",
            "Name: debt_financing, dtype: int64\n",
            "\n",
            "angel\n",
            "False    49437\n",
            "Name: angel, dtype: int64\n",
            "\n",
            "grant\n",
            "False    49437\n",
            "Name: grant, dtype: int64\n",
            "\n",
            "private_equity\n",
            "False    49437\n",
            "Name: private_equity, dtype: int64\n",
            "\n",
            "post_ipo_equity\n",
            "False    49437\n",
            "Name: post_ipo_equity, dtype: int64\n",
            "\n",
            "post_ipo_debt\n",
            "False    49437\n",
            "Name: post_ipo_debt, dtype: int64\n",
            "\n",
            "secondary_market\n",
            "False    49437\n",
            "Name: secondary_market, dtype: int64\n",
            "\n",
            "product_crowdfunding\n",
            "False    49437\n",
            "Name: product_crowdfunding, dtype: int64\n",
            "\n",
            "round_A\n",
            "False    49437\n",
            "Name: round_A, dtype: int64\n",
            "\n",
            "round_B\n",
            "False    49437\n",
            "Name: round_B, dtype: int64\n",
            "\n",
            "round_C\n",
            "False    49437\n",
            "Name: round_C, dtype: int64\n",
            "\n",
            "round_D\n",
            "False    49437\n",
            "Name: round_D, dtype: int64\n",
            "\n",
            "round_E\n",
            "False    49437\n",
            "Name: round_E, dtype: int64\n",
            "\n",
            "round_F\n",
            "False    49437\n",
            "Name: round_F, dtype: int64\n",
            "\n",
            "round_G\n",
            "False    49437\n",
            "Name: round_G, dtype: int64\n",
            "\n",
            "round_H\n",
            "False    49437\n",
            "Name: round_H, dtype: int64\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "y0eiSLNezMME",
        "outputId": "ba112653-7a2b-4e56-9300-7c3892a6fbb4"
      },
      "source": [
        "data4.describe()"
      ],
      "execution_count": 213,
      "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>funding_total_usd</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>4.090600e+04</td>\n",
              "      <td>49437.000000</td>\n",
              "      <td>38481.000000</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "      <td>4.943700e+04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>1.591291e+07</td>\n",
              "      <td>1.696219</td>\n",
              "      <td>2007.359034</td>\n",
              "      <td>2.173254e+05</td>\n",
              "      <td>7.501202e+06</td>\n",
              "      <td>6.163447e+03</td>\n",
              "      <td>1.302239e+05</td>\n",
              "      <td>2.336457e+04</td>\n",
              "      <td>1.888195e+06</td>\n",
              "      <td>6.542030e+04</td>\n",
              "      <td>1.628486e+05</td>\n",
              "      <td>2.074328e+06</td>\n",
              "      <td>6.088860e+05</td>\n",
              "      <td>4.434449e+05</td>\n",
              "      <td>3.845670e+04</td>\n",
              "      <td>7.074370e+03</td>\n",
              "      <td>1.243980e+06</td>\n",
              "      <td>1.492921e+06</td>\n",
              "      <td>1.205380e+06</td>\n",
              "      <td>7.375410e+05</td>\n",
              "      <td>3.424751e+05</td>\n",
              "      <td>1.697726e+05</td>\n",
              "      <td>5.767183e+04</td>\n",
              "      <td>1.423226e+04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>1.686808e+08</td>\n",
              "      <td>1.294222</td>\n",
              "      <td>7.579279</td>\n",
              "      <td>1.056995e+06</td>\n",
              "      <td>2.847139e+07</td>\n",
              "      <td>1.999068e+05</td>\n",
              "      <td>2.981434e+06</td>\n",
              "      <td>1.432060e+06</td>\n",
              "      <td>1.382060e+08</td>\n",
              "      <td>6.582974e+05</td>\n",
              "      <td>5.612145e+06</td>\n",
              "      <td>3.167263e+07</td>\n",
              "      <td>2.678375e+07</td>\n",
              "      <td>3.428204e+07</td>\n",
              "      <td>3.864500e+06</td>\n",
              "      <td>4.282209e+05</td>\n",
              "      <td>5.532027e+06</td>\n",
              "      <td>7.472777e+06</td>\n",
              "      <td>7.993671e+06</td>\n",
              "      <td>9.815317e+06</td>\n",
              "      <td>5.406969e+06</td>\n",
              "      <td>6.277969e+06</td>\n",
              "      <td>5.252365e+06</td>\n",
              "      <td>2.716893e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>1.000000e+00</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1902.000000</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>3.500000e+05</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>2006.000000</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>2.000000e+06</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>2010.000000</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1.000000e+07</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>2012.000000</td>\n",
              "      <td>2.500000e+04</td>\n",
              "      <td>5.000000e+06</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>3.007950e+10</td>\n",
              "      <td>18.000000</td>\n",
              "      <td>2014.000000</td>\n",
              "      <td>1.300000e+08</td>\n",
              "      <td>2.351000e+09</td>\n",
              "      <td>2.500000e+07</td>\n",
              "      <td>2.924328e+08</td>\n",
              "      <td>3.000000e+08</td>\n",
              "      <td>3.007950e+10</td>\n",
              "      <td>6.359026e+07</td>\n",
              "      <td>7.505000e+08</td>\n",
              "      <td>3.500000e+09</td>\n",
              "      <td>4.700000e+09</td>\n",
              "      <td>5.800000e+09</td>\n",
              "      <td>6.806116e+08</td>\n",
              "      <td>7.200000e+07</td>\n",
              "      <td>3.190000e+08</td>\n",
              "      <td>5.420000e+08</td>\n",
              "      <td>4.900000e+08</td>\n",
              "      <td>1.200000e+09</td>\n",
              "      <td>4.000000e+08</td>\n",
              "      <td>1.060000e+09</td>\n",
              "      <td>1.000000e+09</td>\n",
              "      <td>6.000000e+08</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       funding_total_usd  funding_rounds  ...       round_G       round_H\n",
              "count       4.090600e+04    49437.000000  ...  4.943700e+04  4.943700e+04\n",
              "mean        1.591291e+07        1.696219  ...  5.767183e+04  1.423226e+04\n",
              "std         1.686808e+08        1.294222  ...  5.252365e+06  2.716893e+06\n",
              "min         1.000000e+00        1.000000  ...  0.000000e+00  0.000000e+00\n",
              "25%         3.500000e+05        1.000000  ...  0.000000e+00  0.000000e+00\n",
              "50%         2.000000e+06        1.000000  ...  0.000000e+00  0.000000e+00\n",
              "75%         1.000000e+07        2.000000  ...  0.000000e+00  0.000000e+00\n",
              "max         3.007950e+10       18.000000  ...  1.000000e+09  6.000000e+08\n",
              "\n",
              "[8 rows x 24 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 213
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JVJ-JqDzzQC_",
        "outputId": "705d3cf3-02dc-4001-9f9d-3dcb7b90408b"
      },
      "source": [
        "data4.info()"
      ],
      "execution_count": 214,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 49437 entries, 0 to 49437\n",
            "Data columns (total 28 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   market                45469 non-null  object \n",
            " 1   funding_total_usd     40906 non-null  float64\n",
            " 2   status                48123 non-null  object \n",
            " 3   country_code          44165 non-null  object \n",
            " 4   region                44165 non-null  object \n",
            " 5   funding_rounds        49437 non-null  float64\n",
            " 6   founded_year          38481 non-null  float64\n",
            " 7   seed                  49437 non-null  float64\n",
            " 8   venture               49437 non-null  float64\n",
            " 9   equity_crowdfunding   49437 non-null  float64\n",
            " 10  undisclosed           49437 non-null  float64\n",
            " 11  convertible_note      49437 non-null  float64\n",
            " 12  debt_financing        49437 non-null  float64\n",
            " 13  angel                 49437 non-null  float64\n",
            " 14  grant                 49437 non-null  float64\n",
            " 15  private_equity        49437 non-null  float64\n",
            " 16  post_ipo_equity       49437 non-null  float64\n",
            " 17  post_ipo_debt         49437 non-null  float64\n",
            " 18  secondary_market      49437 non-null  float64\n",
            " 19  product_crowdfunding  49437 non-null  float64\n",
            " 20  round_A               49437 non-null  float64\n",
            " 21  round_B               49437 non-null  float64\n",
            " 22  round_C               49437 non-null  float64\n",
            " 23  round_D               49437 non-null  float64\n",
            " 24  round_E               49437 non-null  float64\n",
            " 25  round_F               49437 non-null  float64\n",
            " 26  round_G               49437 non-null  float64\n",
            " 27  round_H               49437 non-null  float64\n",
            "dtypes: float64(24), object(4)\n",
            "memory usage: 10.9+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-7Ec1WPA5THk"
      },
      "source": [
        "##4. Handling other missing value by replacing : \n",
        "\n",
        "1. Replace missing value in 'market' with 'other'\n",
        "2. Replace missing value in 'funding_total_usd' with 0\n",
        "3. Replace missing value in 'status' with majority status 'operating' \n",
        "4. Replace missing value in 'country_code' with OTHER\n",
        "5. Replace missing value in 'region' with other\n",
        "6. Replace missing value in 'founded_year' with mean\n",
        "---\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "00cuwtJowHyX"
      },
      "source": [
        "1. Missing Value in 'market'\n",
        "Will replace the nan value with 'other' category market."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TpTwMgPzwNM1",
        "outputId": "ee6db76e-7bd8-4bdd-df53-6c8fd299ec81"
      },
      "source": [
        "data4['market'].isnull().sum()"
      ],
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3968"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 215
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3fBnKzzAwTyN"
      },
      "source": [
        "data4['market'].replace(np.nan, 'other', inplace=True)"
      ],
      "execution_count": 216,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WBIvU38vLRur",
        "outputId": "f425bf03-e641-4451-f419-e39cb5ebdb91"
      },
      "source": [
        "data4.market.isnull().sum()"
      ],
      "execution_count": 217,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 217
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w_zJ7QTtxKFw"
      },
      "source": [
        "2. Missing Value in 'funding_total_usd' will be replaced by 0."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ryVW0xKFE1Y7",
        "outputId": "fb8cb4ca-c066-4876-b979-0cd06deb4e31"
      },
      "source": [
        "data4['funding_total_usd'].isnull().sum()"
      ],
      "execution_count": 218,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "8531"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 218
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aenoO664F6uk"
      },
      "source": [
        "data4['funding_total_usd'].replace(np.nan, 0, inplace=True)"
      ],
      "execution_count": 219,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eFdJXEOPiloX",
        "outputId": "4a644f2d-aa6e-4477-9268-e18a1d17b408"
      },
      "source": [
        "data4['funding_total_usd'].isnull().sum()"
      ],
      "execution_count": 220,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 220
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GRUI3Cq6xq2l"
      },
      "source": [
        "3. Missing value in status will be replaced by majority status 'operating'"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "i5v7Ze0WxoPh",
        "outputId": "78fe443e-ad6d-466d-d96a-1840b5eb8bad"
      },
      "source": [
        "data4['status'].value_counts()"
      ],
      "execution_count": 221,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "operating    41829\n",
              "acquired      3692\n",
              "closed        2602\n",
              "Name: status, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 221
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bjw3w_8Bx2ik",
        "outputId": "7247cc2f-084a-4856-d9bd-0cc6f7787098"
      },
      "source": [
        "data4['status'].isnull().sum()"
      ],
      "execution_count": 222,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1314"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 222
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "apllV6Kex5dk"
      },
      "source": [
        "data4['status'].replace(np.nan, 'operating', inplace=True)"
      ],
      "execution_count": 223,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZLDhZOZoyCuh",
        "outputId": "06084e4c-89a1-4175-cebe-16989bfdf629"
      },
      "source": [
        "data4['status'].isnull().sum()"
      ],
      "execution_count": 224,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 224
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a1twXDe0yI88"
      },
      "source": [
        "4. Missing value in 'Country_Code' & 'Region' will be replaced with 'other'"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "25CyBMTiyIBb",
        "outputId": "a8e33f3f-d3b4-465f-e181-b1788eac20bf"
      },
      "source": [
