Survival-Analysis-for-Data-Analysis-Introduction-colab.ipynb

survival-analysis-for-data-analysis-introduction-colab.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.2"
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/alonsosilvaallende/4c1372470c096db3121601387872a415/survival-analysis-for-data-analysis-introduction-colab.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aEj365e4uTvD"
      },
      "source": [
        "# Survival Analysis"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "heading_collapsed": true,
        "id": "0Z1wEfUFuTvG"
      },
      "source": [
        "* Historically, survival analysis was developed and used by actuaries\n",
        "and medical researchers to measure the lifetime of populations.\n",
        "* What's the expected lifetime of patients that were given drug A? drug B?\n",
        "* What's the life-expectancy of a baby born today in France?\n",
        "\n",
        "These researchers wanted to measure the duration between *Birth* and *Death*\n",
        "\n",
        "\n",
        "Source: [Lifelines: Survival Analysis in Python](https://www.youtube.com/watch?v=XQfxndJH4UA)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JxSbfi7QuTvI"
      },
      "source": [
        "# Survival function and hazard function"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fXfTmltSuTvK"
      },
      "source": [
        "**Definition:** Let $T$ be a random variable called failure time.\n",
        "\n",
        "- $f(t)$ be its probability density function\n",
        "- $F(t):=\\mathcal{P}(T\\le t)$ its cumulative distribution function\n",
        "\n",
        "Then we define\n",
        "\n",
        "- The *survival function* $S(t):=\\mathcal{P}(T>t)=1-F(t)$.\n",
        "- The *hazard function* (probability of failure between $t$ and $t+\\delta t$ knowing that it was working at time $t$):\n",
        "$$\n",
        "h(t):=\\lim_{\\delta t\\to0}\\frac{\\mathcal{P}(T<t+\\delta t|T>t)}{\\delta t}=\n",
        "\\lim_{\\delta t\\to0}\\frac{F(t+\\delta t)-F(t)}{\\delta t}\\times\\frac{1}{1-F(t)}=\\frac{f(t)}{1-F(t)}.\n",
        "$$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZBY8J3sxuTvM"
      },
      "source": [
        "**Properties:**\n",
        "- $S(t)=\\exp(-\\int_0^t h(s)\\,ds)$.\n",
        "- $h(t)=-\\frac{d}{dt}\\ln(S(t))$."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1VYNZWpmuTvO"
      },
      "source": [
        "# Right censoring\n",
        "\n",
        "By the end of the study, the event of interest (for example, in medicine \"death of a patient\" or \"churn of a customer\") has only occurred for a subset of the observations.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vao5BhTYuTvP"
      },
      "source": [
        "# Modern Survival Analysis\n",
        "\n",
        "+ **Birth:** Customer joins Netflix  \n",
        "**Death:** Customer leaves Netflix  \n",
        "**Censorship:** At the current time, I cannot see all cancelations  \n",
        "  \n",
        "  \n",
        "+ **Birth:** Leader forms government  \n",
        "**Death:** Government dissolves  \n",
        "**Censorship:** Death of leader or current time do not allow me to see all dissolvements  \n",
        "\n",
        "\n",
        "+ **Birth:** Couple starts dating  \n",
        "**Death:** Couple breaks-up  \n",
        "**Censorship:** Some couples never break-up (partner's death comes first)  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EVUDoqXOuTvU"
      },
      "source": [
        "First, let's take a dataset from lifelines to see what does it mean in practice."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IcqsnDunuZ3x",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "935b0677-18e1-421b-ccfe-86c1353fdedb"
      },
      "source": [
        "!pip install -q lifelines"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/349.7 KB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[91m╸\u001b[0m \u001b[32m348.2/349.7 KB\u001b[0m \u001b[31m13.5 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m349.7/349.7 KB\u001b[0m \u001b[31m6.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m77.1/77.1 KB\u001b[0m \u001b[31m5.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Building wheel for autograd-gamma (setup.py) ... \u001b[?25l\u001b[?25hdone\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:10.331529Z",
          "start_time": "2020-01-09T22:37:03.811697Z"
        },
        "id": "A5qmUxSxuTvW"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "plt.style.use('seaborn')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-Hz_oM2IuTve"
      },
      "source": [
        "from lifelines.datasets import load_dd\n",
        "\n",
        "df = load_dd()\n",
        "df = df[['ctryname', 'un_region_name', 'un_continent_name', 'ehead',\\\n",
        "         'democracy', 'regime', 'start_year', 'duration', 'observed']]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xCN-nZXDuTvk"
      },
      "source": [
        "# Democracy and dictatorship\n",