        "data4['country_code'].value_counts()"
      ],
      "execution_count": 225,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "USA    28793\n",
              "GBR     2642\n",
              "CAN     1405\n",
              "CHN     1239\n",
              "DEU      968\n",
              "       ...  \n",
              "MUS        1\n",
              "TTO        1\n",
              "ALB        1\n",
              "MOZ        1\n",
              "BRN        1\n",
              "Name: country_code, Length: 115, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 225
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FUKNoIRxydLq"
      },
      "source": [
        "data4['country_code'].replace(np.nan, 'other', inplace=True)"
      ],
      "execution_count": 226,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nZaVV8RtyF8J",
        "outputId": "fb79262f-dc6b-4894-810e-ef57f21a2012"
      },
      "source": [
        "data4['country_code'].isnull().sum()"
      ],
      "execution_count": 227,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 227
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_wcYFi5RysqS",
        "outputId": "a90e5140-4bcb-4771-8f6e-6741097ddd9c"
      },
      "source": [
        "data4['region'].value_counts()"
      ],
      "execution_count": 228,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SF Bay Area              6804\n",
              "New York City            2577\n",
              "Boston                   1837\n",
              "London                   1588\n",
              "Los Angeles              1389\n",
              "                         ... \n",
              "Shropshire                  1\n",
              "San Sebastian               1\n",
              "Tarragona                   1\n",
              "Scottish Borders, The       1\n",
              "Dki Jakarta                 1\n",
              "Name: region, Length: 1089, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 228
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZH6YX4e9y0Wm",
        "outputId": "bfcfd56b-f490-48d2-b98a-d7e7ac3e3be3"
      },
      "source": [
        "data4['region'].isnull().sum()"
      ],
      "execution_count": 229,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "5272"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 229
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sUMakn2Sy4IW"
      },
      "source": [
        "data4['region'].replace(np.nan, 'other', inplace=True)"
      ],
      "execution_count": 230,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AFAui9z3y9eh"
      },
      "source": [
        "5. Missing Value in Founded_Year will be replaced by mean of their value."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c2AXkDhXzCXS"
      },
      "source": [
        "avg_founded_year = data4['founded_year'].astype(\"float\").mean(axis=0)\n",
        "data4[\"founded_year\"].replace(np.nan, avg_founded_year, inplace=True)\n"
      ],
      "execution_count": 231,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bv6Y4Fkf0jUs",
        "outputId": "e52be313-13b1-4090-b8ed-1af214b211dc"
      },
      "source": [
        "data4.isnull().sum()"
      ],
      "execution_count": 232,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "market                  0\n",
              "funding_total_usd       0\n",
              "status                  0\n",
              "country_code            0\n",
              "region                  0\n",
              "funding_rounds          0\n",
              "founded_year            0\n",
              "seed                    0\n",
              "venture                 0\n",
              "equity_crowdfunding     0\n",
              "undisclosed             0\n",
              "convertible_note        0\n",
              "debt_financing          0\n",
              "angel                   0\n",
              "grant                   0\n",
              "private_equity          0\n",
              "post_ipo_equity         0\n",
              "post_ipo_debt           0\n",
              "secondary_market        0\n",
              "product_crowdfunding    0\n",
              "round_A                 0\n",
              "round_B                 0\n",
              "round_C                 0\n",
              "round_D                 0\n",
              "round_E                 0\n",
              "round_F                 0\n",
              "round_G                 0\n",
              "round_H                 0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 232
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QTaY2MEK19rA"
      },
      "source": [
        "## 5. Add Target Variable"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-ftWLPbM2PQb"
      },
      "source": [
        "data4['M&A'] = data4['status'].map(lambda m :1  if m == 'acquired' else 0)"
      ],
      "execution_count": 233,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dmdQivBm2PQc"
      },
      "source": [
        "data4['ipo'] = data4['post_ipo_equity'].map(lambda i:1 if i>0 else 0)"
      ],
      "execution_count": 234,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "id": "t6YXwbWN2PQc",
        "outputId": "220fbb63-d32a-48a5-8862-3135597410ba"
      },
      "source": [
        "data4['target'] =  data4['M&A'] | data4['ipo']\n",
        "data4.head(3)"
      ],
      "execution_count": 235,
      "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>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>region</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "      <th>M&amp;A</th>\n",
              "      <th>ipo</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>News</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>acquired</td>\n",
              "      <td>USA</td>\n",
              "      <td>New York City</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.000000</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Games</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>Los Angeles</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2007.359034</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Publishing</td>\n",
              "      <td>40000.0</td>\n",
              "      <td>operating</td>\n",
              "      <td>EST</td>\n",
              "      <td>Tallinn</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.000000</td>\n",
              "      <td>40000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         market  funding_total_usd     status  ... M&A ipo  target\n",
              "0         News           1750000.0   acquired  ...   1   0       1\n",
              "1        Games           4000000.0  operating  ...   0   0       0\n",
              "2   Publishing             40000.0  operating  ...   0   0       0\n",
              "\n",
              "[3 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 235
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 173
        },
        "id": "5aPzmLrESlYW",
        "outputId": "4cccd554-bb21-435a-f615-b4d8588ee225"
      },
      "source": [
        "data4 = data4.drop(columns=['ipo','M&A','post_ipo_equity'])\n",
        "data4.head(2)\n"
      ],
      "execution_count": 236,
      "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>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>region</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>News</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>acquired</td>\n",
              "      <td>USA</td>\n",
              "      <td>New York City</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.000000</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Games</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>operating</td>\n",
              "      <td>USA</td>\n",
              "      <td>Los Angeles</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2007.359034</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    market  funding_total_usd     status  ... round_G round_H  target\n",
              "0    News           1750000.0   acquired  ...     0.0     0.0       1\n",
              "1   Games           4000000.0  operating  ...     0.0     0.0       0\n",
              "\n",
              "[2 rows x 28 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 236
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6a1vauEB2hhE",
        "outputId": "5fceb92d-439a-4153-a5a8-b7aae629f316"
      },
      "source": [
        "data4.info()"
      ],
      "execution_count": 237,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 49437 entries, 0 to 49437\n",
            "Data columns (total 28 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   market                49437 non-null  object \n",
            " 1   funding_total_usd     49437 non-null  float64\n",
            " 2   status                49437 non-null  object \n",
            " 3   country_code          49437 non-null  object \n",
            " 4   region                49437 non-null  object \n",
            " 5   funding_rounds        49437 non-null  float64\n",
            " 6   founded_year          49437 non-null  float64\n",
            " 7   seed                  49437 non-null  float64\n",
            " 8   venture               49437 non-null  float64\n",
            " 9   equity_crowdfunding   49437 non-null  float64\n",
            " 10  undisclosed           49437 non-null  float64\n",
            " 11  convertible_note      49437 non-null  float64\n",
            " 12  debt_financing        49437 non-null  float64\n",
            " 13  angel                 49437 non-null  float64\n",
            " 14  grant                 49437 non-null  float64\n",
            " 15  private_equity        49437 non-null  float64\n",
            " 16  post_ipo_debt         49437 non-null  float64\n",
            " 17  secondary_market      49437 non-null  float64\n",
            " 18  product_crowdfunding  49437 non-null  float64\n",
            " 19  round_A               49437 non-null  float64\n",
            " 20  round_B               49437 non-null  float64\n",
            " 21  round_C               49437 non-null  float64\n",
            " 22  round_D               49437 non-null  float64\n",
            " 23  round_E               49437 non-null  float64\n",
            " 24  round_F               49437 non-null  float64\n",
            " 25  round_G               49437 non-null  float64\n",
            " 26  round_H               49437 non-null  float64\n",
            " 27  target                49437 non-null  int64  \n",
            "dtypes: float64(23), int64(1), object(4)\n",
            "memory usage: 10.9+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UhBvHDM8MUJm",
        "outputId": "4b96e0bd-441d-47fd-f8bf-797a2c180f5a"
      },
      "source": [
        "data4.columns"
      ],
      "execution_count": 238,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['market', 'funding_total_usd', 'status', 'country_code', 'region',\n",
              "       'funding_rounds', 'founded_year', 'seed', 'venture',\n",
              "       'equity_crowdfunding', 'undisclosed', 'convertible_note',\n",
              "       'debt_financing', 'angel', 'grant', 'private_equity', 'post_ipo_debt',\n",
              "       'secondary_market', 'product_crowdfunding', 'round_A', 'round_B',\n",
              "       'round_C', 'round_D', 'round_E', 'round_F', 'round_G', 'round_H',\n",
              "       'target'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 238
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jr-NJ50A3RxT"
      },
      "source": [
        "## 6. Change Object Data Type into Numeric "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K2_HbrT12vLt",
        "outputId": "9088d076-82ec-40ea-87b2-e0af3bfbf3a9"
      },
      "source": [
        "data4.describe"
      ],
      "execution_count": 239,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<bound method NDFrame.describe of                       market  funding_total_usd  ... round_H target\n",
              "0                      News           1750000.0  ...     0.0      1\n",
              "1                     Games           4000000.0  ...     0.0      0\n",
              "2                Publishing             40000.0  ...     0.0      0\n",
              "3               Electronics           1500000.0  ...     0.0      0\n",
              "4                   Tourism             60000.0  ...     0.0      0\n",
              "...                      ...                ...  ...     ...    ...\n",
              "49433             Education            320000.0  ...     0.0      0\n",
              "49434   Enterprise Software           1587301.0  ...     0.0      0\n",
              "49435       Web Development             97398.0  ...     0.0      0\n",
              "49436                 Games           9300000.0  ...     0.0      0\n",
              "49437   Enterprise Software          45000000.0  ...     0.0      0\n",
              "\n",
              "[49437 rows x 28 columns]>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 239
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nE-kQdJD6emR"
      },
      "source": [
        "from sklearn.preprocessing import LabelBinarizer, LabelEncoder, OrdinalEncoder\n",
        "le = LabelEncoder()"
      ],
      "execution_count": 240,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cMLjtf_C6JVW"
      },
      "source": [
        "categorical_variables = ['status','market', 'country_code', 'region']"
      ],
      "execution_count": 241,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v1B9eTzp3clq"
      },
      "source": [
        "for column in categorical_variables:\n",
        "    data4[column] = le.fit_transform(data4[column])\n"
      ],
      "execution_count": 242,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QHDEngXz67xq",
        "outputId": "9d07fe66-a87e-4783-caf5-bcfade3f5db2"
      },
      "source": [
        "data4['region'].sample(5)"
      ],
      "execution_count": 243,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "35646    1087\n",
              "21934     946\n",
              "43477     699\n",
              "18556     459\n",
              "9050      260\n",
              "Name: region, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 243
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        },
        "id": "txlUpR-e7hq-",
        "outputId": "46b31e69-dbfa-4b03-ea18-2c41b15b3ec1"
      },
      "source": [
        "data4.head(2)"
      ],
      "execution_count": 244,
      "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>market</th>\n",
              "      <th>funding_total_usd</th>\n",
              "      <th>status</th>\n",
              "      <th>country_code</th>\n",
              "      <th>region</th>\n",
              "      <th>funding_rounds</th>\n",
              "      <th>founded_year</th>\n",
              "      <th>seed</th>\n",
              "      <th>venture</th>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <th>undisclosed</th>\n",
              "      <th>convertible_note</th>\n",
              "      <th>debt_financing</th>\n",
              "      <th>angel</th>\n",
              "      <th>grant</th>\n",
              "      <th>private_equity</th>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <th>secondary_market</th>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <th>round_A</th>\n",
              "      <th>round_B</th>\n",
              "      <th>round_C</th>\n",
              "      <th>round_D</th>\n",
              "      <th>round_E</th>\n",
              "      <th>round_F</th>\n",
              "      <th>round_G</th>\n",
              "      <th>round_H</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>465</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0</td>\n",
              "      <td>110</td>\n",
              "      <td>699</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2012.000000</td>\n",
              "      <td>1750000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>277</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>2</td>\n",
              "      <td>110</td>\n",
              "      <td>570</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2007.359034</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4000000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   market  funding_total_usd  status  ...  round_G  round_H  target\n",
              "0     465          1750000.0       0  ...      0.0      0.0       1\n",
              "1     277          4000000.0       2  ...      0.0      0.0       0\n",
              "\n",
              "[2 rows x 28 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 244
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iwDnoWH57USc",
        "outputId": "429c31da-d607-4b74-9739-9153bea737f8"
      },
      "source": [
        "data4.info() #CHECK if all categories are transformed into numeric variables"
      ],
      "execution_count": 245,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 49437 entries, 0 to 49437\n",
            "Data columns (total 28 columns):\n",
            " #   Column                Non-Null Count  Dtype  \n",
            "---  ------                --------------  -----  \n",
            " 0   market                49437 non-null  int64  \n",
            " 1   funding_total_usd     49437 non-null  float64\n",
            " 2   status                49437 non-null  int64  \n",
            " 3   country_code          49437 non-null  int64  \n",
            " 4   region                49437 non-null  int64  \n",
            " 5   funding_rounds        49437 non-null  float64\n",
            " 6   founded_year          49437 non-null  float64\n",
            " 7   seed                  49437 non-null  float64\n",
            " 8   venture               49437 non-null  float64\n",
            " 9   equity_crowdfunding   49437 non-null  float64\n",
            " 10  undisclosed           49437 non-null  float64\n",
            " 11  convertible_note      49437 non-null  float64\n",
            " 12  debt_financing        49437 non-null  float64\n",
            " 13  angel                 49437 non-null  float64\n",
            " 14  grant                 49437 non-null  float64\n",
            " 15  private_equity        49437 non-null  float64\n",
            " 16  post_ipo_debt         49437 non-null  float64\n",
            " 17  secondary_market      49437 non-null  float64\n",
            " 18  product_crowdfunding  49437 non-null  float64\n",
            " 19  round_A               49437 non-null  float64\n",
            " 20  round_B               49437 non-null  float64\n",
            " 21  round_C               49437 non-null  float64\n",
            " 22  round_D               49437 non-null  float64\n",
            " 23  round_E               49437 non-null  float64\n",
            " 24  round_F               49437 non-null  float64\n",
            " 25  round_G               49437 non-null  float64\n",
            " 26  round_H               49437 non-null  float64\n",
            " 27  target                49437 non-null  int64  \n",
            "dtypes: float64(23), int64(5)\n",
            "memory usage: 10.9 MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 844
        },
        "id": "zEdngYmo72nE",
        "outputId": "d1b45d65-d030-40f9-c975-640173aa3d01"
      },
      "source": [
        "data4.describe().T"
      ],
      "execution_count": 246,
      "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>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>market</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>3.799170e+02</td>\n",
              "      <td>2.375538e+02</td>\n",
              "      <td>0.0</td>\n",
              "      <td>184.0</td>\n",
              "      <td>351.0</td>\n",
              "      <td>622.0</td>\n",
              "      <td>7.530000e+02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>funding_total_usd</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.316693e+07</td>\n",
              "      <td>1.535556e+08</td>\n",
              "      <td>0.0</td>\n",
              "      <td>50000.0</td>\n",
              "      <td>1000000.0</td>\n",
              "      <td>6772388.0</td>\n",
              "      <td>3.007950e+10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>status</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.798006e+00</td>\n",
              "      <td>5.572798e-01</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>country_code</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>8.903412e+01</td>\n",
              "      <td>3.525081e+01</td>\n",
              "      <td>0.0</td>\n",
              "      <td>51.0</td>\n",
              "      <td>110.0</td>\n",
              "      <td>110.0</td>\n",
              "      <td>1.150000e+02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>region</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>6.444210e+02</td>\n",
              "      <td>3.241698e+02</td>\n",
              "      <td>0.0</td>\n",
              "      <td>377.0</td>\n",
              "      <td>700.0</td>\n",
              "      <td>858.0</td>\n",
              "      <td>1.089000e+03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>funding_rounds</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.696219e+00</td>\n",
              "      <td>1.294222e+00</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.800000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>founded_year</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>2.007359e+03</td>\n",
              "      <td>6.686883e+00</td>\n",
              "      <td>1902.0</td>\n",
              "      <td>2007.0</td>\n",
              "      <td>2008.0</td>\n",
              "      <td>2011.0</td>\n",
              "      <td>2.014000e+03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>seed</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>2.173254e+05</td>\n",
              "      <td>1.056995e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>25000.0</td>\n",
              "      <td>1.300000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>venture</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>7.501202e+06</td>\n",
              "      <td>2.847139e+07</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5000000.0</td>\n",
              "      <td>2.351000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>6.163447e+03</td>\n",
              "      <td>1.999068e+05</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.500000e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>undisclosed</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.302239e+05</td>\n",
              "      <td>2.981434e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.924328e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>convertible_note</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>2.336457e+04</td>\n",
              "      <td>1.432060e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.000000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>debt_financing</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.888195e+06</td>\n",
              "      <td>1.382060e+08</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.007950e+10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>angel</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>6.542030e+04</td>\n",
              "      <td>6.582974e+05</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>6.359026e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>grant</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.628486e+05</td>\n",
              "      <td>5.612145e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.505000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>private_equity</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>2.074328e+06</td>\n",
              "      <td>3.167263e+07</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.500000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>4.434449e+05</td>\n",
              "      <td>3.428204e+07</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5.800000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>secondary_market</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>3.845670e+04</td>\n",
              "      <td>3.864500e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>6.806116e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>7.074370e+03</td>\n",
              "      <td>4.282209e+05</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.200000e+07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_A</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.243980e+06</td>\n",
              "      <td>5.532027e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.190000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_B</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.492921e+06</td>\n",
              "      <td>7.472777e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5.420000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_C</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.205380e+06</td>\n",
              "      <td>7.993671e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4.900000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_D</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>7.375410e+05</td>\n",
              "      <td>9.815317e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.200000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_E</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>3.424751e+05</td>\n",
              "      <td>5.406969e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4.000000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_F</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.697726e+05</td>\n",
              "      <td>6.277969e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.060000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_G</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>5.767183e+04</td>\n",
              "      <td>5.252365e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_H</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>1.423226e+04</td>\n",
              "      <td>2.716893e+06</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>6.000000e+08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>target</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>8.093129e-02</td>\n",
              "      <td>2.727323e-01</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000e+00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                        count          mean  ...        75%           max\n",
              "market                49437.0  3.799170e+02  ...      622.0  7.530000e+02\n",
              "funding_total_usd     49437.0  1.316693e+07  ...  6772388.0  3.007950e+10\n",
              "status                49437.0  1.798006e+00  ...        2.0  2.000000e+00\n",
              "country_code          49437.0  8.903412e+01  ...      110.0  1.150000e+02\n",
              "region                49437.0  6.444210e+02  ...      858.0  1.089000e+03\n",
              "funding_rounds        49437.0  1.696219e+00  ...        2.0  1.800000e+01\n",
              "founded_year          49437.0  2.007359e+03  ...     2011.0  2.014000e+03\n",
              "seed                  49437.0  2.173254e+05  ...    25000.0  1.300000e+08\n",
              "venture               49437.0  7.501202e+06  ...  5000000.0  2.351000e+09\n",
              "equity_crowdfunding   49437.0  6.163447e+03  ...        0.0  2.500000e+07\n",
              "undisclosed           49437.0  1.302239e+05  ...        0.0  2.924328e+08\n",
              "convertible_note      49437.0  2.336457e+04  ...        0.0  3.000000e+08\n",
              "debt_financing        49437.0  1.888195e+06  ...        0.0  3.007950e+10\n",
              "angel                 49437.0  6.542030e+04  ...        0.0  6.359026e+07\n",
              "grant                 49437.0  1.628486e+05  ...        0.0  7.505000e+08\n",
              "private_equity        49437.0  2.074328e+06  ...        0.0  3.500000e+09\n",
              "post_ipo_debt         49437.0  4.434449e+05  ...        0.0  5.800000e+09\n",
              "secondary_market      49437.0  3.845670e+04  ...        0.0  6.806116e+08\n",