        "\n",
        "This dataset contains a classification of political regimes as democracy and dictatorship.\n",
        "* Classification of democracies as \n",
        "  + parliamentary, \n",
        "  + semi-presidential (mixed), and \n",
        "  + presidential. \n",
        "  \n",
        "* Classification of dictatorships as \n",
        "  + military, \n",
        "  + civilian, and \n",
        "  + royal. \n",
        "  \n",
        "Coverage: 202 countries, from 1946 or year of independence to 2008.\n",
        "\n",
        "**References**\n",
        "\n",
        "José Antonio Cheibub, Jennifer Gandhi, and James Raymond Vreeland. [\"Democracy and Dictatorship Revisited.\"](https://doi.org/10.1007/s11127-009-9491-2) Public Choice, vol. 143, no. 2-1, pp. 67-101, 2010."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:10.515360Z",
          "start_time": "2020-01-09T22:37:10.466697Z"
        },
        "id": "9b51kQDjuTvm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "outputId": "75cd5f50-8acf-4f2e-cdb2-1f78262c28aa"
      },
      "source": [
        "df.tail(10).style.hide_index()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<pandas.io.formats.style.Styler at 0x7f84880718b0>"
            ],
            "text/html": [
              "<style type=\"text/css\">\n",
              "</style>\n",
              "<table id=\"T_ca6b0_\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th class=\"col_heading level0 col0\" >ctryname</th>\n",
              "      <th class=\"col_heading level0 col1\" >un_region_name</th>\n",
              "      <th class=\"col_heading level0 col2\" >un_continent_name</th>\n",
              "      <th class=\"col_heading level0 col3\" >ehead</th>\n",
              "      <th class=\"col_heading level0 col4\" >democracy</th>\n",
              "      <th class=\"col_heading level0 col5\" >regime</th>\n",
              "      <th class=\"col_heading level0 col6\" >start_year</th>\n",
              "      <th class=\"col_heading level0 col7\" >duration</th>\n",
              "      <th class=\"col_heading level0 col8\" >observed</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row0_col0\" class=\"data row0 col0\" >Yugoslavia</td>\n",
              "      <td id=\"T_ca6b0_row0_col1\" class=\"data row0 col1\" >Southern Europe</td>\n",
              "      <td id=\"T_ca6b0_row0_col2\" class=\"data row0 col2\" >Europe</td>\n",
              "      <td id=\"T_ca6b0_row0_col3\" class=\"data row0 col3\" >Stipe Suvar</td>\n",
              "      <td id=\"T_ca6b0_row0_col4\" class=\"data row0 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row0_col5\" class=\"data row0 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row0_col6\" class=\"data row0 col6\" >1988</td>\n",
              "      <td id=\"T_ca6b0_row0_col7\" class=\"data row0 col7\" >1</td>\n",
              "      <td id=\"T_ca6b0_row0_col8\" class=\"data row0 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row1_col0\" class=\"data row1 col0\" >Yugoslavia</td>\n",
              "      <td id=\"T_ca6b0_row1_col1\" class=\"data row1 col1\" >Southern Europe</td>\n",
              "      <td id=\"T_ca6b0_row1_col2\" class=\"data row1 col2\" >Europe</td>\n",
              "      <td id=\"T_ca6b0_row1_col3\" class=\"data row1 col3\" >Milan Pancevski</td>\n",
              "      <td id=\"T_ca6b0_row1_col4\" class=\"data row1 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row1_col5\" class=\"data row1 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row1_col6\" class=\"data row1 col6\" >1989</td>\n",
              "      <td id=\"T_ca6b0_row1_col7\" class=\"data row1 col7\" >1</td>\n",
              "      <td id=\"T_ca6b0_row1_col8\" class=\"data row1 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row2_col0\" class=\"data row2 col0\" >Yugoslavia</td>\n",
              "      <td id=\"T_ca6b0_row2_col1\" class=\"data row2 col1\" >Southern Europe</td>\n",
              "      <td id=\"T_ca6b0_row2_col2\" class=\"data row2 col2\" >Europe</td>\n",
              "      <td id=\"T_ca6b0_row2_col3\" class=\"data row2 col3\" >Borisav Jovic</td>\n",
              "      <td id=\"T_ca6b0_row2_col4\" class=\"data row2 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row2_col5\" class=\"data row2 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row2_col6\" class=\"data row2 col6\" >1990</td>\n",
              "      <td id=\"T_ca6b0_row2_col7\" class=\"data row2 col7\" >1</td>\n",
              "      <td id=\"T_ca6b0_row2_col8\" class=\"data row2 col8\" >0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row3_col0\" class=\"data row3 col0\" >Zambia</td>\n",
              "      <td id=\"T_ca6b0_row3_col1\" class=\"data row3 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row3_col2\" class=\"data row3 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row3_col3\" class=\"data row3 col3\" >Kenneth Kaunda</td>\n",