              "product_crowdfunding  49437.0  7.074370e+03  ...        0.0  7.200000e+07\n",
              "round_A               49437.0  1.243980e+06  ...        0.0  3.190000e+08\n",
              "round_B               49437.0  1.492921e+06  ...        0.0  5.420000e+08\n",
              "round_C               49437.0  1.205380e+06  ...        0.0  4.900000e+08\n",
              "round_D               49437.0  7.375410e+05  ...        0.0  1.200000e+09\n",
              "round_E               49437.0  3.424751e+05  ...        0.0  4.000000e+08\n",
              "round_F               49437.0  1.697726e+05  ...        0.0  1.060000e+09\n",
              "round_G               49437.0  5.767183e+04  ...        0.0  1.000000e+09\n",
              "round_H               49437.0  1.423226e+04  ...        0.0  6.000000e+08\n",
              "target                49437.0  8.093129e-02  ...        0.0  1.000000e+00\n",
              "\n",
              "[28 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 246
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lN9zr4TQ-NQu"
      },
      "source": [
        "7. Feature Scaling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GPjWL4ycATMt"
      },
      "source": [
        "from sklearn.preprocessing import MinMaxScaler\n",
        "mm = MinMaxScaler()\n",
        "\n",
        "fields = list(data4.columns[:-1])\n",
        "data4[fields] = np.asarray(data4[fields])"
      ],
      "execution_count": 247,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gebLFFbP-gTC"
      },
      "source": [
        "for column in fields:\n",
        "    data4[column]= mm.fit_transform(data4[[column]])"
      ],
      "execution_count": 248,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U1Egtu8VNZOX",
        "outputId": "159ca16d-fb7e-417e-f47f-90e93a015fb4"
      },
      "source": [
        "fields"
      ],
      "execution_count": 249,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['market',\n",
              " 'funding_total_usd',\n",
              " 'status',\n",
              " 'country_code',\n",
              " 'region',\n",
              " 'funding_rounds',\n",
              " 'founded_year',\n",
              " 'seed',\n",
              " 'venture',\n",
              " 'equity_crowdfunding',\n",
              " 'undisclosed',\n",
              " 'convertible_note',\n",
              " 'debt_financing',\n",
              " 'angel',\n",
              " 'grant',\n",
              " 'private_equity',\n",
              " 'post_ipo_debt',\n",
              " 'secondary_market',\n",
              " 'product_crowdfunding',\n",
              " 'round_A',\n",
              " 'round_B',\n",
              " 'round_C',\n",
              " 'round_D',\n",
              " 'round_E',\n",
              " 'round_F',\n",
              " 'round_G',\n",
              " 'round_H']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 249
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4xDOUA64NPMP",
        "outputId": "39cb6880-bab6-4e0f-a6e3-22753992b858"
      },
      "source": [
        "feature_cols = data4.columns[:-1]\n",
        "feature_cols "
      ],
      "execution_count": 250,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['market', 'funding_total_usd', 'status', 'country_code', 'region',\n",
              "       'funding_rounds', 'founded_year', 'seed', 'venture',\n",
              "       'equity_crowdfunding', 'undisclosed', 'convertible_note',\n",
              "       'debt_financing', 'angel', 'grant', 'private_equity', 'post_ipo_debt',\n",
              "       'secondary_market', 'product_crowdfunding', 'round_A', 'round_B',\n",
              "       'round_C', 'round_D', 'round_E', 'round_F', 'round_G', 'round_H'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 250
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 844
        },
        "id": "yWt7qOYJMpkX",
        "outputId": "3aa0cbfd-dea4-40f2-a0ce-65d358cf81fc"
      },
      "source": [
        "round(data4.describe().T, 3)"
      ],
      "execution_count": 251,
      "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>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>market</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.505</td>\n",
              "      <td>0.315</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.244</td>\n",
              "      <td>0.466</td>\n",
              "      <td>0.826</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>funding_total_usd</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.005</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>status</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.899</td>\n",
              "      <td>0.279</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>country_code</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.774</td>\n",
              "      <td>0.307</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.443</td>\n",
              "      <td>0.957</td>\n",
              "      <td>0.957</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>region</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.592</td>\n",
              "      <td>0.298</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.346</td>\n",
              "      <td>0.643</td>\n",
              "      <td>0.788</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>funding_rounds</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.041</td>\n",
              "      <td>0.076</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.059</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>founded_year</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.941</td>\n",
              "      <td>0.060</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.938</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.973</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>seed</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.002</td>\n",
              "      <td>0.008</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>venture</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.003</td>\n",
              "      <td>0.012</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.002</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>equity_crowdfunding</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.008</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>undisclosed</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.010</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>convertible_note</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.005</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>debt_financing</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.005</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>angel</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.001</td>\n",
              "      <td>0.010</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>grant</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.007</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>private_equity</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.001</td>\n",
              "      <td>0.009</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>post_ipo_debt</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.006</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>secondary_market</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.006</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>product_crowdfunding</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.006</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_A</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.004</td>\n",
              "      <td>0.017</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_B</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.003</td>\n",
              "      <td>0.014</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_C</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.002</td>\n",
              "      <td>0.016</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_D</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.001</td>\n",
              "      <td>0.008</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_E</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.001</td>\n",
              "      <td>0.014</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_F</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.006</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_G</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.005</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>round_H</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.005</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>target</th>\n",
              "      <td>49437.0</td>\n",
              "      <td>0.081</td>\n",
              "      <td>0.273</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                        count   mean    std  min    25%    50%    75%  max\n",
              "market                49437.0  0.505  0.315  0.0  0.244  0.466  0.826  1.0\n",
              "funding_total_usd     49437.0  0.000  0.005  0.0  0.000  0.000  0.000  1.0\n",
              "status                49437.0  0.899  0.279  0.0  1.000  1.000  1.000  1.0\n",
              "country_code          49437.0  0.774  0.307  0.0  0.443  0.957  0.957  1.0\n",
              "region                49437.0  0.592  0.298  0.0  0.346  0.643  0.788  1.0\n",
              "funding_rounds        49437.0  0.041  0.076  0.0  0.000  0.000  0.059  1.0\n",
              "founded_year          49437.0  0.941  0.060  0.0  0.938  0.946  0.973  1.0\n",
              "seed                  49437.0  0.002  0.008  0.0  0.000  0.000  0.000  1.0\n",
              "venture               49437.0  0.003  0.012  0.0  0.000  0.000  0.002  1.0\n",
              "equity_crowdfunding   49437.0  0.000  0.008  0.0  0.000  0.000  0.000  1.0\n",
              "undisclosed           49437.0  0.000  0.010  0.0  0.000  0.000  0.000  1.0\n",
              "convertible_note      49437.0  0.000  0.005  0.0  0.000  0.000  0.000  1.0\n",
              "debt_financing        49437.0  0.000  0.005  0.0  0.000  0.000  0.000  1.0\n",
              "angel                 49437.0  0.001  0.010  0.0  0.000  0.000  0.000  1.0\n",
              "grant                 49437.0  0.000  0.007  0.0  0.000  0.000  0.000  1.0\n",
              "private_equity        49437.0  0.001  0.009  0.0  0.000  0.000  0.000  1.0\n",
              "post_ipo_debt         49437.0  0.000  0.006  0.0  0.000  0.000  0.000  1.0\n",
              "secondary_market      49437.0  0.000  0.006  0.0  0.000  0.000  0.000  1.0\n",
              "product_crowdfunding  49437.0  0.000  0.006  0.0  0.000  0.000  0.000  1.0\n",
              "round_A               49437.0  0.004  0.017  0.0  0.000  0.000  0.000  1.0\n",
              "round_B               49437.0  0.003  0.014  0.0  0.000  0.000  0.000  1.0\n",
              "round_C               49437.0  0.002  0.016  0.0  0.000  0.000  0.000  1.0\n",
              "round_D               49437.0  0.001  0.008  0.0  0.000  0.000  0.000  1.0\n",
              "round_E               49437.0  0.001  0.014  0.0  0.000  0.000  0.000  1.0\n",
              "round_F               49437.0  0.000  0.006  0.0  0.000  0.000  0.000  1.0\n",
              "round_G               49437.0  0.000  0.005  0.0  0.000  0.000  0.000  1.0\n",
              "round_H               49437.0  0.000  0.005  0.0  0.000  0.000  0.000  1.0\n",
              "target                49437.0  0.081  0.273  0.0  0.000  0.000  0.000  1.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 251
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XtCHsUeaUW_8"
      },
      "source": [
        "data4.to_csv('data_final.csv', index=True)"
      ],
      "execution_count": 252,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "de68CW3UCkg3"
      },
      "source": [
        "# Machine Learning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6F9B7WdDN7U6"
      },
      "source": [
        "To tackle imbalance data set, we use Stratified Shuffle Split Sample."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CkQOFlMRUrt9"
      },
      "source": [
        "data = pd.read_csv('data_final.csv')"
      ],
      "execution_count": 253,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wF0B1jtqM6C1"
      },
      "source": [
        "feature_cols = data.columns[:-1]\n",
        "\n",
        "### BEGIN SOLUTION\n",
        "from sklearn.model_selection import StratifiedShuffleSplit\n",
        "\n",
        "# Get the split indexes\n",
        "strat_shuf_split = StratifiedShuffleSplit(n_splits=1, \n",
        "                                          test_size=0.3, \n",
        "                                          random_state=42)\n",
        "\n",
        "train_idx, test_idx = next(strat_shuf_split.split(data[feature_cols], data.target))\n",
        "\n",
        "# Create the dataframes\n",
        "X_train = data.loc[train_idx, feature_cols]\n",
        "y_train = data.loc[train_idx, 'target']\n",
        "\n",
        "X_test  = data.loc[test_idx, feature_cols]\n",
        "y_test  = data.loc[test_idx, 'target']"
      ],
      "execution_count": 254,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t6Bsi-VxNuwr",
        "outputId": "aac63202-da74-470a-9daf-e9e09db1250b"
      },
      "source": [
        "y_train.value_counts(normalize=True).sort_index()"
      ],
      "execution_count": 255,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    0.919058\n",
              "1    0.080942\n",
              "Name: target, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 255
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dzlCHCixN3JB",
        "outputId": "81622291-5690-4263-9879-290d9869bc78"
      },
      "source": [
        "y_test.value_counts(normalize=False).sort_index()"
      ],
      "execution_count": 256,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    13632\n",
              "1     1200\n",
              "Name: target, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 256
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mTkNokQPxCuj",
        "outputId": "80267194-a4f9-459b-c8b2-791cf06ccdc7"