              "      <td id=\"T_ca6b0_row3_col4\" class=\"data row3 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row3_col5\" class=\"data row3 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row3_col6\" class=\"data row3 col6\" >1964</td>\n",
              "      <td id=\"T_ca6b0_row3_col7\" class=\"data row3 col7\" >27</td>\n",
              "      <td id=\"T_ca6b0_row3_col8\" class=\"data row3 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row4_col0\" class=\"data row4 col0\" >Zambia</td>\n",
              "      <td id=\"T_ca6b0_row4_col1\" class=\"data row4 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row4_col2\" class=\"data row4 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row4_col3\" class=\"data row4 col3\" >Frederick Chiluba</td>\n",
              "      <td id=\"T_ca6b0_row4_col4\" class=\"data row4 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row4_col5\" class=\"data row4 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row4_col6\" class=\"data row4 col6\" >1991</td>\n",
              "      <td id=\"T_ca6b0_row4_col7\" class=\"data row4 col7\" >11</td>\n",
              "      <td id=\"T_ca6b0_row4_col8\" class=\"data row4 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row5_col0\" class=\"data row5 col0\" >Zambia</td>\n",
              "      <td id=\"T_ca6b0_row5_col1\" class=\"data row5 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row5_col2\" class=\"data row5 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row5_col3\" class=\"data row5 col3\" >Levy Patrick Mwanawasa</td>\n",
              "      <td id=\"T_ca6b0_row5_col4\" class=\"data row5 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row5_col5\" class=\"data row5 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row5_col6\" class=\"data row5 col6\" >2002</td>\n",
              "      <td id=\"T_ca6b0_row5_col7\" class=\"data row5 col7\" >6</td>\n",
              "      <td id=\"T_ca6b0_row5_col8\" class=\"data row5 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row6_col0\" class=\"data row6 col0\" >Zambia</td>\n",
              "      <td id=\"T_ca6b0_row6_col1\" class=\"data row6 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row6_col2\" class=\"data row6 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row6_col3\" class=\"data row6 col3\" >Rupiah Bwezani Banda</td>\n",
              "      <td id=\"T_ca6b0_row6_col4\" class=\"data row6 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row6_col5\" class=\"data row6 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row6_col6\" class=\"data row6 col6\" >2008</td>\n",
              "      <td id=\"T_ca6b0_row6_col7\" class=\"data row6 col7\" >1</td>\n",
              "      <td id=\"T_ca6b0_row6_col8\" class=\"data row6 col8\" >0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row7_col0\" class=\"data row7 col0\" >Zimbabwe</td>\n",
              "      <td id=\"T_ca6b0_row7_col1\" class=\"data row7 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row7_col2\" class=\"data row7 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row7_col3\" class=\"data row7 col3\" >Ian Smith</td>\n",
              "      <td id=\"T_ca6b0_row7_col4\" class=\"data row7 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row7_col5\" class=\"data row7 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row7_col6\" class=\"data row7 col6\" >1965</td>\n",
              "      <td id=\"T_ca6b0_row7_col7\" class=\"data row7 col7\" >14</td>\n",
              "      <td id=\"T_ca6b0_row7_col8\" class=\"data row7 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row8_col0\" class=\"data row8 col0\" >Zimbabwe</td>\n",
              "      <td id=\"T_ca6b0_row8_col1\" class=\"data row8 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row8_col2\" class=\"data row8 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row8_col3\" class=\"data row8 col3\" >Abel Muzorewa</td>\n",
              "      <td id=\"T_ca6b0_row8_col4\" class=\"data row8 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row8_col5\" class=\"data row8 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row8_col6\" class=\"data row8 col6\" >1979</td>\n",
              "      <td id=\"T_ca6b0_row8_col7\" class=\"data row8 col7\" >1</td>\n",
              "      <td id=\"T_ca6b0_row8_col8\" class=\"data row8 col8\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_ca6b0_row9_col0\" class=\"data row9 col0\" >Zimbabwe</td>\n",
              "      <td id=\"T_ca6b0_row9_col1\" class=\"data row9 col1\" >Eastern Africa</td>\n",
              "      <td id=\"T_ca6b0_row9_col2\" class=\"data row9 col2\" >Africa</td>\n",
              "      <td id=\"T_ca6b0_row9_col3\" class=\"data row9 col3\" >Robert Mugabe</td>\n",
              "      <td id=\"T_ca6b0_row9_col4\" class=\"data row9 col4\" >Non-democracy</td>\n",
              "      <td id=\"T_ca6b0_row9_col5\" class=\"data row9 col5\" >Civilian Dict</td>\n",
              "      <td id=\"T_ca6b0_row9_col6\" class=\"data row9 col6\" >1980</td>\n",
              "      <td id=\"T_ca6b0_row9_col7\" class=\"data row9 col7\" >29</td>\n",
              "      <td id=\"T_ca6b0_row9_col8\" class=\"data row9 col8\" >0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-o57OZxvuTvu"
      },
      "source": [
        "Let's look at right-censored samples."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:12.185066Z",
          "start_time": "2020-01-09T22:37:10.519270Z"
        },
        "scrolled": false,
        "id": "vzaXN9QWuTvv",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 394
        },
        "outputId": "766d33ac-4f02-4252-b0e8-5842a0ca0af6"
      },
      "source": [
        "format_dict = {'ehead':'{}','duration':'{}', 'observed':'{}'}\n",
        "(df.query('ctryname == \"United States of America\"')[['ehead', 'duration', 'observed']].style.format(format_dict)\n",
        "  .hide_index()\n",
        "  .highlight_min('observed', color='lightgreen'))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<pandas.io.formats.style.Styler at 0x7f84a80f4df0>"
            ],
            "text/html": [
              "<style type=\"text/css\">\n",
              "#T_f10c3_row2_col2, #T_f10c3_row10_col2 {\n",
              "  background-color: lightgreen;\n",
              "}\n",
              "</style>\n",
              "<table id=\"T_f10c3_\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th class=\"col_heading level0 col0\" >ehead</th>\n",
              "      <th class=\"col_heading level0 col1\" >duration</th>\n",
              "      <th class=\"col_heading level0 col2\" >observed</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row0_col0\" class=\"data row0 col0\" >Harry Truman</td>\n",
              "      <td id=\"T_f10c3_row0_col1\" class=\"data row0 col1\" >7</td>\n",
              "      <td id=\"T_f10c3_row0_col2\" class=\"data row0 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row1_col0\" class=\"data row1 col0\" >Dwight D. Eisenhower</td>\n",
              "      <td id=\"T_f10c3_row1_col1\" class=\"data row1 col1\" >8</td>\n",
              "      <td id=\"T_f10c3_row1_col2\" class=\"data row1 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row2_col0\" class=\"data row2 col0\" >John Kennedy</td>\n",
              "      <td id=\"T_f10c3_row2_col1\" class=\"data row2 col1\" >2</td>\n",
              "      <td id=\"T_f10c3_row2_col2\" class=\"data row2 col2\" >0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row3_col0\" class=\"data row3 col0\" >Lyndon Johnson</td>\n",
              "      <td id=\"T_f10c3_row3_col1\" class=\"data row3 col1\" >6</td>\n",
              "      <td id=\"T_f10c3_row3_col2\" class=\"data row3 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row4_col0\" class=\"data row4 col0\" >Richard Nixon</td>\n",
              "      <td id=\"T_f10c3_row4_col1\" class=\"data row4 col1\" >5</td>\n",
              "      <td id=\"T_f10c3_row4_col2\" class=\"data row4 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row5_col0\" class=\"data row5 col0\" >Gerald Ford</td>\n",
              "      <td id=\"T_f10c3_row5_col1\" class=\"data row5 col1\" >3</td>\n",
              "      <td id=\"T_f10c3_row5_col2\" class=\"data row5 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row6_col0\" class=\"data row6 col0\" >Jimmy Carter</td>\n",
              "      <td id=\"T_f10c3_row6_col1\" class=\"data row6 col1\" >4</td>\n",
              "      <td id=\"T_f10c3_row6_col2\" class=\"data row6 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row7_col0\" class=\"data row7 col0\" >Ronald Reagan</td>\n",
              "      <td id=\"T_f10c3_row7_col1\" class=\"data row7 col1\" >8</td>\n",
              "      <td id=\"T_f10c3_row7_col2\" class=\"data row7 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row8_col0\" class=\"data row8 col0\" >George Bush</td>\n",
              "      <td id=\"T_f10c3_row8_col1\" class=\"data row8 col1\" >4</td>\n",
              "      <td id=\"T_f10c3_row8_col2\" class=\"data row8 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row9_col0\" class=\"data row9 col0\" >Bill Clinton</td>\n",
              "      <td id=\"T_f10c3_row9_col1\" class=\"data row9 col1\" >8</td>\n",
              "      <td id=\"T_f10c3_row9_col2\" class=\"data row9 col2\" >1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td id=\"T_f10c3_row10_col0\" class=\"data row10 col0\" >George W. Bush</td>\n",
              "      <td id=\"T_f10c3_row10_col1\" class=\"data row10 col1\" >8</td>\n",
              "      <td id=\"T_f10c3_row10_col2\" class=\"data row10 col2\" >0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:12.216114Z",
          "start_time": "2020-01-09T22:37:12.189453Z"
        },
        "id": "AVHDxLziuTv2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3923b2da-86ea-4909-f525-41bc0d458f6e"
      },
      "source": [
        "print(f'samples: {len(df)}\\n')\n",
        "print(f'right censored samples: {len(df.query(\"observed == 0\"))}')\n",
        "print(f'right censored samples (%): {100*len(df.query(\"observed == 0\"))/len(df):.1f}%')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "samples: 1808\n",