      },
      "source": [
        "data['target'].value_counts(normalize=False).sort_index()"
      ],
      "execution_count": 257,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    45436\n",
              "1     4001\n",
              "Name: target, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 257
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "id": "LRpcTja-xLSt",
        "outputId": "55912f05-089c-44a8-a4fa-89994dd4343d"
      },
      "source": [
        "ax = data[data['target'] == 1][0:50].plot(kind='scatter', x='founded_year', y='funding_rounds', color='DarkBlue', label='Successful');\n",
        "data[data['target'] == 0][0:50].plot(kind='scatter', x='founded_year', y='funding_rounds', color='Yellow', label='Not', ax=ax);\n",
        "plt.show()"
      ],
      "execution_count": 258,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ0AAAEcCAYAAAAcM2nfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeVxU9f748deMbKEpKugVVxIYVBb3Fb2JmZQSWOKSSqb5vWmLaX1dv7d7r5WWWVlq+rPlumQuGIvmkltquXHVcskN3BEXFBVBhwHm/P7gMjnMgDMwMwi8n48HD53P+ZzPeX/OgXnPOecz56NSFEVBCCGEcAB1eQcghBCi6pCkI4QQwmEk6QghhHAYSTpCCCEcRpKOEEIIh3Eq7wAeJS1btkSv11OjRo3yDkUIISqMrKws1Go1x48ff2hdOdN5gF6vR0aQCyGEdRRFQa/XW1RXznQeUHiGc+DAgXKORAghKo727dtbXFfOdIQQQjiMJB0hhBAOI0lHCCGEw0jSEUII4TAykEAIUUXogQQgBfAFopDP3Y4nSacUsrOzyczMJC8vz+JhgkIAqNVq3Nzc8PT0RKVSlXc4VYgeGAzEPlAWDaxEEo9jyd62gl6vJzU1lYsXL5KZmUlubm55hyQqmNzcXG7cuMHly5flO2EOlYBxwuG/rxPLIZaqTc50rHDnzh3u3r2Lp6cndevWRa2WnC2sl5GRwbVr17hx4wZeXl7lHU4VkVJMebJDoxCPwJmOTqfj448/JjQ0lODgYAYOHMjevXutbmf06NFoNBo++OADO0RZICsrCxcXFzw9PSXhiFKrU6cOrq6uaLXa8g6lCvEtptzPoVGIRyDpTJ48mSVLlvDcc88xbdo01Go1o0eP5rfffrO4jR07djjkKQJ6vR4nJye5Fi/KrFq1anI/0KGiKLiH86CBQGQ5xFK1lWvSOXLkCOvXr+edd95h4sSJDBo0iCVLltCgQQNmz55tURs6nY6ZM2cyatQoO0crhKi41BQMGogDPvrvvyt4BD53Vznlusc3bdqEs7Mz0dF/fgJxdXVlwIABHDx4kOvXrz+0jaVLl6LVaiXpCCEeQg30Byb+919JOOWhXPf6iRMn8PHxoXr16kblwcHBKIrCiRMnSlw/PT2dL7/8kvHjx/PYY4/ZM1QhhBA2UK5JJz09nXr16pmUF47oediZzqeffoqPjw+RkXJdVthGVlYWU6ZMoWvXrmg0GubOnWvV+sOHD2f48OF2ik6Iiq9ch0xrtVqcnZ1Nyl1dXQHIyckpdt0jR46QkJDAsmXL5Ma+jZw6dYr58+dz9OhRbty4gYeHB76+voSFhVWZN9KFCxeybt06xo4dS8OGDdFoNOUdkhCVSrkmHTc3N7NfsCxMNoXJpyhFUfjggw94+umnrZrHQRTv0KFDxMTE4O3tTXR0NF5eXly5coXDhw+zdOnSKpN0kpKSaNOmDWPHji3vUISolMo16Xh5eZm9hJaeng5g9tIbwJYtWzhy5Ajjx48nNTXVaFlWVhapqal4enri5uZm+6ArqYULF1KrVi3WrFlDzZo1jZbdvHmznKJyvJs3b8rZjRB2VK73dAICAjh37hzZ2dlG5YcPHzYsNyctLQ29Xs9LL71Er169DD8AcXFx9OrVi6SkJPsGX8lcvHgRf39/k4QDULduXQBSU1PRaDTExcWZ1DF3/+PKlStMmTKF0NBQgoKC6N27N++//36p6kycOJEuXboQGBhIREQEP/74o1Gd3Nxc5s2bx9NPP01QUBCdOnViyJAh7N6921Dn/PnzvPHGG3Tr1o2goCB69OjB+PHjuXv3Lvv370ej0ZCamsq2bdvQaDSG13FxcYb/P6hwnf3791uwh4UQUM5nOuHh4Xz77bfExsYyYsQIoOB7N3FxcbRt25b69esDBUnm/v37NG/eHICwsDAaNWpk0t5rr71Gz549GTBgAK1atXJYPyqDhg0bcvjwYVJSUvD1Le7b25a7du0a0dHRZGdnM2jQIHx8fEhLS2PDhg383//9n8V1rl+/zsCBA3F2diYmJoZatWqxbds23n77bXQ6Hc8//zwA8+bN45tvvuHFF1/Ez8+Pu3fvcvToUf744w+6deuGTqdj1KhRVKtWjREjRuDh4cHVq1f5+eefyczMpHnz5syaNYuZM2fi7e3NSy+9BBQ8PUAIYTvlmnRCQkIIDw9n9uzZpKen06RJE+Lj40lLS2PmzJmGepMmTSIpKYlTp04B0KRJE5o0aWK2zcaNG/PUU085JH5b0usVEhKSSUm5ja+vB1FRfqjVjhsgMXLkSEaPHs1zzz1HcHAw7du3p0uXLnTs2NHsYI+HmT17NhkZGfzwww+0aNHCUP7WW29ZVWfOnDmo1WoSEhIMZ2Evvvgir7zyCp9++ilRUVGo1Wp27NhBdHQ0U6dONRvPmTNnSE1NJTY2luDgYEP5G2+8Yfh/ZGQkn3/+OX/5y19kRKQQdlLu346aNWsWw4cPJzExkffff5+8vDwWLVpEu3btyjs0h9HrFQYPXscLL6xl0qRdvPDCWgYPXode77inEHfr1o2VK1cSFhbGyZMn+eqrrxg5ciRPPvkkP//8s1Vt6fV6tm3bxlNPPWWUTADDSENL6iiKwpYtWwgLCyMvL4+MjAzDT/fu3UlPT+fcuXMA1KxZk8OHD3P16lWzMdWoUQOAn3/+GZ1OZ1V/hBC2U+5PmXZ1dWXSpElMmjSp2DrLli2zqK3CM6GKJiEhmdjY00ZlsbGnGTIkhf79HfdAwuDgYObNm4dOp+PkyZNs3bqVxYsX88Ybb5CYmFjsaMKiMjIyyM7Oxs+v+NgtrZOZmcn333/P999/b7bOrVu3AHjzzTcZO3YsTz75JC1btqR79+5EREQYLhU2btyYl19+mS+//JLFixfTsWNHevbsSb9+/QwJSQhhf+WedASkpNw2W56cfMvBkRRwcXEhODiY4OBgmjVrxpQpU9i4cSP9+/c3Wz8/P98ucRQ+EPP5558nIiLCbJ3CpNWhQwe2bNnC9u3b2b17NytWrODrr79m+vTpvPDCC0DBw2Wff/55tm3bxq+//sq//vUvFi5cyKpVqwz3D80p7ntg8sBOIawnSecR4OvrYbbcz6+2gyMxFRgYCBTc0K9VqxYAmZmZRnXS0tKMXtepU4fq1auTnFz8XCXW1FEUha5duz40Vg8PD55//nmef/557t27x/Dhw/niiy8MSQfA398ff39/xowZw+HDhxk4cCArVqwwuo9UVOG9pLt37xqVX758+aExCSGMlfs9HQFRUX5ER/sblQ0cqCEysuyjyCy1b98+szNZ7ty5E4AnnniCGjVqULt2bZNpJIpe+lKr1fTq1YutW7dy/Phxo2WF27CkTrVq1ejduzcbNmzg7NmzJrFlZGQY/l94ma2Qu7s7zZo1M3zROCsri7y8PKM6fn5+ODk5lfjkC8AwaOU///mPoSw/P5/Vq1eXuJ4QwpSc6TwC1GoVK1dGMGRICsnJt/Dzq01kpK9DR6+9//773L9/n969e/PEE0+Qm5vLoUOH2LhxIw0bNjQMTY6OjmbRokVMmzaNwMBADhw4YLiZ/6AJEyawe/duhg4dyuDBg/Hx8eHKlSts2LCBn376yeI6b7/9Nvv37+eFF15g0KBBPPHEE9y6dYujR49y/Phxtm/fDkDfvn3p0KEDgYGBeHh4cOzYMTZs2MDQoUOBgqQ6ffp0+vTpg4+PD3q9nrVr16JSqejTp0+J+8bPz4/WrVvz6aefcufOHWrVqsWGDRtMkpgQ4uEk6Twi1GqVQwcNFDVx4kQ2bdrEzp07WbVqFbm5uXh7e/Piiy8yZswYwyWm1157jYyMDH766Sc2btxIjx49+Prrr+nSpYtRew0aNGD16tXMmTOH+Ph4srOzadCgAU8++aRVderVq0dsbCzz5s1j48aN3Lx5Ew8PDzQaDePGjTPUGz58ONu3b2fPnj3odDq8vb0ZN26cYcoLjUZDaGgoO3bsYNWqVTz22GNoNBq++uorWrdu/dD9M3v2bN59910WLVpEzZo1GTBgAJ06deLll18uw14XoupRKeauqVRRhc9xK24W0gsXLgDQtGlTh8UkKif5XRKVycPeOx8k93SEEEI4jCQdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdIYQQDiNJRwghhMPYLOnIFMBCCCEexqqks3PnTubOnWtUtnz5ctq2bUvr1q15++23yc3NtWmAwnHi4uLQaDSEhISQnp5usjwyMpLhw4db1WZOTg5z585l//79tgpTCFGBWZV0vvnmG6PJtM6cOcOMGTOoV68eXbt2ZcOGDSxfvtzmQQrH0mq1fPPNNzZpKycnh3nz5pGUlGST9oQQFZtVSefs2bOG6YsBNmzYgKurK2vWrOHrr7/m2WefJSEhweZBCsdq0aIFK1eu5ObNm+UdihCikrEq6dy5c4fatWsbXu/Zs4fOnTtTo0YNADp27EhqaqptIxQO97e//Q2dTvfQs53s7GxmzJhB9+7dCQwM5NlnnzWaujo1NZUOHToAMG/ePDQaDRqNxuQSrRCi6rAq6dSuXZu0tDSgYM75o0ePGibvAcjLyyM/P9+2EQqHa9asGX379mXFihVkZGSYraMoCmPGjGHp0qX07NmTyZMn06BBA/71r3+xYMECAOrUqcP06dMB6N27N7NmzWLWrFn07t3bYX0RQjxarEo6rVu3ZuXKlWzatIkZM2aQn59Pjx49DMsvXLhAvXr1bB5k1aAH4oBZ//1XX67RjBkzBq1Wy7///W+zy7dt28b+/fuZMGEC06dPZ9iwYXz99df06NGDL7/8koyMDNzd3XnmmWeAgumiIyMjiYyMJCAgwJFdEUI8QqxKOm+++SZ6vZ633nqLuLg4oqKi8PX1BQo++W7dupW2bdvaJdDKTQ8MBl4AJv3338GUZ+J54oknePbZZ/nuu++4ffu2yfJdu3bh7OzMsGHDDGUqlYqYmBh0Oh179+51ZLhCiArCyZrKvr6+bNiwgUOHDvH4448brtcDZGZm8tJLL9GpUyebB1n5JQCxRcpigSFAf8eH819jx45lw4YN/Pvf/2b8+PFGyy5fvkz9+vVxd3c3Kn/iiScMy4UQoiirkg6Ah4cHYWFhJuW1atXipZdesklQVU9KMeXJDo2iqObNm/PMM8/w3XffMXLkyHKNRQhROchjcB4JvsWU+zk0CnPGjh3LvXv3WLJkiVF5w4YNuXbtGvfu3TMqP3funGE5FFxyE0KIQiWe6QQEBFj9pqFSqTh+/HiZgqp6ooBojC+xDQQiyyecB/j6+tKnTx+WLl1KzZo1qVmzJgA9evRg1apVfP/997zyyitAwX29ZcuW4eLiQpcuXQBwdXVFpVKRmZlZbn0QQjw6Skw6UVFRJknn2LFjJCcn4+PjQ/PmzQFISUnh/Pnz+Pn5GX15VFhKDayk4B5OMgVnOJE8KieiY8eOZdOmTdy9e9dwBhMWFkanTp345JNPSE1Nxc/Pjx07drBr1y7GjRtHnTp1AHBxccHf35+NGzfSrFkzPDw88PPzw9/fvzy7JIQoJyUmnQ8//NDo9e7du9m0aRPz58+nV69eRsu2bt3K//7v/zJlyhTbR1klqCnPQQMl8ff35+mnn+ann34ylKnVahYsWMCcOXPYuHEja9asoXHjxrz77rsMHTrUaP333nuP6dOn8+GHH6LT6Xj99dcl6QhRRakURVEsrTxw4EDatWvHpEmTzC7/8MMPOXToEKtXr7ZZgI5U+EXXAwcOmF1+4cIFAJo2beqwmETlJL9LojJ52Hvng6y6fnPq1CkaN25c7PImTZpw+vRpa5oUQghRhViVdGrWrMnu3buLXf7LL78YnsMmhBBCFGVV0unXrx/btm1j6tSpnDlzhvz8fPLz8zlz5gxTpkxhx44dREREWBWATqfj448/JjQ0lODgYAYOHGjRt9nXrl1LTEwM3bp1IzAwkLCwMKZMmSJfShRCiEeYVV8OHT9+PBcvXiQuLo74+HjU6oKcpdfrURSFnj17mnxz/WEmT57M5s2biYmJoWnTpsTHxzN69GiWLVtGmzZtil3v5MmT1K9fn7/+9a/UqlWLtLQ0Vq9ezY4dO1i7di1eXl5WxSGEEML+rBpIUOjXX39l27ZtXLp0CYDGjRvTq1cvQkNDrWrnyJEjREdHM2XKFEaMGAEUTPrVr18/6tWrZ/WEcH/88QfPP/88EydOZNSoUVatCzKQQDiO/C6JysSagQRWPwYHIDQ01OoEY86mTZtwdnYmOjraUObq6sqAAQP47LPPuH79ulVPrfb29gaQLyIKIcQjqlRJx1ZOnDiBj48P1atXNyoPDg5GURROnDjx0KRz+/Zt8vPzSUtLY/78+QCGb8PbmlqtRqfToSiKPN5FlEl+fj7Ozs7lHYYQDmd10klLS2PVqlWcP3+e27dvU/TqnEqlMnlOV3HS09OpX7++SXnh/Zjr168/tI0+ffoYHr3v4eHBu+++S+fOnS3avrVq1KjB1atXSU9Px9PT03BPSwhrZGRkkJOTw+OPP17eoTicXq+QkJBMSsptfH09iIryQ602/QBnaT1r61oYJQVPfk+h4LmIUZTl6SCWx2fb7T6qrEo6O3fu5PXXXyc3Nxd3d3c8PDzKtHGtVmv2056rqytQcH/nYebNm8e9e/c4d+4ca9euJTs7u0wxlaRWrVrcu3ePmzdvcuvWLZydnalWrZrdticqn/z8fEPC8fT0LO9wHEqvVxg8eB2xsX9+ly862p+VKyOM3oQtrWdtXQujpGAuqwefgxhNwWOqrE8Alsdn2+0+yqxKOp9++im1a9dm/vz5BAUFlXnjbm5u5ObmmpQXJpvC5FOSwjl9/vrXv9KrVy8iIiJwd3c3mlzMVtRqNQ0bNsTDw4PMzEzy8vLQ68t3hk9RsTg7OxsSTlW7RJuQkGz05gsQG3uaIUNS6N/fz+p61ta1MEpsObeV5fE9mnNq2YNVSefs2bO89dZbNkk4UHAZzdwltPT0dACrp75u3LgxrVq1Yt26dXZJOoWqV69uch9KCFGylBTTGWgBkpNvlaqetXUtY9u5rSyP79GcU8serDpvq1Onjk1vfgYEBHDu3DmTS2KHDx82LLeWVqvl7t27NolPCGE7vr7mL8f7+dUuVT1r61rGtnNbWR7fozunlq1ZlXQiIyPZvHmzzTYeHh5Obm4usbF/nlbqdDri4uJo27atYZBBWloaZ86cMVo3IyPDpL1jx45x8uRJWrVqZbMYhRC2ERXlR3S08dPFBw7UEBnpW6p61ta1MEoK7qUYtUhp57ayPD7bbvdRZtWXQ8+dO8fkyZOpU6cOMTExNGrUyOyN9MLvy1hi3LhxbNu2jZdeeokmTZoQHx/PsWPHWLJkCe3atQNg+PDhJCUlcerUKcN6ISEhPPPMM/j7++Pu7k5KSgo//PADzs7OrFq1Ch8fH4tjKGTNF5yEENbT6xUSE1NITr6Fn19tIiN9ix29Zkk9a+taGCWQiK3mtrI8Pttu15Gsee+0KukUziT6sO+pnDhxwtImycnJYc6cOaxbt447d+6g0WiYMGECXbt2NdQxl3Q++ugj9u7dS2pqKlqtFi8vLzp37szYsWNLfBJ2SSTpCCGE9eyWdObOnWvRiJvXX3/d0iYfKZJ0hBDCenZ7DM4bb7xRuoiEEEIIKsoFQyGEEJWCVWc6//nPfyyqV/iFTSGEEOJBViWd4cOHW3RPx5qBBEIIIaoOq5LOzJkzTcry8vK4dOkScXFxNGrUiEGDBtksOCGEEJWLVUmnf//inwE0atSoEpcLIYQQNhtIUKtWLaKjo/n6669t1aQQQohKxqaj12rWrGmYwloIIYQoymZJJycnh7Vr11a5OUKEEEJYzqp7OlOmTDFbfufOHX7//XcyMjKYOHGiTQITQghR+ViVdOLj482W16pVCx8fH6ZMmUJERIRNAhNCCFH5WJV0Tp48aa84hBBCVAHyGBwhhBAOY9WZTqGsrCz27NljGKnWuHFjunbtSo0aNWwanBBCiMrF6qQTGxvLhx9+yL179yicFUGlUuHu7s7kyZOJji46+50QQghRwKqks23bNv7+97/TuHFjxo0bh59fwfzdycnJfPfdd7z77rvUrVuXsLAwuwQrhBCiYrNqErchQ4aQmZnJ6tWrqV69utGyrKwsBg0aRM2aNVmxYoXNA3UEmcRNCCGsZ817p1UDCU6ePEn//v1NEg5AjRo1iIqKkhFuQgghimXT0WuWTHsghBCi6rIq6Wg0GuLj47l3757JsuzsbOLj4wkICLBZcEIIISoXqwYSvPLKK7z++uv079+fmJgYmjdvDkBKSgrLli3j4sWLzJ071y6BCiGEqPisSjpPPfUUf//735k9ezbvvfee4XKaoig89thj/P3vf+epp56yS6BCCCEqPqu/pzN06FAiIiLYvXs3qampQMGXQ7t168bjjz9u8wCFEEJUHhYnnezsbMaMGUNERATR0dE888wz9oxLCCFEJWTxQILq1atz9OhRe8YihBCikrNq9FqLFi04e/asvWIRQghRyVmVdN544w1Wr17Nvn377BWPEEKISsyqgQRr167F29ubl19+mYCAAJo1a4abm5tRHZVKxYwZM2wapBBCiMqh1DOHnjhxghMnTpjUkaQjhBCiODJzqBBCCIex68yh9+7dY968eYbv8wghhKja7J505s+fb5hhVAghRNVm16QDYMV0PUIIISo5uycdIYQQopDVz16zNZ1Ox+eff05iYiKZmZkEBAQwfvx4unTpUuJ6mzdvZsOGDRw5coSbN2/SoEEDevbsydixY+UZcEII8Ygq96QzefJkNm/eTExMDE2bNiU+Pp7Ro0ezbNky2rRpU+x6f//736lXrx6RkZF4e3tz6tQpli1bxi+//MIPP/yAq6urA3shhBDCEuWadI4cOcL69euZMmUKI0aMACAqKop+/foxe/Zsli9fXuy6X3zxBZ06dTIqCwwMZNKkSaxfv57nn3/enqELIYQohXK9p7Np0yacnZ2Jjo42lLm6ujJgwAAOHjzI9evXi123aMIBDHP5nDlzxvbBCiGEKLNyTTonTpzAx8eH6tWrG5UHBwejKIrZJx6U5MaNGwDUrl3bZjEKIYSwHbsmHbVajbe3t8nz2Qqlp6dTr149k3IvLy+AEs90zPnqq6+oVq0aTz/9tPXBCiGEsDu73tOpU6cO27dvL3a5VqvF2dnZpLxwEEBOTo7F21q3bh1r1qzhb3/7G02aNLE+WCGEEHZnVdKZMmVKictVKhVubm40aNCAbt260bJlyxLru7m5kZuba1JemGwsHYF24MABpk2bxpNPPsm4ceMsWkcIIYTjWf2UaZVKBZg+aaBo+aeffsqzzz7LrFmzqFatmtn2vLy8zF5CS09PBzB76a2okydPMmbMGDQaDZ999lmx2xJCCFH+rLqns3fvXlq2bEl4eDirV6/mwIEDHDhwgNWrV9OnTx8CAwPZtm0ba9asoU+fPmzYsIGvv/662PYCAgI4d+4c2dnZRuWHDx82LC/JxYsXeeWVV6hTpw7/7//9P9zd3a3pjhBCCAezKunMmjULT09PPvvsM4KDg6lRowY1atQgODiYOXPmULt2bebNm0dgYCBz5syhbdu2rF27ttj2wsPDyc3NJTY21lCm0+mIi4ujbdu21K9fH4C0tDSTYdDp6emMHDkSlUrFN998Q506dazpihBCiHJg1eW17du38+abbxa7/K9//StffPGF4XVYWJjR66JCQkIIDw9n9uzZpKen06RJE+Lj40lLS2PmzJmGepMmTSIpKYlTp04Zyl555RUuXbrEK6+8wsGDBzl48KBhWZMmTUp8moEQQojyYVXS0el0JQ5jvnbtGjqdzvD6sccee+g9llmzZjFnzhwSExO5c+cOGo2GRYsW0a5duxLXK5xQztzlu/79+0vSEUKIR5BVSadNmzZ899139OzZk9atWxst++233/juu++M3uxPnz5NgwYNSmzT1dWVSZMmMWnSpGLrLFu2zKTswbMeIYS19EACkAL4AlGYu9qu1yskJCSTknIbX18PoqL8UKtVpd5qXp6eqVN3cezYTQID6zJjRg+cnCrmw+4t3zeW7euqwqqkM3HiRIYOHcqQIUMIDg7Gx8cHgHPnznHkyBHc3d2ZOHEiUDDsed++fYZH0wghHhV6YDAQ+0BZNLCSB98M9XqFwYPXERt7+s9a0f6sXBlRqsSTl6fHx2cRqalZAGzceI4VK05y7tz/VLjEY/m+sWxfVyVW9TogIIC4uDiefvppTp8+TUJCAgkJCZw+fZqnn36aNWvWGEacubq6smnTJt555x27BC6EKK0EjN8E+e/rRONaCclGb6oAsbGnSUxMKdVWp07dZUg4hVJTs5g27ZdStVeeLN83lu3rqsTqJxI0bdqUzz//HL1eT0ZGBlDw5AG1umpmbSEqnuKSRrJxrZTb5msl3yrVVo8du2m2/OjRG6VqrzxZvm8s29dVSakzhVqtxtPTE09PT0k4QlQovsWU+xnX8vUwX8uvdA/UDQysa7Y8KMizVO2VJ8v3jWX7uipRKUUfLWCB8+fPc+HCBW7dMv+JJyoqqsyBlYf27dsDBY/VEaLyMnefYSCwgofd0xk4UMOKFf1sck8HoHHjxzl7dnSluKdjft9Ytq8rOmveO61KOjdu3GDSpEns2bMHMH0UDhQ8DsfaKQkeFZJ0RNWhp+C+QjIFn7ojKW70WmJiCsnJt/Dzq01kpG+ZR69Nm/YLR4/eICjIkw8+6F7hEk4hy/eNZfu6IrNb0nnzzTfZsmULQ4YMoXPnznh4mD/F7Nixo6VNPlIk6QghhPWsee+0aiDBnj17GDx4MO+++27pIhNCCFGlWXWOp9frH/oQTiGEEKI4ViWd9u3bGx4/I4QQQljLqqQzefJktmzZwk8//WSveIQQQlRiVt3T+ec//0n16tV56623qFevHo0bNzb5jo5KpWLJkiU2DVIIIUTlYFXSSU1NBTA8xDMtLc32EQkhhKi0rJ5PRwghhCityvUNJSGEEI80STpCCCEcpsTLazExMbtqRo8AACAASURBVKhUKr755hucnJyIiYl5aIMykEAIIURxSkw6qampqFQqwzPWCgcSCCGEEKVRYtIpOnBABhIIIYQoC7mnI4QQwmEk6QghhHCYhw4ksJYMJBBCCFGchw4kKOr+/fuGGUNr1qwJQGZmJgC1a9fG3d3d1jEKIYSoJKwaSHDp0iViYmKIiYlh9OjReHl5AZCens6iRYvYtm0bixcvtluwQgghKjar7unMmDGDNm3aMHXqVEPCAfDy8mLatGmEhIQwc+ZMmwcphBCicrAq6SQlJdGhQ4dil3fs2JGkpKQyByWEEKJysirpqFQqzpw5U+zylJSUMgckhBCi8rIq6XTr1o2VK1eSkJBgeEoBgKIoxMfHs2rVKrp162bzIIUQQlQOKuXB7PEQV69e5cUXX+TKlSvUrVuXZs2aAXD+/Hlu3rxJgwYN+P777/nLX/5ir3jtqn379gAcOHCgnCMRwpher5CQkExKym18fT2IivJDrVbZvb3y2q495OXpmTp1F8eO3SQwsC4zZvTAycn0c7fl+yafpKQFaLWncHPT0LHjGNTqao7oyiPHmvdOq5IOwN27d/nqq6/Ytm0bly5dAqBx48b06tWLV155xTCMuiKSpCMeRXq9wuDB64iNPW0oi472Z+XKiFK9YVvaXnlt1x7y8vT4+CwiNTXLUNaoUQ3Onfsfo8Rj+b7JZ9++HnTtusdQtmdPNzp33lklE481751WP5Hg8ccfZ8KECaxfv54jR45w5MgR1q9fz4QJEyp0whHiUZWQkGz0JggQG3uaxMTS3UO1tL3y2q49TJ26yyjhAKSmZjFt2i+lijEpaYFRwgHo2nU3SUkLbRh15SSPwRHiEZeScttseXLyLbu2V17btYdjx26aLT969IbRa0tj1GpPma2n1Z4sRXRVi1XTVUPBoIE9e/Zw/vx5bt++TdGrcyqVitdee81mAQpR1fn6epgt9/Orbdf2ymu79hAYWJeNG8+ZlAcFeRq9tjRGNzeN2XpubgGljLDqsOqezvnz53nttdc4e/asSbIxNKhSceLECZsF6EhyT0c8iszdZxg4UMOKFf1sdm/FXHvltV17MHdPp3Hjxzl7dvRD7+mY3zf57Nv3V7p23W0o27MnlM6dd8g9nYewKumMGjWKpKQk3nrrLTp37oyHh/lPBQ0bNrS0SXQ6HZ9//jmJiYlkZmYSEBDA+PHj6dKlS4nrHTlyhLi4OI4cOcLp06fJzc3l1Cnzp7yWkqQjHlV6vUJiYgrJybfw86tNZKRvmUeRWdJeeW3XHvLy9Eyb9gtHj94gKMiTDz7oXuzoNcv2TT5JSQvRak/i5hZAx46vVsmEA3ZMOq1bt2bYsGG88847pY+uiAkTJrB582ZiYmJo2rQp8fHxHDt2jGXLltGmTZti15s7dy4LFy5Eo9Fw//59zp49K0lHCCHKgd1Gr7m4uNCoUaPSRWVG4ci3d955h4kTJzJo0CCWLFlCgwYNmD17donrDhkyhIMHDxIXF0doaKjNYhJCCGE/ViWd0NBQDh06ZLONb9q0CWdnZ6Kjow1lrq6uDBgwgIMHD3L9+vVi1/X09MTNzc1msQghhLA/q5LO5MmT+f333/n222/R6XRl3viJEyfw8fGhevXqRuXBwcEoilJhByQIIYQwz6oh00OGDOH+/ft8/PHHfPLJJ9SrVw+12jhvqVQqtm7dalF76enp1K9f36S8cNqEks50hBBCVDxWJR1vb2+bblyr1eLs7GxS7urqCkBOTo5NtyeEEKJ8WZV0li1bZtONu7m5kZuba1JemGwKk48QQojKoVwfg+Pl5WX2Elp6ejoA9erVc3RIQggh7Khck05AQADnzp0jOzvbqPzw4cOG5UIIISoPqy6vBQQEoFKV/O1hlUrF8ePHLWovPDycb7/9ltjYWEaMGAEUPKEgLi6Otm3bGgYZpKWlcf/+fZo3b25NuEIIIR4xViWdqKgok6STl5fHpUuXOHz4MBqNhhYtWljcXkhICOHh4cyePZv09HSaNGlCfHw8aWlpzJw501Bv0qRJJCUlGT1x4PLlyyQmJgJw9OhRAL788kugIDmGhYVZ0zUhhBAOYFXS+fDDD4tddujQIcaMGcM///lPqwKYNWsWc+bMITExkTt37qDRaFi0aBHt2rUrcb3U1FQ+//xzo7LC1/3795ekI4QQjyCrZw4tyUcffcTRo0f57rvvbNWkQ8mz14QQwnp2nTm0JM2aNeOPP/6wZZNCCCEqEZsmnaSkJPlujRBCiGJZdU8nISHBbPnt27fZu3cvu3btYsCAATYJTAghROVjVdKZPHkyKpXK7KyhTk5ODBgwgClTptgsOCGEEJVLiUlnypQpDB48mJCQEACWLl3KjRs38PT8c15xlUpFrVq1aNSoEe7u7vaNVgghRIVWYtKJj4+na9euhqTz0ksvMWvWLDp27