            "\n",
            "right censored samples: 340\n",
            "right censored samples (%): 18.8%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3v0OGKU_uTv9"
      },
      "source": [
        "# How can we estimate the probability of a government survival?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J0UzmdpeuTv9"
      },
      "source": [
        "**Example:** I want to estimate the survival function of a new machine and I have 100 of these new machines. After the first year:\n",
        "\n",
        "Samples | I \n",
        "--- | --- \n",
        "Initial numbers | 100 \n",
        "Deaths in first year of age | 70 \n",
        "One-year survivors | `30` \n",
        "\n",
        "Therefore, a reasonable estimate of the survival probability of 1 year is 0.3.\n",
        "\n",
        "I have increased my production. So now I have 1000 new machines.\n",
        "\n",
        "Samples | I | II\n",
        "--- | --- | ---\n",
        "Initial numbers | 100 | 1000\n",
        "Deaths in first year of age | 70 | 750\n",
        "One-year survivors | `30` | `250`\n",
        "Deaths in second year of age | 15 | \n",
        "Two-year survivors | `15` | \n",
        "\n",
        "What would be a good estimate of the survival probability of 1 year? and 2 years?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RVevvzzCuTv-"
      },
      "source": [
        "The estimate of the probability of survival of 1 year would be \n",
        "$\\hat{P}(1)=(30+250)/(100+1000) \\sim 0.255$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BlBS5dveuTwA"
      },
      "source": [
        "$\\hat{P}(2|1) = 15/30 = 0.5$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zVja8-P5uTwA"
      },
      "source": [
        "$\\hat{P}(2)=0.255\\times0.5=0.127$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pG5yhITUuTwB"
      },
      "source": [
        "# Kaplan-Meier estimator"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E6hn2l6PuTwC"
      },
      "source": [
        "**Definition:** Kaplan-Meier estimator of the survival function is given by\n",
        "\n",
        "$$\n",
        "\\hat{S}(t):=\\prod_{i:t_i\\le t}\\left(1-\\frac{d_i}{n_i}\\right)\n",
        "$$\n",
        "where $t_i$ is a time where at least one event happened, $d_i$ the number of events that happened at time $t_i$, and $n_i$ the individuals known to have survived up to time $t_i$."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cojEUEjsuTwD"
      },
      "source": [
        "We use the [Kaplan-Meier estimator](https://en.wikipedia.org/wiki/Kaplan?Meier_estimator) to estimate the probability of a government survival."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:16.155051Z",
          "start_time": "2020-01-09T22:37:12.219801Z"
        },
        "scrolled": false,
        "id": "LCGcjoexuTwE",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "6a0e3fd6-393b-48d3-c917-4c8f17f25ba8"
      },
      "source": [
        "from lifelines import KaplanMeierFitter\n",
        "kmf = KaplanMeierFitter()\n",
        "kmf.fit(df['duration'],df['observed'], label='Estimate for average government')\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(10,7))\n",
        "kmf.plot(ax=ax)\n",
        "plt.title('Estimated probability of government survival vs number of years')\n",
        "plt.xlabel('Time (in years)')\n",
        "plt.ylabel('Estimated probability of government survival')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAG4CAYAAAB7FoK6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeXwM9/8H8NdsNkQkQchR99E6EuKmrkYikYTUTYOQ0lbdCaUI6kxQqkpdbd20paRE666jou6rFUeLIkpE5BaSze7n90d+ma+IZNaxQ9br+Xh4yOx+5jPvnc8keeUzs7OSEEKAiIiIiFShedkFEBEREb1OGL6IiIiIVMTwRURERKQihi8iIiIiFTF8EREREamI4YuIiIhIRQxfZDI1atSAt7c3fH19c/37888/C1xvw4YN8tdBQUGIjo5+IfVkZmZi8+bNT73eokWLMHbs2BdSw+Nu3rwJFxeXp16vT58+2LJlS57H79y5A39/fwDAggULMH78eAC59+Oj+/d57dmzBy1btsSkSZNeWJ+F1cGDB3Hr1i1Vt7l27VrMmzfvmdZ91mPvZYqIiMD777+v2vbWrVuHFi1aYPHixaptk14P2pddAJm3NWvWwNnZ2ej2er0en3/+OXr06AEAWLVq1Qur5fz589i8eTM6der0wvp81Tg5OeGXX37J83jOfnx8/z6vvXv3olu3bggJCXkh/RVmK1euxKBBg1C2bFnVthkYGKjatl5Hu3btQkhICLp37/6ySyEzw5kveimysrIwfvx4+Pj4wNvbG0OHDkVaWhr69euH1NRU+Pr6IiYmBp6enjhx4gRu3ryJli1b4ttvv4WPjw98fHxw5swZDBgwAK1atcK4cePkvn/66Sf4+fmhbdu26N27N/777z/Ex8dj6NChOHPmDHr16gUAOHnyJLp27Qpvb2/06NEDMTExAICHDx8iJCQEHh4eCAwMRGxs7BNfw4IFCzB27Fh8/PHH8PDwQEBAAO7duwcge2bqyy+/hJ+fH06dOoWkpCQEBwfDx8cH7dq1wzfffJOrrxUrVsDPzw+enp7Ys2cPAMBgMGDKlCnw8fGBp6cnRo8eDZ1OJ6/z999/o1u3bnB3d8eECROg1+vznc3I2Y+P7t9169bh448/ltsYDAY0b94cFy5cyLWuwWDAl19+Kc9cjh07Funp6Vi1ahV27tyJH3/8ERMmTMizzYiICLRo0QIdOnRAREQEatSoUWB/69atw8CBA+X19Xo9mjZtiitXriA2NhYDBw6Ux/7AgQMAIB8X4eHhchCpUaOGHLJbtmyJlStXAgCOHj2K9957D2FhYWjTpg26dOmCs2fPok+fPmjRogXmz58vb3v9+vXw9fWFp6cnRo4ciYcPHwIAxo4di/nz56Nfv37w8PBAv3798ODBA8ybNw9HjhzB6NGjsW3btlz7Ib9j/ejRo/D29pbbPbq8YMECTJgwAd26dcPKlSvh5uaGhIQEuW1YWBjmzJkjz24WtO+uXr2Knj17ws/PD97e3k8M548KDg7G8uXL5eULFy6gZcuW8rjljEHfvn1x586dPOvnt49yxubR76ec5acZG71ej9GjR8PLywudO3fG1atXAQApKSkYPXo0fHx80KZNG2zatCnXdpYuXQofHx/o9fpc9WZkZOCzzz6Dj48P/Pz8MHPmTPmPlDNnzuCrr77CggULcq3TtWtX7NixQ17et28fOnbsCCB7Nvjdd99FmzZt0L9/f3ncHjx4gJCQEPn7edasWfL6j/+8OHbsGDp37ox27drBz88P27dvL3DMqBASRCZSvXp1cfv27Sc+t2/fPtG3b19hMBiEwWAQX375pfj9999FTEyMqFWrltzOw8NDHD9+XMTExAgXFxfx888/CyGEGDZsmGjdurW4d++eSEhIELVr1xbXr18X8fHxonbt2vJ2x44dK0JDQ4UQQmzatEkEBQUJIYRITU0VjRs3FlFRUUIIIbZu3So6d+4shBBi7dq1onfv3kKn04mEhATh4eEhxowZk+c1zJ8/X9SvX1/cuHFDCCHEqFGjRFhYmBBCiMDAQNG/f3+h1+uFEEJMnDhRTJw4UQghRGJiomjdurX8uqpXry6+/fZbIYQQUVFR4u233xaZmZlix44dwt/fX2RmZoqHDx8KPz8/sXnzZrn/rl27ivT0dJGeni7atm0rdu/enWv/zZ8/X37tj+7HnOfj4+OFm5ubSEhIEEIIcfz4ceHj45Pndf7yyy+iU6dO4v79+yIrK0sMGjRILFy4UAghxJgxY+SvH5WYmCjc3NzEpUuXhF6vFyNGjBDVq1cvsL+4uDhRr149kZ6eLoQQ4vDhw8Lf318IIUTfvn3Fl19+KYQQ4tq1a6JJkyYiISFBxMTECFdXVxERESFvu3r16mL27NlCCCHOnj0r6tSpI7KyssSRI0eEq6urOHLkiDAYDKJr166iS5cuIj09XVy6dEm4uLiIhw8fiuPHj4tmzZqJ2NhYeexmzpwpv14/Pz+RmJgodDqd6NChg9iyZUuuffy4/I71I0eOCC8vL7ndo8vz588XLVu2FPfu3RNCCPHhhx+KjRs3ym09PDzEuXPn5DEuaN99/PHHYunSpUIIIY4dOybc3NxEZmZmnu+1HL/++qvo3bu3vPzVV1+JadOmib///lu0bdtWZGZmCiGEWL16tfz9+KiC9tHjPxNylo0dm02bNgkXFxdx6tQpIYQQc+fOFYMHDxZCCDFu3Djx6aefCr1eL+7duyfc3d3FpUuX5O0sXrw4T61CCLF06VLx0UcfCZ1OJx48eCC6du2a6/ss5+tHLV++XAwZMkReHjdunFi6dKm4ceOGqF+/vrzdJUuWiGHDhgkhhFi2bJn48MMPhcFgEElJSaJJkyby8fL4z4suXbqIo0ePCiGE+Pfff8XIkSOfWDsVXpz5IpPq06dPruu9cmad7O3tceXKFezevVv+i7BVq1YF9pWVlQVfX18AQPXq1VGnTh3Y29ujVKlScHBwQFxcHEqXLo2TJ0/KpzobNWokz2g96uTJk3ByckKLFi0AAP7+/rhx4wZu3bqFEydOwNvbG1qtFqVKlYKHh0e+NTVt2hQVKlQAALRt2xanT5+Wn3N3d4dGk/0tduDAAfm1lyxZEt7e3jh06JDctnPnzgCAFi1aICsrCzdu3ICPjw82bdoES0tLFC1aFHXq1Mn1Wnx8fFCsWDEUK1YM7u7uOHPmTIH773GlS5dGo0aNsHPnTgDA7t270a5duzzt9u/fj06dOsHa2hoWFhbo0qVLrtqf5OzZs6hcuTKqV68OjUaDnj17Kvbn4OAAFxcXue89e/bAz88P6enpOHr0qHytT6VKldCwYUN59kun0+WaQQIgz0K4uroiIyNDnpG0s7ND06ZNIUkS3nrrLTRp0gTFihXDW2+9Bb1ej4SEBOzduxft2rWDk5MTAKBnz57YtWuX3Le7uztKliwJrVaL6tWr4/bt2wXui2c51gGgbt26sLe3B5A91nv37gUAREdHQ6vVwtXVVW6b374Dsq9Z/OCDDwAADRs2REZGBu7evZvvdlu3bo3z588jKSkJQPZx4evrCzs7OyQkJGDr1q1ITk5Gnz598j2F/7T7CDBubIDs8a9fvz4AwM/PTz7u9+3bh759+0Kj0cDe3h7e3t65xq1169ZP3O7+/fvRo0cPaLVaWFlZ4d1331U8vtu1a4eDBw8iNTUVer0e+/btg5+fH37//Xc0adIE1atXBwAEBARg79690Ov16N+/PxYtWgRJklCiRAm89dZbuHnzZq59lvPzonTp0ti8eTOuXLmCypUr44svvlDcf1S48JovMqn8rvlyc3PDhAkTsGbNGowZMwaenp6KF21bWFjAysoKAKDRaGBtbZ3rOb1eD71ej/nz58s/8O7fv48qVark6SslJQUxMTFymAOAIkWKICEhAcnJybC1tZUft7Ozw/37959YU8mSJXO1S0lJkZdLlCghf52QkAA7O7tcbePi4uTlUqVKyV/b2toiJSUFCQkJmDZtGs6fPw9JkhAfH4+goCC5Xc4v5px1CvqFmp/27dsjIiICAQEB+O2337BkyZI8bRISEnK9lhIlSshhJj8pKSm51skJMkr95YQMLy8v/Pbbb1ixYgVSU1MhhEBAQIC8Tnp6Ot5++20A2WNvY2OTa/s542dhYQEg+1QnABQvXlxu8+gxJEkSNBoN9Ho9UlNTsXv3bkRFRQEAhBC5Tvc+emzkHHcFeZZjPWe/5PDy8sLMmTORkZGRK1g96kn7Dsh+I8DixYuRmJgISZIghJD3x5NYW1ujefPm2L9/Pxo2bIiUlBQ0bNgQkiRhwYIFWL58OaZNm4bGjRtjypQpeOONN/L08bT7CDBubIDcx72NjQ2Sk5MBAKmpqQgJCZHHPCMjI9f396Pfq496luPbyckJbm5u2LVrFypWrIhy5cqhQoUKSE1NxYkTJ3Jt18bGBklJSUhNTcXMmTNx9epVaDQaxMbGokuXLrm2myM8PByLFy9Gv379YGVlhZEjR+bqkwo/hi96aXJmw5KSkhAaGoply5Y994Wt27Ztw969e7F27VrY29tjw4YN2Lp1a552jo6OqFq1KiIiIvI8Z2dnh9TUVHn50WttHpeYmCh/nZycnOsH6KPKlCmDpKQk+WLspKQklClTJte6OQEsp58vv/wSWq0WW7duRZEiRfDJJ5/k6jPnl47Stgvi7e2NqVOn4sCBAyhWrBjefPPNfGvP8XjtT2JjY4P09HR5+dGgWVB/Pj4+WLp0Kf766y+UKFEClStXRlZWFiwsLLBp06Zcv6AB5Jo5eFEcHR3RuXNnjBkz5oX1+aRjvVWrVrlCyaPB/XElS5aEm5sbDh8+jD179mD27Nl52jxp3+l0OoSEhGDevHlwd3dHZmYm3NzcFOv18fHB7t27kZiYCB8fH0iSBAB4++238fbbbyM9PR2zZs3CnDlznmpW5tEQ9ejx+zQeXS8lJUUOVY6Ojli4cKE862SsZzm+gew/XHbs2IFKlSrJM8aOjo5o3rx5rmvUcowePRqurq5YuHAhLCwscv0x8aSaJk6ciIkTJyIqKgrDhg1Dq1at8hz/VHjxtCO9FJs2bcLChQsBZP9iqVq1KgDA0tISBoMBaWlpz9TvvXv3UK5cOdjb2yMxMRHbt2+XZ620Wi3S0tIghEDdunVx9+5dnD17FgAQExOD0aNHQwiBevXqyTNnCQkJ+P333/Pd3smTJ+VTKjt37kTDhg2f2K5169ZYv349gOwwt3v37lynQXIC4qFDh1CsWDFUrFgR9+7dQ/Xq1VGkSBFcvHgRp0+fzhVodu3ahYyMDKSnp+PgwYNo1KiR4v55fP/a2tqiVatWmDJlyhNnU3Jqj4yMxIMHD5CVlYWNGzfC3d29wO24urri0qVLuH79OgwGAzZu3GhUf05OTqhQoQKWLFki16PVauHu7o4ff/wRQPaFy+PGjTPqVNaz8PT0xK5du+TQvWfPnjxvkHgSrVabK7TnyO9Yd3BwwN27d3Hv3j3o9fon/pHwKB8fH2zYsAE6nQ41a9bM8/yT9t2DBw+Qnp6O2rVrA8h+16ulpWWu4+hJPDw8cPr06VyzbFFRUZgyZQoMBgOsra1Rs2ZNOZQZy8HBARcvXgSQvV9yTrM9jX///Rfnzp0DkPt7ztPTUz5GsrKyEB4ebtRtalq3bo2NGzdCr9cjPT0dW7ZsUTy+gexAffLkSezYsUPeRy1btsSJEyfkywP+/PNPTJ8+HUD2z6ZatWrBwsIChw4dwvXr1584DjqdDn369JH/YHF1dYVWq32mfUWvLo4mmdTj13z5+vpi7dq1aNOmDaKjo9G2bVv4+fnh8uXL6NevHxwcHNCwYUN4eHjg1KlTT709f39/JCUlwdvbG5988glCQkIQGxuLmTNnomHDhoiLi0OrVq1gaWmJ+fPnY9q0afDz88OQIUPg6+sLSZLQo0cP2NrawsvLC8OGDYOXl1e+22vevDmmTJkCd3d33Lp1Cx999NET24WEhCAlJQW+vr4IDAzEgAED5BkIa2trGAwG+Pv7Y9KkSQgLC4NWq0X//v3x448/ws/PD+vWrcOYMWPw008/ye98at68Ofr27Yt27dqhWbNmRl1H9KT92759e/z3339PvN4LyP4l884776BLly7w9/eHs7Mz+vbtW+B2HB0dMXLkSPTt2xfdu3fPFUqV+vPx8clzam3y5Mk4fvw4fH190blzZ1SoUOGJp7teBFdXVwwcOBB9+vSBn58fVq5ciTZt2iiu5+Pjg5EjR8qn+3Lkd6xXqlQJXbt2RadOndCrVy/5NGp+vL29sX///gJPPz2+7+zs7PDhhx+iU6dO6NSpEypWrAgvLy8MHDhQfgfik9jY2MDV1RW3bt1CvXr1AACNGzfGw4cP4ePjg/bt22Pbtm0IDg5W3C+PGjFiBCZPnoyOHTuiWLFieU4XG6Np06ZYs2YN2rZti6ioKIwaNQpA9vdYamqqXJ/BYJDfYVuQPn36wNnZGe3bt0fXrl3RunXrfP8QeVTJkiXRuHFjlC9fXj4WHR0dMW3aNAwZMgR+fn6YOnWq/H01aNAgzJo1C/7+/jh27BiGDh2KBQsW4OTJk7n6tbS0RLdu3fD++++jXbt26NOnDyZMmIBixYo97a