OiQ4IQQQlQuJQ4kqF27Njdv3jS8tuHoaiGEEFVQiWc6bdq0YcGCBaSlpVGzZk0AtmzZwoULF4pdR6VS8dprr9k2SiGEEJVCiV8OTU1NZfLkyRw8eBBFUYodRGDUoEpVYWf8lC+HCmFfer1CQkIyKSm38fX1ICrKD7Xa9HmOltYr7bafeMKdzp3f4rHHUtBqfWnQYD1qtatdYtTr80lKWoBWewo3Nw0dO45Bra5Wxj7nAVOBY0AgMANz5xB6fS7JyWNwcjpBXl4L/PwWoFabzmEGeiABSAF8gSis+UaNNe+dFj2RQKfTcePGDcLCwpg6dSq9evUqsX7Dhg0tDPXRIklHCPvR6xUGD15HbOxpQ1l0tD8rV0YYvblaWq+0265WTUdm5j9wd88zLM/JccHZORNwsWmMen0++/b1oGvXPYayPXu60bnzTqPEY12f8wAfIPWBskbAOR5MPHp9LhkZDfD0/PMWyY0bdalT50qRxKMHBgOxD5RFAyuxNPHY/IkELi4ueHt7079/f0JCQmjYsGGJP0IIUVRCQrLRmypAbOxpEhNTSlWvtNuOjV1ulHAAXF11XLnSz+YxJiUtMEo4AF277iYpaWGp2iswFeOEw39fTzMqSU4eY5RwADw9b5KcXPT2RwLGCYf/vk40s+2ys+p7Og8++VkIIayRknLbbHly8q1S1SvttgMCTCeOBHBzS7Z5jFrtKbP1tNqTpWqvwDGzdeGo0SsnJ/O3OZycij6qrLhknlxMedmU6yRuQoiqw9fXw2y5n1/tUtUr7bZPnjQ/I7FW62fzGN3cNGbrubkZT1BpXZ8DzdaFIKNXeXnmp5nJy2tVpMS3mPb8iikvG0k6QgiHiIryIzra36hs4EANkZG+papX2m1HRw8lO9v4Ik9OjgsNGvxo8xg7dhzDnj3djMr27AmlY8dXS9VegRkU3MN5UGPgA6MSP78F3LhR16jsxg1P/PzmF1k3ioJ7OEZbByLNbLvsbDq1QUUnAwmEsC+9XiExMYXk5Fv4+dUmMtK32JFhltQr7babN3enc+fxuLklo9X60aDBj0aj12wZY8HotYVotSdxcwugY8dXix29Znmf8yi4h3OUgjOcDyh+9NprODn9QV5eK/z85pcwei2RgktqfhQknHIcvVZVSNIRQgjrldt8OkIIIURJJOkIIYRwGEk6QgghHEaSjhBCCIeRpCOEEMJhJOkIIYRwGEk6QgghHEaSjhBCCIeRpCOEEMJhJOkIIYRwGEk6QgghHEaSjhBCCIeRpCOEEMJhJOkIIYRwGEk6QgghHEaSjhBCCIeRpCOEEMJhJOkIIYRwmHJPOjqdjo8//pjQ0FCCg4MZOHAge/futWjda9euMW7cONq3b0/btm0ZO3Ysly5dsnPEQgghSqvck87kyZNZsmQJzz33HNOmTUOtVjN69Gh+++23EtfLzs4mJiaGgwcP8uqrr/Lmm29y/PhxYmJiuHPnjoOiF0IIYQ2n8tz4kSNHWL9+PVOmTGHEiBEAREVF0a9fP2bPns3y5cuLXff777/nwoULxMXF0bJlSwC6d+9OREQEixcvZty4cY7oAgB6fT5JSQvQak/h5qahY8cxqNXVytoqkACkAL5AFOY+I+j1CgkJyaSk3MbX14OoKD/UalXpt2phe7bebmWSl6dn6tRdHDt2k8DAusyY0QMnp7IdO8uPi2W/i1XzOP/5N5WTUwtFeR0Xlzx0Oif0+lTc3esX1NLncOVKX9zcUtBqfWnQYD1qtatJa3l5On79NYbq1U+Tne1PaOhSnJxczNS7R1paO2rWvExmZkO8vQ/i5ORuJj4t0AM4B/gAuwA3sz3Rau+xb18o3t6XSUtrSOfOv+LmVpY284CpwDEgEJiB3dKDUo4++ugjpVWrVkpWVpZR+cKFCxWNRqNcu3at2HVfeOEFZdCgQSblI0eOVPr06VOqeNq1a6e0a9fOqnXy8/OU3bu7KoqC4Wf37m5Kfn5eqWL4b6uKokQbtVnwOr/ItvVKdHSiAh8bfqKjE5X8fH3ptmphe7bebmWSm5uvNGq0wGjfNGq0QMnNLf2xs/y4WPa7WDWPs7m/qT9/9HqU7OyrSn6+VtFqXYyWabUuSn6+1qi13NwcJTXVw6heaqqHkpubU6RetpKbqzKql5urUnJzs4vEd19RFHWRuNT/LS9S8362kp3tZFQ3K8tJuX+/tG3mKorSqEi9Rv8tt4w1753lenntxIkT+Pj4UL16daPy4OBgFEXhxIkTZtfT6/WcOnWKwMBAk2VBQUGcP3+e+/fv2yXmopKSFtC16x6jsq5dd5OUtLAMrSYAsUXKYoFE41oJycTGnjauFXuaxMSU0m3VwvZsvd3KZOrUXaSmZhmVpaZmMW3aL0Zl1uxDS+ta+rtYNY+zub+pP6lUoFI14sqVvri66oyWubrquHKln1HZr7/G0LDhbaOyhg1v88svI4zK0tLa4eSkGJU5OSlcvtyuSAQ9KDgTe5Ae+KtJrPv2heLunmdUVr16Hnv39ihlm1OB1CJlqcA0k23bQrkmnfT0dOrVq2dS7uXlBcD169fNrnf79m10Op2hXtF1FUUhPT3dtsEWQ6s9VUz5yTK0WtwfdbJxrZTb5msl3yrdVi1sz9bbrUyOHbtptvzo0RtGr63Zh5bWtfR3sWoe54cnSlfXPNzczNdzczP+26te/bTZejVqGO/rmjUvm61Xq1bR8nPFRHXWpMTb23ybDRsWHURlaZvHiql3tJjysinXpKPVanF2djYpd3UtuH6ak5Njdr3CchcX0+unhetqtVpbhVkiNzdNMeUBZWjVt5hyP+Navh7ma/nVLt1WLWzP1tutTAID65otDwryNHptzT60tK6lv4tV8zgX9zf1p5wcJ7Ra8/W0WuO/vexsf7P1srKM93VmZkOz9e7cKVruU0xUT5iUpKWZb/Py5calbNP0ilGBoGLKy6Zck46bmxu5ubkm5YVJpTCBFFVYrtPpTJYVruvmZv4GnK117DiGPXu6GZXt2RNKx46vlqHVKCC6SNlAINK4VpQf0dHGv/wDB2qIjHz4H5jZrVrYnq23W5nMmNGDRo1qGJU1bvw4H3zQ3ajMmn1oaV1Lfxer5nE29zf1J0UBRUmlQYP15OQYf5jNyXGhQYMfjcpCQ5dy+bJxUr58uTbduy82KvP2PkhenvHAi7w8FQ0bHiwSwS5M347VwE6TWDt3/pXsbOOb/NnZTnTpsquUbc4AGhUpawx8YLJtWyjX0WteXl5mL6EVXhozd+kNwMPDAxcXF7OX0NLT01GpVGYvvdmDWl2Nzp13sm/fQrTak7i5BdC586tlHL2mBlYCQyi4pOZHQcIx/gVSq1WsXBnBkCEpJCffws+vNpGRvqUeXWRpe7bebmXi5KTm3Ln/Ydq0Xzh69AZBQZ588EF3k9Fr1uxDy4+LZb+LVfM4G/9N5eR4oCiv4eKSR06OE4ry5+g1Z+dMLl/uh5tbMlqtHw0a/Ggyes3JyYX69a/x888jqFHjJFlZAXTvvthk9JqTkzt5eVlcuNCOWrUuc+dOQxo2NDd6zQ3IpuB+y1kKzkZ2Ym6kmZubO1rtHX7+uQcNG17i8uXGdOmyy8zoNUvbdKLgUtw0Ci6pBVGQcOyTHlSKoigPr2YfH330EcuWLWP//v1GgwkWLlzIZ599xq5du6hfv77ZdV944QWcnZ1ZuXKlUfnIkSO5fPkyP/30k9XxtG/fHoADBw5Yva4QQlRV1rx3luvltfDwcHJzc4mN/XNUiU6nIy4ujrZt2xoSTlpaGmfOnDFat0+fPvz+++8cP37cUHb27Fn27dtHeHi4YzoghBDCKuV6eS0kJITw8HBmz55Neno6TZo0IT4+nrS0NGbOnGmoN2nSJJKSkjh16s/ROS+++CKxsbH8z//8Dy+//DLVqlVj8eLFeHl5Gb5oKoQQ4tFSrkkHYNasWcyZM4fExETu3LmDRqNh0aJFtGtXdBy7sRo1arBs2TJmzJjBl19+iV6vp1OnTkybNo3atSvi6BohhKj8yvWezqNG7ukIIYT1rHnvLPcznUdJVlYWiqIYdqAQQoiHu3v3LiqVZaMay/0p048StVpt8Y4TQghRQKVSoVZblk7k8poQQgiHkTMdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdO9LpdHz88ceEhoYSHBzMwIED2bt370PXCwsLQ6PRmP15+umnjeoWV2/FihX26paR0vYRYM+ePQwfPpxOnTrRoUMHBg0axIYNG8zWjY2N5ZlnniEoKIg+ffqwfPlyW3ajRI7oY0U+jgkJCURERBAUFERoaCjvv/8+2dnZJvX0ej1fffUVYWFhBAUFERERUezxtgd79zE1NbXY47hr1y57dMnI9evXmT17NsOHD6dNmzZoNBr2799v8fpnzpxh1KhRtGnTho4dOzJp0iQyMjJM6pX1OMp8OnY0efJkNm/eTExMDE2bNiU+Pp7Ro0ezbNky2rRpU+x6U6dONfmFTktLY86cOXTr1s2kfmhoKM8995xRWUhIiG068RCl7ePPP//MmDFjaNOmDW+88QYA69evZ/z48WRnZxMdHW2ou3LlSv7xj38QHh7Oyy+/zIEDB5g+fTo5OTmMHDmyUvQRKuZxXLJkCTNmzKBbt24MHjyYa9eusXTpUpKTk1m8eLHRVCGfffYZixYtYtCgQQQGBrJt2zbGjx+PWq0mPDy8UvQR4LnnniM0NNSoLCAgwC59etC5c+f46quvaNq0KRqNht9++83ida9evcrQoUOpWbMm48eP5969e3z77becPn2a1atX4+zsbKhb5uOoCLs4fPiw4u/vr/z73/82lGm1WuWpp55SXnzxRavbmz9/vuLv768cPHjQqNzf3195//33yxpuqZSlj6NGjVJCQ0OVnJwcQ1lOTo4SGhqqDB061FB2//59pWPHjsqYMWOM1n/77beVNm3aKJmZmbbpTDEc0UdFqZjHMScnR2nXrp0SExOj6PV6Q/n27dsVf39/ZcuWLYayq1evKq1atTLqo16vV1588UWlZ8+eSn5+vm07VYQj+njp0iWTbTjS3bt3lYyMDEVRFGXLli2Kv7+/sm/fPovW/cc//qG0bt1auXr1qqFs9+7dir+/vxIbG2sos8VxlMtrdrJp0yacnZ2NPs26uroyYMAADh48yPXr161q78cff6RRo0a0bdvW7HKtVktOTk6ZYrZWWfqYlZVFrVq1cHFxMZS5uLhQq1YtXF1dDWX79+/n9u3bvPjii0brDx06lOzsbLtftnBEHx9UkY5jcnIyd+/e5dlnnzX6tN+zZ0/c3d2NLrls3bqV3Nxco+OoUqkYMmQIly9f5siRI3bo2Z8c0ccH3bt3D51OZ9tOPESNGjWoXbt2qdbdvHkzYWFh1K9f31DWtWtXmjVrxsaNGw1ltjiOknTs5MSJE/j4+FC9enWj8uDgYBRF4cSJExa3dfz4cc6cOUO/fv3MLl+zZg2tW7cmODiYiIgItmzZUqbYLVWWPnbs2JHk5GTmzJnDxYsXuXjxInPmzOH8+fNGl8yOHz8OQGBgoNH6rVq1Qq1WG5bbiyP6WKiiHcfCN1VzCdTNzY0//vjDaBs1atTAx8fHZBvAI3screljoc8//5w2bdoQHBzMoEGD+M9//mODHtjPtWvXuHnzpsnfGBTsnwf3jS2Oo9zTsZP09HSjTw2FvLy8AKw601m3bh2AyfV+gDZt2vDss8/SqFEjrly5wtKlS3n99df55JNPik1StlKWPr766qtcvHiRhQsXsmDBAgDc3d358ssvje5bpaen4+LigoeHh9H6hWXWnjFayxF9hIp5HJs2bYpKpeLQoUNERUUZys+ePUtGRgZardZoG56enlZvw1Yc0Ue1Wk1oaCi9e/emXr16XLhwgW+++YaXX36ZxYsX0759exv3yjYK+164Lx7k5eXFzZs3yc/Pp1q1ajY5jpJ07ESr1RrdfCtU+InJ0ksoer2e9evX07JlS5o3b26yfOXKlUav+/fvT79+/fj444/p27evyU1OWypLH11cXGjWrBnh4eH07t2b/Px8Vq9ezVtvvcXixYsNn5yK20bhdux9KcoRfYSKeRzr1KnDM888ww8//MATTzxBr169uHbtGu+99x7Ozs5G62m1WqPLjJZuw1Yc0Udvb2+++eYbo/WfffZZ+vbty+zZs02O8aOisA8lHR+tVkv16tVtchzl8pqduLm5kZuba1JeeFCKu6ZfVFJSEteuXSMiIsKi+u7u7gwePJirV69y9uxZywMuhbL08b333mPXrl18+umn9O3bl+eee45///vfeHl5MWPGDKNtFHdtPCcnx+L9WFqO6KM5FeU4Tp8+nR49ejBz5kyeeuophg4dir+/v+Gex4PbMHccrf17KC1H9NGc+vXr07dvXw4fPsz9+/fL1gk7Kex7ScfHzc3N8G9Zj6Oc6diJl5eX2VPN9PR0AOrVq2dRO+vWrUOtVtO3b1+Lt92gQQMA7ty5Y/E6pVHaPup0OtasWcPf/vY31Oo/P/c4OzvTvXt3VqxYQV5eHk5OTnh5eZGbm8vt27eNLrHpdDpu375t8X4sLUf0sTiP+nEEePzxx1mwYAFpaWlcvnwZb29vGjZsyODBg2natKnRNg4cOFCqbdiCI/pYnAYNGqDX68nMzOSxxx4rfSfspLDvhfviQenp6dStW5dq1aoBtjmOcqZjJwEBAZw7d87k+zaHDx82LH8YnU7H5s2b6dixo9nr0cW5dOkSUHBpwJ5K28fbt2+Tl5dHfn6+ybK8vDzy8vJQFAWAFi1aAHDs2DGjeseOHUOv1xuW24sj+licR/04Psjb25sOHTrQsGFDMjMzOXbsGF26dDEsb9GiBVlZWZw7d87sNh7V4/igh/WxOJcuXaJatWrUqlWrdMHbWf369alTp47J3xjAkSNHjI6NLY6jJB07CQ8PJzc3l9jYWEOZTqcjLi6Otm3bGpJIWloaZ86cMdvGzp07yczMLPbSmrlvC9+6dYvvv/+eRo0a0axZs7J3pASl7WPdunWpWbMmW7ZsMbrkkZ2dzc8//4y/v7/h+nvnzp3x8PDg+++/N9r2ihUrcHd3p0ePHvbsokP6WFGPY3E++eQT1Go1gwYNMpT16tULZ2dno+OoKAorV67E29vb7l+CdUQfzR3HCxcusH79etq3b2+4RFXeCkdSPujpp59m+/btXLt2zVC2d+9ezp8/b/SFT1scR7m8ZichISGEh4cze/Zs0tPTadKkCfHx8aSlpTFz5kxDvUmTJpGUlMSpU6dM2li3bh0uLi706dPH7DaWL1/Otm3bePLJJ/H29ubatWusWrWKjIwM5s+fb7e+FSptH6tVq8bIkSOZM2cOgwYN4rnnnkOv17NmzRquXr3KpEmTDOu6ubnx5ptvMn36dMaNG0doaCgHDhxg7dq1vPPOO9SsWbPC97GiHkeABQsWcObMGUJCQqhWrRrbtm3j119/Zfr06TRu3NhQ7y9/+QsxMTF8++235OTkEBQUxNatWzlw4ACfffaZ0SXIitrHjz/+mEuXLtG5c2fq1avHxYsXDYMHHjze9vTll18CGBJnYmIiBw8epGbNmgwbNgyAESNGALB9+3bDeq+++iqbNm0iJiaGYcOGce/ePb755hsCAgKIjIw01LPJcbTo66qiVLRarfLhhx8q3bp1UwIDA5UXXnhB2b17t1GdYcOGKf7+/ibr3r17VwkKClJef/31Ytv/5ZdflBEjRihdu3ZVWrVqpXTo0EEZPXq0cuDAAZv3pThl6ePatWuVAQMGKO3bt1eCg4OV6OhoZfPmzWa3s2rVKqVPnz5Kq1atlKeeekpZsmSJXfpjjr37WJGP47Zt25QXXnhBad26tdK6dWtl6NChys6dO81uIz8/X1m4cKHSs2dPpVWrVkq/fv2UdevW2a1PRdm7j+vWrVOGDh2qdOrUSWnZsqXSuXNn5c0331ROnz5t1349yN/f3+xPz549DXV69uxp9LrQ6dOnlZEjRyohISFK+/btlXfeeUe5efOmSb2yHkeVojzkwrIQQghhI3JPRwghhMNI0hFCCOEwknSEEEI4jCQdIYQQDiNJRwghhMNI0hFCCOEwknSEEEI4jCQdUaFdunSJsWPH0rlzZzQaDZMnTy7vkB4qLCyM4cOH27zduLg4NBoN+/fvt3nbQtiKPAZHVGhTpkzh1KlTvPrqq3h6etKkSZPyDkkIUQJJOqLC0ul0HDhwgGHDhjFq1KjyDkcIYQG5vCYqrBs3bqAoyiP7yHhRNrm5uXafUVQ4niQdUSFNnjyZnj17AjBv3jw0Go3hfkZeXh6LFi3i2WefJSgoiE6dOvHaa6+ZPMl7//79aDQa4uLizLav0WiMyoYPH05YWBjXrl1jwoQJdOjQgZCQEEaNGmUyvwjAlStXGDduHO3ataNt27a8+uqr42MFxQAACJdJREFUJo+Uf9CePXsYOXIk7du3JygoiIiICFasWGG27urVqwkPDycwMJDevXuzePHih87PY86WLVvQaDSsXr3a7PK+ffvSu3dvo7bPnz/P//7v/xIaGkpgYCBhYWF89NFH3Lt3z2jdM2fO8M9//pO+ffvSpk0bQkJCeP75542mFyg0d+5cNBoNycnJzJw5kx49ehAcHMzvv/9udZ/Eo00ur4kKadCgQQQEBDBz5kx69+5N7969AWjevDnvvPMOGzdupFu3bgwZMoQbN26wfPlyBg8ezPLly2nZsmWpt3vv3j2GDRtGSEgI48ePJzU1laVLlzJ27Fh+/PFHwwyLmZmZDB06lKtXrzJ48GCaN2/Of/7zH2JiYtBqtSbtrlq1in/84x+0bt2aV199lccee4w9e/bwz3/+k4sXLxo9Gn/x4sXMnDmTgIAAJkyYwP379/n222+pW7eu1f3p2bMnXl5e/PDDDwwcONBo2e+//05KSgrjx49HpVIBBZPnvfTSS9SsWZNBgwZRv359Tp48ybJly/jtt99YtmyZYZ6gpKQkDhw4wJNPPkmjRo24f/8+mzZt4v/+7//IyMjgb3/7m0k877zzDm5ubowcORIomKlSVDJWPztbiEfEpUuXFH9/f+WLL74wlP3666+Kv7+/Mm7cOEWv1xvKT5w4obRo0UIZMmSIoWzfvn2Kv7+/8sMPP5i0PWnSJJNH3Bc+9n7RokVG5V999ZXi7++v7Nq1y1D2ySefKP7+/sqaNWuM6r7//vuKv7+/MmzYMEPZtWvXlMDAQGXChAkmcbz33ntKQECAcvHiRUVRFOXOnTtKSEiI8swzzyj37t0z1Lty5YrSunVrxd/fX9m3b5/5HVaMwliTk5ONyqdNm6a0aNFCuXr1qqEsIiJC6dOnj3L37l2jups3bzbZl9nZ2Sbbys/PV4YNG6a0bdtW0el0hvIvvvjCsF9yc3Otil9ULHJ5TVQqW7ZsAQompSr8dA4F0xH37NmTgwcPmp3h0VJqtZqYmBijss6dOwMFs0QW2rp1K56enkRFRRnVHT16tEmbP/30EzqdjgEDBpCRkWH0ExYWhl6vZ8+ePQD8+uuv3L9/n6FDh/LYY48Z2vjLX/5S7AyzDxMdHY1KpWLNmjWGsnv37rFhwwZ69OhhmFXz1KlTnDp1in79+qHT6YzibNeuHe7u7uzevdvQhru7u+H/OTk53Lp1i9u3b9OtWzeysrI4e/asSSwvvfQSTk5yAaYyk6MrKpXU1FTUajXNmzc3Webr68vWrVtJTU2lTp06pWq/Xr16uLq6GpV5eHgAcPv2bUPZpUuXCAoKMlxue3D9orOdFs7yWDijozk3btwACvoH8MQTT5jUMddnSzRu3JiuXbuSmJjI22+/jbOzMxs3biQ7O5sBAwaYxDl37lzmzp1bYpxQMDX3vHnz2LhxI1euXDGpm5mZaVJm76m5RfmTpCOqrAfPhIrKy8szW140iTxIKeV8iIXrffTRR9SrV89snQenRbaHgQMHMm7cOLZv306fPn1Ys2YNXl5ePPnkkyZ1R44cSffu3c2282BCffvtt9mxYwcDBw6kQ4cOeHh4UK1aNXbu3MnixYvR6/Um67u5udmsT+LRJElHVCqNGzdGr9dz5swZAgICjJYVflJv1KgRgGGo9Z07d0zaKTyjKEscFy5cID8/3yhRXb9+3eQTfuGn+9q1a9O1a9cS2y2M/ezZs3Tp0sVoWWH/SqNXr17UrVuXNWvW4Ofnx6FDhxg9erTRpa6mTZsCBZcYHxZnZmYmO3bsIDIykunTpxstK7xUKKomuacjKpWnnnoKgEWLFhmdeZw+fZrt27fTrl07w6W1Ro0a4eTkZPImeOjQoTIP1e3Vqxc3btwgISHBqPyrr74yqfvMM8/g4uLC3LlzzY5su3v3LjqdDoBu3brh5ubG8uXLuX//vqHO1atXWbduXanjdXZ2pn///vz666/Mnz8fwOjSGkDLli3x9/dn5cqVXLp0yaSNvLw8wyVGtbrgraXo2d/169fNDpkWVYec6fz/du4YJL0+CuP4NwgKLaymQqMbuFQ0RbfBqbGGS3Yh2iKp+ZKbYdBopXM0VBBUWoMROqUuNURLDhkRNDg0BYE0FC39hz8IvQUv+JLw2vOZz3C2h3vP+R1pKIFAgImJCbLZLJVKhfHxcZ6enjg4OKClpYVoNFqtdbvdBINBjo+PCYfDmKZJuVyu3jC7u7uruY+FhQUymQwrKyuUSiX8fj9XV1cUi0U6Ozs/1XZ3d7O6uko0GmVychLLsvB6vTw/P3N/f08ulyObzeLz+fB4PDiOw9raGrOzs0xNTfH6+koymcQwDG5vb2vueWZmhu3tbTKZDKZpfpmvNDU1sb6+ztzcHJZlYds2fr+ft7c3yuUyZ2dnhMNhpqenaWtrIxAIcHp6SmtrK8PDwzw+PpJKpfD5fJ/mX/K7KHSk4cTjcQYHB0mn08RiMVwuF6OjoziO8+XBZyQS4ePjg1wuRz6fZ2hoiM3NTY6Ojv5T6Hg8Hvb394nFYtWvHdM02dvb+3ZhwLZtDMNgZ2eHVCrFy8sLHR0d9Pf34zjOp/cqoVAIl8vF7u4uiUSCnp4eQqEQ7e3tLC8v19xzX18fY2NjXF5eYtv2tzUDAwOk02m2trYoFAokk0ncbjder5dgMPjpl9/GxgaJRIJCoUA6ncYwDJaWlmhubiYSidTcp/y/NX3UOv0UkYazuLhIsVjk/PxcQ335EZrpiAjw953RxcUFlmUpcOTH6PeaSAN6f3//divvn7q6uri5ueHh4aF6wmZ+fr4OHcpvpdARaUDX19dfLid8J5/Pc3h4yMnJCb29vcTj8epatshP0ExHpAFVKhVKpdK/1o2MjHy5sCDykxQ6IiJSN1okEBGRulHoiIhI3Sh0RESkbhQ6IiJSNwodERGpmz+46SdeEs+AZgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RANniGG6OvAP"
      },
      "source": [
        "## 1. Support Vector Machine"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VWUwuEeQQfJi",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cafdab2d-5be1-49ea-d7cf-5c569468eeb6"
      },
      "source": [
        "from sklearn import svm\n",
        "clf = svm.SVC(kernel='rbf')\n",
        "clf.fit(X_train, y_train) "
      ],
      "execution_count": 259,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 259
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Yt3SdjMCQk1b",
        "outputId": "74b7c3e1-112a-4031-f1a7-5da6954c2d18"
      },
      "source": [
        "yhat = clf.predict(X_test)\n",
        "yhat [0:5]"
      ],
      "execution_count": 260,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 260
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ga2SiXvkO5M4"
      },
      "source": [
        "from sklearn.metrics import classification_report, confusion_matrix\n",
        "import itertools\n",
        "from sklearn.metrics import classification_report, accuracy_score, precision_score, recall_score\n",
        "from sklearn.metrics import f1_score, roc_auc_score"
      ],
      "execution_count": 261,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EAVuC1adYlEB",
        "outputId": "48597ffd-b15e-4e70-b96f-b8ac0145c6e8"
      },
      "source": [
        "score_df = pd.DataFrame({'accuracy': accuracy_score(y_test, yhat),\n",
        "                         'precision': precision_score(y_test, yhat),\n",
        "                         'recall': recall_score(y_test, yhat),\n",
        "                         'f1': f1_score(y_test, yhat),\n",
        "                         'auc': roc_auc_score(y_test, yhat)},\n",
        "                         index=pd.Index([0]))\n",
        "\n",
        "# Compute confusion matrix\n",
        "cnf_matrix = confusion_matrix(y_test, yhat, labels=[0,1])\n",
        "np.set_printoptions(precision=2)\n",
        "\n",
        "print (classification_report(y_test, yhat))\n",
        "\n",
        "print(score_df)\n"
      ],
      "execution_count": 262,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.92      1.00      0.96     13632\n",
            "           1       0.00      0.00      0.00      1200\n",
            "\n",
            "    accuracy                           0.92     14832\n",
            "   macro avg       0.46      0.50      0.48     14832\n",
            "weighted avg       0.84      0.92      0.88     14832\n",
            "\n",
            "   accuracy  precision  recall   f1  auc\n",
            "0  0.919094        0.0     0.0  0.0  0.5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "xDJBf5ec513_",
        "outputId": "675d3de2-a09c-459b-dd7b-5004259628ab"
      },
      "source": [
        "sns.set_context('talk')\n",
        "cm = confusion_matrix(y_test, yhat)\n",
        "ax = sns.heatmap(cm, annot=True, fmt='d', cmap=\"YlGnBu\")\n",
        "\n",
        "labels = ['0', '1']\n",
        "ax.set_xticklabels(labels, fontsize=15);\n",
        "ax.set_yticklabels(labels[::1], fontsize=15);\n",
        "ax.set_ylabel('Ground Truth', fontsize=20);\n",
        "ax.set_xlabel('Prediction', fontsize=20)\n",
        "ax.set_title('\\n Confusion Matrix for Support Vector Machine \\n')\n",
        "### END SOLUTION"
      ],
      "execution_count": 263,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, '\\n Confusion Matrix for Support Vector Machine \\n')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 263
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SbCBtEXxSymh"
      },
      "source": [
        "## 2. Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XmCOjV66S1ih"
      },
      "source": [
        "# Suppress warnings about too few trees from the early models\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\", category=UserWarning)\n",
        "warnings.filterwarnings(\"ignore\", category=RuntimeWarning)"
      ],
      "execution_count": 289,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 360
        },
        "id": "ITANmBdaTGqY",
        "outputId": "51a89cb4-48d1-4e03-cc05-137516b9deb8"
      },
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "# Initialize the random forest estimator\n",
        "# Note that the number of trees is not setup here\n",
        "RF = RandomForestClassifier(oob_score=True, \n",
        "                            random_state=42, \n",
        "                            warm_start=True,\n",
        "                            n_jobs=-1)\n",
        "\n",
        "oob_list = list()\n",
        "\n",
        "# Iterate through all of the possibilities for \n",
        "# number of trees\n",
        "for n_trees in [15, 20, 30, 40, 50, 100, 150, 200, 300, 400]:\n",
        "    \n",
        "    # Use this to set the number of trees\n",
        "    RF.set_params(n_estimators=n_trees)\n",
        "\n",
        "    # Fit the model\n",
        "    RF_ = RF.fit(X_train, y_train)\n",
        "\n",
        "    # Get the oob error\n",
        "    oob_error = 1 - RF.oob_score_\n",
        "    \n",
        "    # Store it\n",
        "    oob_list.append(pd.Series({'n_trees': n_trees, 'oob': oob_error}))\n",
        "\n",
        "rf_oob_df = pd.concat(oob_list, axis=1).T.set_index('n_trees')\n",
        "\n",
        "rf_oob_df"
      ],
      "execution_count": 292,
      "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>oob</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>n_trees</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>15.0</th>\n",
              "      <td>0.006271</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20.0</th>\n",
              "      <td>0.005577</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30.0</th>\n",
              "      <td>0.005548</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>40.0</th>\n",
              "      <td>0.005606</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50.0</th>\n",
              "      <td>0.005548</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>100.0</th>\n",
              "      <td>0.005462</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>150.0</th>\n",
              "      <td>0.005433</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>200.0</th>\n",
              "      <td>0.005375</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>300.0</th>\n",
              "      <td>0.005433</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>400.0</th>\n",
              "      <td>0.005375</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "              oob\n",
              "n_trees          \n",
              "15.0     0.006271\n",
              "20.0     0.005577\n",
              "30.0     0.005548\n",
              "40.0     0.005606\n",
              "50.0     0.005548\n",
              "100.0    0.005462\n",
              "150.0    0.005433\n",
              "200.0    0.005375\n",
              "300.0    0.005433\n",
              "400.0    0.005375"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 292
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 463
        },
        "id": "NDOuekxJTG6G",
        "outputId": "106b4427-5138-4a1d-aee5-036f72e8f97e"
      },
      "source": [
        "sns.set_context('talk')\n",
        "sns.set_style('white')\n",
        "\n",
        "ax = rf_oob_df.plot(legend=False, marker='o', figsize=(14, 7), linewidth=5)\n",
        "ax.set(ylabel='out-of-bag error');"
      ],
      "execution_count": 294,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H4DMJCECTG9p"
      },
      "source": [
        "model = RF.set_params(n_estimators=100)\n",
        "\n",
        "y_predRF = model.predict(X_test)"
      ],
      "execution_count": 295,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WePr2zxXcY91",
        "outputId": "ba12d6b4-bc53-487e-83aa-53ea1ab44bf9"
      },
      "source": [
        "# Plot non-normalized confusion matrix\n",
        "cr = classification_report(y_test, y_predRF)\n",
        "print(cr)\n",
        "\n",
        "score_df = pd.DataFrame({'accuracy': accuracy_score(y_test, y_predRF),\n",
        "                         'precision': precision_score(y_test, y_predRF),\n",
        "                         'recall': recall_score(y_test, y_predRF),\n",
        "                         'f1': f1_score(y_test, y_predRF),\n",
        "                         'auc': roc_auc_score(y_test, y_predRF)},\n",
        "                         index=pd.Index([0]))\n",
        "\n",
        "print(score_df)"
      ],
      "execution_count": 296,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00     13632\n",
            "           1       1.00      0.93      0.96      1200\n",
            "\n",
            "    accuracy                           0.99     14832\n",
            "   macro avg       1.00      0.96      0.98     14832\n",
            "weighted avg       0.99      0.99      0.99     14832\n",
            "\n",
            "   accuracy  precision    recall        f1       auc\n",
            "0  0.993999        1.0  0.925833  0.961489  0.962917\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "3a5RT_9K5-mt",
        "outputId": "f8f82218-2284-4863-f190-b1553f96c97e"
      },
      "source": [
        "sns.set_context('talk')\n",
        "cm = confusion_matrix(y_test, y_predRF)\n",
        "ax = sns.heatmap(cm, annot=True, fmt='d', cmap=\"YlGnBu\")\n",
        "### END SOLUTION\n",
        "\n",
        "labels = ['0', '1']\n",
        "ax.set_xticklabels(labels, fontsize=15);\n",
        "ax.set_yticklabels(labels[::1], fontsize=15);\n",
        "ax.set_ylabel('Ground Truth', fontsize=20);\n",
        "ax.set_xlabel('Prediction', fontsize=20)\n",
        "\n",
        "ax.set_title('\\n Confusion Matrix for Random Forest \\n')\n",
        "### END SOLUTION"
      ],
      "execution_count": 297,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, '\\n Confusion Matrix for Random Forest \\n')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 297
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5Xv2gUDeVXrY"
      },
      "source": [
        "## 3. Gradient Boosting"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "OpjZX7zDTHOe",
        "outputId": "a143ba1a-7a48-4df0-be11-95878e2a6195"
      },
      "source": [
        "### BEGIN SOLUTION\n",
        "from sklearn.ensemble import GradientBoostingClassifier\n",
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "error_list = list()\n",
        "\n",
        "# Iterate through various possibilities for number of trees\n",
        "tree_list = [15, 20, 30, 50, 100, 200]\n",
        "for n_trees in tree_list:\n",
        "    \n",
        "    # Initialize the gradient boost classifier\n",
        "    GBC = GradientBoostingClassifier(n_estimators=n_trees, random_state=42)\n",
        "\n",
        "    # Fit the model\n",
        "    print(f'Fitting model with {n_trees} trees')\n",
        "    GBC.fit(X_train.values, y_train.values)\n",
        "    y_predG = GBC.predict(X_test)\n",
        "\n",
        "    # Get the error\n",
        "    error = 1.0 - accuracy_score(y_test, y_predG)\n",
        "    \n",
        "    # Store it\n",
        "    error_list.append(pd.Series({'n_trees': n_trees, 'error': error}))\n",
        "\n",
        "error_df = pd.concat(error_list, axis=1).T.set_index('n_trees')\n",
        "\n",
        "error_df"
      ],
      "execution_count": 298,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fitting model with 15 trees\n",
            "Fitting model with 20 trees\n",
            "Fitting model with 30 trees\n",
            "Fitting model with 50 trees\n",
            "Fitting model with 100 trees\n",
            "Fitting model with 200 trees\n"
          ],
          "name": "stdout"
        },
        {
          "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>error</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>n_trees</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>15.0</th>\n",
              "      <td>0.006540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20.0</th>\n",
              "      <td>0.006540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30.0</th>\n",
              "      <td>0.006540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50.0</th>\n",
              "      <td>0.006607</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>100.0</th>\n",
              "      <td>0.006270</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>200.0</th>\n",
              "      <td>0.005663</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            error\n",
              "n_trees          \n",
              "15.0     0.006540\n",
              "20.0     0.006540\n",
              "30.0     0.006540\n",
              "50.0     0.006607\n",
              "100.0    0.006270\n",
              "200.0    0.005663"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 298
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EbLkYcmaTHV-"
      },
      "source": [
        "### BEGIN SOLUTION\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "tree_list = [15, 20]\n",
        "# The parameters to be fit\n",
        "param_grid = {'n_estimators': tree_list,\n",
        "              'learning_rate': [0.1, 0.01, 0.001, 0.0001],\n",
        "              'subsample': [1.0, 0.5],\n",
        "              'max_features': [1, 2, 3, 4]}\n",
        "\n",
        "# The grid search object\n",
        "GV_GBC = GridSearchCV(GradientBoostingClassifier(random_state=42), \n",
        "                      param_grid=param_grid, \n",
        "                      scoring='accuracy',\n",
        "                      n_jobs=-1)\n",
        "\n",
        "# Do the grid search\n",
        "GV_GBC = GV_GBC.fit(X_train, y_train)"
      ],
      "execution_count": 299,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uc50LdXWTHZH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "bee7f556-039c-4873-fea9-40a6d1232eee"
      },
      "source": [
        "# The best model\n",
        "GV_GBC.best_estimator_"
      ],
      "execution_count": 300,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GradientBoostingClassifier(ccp_alpha=0.0, criterion='friedman_mse', init=None,\n",
              "                           learning_rate=0.1, loss='deviance', max_depth=3,\n",
              "                           max_features=4, max_leaf_nodes=None,\n",
              "                           min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                           min_samples_leaf=1, min_samples_split=2,\n",
              "                           min_weight_fraction_leaf=0.0, n_estimators=20,\n",
              "                           n_iter_no_change=None, presort='deprecated',\n",
              "                           random_state=42, subsample=0.5, tol=0.0001,\n",
              "                           validation_fraction=0.1, verbose=0,\n",
              "                           warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 300
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f9av8FPIV3tW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "68616d14-e11b-4b19-b3d0-7582f5e8224c"
      },
      "source": [
        "# Plot non-normalized confusion matrix\n",
        "\n",
        "\n",
        "from sklearn.metrics import classification_report, accuracy_score, precision_score, recall_score\n",
        "from sklearn.metrics import f1_score, roc_auc_score\n",
        "\n",
        "cr = classification_report(y_test, y_predG)\n",
        "print(cr)\n",
        "\n",
        "score_df1 = pd.DataFrame({'accuracy': accuracy_score(y_test, y_predG),\n",
        "                         'precision': precision_score(y_test, y_predG),\n",
        "                         'recall': recall_score(y_test, y_predG),\n",
        "                         'f1': f1_score(y_test, y_predG),\n",
        "                         'auc': roc_auc_score(y_test, y_predG)},\n",
        "                         index=pd.Index([0]))\n",
        "\n",
        "print(score_df1)"
      ],
      "execution_count": 301,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00     13632\n",
            "           1       0.99      0.94      0.96      1200\n",
            "\n",
            "    accuracy                           0.99     14832\n",
            "   macro avg       0.99      0.97      0.98     14832\n",
            "weighted avg       0.99      0.99      0.99     14832\n",
            "\n",
            "   accuracy  precision    recall        f1      auc\n",
            "0  0.994337   0.992933  0.936667  0.963979  0.96804\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "sflpAz1o6ECO",
        "outputId": "87f5753a-9f22-4da2-9406-7aa9bae47419"
      },
      "source": [
        "sns.set_context('talk')\n",
        "cm = confusion_matrix(y_test, y_predG)\n",
        "ax = sns.heatmap(cm, annot=True, fmt='d', cmap=\"YlGnBu\")\n",
        "### END SOLUTION\n",
        "labels = ['0', '1']\n",
        "ax.set_xticklabels(labels, fontsize=15);\n",
        "ax.set_yticklabels(labels[::1], fontsize=15);\n",
        "ax.set_ylabel('Ground Truth', fontsize=20);\n",
        "ax.set_xlabel('Prediction', fontsize=20)\n",
        "\n",
        "ax.set_title('\\n Confusion Matrix for Gradient Boosting \\n')\n",
        "### END SOLUTION"
      ],
      "execution_count": 302,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, '\\n Confusion Matrix for Gradient Boosting \\n')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 302
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 568
        },
        "id": "clmiM4j-ewLZ",
        "outputId": "14f74e09-441d-4960-973c-0cf814a70383"
      },
      "source": [
        "\n",
        "from sklearn.metrics import roc_curve, precision_recall_curve, confusion_matrix\n",
        "sns.set_context('talk')\n",
        "\n",
        "fig, axList = plt.subplots(ncols=2)\n",
        "fig.set_size_inches(16, 8)\n",
        "\n",
        "# Get the probabilities for each of the two categories\n",
        "y_prob = GV_GBC.predict_proba(X_test)\n",
        "\n",
        "# Plot the ROC-AUC curve\n",
        "ax = axList[0]\n",
        "\n",
        "fpr, tpr, thresholds = roc_curve(y_test, y_prob[:,1])\n",
        "ax.plot(fpr, tpr, color='blue', linewidth=5)\n",
        "# It is customary to draw a diagonal dotted line in ROC plots.\n",
        "# This is to indicate completely random prediction. Deviation from this\n",
        "# dotted line towards the upper left corner signifies the power of the model.\n",
        "ax.plot([0, 1], [0, 1], ls='--', color='black', lw=.3)\n",
        "ax.set(xlabel='False Positive Rate',\n",
        "       ylabel='True Positive Rate',\n",
        "       xlim=[-.01, 1.01], ylim=[-.01, 1.01],\n",
        "       title='ROC curve')\n",
        "ax.grid(True)\n",
        "\n",
        "# Plot the precision-recall curve\n",
        "ax = axList[1]\n",
        "\n",
        "precision, recall, _ = precision_recall_curve(y_test, y_prob[:,1])\n",
        "ax.plot(recall, precision, color='green', linewidth=5)\n",
        "ax.set(xlabel='Recall', ylabel='Precision',\n",
        "       xlim=[-.01, 1.01], ylim=[-.01, 1.01],\n",
        "       title='Precision-Recall curve')\n",
        "ax.grid(True)\n",
        "\n",
        "plt.tight_layout()"
      ],
      "execution_count": 303,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R_5kpH73259I"
      },
      "source": [
        "## 4. Logistic Regresion"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0QXtmjEl3PAC"
      },
      "source": [
        "### BEGIN SOLUTION\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "# L2 regularized logistic regression\n",
        "LR_L2 = LogisticRegression(penalty='l2', max_iter=500, solver='saga').fit(X_train, y_train)"
      ],
      "execution_count": 304,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bdezb4w-3djs",
        "outputId": "a812a0d2-4e8d-4b6d-ade5-c88488683d3b"
      },
      "source": [
        "y_predLR = LR_L2.predict(X_test)\n",
        "print(classification_report(y_predLR, y_test))\n",
        "\n",
        "score_df = pd.DataFrame({'accuracy': accuracy_score(y_test, y_predLR),\n",
        "                         'precision': precision_score(y_test, y_predLR),\n",
        "                         'recall': recall_score(y_test, y_predLR),\n",
        "                         'f1': f1_score(y_test, y_predLR),\n",
        "                         'auc': roc_auc_score(y_test, y_predLR)},\n",
        "                         index=pd.Index([0]))\n",
        "\n",
        "print(score_df)"
      ],
      "execution_count": 305,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.92      0.96     14832\n",
            "           1       0.00      0.00      0.00         0\n",
            "\n",
            "    accuracy                           0.92     14832\n",
            "   macro avg       0.50      0.46      0.48     14832\n",
            "weighted avg       1.00      0.92      0.96     14832\n",
            "\n",
            "   accuracy  precision  recall   f1  auc\n",
            "0  0.919094        0.0     0.0  0.0  0.5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "hm0gqt_Q9lj2",
        "outputId": "919d4335-edc1-4973-c864-2ba3b20ced58"
      },
      "source": [
        "sns.set_context('talk')\n",
        "cm = confusion_matrix(y_test, y_predLR)\n",
        "ax = sns.heatmap(cm, annot=True, fmt='d', cmap=\"YlGnBu\")\n",
        "### END SOLUTION\n",
        "\n",
        "labels = ['0', '1']\n",
        "ax.set_xticklabels(labels, fontsize=15);\n",
        "ax.set_yticklabels(labels[::1], fontsize=15);\n",
        "ax.set_ylabel('Ground Truth', fontsize=20);\n",
        "ax.set_xlabel('Prediction', fontsize=20)\n",
        "\n",
        "ax.set_title(' \\n Confusion Matrix for Logistic Regression \\n')\n",
        "### END SOLUTION"
      ],
      "execution_count": 306,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, ' \\n Confusion Matrix for Logistic Regression \\n')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 306
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YGiPnouD934Q"
      },
      "source": [
        "## 5. Stacking Random Forest & Gradient Boosting (using Voting Classifier)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dK02204j25vh"
      },
      "source": [
        "from sklearn.ensemble import VotingClassifier\n",
        "\n",
        "# The combined model--logistic regression and gradient boosted trees\n",
        "estimators = [('RF', RF_), ('GBC', GV_GBC)]\n",
        "\n",
        "# Though it wasn't done here, it is often desirable to train \n",
        "# this model using an additional hold-out data set and/or with cross validation\n",
        "VC = VotingClassifier(estimators, voting='soft')\n",
        "VC = VC.fit(X_train, y_train)\n"
      ],
      "execution_count": 279,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7ah0vqVB3sxU",
        "outputId": "abc48db2-4e70-4a3a-f714-da1bddfebc79"
      },
      "source": [
        "y_predSt = VC.predict(X_test)\n",
        "print(classification_report(y_test, y_predSt))\n",
        "\n",
        "score_df1 = pd.DataFrame({'accuracy': accuracy_score(y_test, y_predSt),\n",
        "                         'precision': precision_score(y_test, y_predSt),\n",
        "                         'recall': recall_score(y_test, y_predSt),\n",
        "                         'f1': f1_score(y_test, y_predSt),\n",
        "                         'auc': roc_auc_score(y_test, y_predSt)},\n",
        "                         index=pd.Index([0]))\n",
        "\n",
        "print(score_df1)"
      ],
      "execution_count": 280,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00     13632\n",
            "           1       1.00      0.92      0.96      1200\n",
            "\n",
            "    accuracy                           0.99     14832\n",
            "   macro avg       1.00      0.96      0.98     14832\n",
            "weighted avg       0.99      0.99      0.99     14832\n",
            "\n",
            "   accuracy  precision    recall        f1       auc\n",
            "0   0.99346        1.0  0.919167  0.957881  0.959583\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "XwrvKmdv3t3z",
        "outputId": "b5f25921-f1ea-46fe-ebe1-1cbabc3e6396"
      },
      "source": [
        "sns.set_context('talk')\n",
        "cm = confusion_matrix(y_test, y_predSt)\n",
        "ax = sns.heatmap(cm, annot=True, fmt='d', cmap=\"YlGnBu\")\n",
        "### END SOLUTION\n",
        "\n",
        "labels = ['0', '1']\n",
        "ax.set_xticklabels(labels, fontsize=15);\n",
        "ax.set_yticklabels(labels[::1], fontsize=15);\n",
        "ax.set_ylabel('Ground Truth', fontsize=20);\n",
        "ax.set_xlabel('Prediction', fontsize=20)\n",
        "\n",
        "ax.set_title(' \\n Confusion Matrix for Stacking Random Forest & Gradient Boosting \\n ')"
      ],
      "execution_count": 281,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, ' \\n Confusion Matrix for Stacking Random Forest & Gradient Boosting \\n ')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 281
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "99zGQoU28M-j"
      },
      "source": [
        "## Addition : Check the correlation of each variable\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 787
        },
        "id": "KvT2Q2gU8SAu",
        "outputId": "849c8281-c4fe-4f26-b5b3-cc4cad4aee19"
      },
      "source": [
        "corrmat = data.corr(method='spearman')\n",
        "f, ax = plt.subplots(figsize=(12, 10))\n",
        "\n",
        "matrix = np.triu(corrmat)\n",
        "sns.heatmap(corrmat, ax=ax, cmap=\"YlGnBu\", linewidths=0.1, mask=matrix, annot = False)"
      ],
      "execution_count": 307,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fa9d8034278>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 307
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AG4XNWNHwWxO",
        "outputId": "224ea920-ce51-40b3-c673-39777a69f814"
      },
      "source": [
        "y = data4['target']\n",
        "fields = list(data4.columns[:-1])  # everything except \"color\"\n",
        "correlations = data4[fields].corrwith(y)\n",
        "correlations.sort_values(inplace=True)\n",
        "correlations\n"
      ],
      "execution_count": 308,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "status                 -0.875445\n",
              "founded_year           -0.134102\n",
              "seed                   -0.013787\n",
              "grant                  -0.006466\n",
              "equity_crowdfunding    -0.005853\n",
              "product_crowdfunding   -0.004435\n",
              "secondary_market       -0.002787\n",
              "round_G                -0.001910\n",
              "round_H                -0.001554\n",
              "angel                  -0.001353\n",
              "convertible_note       -0.001075\n",
              "debt_financing          0.000162\n",
              "round_F                 0.000280\n",
              "market                  0.004286\n",
              "undisclosed             0.004894\n",
              "private_equity          0.006603\n",
              "post_ipo_debt           0.007197\n",
              "round_E                 0.009743\n",
              "region                  0.011954\n",
              "round_D                 0.025184\n",
              "funding_total_usd       0.028143\n",
              "round_A                 0.046159\n",
              "round_C                 0.052987\n",
              "country_code            0.053599\n",
              "venture                 0.064762\n",
              "round_B                 0.072362\n",
              "funding_rounds          0.084266\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 308
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BZ_ZWrU31xb3",
        "outputId": "7ea9c39a-0ddf-4ae7-acc8-839437a24d2f"
      },
      "source": [
        "correlations.map(abs).sort_values(ascending=False)"
      ],
      "execution_count": 309,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "status                  0.875445\n",
              "founded_year            0.134102\n",
              "funding_rounds          0.084266\n",
              "round_B                 0.072362\n",
              "venture                 0.064762\n",
              "country_code            0.053599\n",
              "round_C                 0.052987\n",
              "round_A                 0.046159\n",
              "funding_total_usd       0.028143\n",
              "round_D                 0.025184\n",
              "seed                    0.013787\n",
              "region                  0.011954\n",
              "round_E                 0.009743\n",
              "post_ipo_debt           0.007197\n",
              "private_equity          0.006603\n",
              "grant                   0.006466\n",
              "equity_crowdfunding     0.005853\n",
              "undisclosed             0.004894\n",
              "product_crowdfunding    0.004435\n",
              "market                  0.004286\n",
              "secondary_market        0.002787\n",
              "round_G                 0.001910\n",
              "round_H                 0.001554\n",
              "angel                   0.001353\n",
              "convertible_note        0.001075\n",
              "round_F                 0.000280\n",
              "debt_financing          0.000162\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 309
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 553
        },
        "id": "hF3l6vKq1ttV",
        "outputId": "89918cb5-19c2-4b8a-91f2-7555099bd130"
      },
      "source": [
        "\n",
        "all = list(data4.columns[::])\n",
        "feature_imp = pd.Series(GV_GBC.best_estimator_.feature_importances_, index=all).sort_values(ascending=False)\n",
        "ax = feature_imp.plot(kind='bar', figsize=(16, 6))\n",
        "ax.set(ylabel='Relative Importance');\n",
        "ax.set(ylabel='Feature');"
      ],
      "execution_count": 310,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        },
        "id": "_lJlOZca65i5",
        "outputId": "654c14a7-3b80-46bb-bebc-20e86923c223"
      },
      "source": [
        "ax = correlations.plot(kind='bar', color='blue', fontsize=10)\n",
        "ax.set(ylim=[-1, 1], ylabel='pearson correlation');"
      ],
      "execution_count": 311,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ydd24I3jyLa3",
        "outputId": "13b03c9d-7ef2-4678-ee47-5bec12842c9c"
      },
      "source": [
        "fields = correlations.map(abs).sort_values().iloc[-2:].index\n",
        "print(fields)\n",
        "X = data4[fields]\n",
        "X = pd.DataFrame(X, columns=['%s_scaled' % fld for fld in fields])\n",
        "print(X.columns)"
      ],
      "execution_count": 312,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Index(['founded_year', 'status'], dtype='object')\n",
            "Index(['founded_year_scaled', 'status_scaled'], dtype='object')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FXspfid3ODiC"
      },
      "source": [
        ""
      ],
      "execution_count": 288,
      "outputs": []
    }
  ]
}