6iV5gkhBAvuwiiwmjBggWIjY1FWFjYyy7lufz555+YOnVqrtmpF0EIIc+M/PPPP+jVqxeOHz/+QrdB9DJNnjwZb731Fnr37v2yS6FChjNfRK+xrKwsLFy4EH369Hnh/bZq1Uo+rbtt2zZ5BoXIHFy7dg2///47OnTo8LJLoUKIF9wTvabOnz+PIUOGoGXLli/8F4hWq8WkSZMwZswYCCHg4OBQ6GcIiXJ89dVX2LJlCyZOnJjrnZ1ExuJpRyIiIiIV8bQjERERkYoKzWnHu3fzvoXbFEqVskZiYsFvw6ZXA8eqcOF4FS4cr8KF4/XqcXDI/5Q0Z74eo9VavOwSyEgcq8KF41W4cLwKF45X4cLwRURERKQihi8iIiIiFTF8EREREamI4YuIiIhIRQxfRERERCpi+CIiIiJSEcMXERERkYoYvoiIiIhUVGjucE9ERPQ8bt++hb59A1CjRs1cj4eHz4adXYk87WNjY5GQEA8Xl9r46qsv0L17AMqWLfdM2758+R8UKVIEFStWUmyblJSEoUMHoGXLdzBw4NBn2h6ZxqPHxPNg+CIiotdGxYqV8PXX3xjV9tSp43jwIB0uLrURHPzJc233wIG9qFnTxajwde3aVVSoUIHB6xX06DHxPBi+iIjotXfs2BF8++0iFC1qhVKl7DFy5BgsX/4NtFotnJyc8eOP6zBy5KfYt+83JCcn4ebNm7h16z989NEg/PprJGJjb2H27K/g5OSMsLDJuHs3Dg8ePED//gPg7PwGtmyJwIEDe1GqVCnodDosXboQWq0Wjo5OGDNmAiwtLeVa5s+fi7i4WCxZ8jW6dOmOGTOmQqfTQaPRYOzYiZAkCVOnTkSxYtbo2rUHWrRoJa+7YMFcnD8fjczMTHTq1BXvvNMaH3/cHz/+GAEA2L79F1y+/Dd69uyDGTOmISsru98xYybC2dkZAQGdUb16TTRp0hROTm/gu++WwNLSEra2tpg6daa87djY26hTxw179+7Bzz9vw7//XsWXX34OSZJgbW2N0NDJsLX932cbpqWlYcKET5GRkYFmzVpg69bN+OmnSJw6dQLffLMIWq0WDg6OGDfuMwwc2A/h4V/A2dkZsbG3ERo6Gt9+uwqffx6GW7f+Q1ZWFj78cCAaNmyMoUMHoGrVagCAEiVK4v79NNy4cR3//XcTw4d/gmbNWqBHj45o2fIdnDhxDG+/3RwGg8Dx40fx9tvNMWjQsCfWnpaWirCwyShbthwuX/4H1avXwMcfD811TLRs6f7MxxvDFxERqW7D3ss4fjHuhfbZuKYjeni++Uzrbtq0HkOHjkDduvVx4MBeGAx6+Pn5o2TJkmjZ0h0//rhObpuSkoK5cxdg6dKF2LHjF8yduwDffrsYhw79Dm9vXzRp8jb8/Pzx3383MXHiWCxfvhZNmzZD69Zt4OJSG/369cJXXy2GnV0JLFr0Ffbt24O2bf3k/ocODUFExAYMHDgU4eFT4O/fEW3atMW+fXuwfPk3+OCDj/HPP5ewadMvKFGipLxeRkYGnJ3LYtiwkcjIeIgePTrh3Xc7wcnJCVevXkHVqtVw8OAB9OwZiG+/XYyAgN5o3LgpDh+OwqpV32HMmAm4des/hIfPQdWq1bB37x5MmjQdZcuWw7Rpn+Ho0cMAgMzMDHzzzUocOnQQGzb8AACYN282Ro8ORYUKFRER8RMiIjYgKOgDubYdO35B5cpVERIyChERP0EIAQCYM2cGvvxyIZycnDF37izs3r0D77zjgUOHfkfXrj1w8OABtG7tid27d6B06TIYN+4zJCUlITh4IFat+hEAULVqNXTq1A3Lli1FXNwdzJkzH0eO/IEtWzahWbMWuH37Fjp27IoBA4agXTtPLFjwDT76aCC6dn0XgwYNe2Ltbdv64dKlC5gyJRylStmjc+d2GDIkJNcx8TxMGr7+/vtvDB48GO+//z4CAwNzPffHH39g7ty5sLCwwDvvvIMhQ4aYshQiIiLcuHEdQ4cOkJcrVqyETz8dDw8PL8yePQNt2/rCy8sHpUuXybcPFxdXAECZMmUgSRIAwN7eHsnJybC1tcOFC9GIjIyAJGmQkpKca92EhHu4eTMGoaGjAQAPHz7MFaAed+nSBfn0Y4MGjbBy5XcAgHLlyudZr2jRokhJScbAgf2h1WqRlJQIAP8fZg6iXLny+PffK6hd2w0zZ07DjRvXsWrVMhgMBpQsWQoAYGVVTJ5JKlmyJGbNmg69Xo9bt/5Dw4aNkZiYgDp16gIAmjVrAQuL7A/0Pn8+GrNmTQcA6HQ61Krlkqu2a9euoX79hgCAli3fwfffr0ZKSjIkSYKTk7P8+s6cOYUOHbrg66/noWvXHoiKOoBPPhmLDRu+x9mzp/Hnn2cAZAdNnU4HAKhV63+nAN3c6gEAHB0dkZaWBgAoXrw4KlWqDAAoVqwYatSoCa1WCyEMBdZerlwF+TgoU8YB9++n5TtOT8tk4Ss9PR3Tpk1Ds2bNnvj89OnTsWzZMjg5OSEwMBA+Pj54881n+4uFiIgKlx6ebz7zLNXzyO+aL1/f9mjatBl+/30/xowZgenTP8+3j5zA8fjXQgjs3r0DKSkpWLjwO6SkpODDD/vkWlertUSZMg5GX3cGSPIskU6XBUnSyP087vTpkzh16gS+/jr71Ji3d/bpSHd3D0ycOBZVq1ZD06bNIEkStFpLTJs2C2XK5A6Zlpb/iwUzZkzD7NnzULlyFcydO0t+jRpN9muWJEkOn1ZWVliwYKm8nJeARiPJ6z3+2rJfnw6SpEHVqtVw795d3LkTi9TUVFSsWAlarSX69u0Pb2/fPD0/WvPj4/H4Y9n7Lnf0eVLtt2/fyrPeo7U+L5PdaqJIkSL49ttv4ejomOe5mJgYlChRAm+88QY0Gg3c3d1x+PBhU5VCRERUoJUrv4OFhRYdO3ZBmzZtce3aVWg0Guj1+qfqJykpCW+8URYajQYHDuyVZ2ckSYJer4ednR0A4N9/rwIANm78EZcv/5Nvf7VqueDUqRMAgDNnTqJmzVr5tk1OToKjoxO0Wi2iog5ArzdAp9OhTBkHSJKEPXt2onXrNgAAF5faOHhwPwDg5Mnj2LVrR57+7t9Pg5OTM1JTU3Hq1EnodDqUK1cely6dB5B9nVzO/nnzzbdw5MgfAIA9e3bixIljufoqW7Y8Ll68AAByOzs7O0iShNjY2P9/fafk19esWUt8880itGrlLtcbFXUAAJCYmIClSxfmux+ellLtj3qWY+JJTDbzpdVq86TLHHfv3oW9vb28bG9vj5iYGFOVYrQNey/j1D93odcrp9vnubaAiIhejsdPOwLA4MHD4eTkjJCQwbC1tYOtrS0CAgJhbW2N6dMny6fkjNG6tSfGjh2J8+fPoX37DnB0dMSKFd+ibt36mDdvNqytrTF27GcID58CS8vsWbAOHbrk29+HHw7EjBnTsHXrZmi1lhg3biKysrKe2LZRo6ZYt24Vhg4dgFat3NG8eUvMmTMD48Z9hpYt38FPP/2IiROnAgA++GAAwsOnYM+enZAkCaGhk/L016VLdwwa9AEqVKiI3r37Yvnyb7B48XL8+mskBg36APXrN5Rv0REcPAqffx6GdetWoUiRopg8eXquvtq1exfjxo3E0KED0LhxU2g02XM/n346AVOmjIeFhQXKlSuPNm3aAsierRs4sD9Wrsy+pszT0wunTh3HwIH9odfr0b9/7jF8Hk+q/f79+09sW7t2HfmYePQ6vacliRc5j/YECxYsQKlSpXJd83Xq1CksW7YMCxdmJ9effvoJMTExGDlyZL79ZGXpodVa5Pv8i7B8azT2nVQOgSlpGbCzKYo1k/NOfxIREZmrpKQkHD16FD4+Prhz5w6CgoKwY0feWbPH/ffff7h69SpatWqF06dPY8GCBVi+fLkKFb+aXsq7HR0dHREfHy8v37lz54mnJx+VmJhu6rLw7tsV0f9dV9y9m1pgu9GL/oBOb1BsR6bl4GDLMShEOF6FC8ercFFrvLKyBDZv3oolS76BEAYMHhxi1HYzMiR88813+Oqr+RACCAkZZfbHl4ODbb7PvZTwVb58eaSlpeHmzZtwdnbGvn37MGfOnJdRChERERlJq9Vi6tQZT72era0t5s792gQVFU4mC1/nzp3DrFmz8N9//0Gr1WLnzp3w9PRE+fLl4e3tjcmTJ+OTT7LvGNyuXTtUqVLFVKUQERERvTJMFr5q166NNWvW5Pt848aNsX79elNtnoiIiOiVZLJbTRARERFRXgxfRERERCpi+CIiIiJSEcMXERERkYoYvoiIiIhUxPBFREREpCKGLyIiIiIVMXwRERERqYjhi4iIiEhFDF9EREREKmL4IiIiIlIRwxcRERGRihi+iIiIiFTE8EVERESkIoYvIiIiIhUxfBERERGpiOGLiIiISEUMX0REREQqYvgiIiIiUhHDFxEREZGKGL6IiIiIVH9L/LIAACAASURBVMTwRURERKQihi8iIiIiFTF8EREREamI4YuIiIhIRQxfRERERCpi+CIiIiJSEcMXERERkYoYvoiIiIhUxPBFREREpCKGLyIiIiIVMXwRERERqYjhi4iIiEhFDF9EREREKmL4IiIiIlIRwxcRERGRihi+iIiIiFTE8EVERESkIoYvIiIiIhUxfBERERGpiOGLiIiISEUMX0REREQqYvgiIiIiUhHDFxEREZGKGL6IiIiIVMTwRURERKQihi8iIiIiFTF8EREREamI4YuIiIhIRQxfRERERCpi+CIiIiJSEcMXERERkYoYvoiIiIhUxPBFREREpCKGLyIiIiIVMXwRERERqYjhi4iIiEhFDF9EREREKmL4IiIiIlIRwxcRERGRihi+iIiIiFSkze+JmJiYAlesUKGCYufh4eE4e/YsJElCaGgo3Nzc5OfWrVuHyMhIaDQa1K5dG+PHj3+KsomIiIgKp3zDV1BQECRJghAiz3OSJOG3334rsONjx47h+vXrWL9+Pa5cuYLQ0FCsX78eAJCWloZly5Zh165d0Gq16N+/P86cOYN69eo958shIiIierXlG7727t2b70onT55U7Pjw4cPw8vICAFSrVg3JyclIS0uDjY0NLC0tYWlpifT0dFhbW+PBgwcoUaLEM5RPREREVLjkG75ypKWlYcuWLUhMTAQA6HQ6bNq0CVFRUQWuFx8fD1dXV3nZ3t4ed+/ehY2NDYoWLYohQ4bAy8sLRYsWRfv27VGlSpUC+ytVyhparYUxr+m5OTjYFvi8hYUEvZAU25HpcQwKF45X4cLxKlw4XoWHYvgKCQlB2bJlERUVBR8fHxw6dAiTJ09+6g09evoyLS0NS5cuxY4dO2BjY4OgoCBcvHgRNWvWzHf9xMT0p97ms3BwsMXdu6kFttHrBQwGodiOTMuYsaJXB8ercOF4FS4cr1dPQWFY8d2OGRkZmDp1KsqVK4cxY8Zg9erV2L59u+JGHR0dER8fLy/HxcXBwcEBAHDlyhVUqFAB9vb2KFKkCBo1aoRz584Z81qIiIiICjXF8KXT6ZCeng6DwYDExESULFlS8Z2QANCiRQvs3LkTABAdHQ1HR0fY2NgAAMqVK4crV67g4cOHAIBz586hcuXKz/EyiIiIiAoHxdOOHTt2xIYNG9C9e3e0a9cO9vb2qFSpkmLHDRo0gKurKwICAiBJEiZNmoSIiAjY2trC29sbH3zwAfr27QsLCwvUr18fjRo1eiEviIiIiOhVJokn3UviETnvUASAO3fu4N69e6hVqxYkSVKlwBxqncs25rz56EV/QKc3YN6wlqrURE/GaxwKF45X4cLxKlw4Xq+egq75Upz58vb2RpMmTdChQwe4u7vDycnphRZHRERE9DpRvOZr//79aN++PbZu3QovLy9MnToVZ8+eVaM2IiIiIrOjGL6KFi2Ktm3bYt68eYiMjIQQAr1791ajNiIiIiKzo3jaUafTISoqCjt37sSxY8fQtGlTfPPNN2rURkRERGR2FMPXO++8g4YNG8Lf3x9Tp05FkSJF1KiLiIiIyCzlG76EEJAkCb/++itKliwpP24wGAAAGo3iGUsiIiIieky+4SsoKAirV69G8+bNc91WIieUXbhwQZUCiYiIiMxJvuFr9erVAIDjx4/D1pYf1klERET0IiieO2zbti2Cg4Oxd+9eZGVlqVETERERkdky+j5fkZGR8n2+zpw5o0ZtRERERGbnme7zFRgYqEZtRERERGaH9/kiIiIiUhHv80VERESkIsXwNWDAAPTr10+NWoiIiIjMnmL4Onr0KLp168bbTTxGCIGj0bFGtbW20qJOtTImroiIiIgKA8Xw9fDhQ3h6eqJKlSqwtLSUH1+3bp1JC3vVCQGUsClqVNvktAwTV0NERESFhWL4Gjx4sBp1EBEREb0WFMOXXq9Xow4iIiKi14Ji+Fq0aJH8tU6nw+XLl9GgQQM0a9bMpIURERERmSPF8LVmzZpcy/fu3cMXX3xhsoKIiIiIzJli+Hpc6dKlcfXqVVPUUqjcf6jDkshoo9pWdrZFU1dnE1dEREREhYFi+Bo9ejQkSZKXb9++DY1G8VOJzFrjmo6I+uu2UW1T0zNx9VaKiSsiIiKiwkIxfDVv3lz+WpIk2NjYoEWLFiYt6lXXw/NNVHKyMepWE0sio2EwCBWqIiIiosJAMXz5+/vj3r17cHZ2xsWLF3Hx4kU16iIiIiIyS4rnD8eOHYszZ87gzp07GD58OP7++2+MHTtWjdqIiIiIzI5i+Lpz5w58fX2xbds29OzZE59++imSk5PVqI2IiIjI7CiGr8zMTAghsHv3brRu3RoAkJ6ebuq6iIiIiMySYvhq0qQJGjZsCAcHB1SpUgUrV65ElSpV1KiNiIiIyOwoXnA/atQoDBgwAHZ2dgCANm3aoHfv3iYv7FVnbaU16gOzDQaBR+7UQURERK85o26ymhO8AKBChQomK6YwqVOtjFHtNh64Cp3eYOJqiIiIqLB4ve+WSkRERKQyxfB15cqVPI+dOXPGJMUQERERmbt8w1dKSgpu3LiB0NBQxMTEyP+uXr2KMWPGqFkjERERkdnI95qv06dPY9WqVbhw4QKCgoLkxzUaDVq2bKlKcURERETmJt/w5e7uDnd3d/zwww/o2bOnmjURERERmS3Fdzt6eXlh1apVSE5OhhD/+4Do4OBgkxZGREREZI4UL7j/+OOPcfHiRWg0GlhYWMj/iIiIiOjpKc58WVtbY8aMGWrUQkRERGT2FGe+6tat+8TbTRARERHR01Oc+Tp48CBWrlyJUqVKQavVQggBSZKwf/9+FcojIiIiMi+K4Wvx4sVq1EFERET0WlA87ejg4ID9+/fjhx9+QLly5RAfH48yZYz7XEMiIiIiyk0xfE2ePBk3btzA0aNHAQDR0dEYO3asyQsjIiIiMkeK4evq1asYN24crKysAAC9evVCXFycyQsjIiIiMkeK4Uurzb4sTJIkAEB6ejoePnxo2qqIiIiIzJTiBfe+vr4ICgrCzZs3MX36dPz+++/o1auXGrURERERmR3F8BUYGAg3NzccO3YMRYoUwdy5c1G7dm01aiMiIiIyO4qnHQGgSJEiqFevHmrVqoUHDx7g+PHjpq6LiIiIyCwpznwNHDgQ//zzD5ycnOTHJEnCunXrTFoYERERkTlSDF93797Fb7/9pkYtRERERGZP8bRj7dq1cfPmTTVqISIiIjJ7ijNftWrVgq+vL8qUKQMLCwv5sx05G0ZERET09BTD13fffYfly5fD2dlZjXqIiIiIzJpi+KpRowaaNGmiRi1EREREZk8xfJUpUwZ9+vRB/fr1YWFhIT8eHBxs0sKIiIiIzJFi+HJwcICDg4MatZgtIQSORsca1dbaSos61cqYuCIiIiJ6WRTD19ChQwFkBwghhMkLMkdCACVsihrVNjktw8TVEBER0cukGL6WLVuGxYsX4/79+wAgv9vxwoULJi+OiIiIyNwohq+NGzciMjISZcuWVaMeIiIiIrOmeJPVSpUqMXgRERERvSBG3Wrik08+QZMmTXK927Fbt26KnYeHh+Ps2bOQJAmhoaFwc3OTn7t9+zZGjhwJnU4HFxcXTJ069RlfAhEREVHhoTjzFRcXhyJFiuDMmTM4efKk/E/JsWPHcP36daxfvx5hYWEICwvL9fzMmTPRv39/bNy4ERYWFrh169azvwoiIiKiQkJx5svX1xfu7u5P3fHhw4fh5eUFAKhWrRqSk5ORlpYGGxsbGAwGnDx5EnPnzgUATJo06an7L0zuP9RhSWS0UW0rO9uiqSs/TYCIiMhcKYavlStXokWLFtBqFZvmEh8fD1dXV3nZ3t4ed+/ehY2NDRISElC8eHHMmDED0dHRaNSoET755JMC+ytVyhparUWBbV4UBwfbF9bXOw3K47fjN6DRSIptU9IycC029YVu39xxXxUuHK/CheNVuHC8Cg/FRGVra4v27dvDxcUFlpaW8uOff/75U23o0XuECSFw584d9O3bF+XKlcOAAQOwf/9+tG7dOt/1ExPTn2p7z8rBwRZ376a+sP7efbsiHG2LGHWfryWR0TAYxAvdvjl70WNFpsXxKlw4XoULx+vVU1AYVgxfHh4e8PDweOqNOjo6Ij4+Xl6Oi4uT75RfqlQplC1bFhUrVgQANGvWDP/880+B4YuIiIjIHChecN+5c2e4urrC1tYWnTt3Rps2bdC5c2fFjlu0aIGdO3cCAKKjo+Ho6AgbGxsAgFarRYUKFXDt2jX5+SpVqjzHyyAiIiIqHIy65uuXX35BZmYmvLy8sGjRItjZ2WHw4MEFrtegQQO4uroiICAAkiRh0qRJiIiIgK2tLby9vREaGoqxY8dCCIHq1avD09Pzhb0oIiIioleVYvj65ZdfsGHDBgQFBQEAPv30UwQEBCiGLwAYNWpUruWaNWvKX1eqVAk//PDD09ZLREREVKgpnnYsXrw4NJr/NdNoNLmWiYiIiMh4ijNfFStWxNdff42UlBTs2rUL27ZtQ7Vq1dSojYiIiMjsKIavzz77DKtXr4aTkxMiIyPRsGFD9O7dW43azIa1lRbJaRmK7QwGAUn5dmBERERUiCmGr/nz56Njx4744IMP1KjHLNWpVsaodhsPXIVObzBxNURERPQyKYYva2trjBgxApaWlujQoQP8/f1RpoxxYYKIiIiIclO8cn7QoEHYunUrZs+ejdTUVAwYMAAfffSRGrURERERmR2j37ZYtGhRFCtWDMWKFcODBw9MWRMRERGR2VI87bh06VLs3LkTOp0O/v7+mDVrFsqXL69GbURERERmRzF8JScnIzw8PNcNUomIiIjo2SiGr8GDB2PVqlWYN28eJElCvXr1EBQUBCsrKzXqIyIiIjIritd8ffbZZ0hLS0NAQAB69OiB+Ph4TJgwQY3aiIiIiMyO4sxXfHw85s6dKy97eHigT58+Ji2KiIiIyFwpznw9ePAg17sb09PTkZGhfLd2IiIiIspLcebrvffeg5+fH2rXrg0hBM6fP4/g4GA1aiMiIiIyO4rhq1u3bmjRogWio6MhSRI+++wzODk5qVEbERERkdlRDF8bN27MtXzw4EFotVpUqVIFdevWNVlhREREROZIMXwdOnQIhw4dQoMGDWBhYYGTJ0+icePGiImJgbu7O0aMGKFGnURERERmQTF86fV6bNu2Tf4w7Xv37mHGjBn4+eefERAQYPICXzdCCByNjjWqrbWVFnWq8UPOiYiIChPF8HXnzh05eAFA6dKlcfPmTUiSBIPBYNLiXkdCACVsihrVNjmN7zolIiIqbBTDV9myZTF8+HA0adIEkiTh9OnTKF68OHbs2IE33nhDjRqJiIiIzIZi+Jo1axa2bNmCixcvwmAwoG7duujcuTPu378Pd3d3NWokIiIiMhuK4atIkSLo3r17nsdtbGxMUhARERGROVO8wz0RERERvTgMX0REREQqyve049ChQ/H1119j+PDhmD9/vpo1vdbuP9RhSWS0UW0rO9uiqauziSsiIiKiFynf8HX9+nW89957uHr1Knr37p3n+XXr1pm0sNdR45qOiPrrtlFtU9MzcfVWiokrIiIiohct3/D1/fff49KlS5g+fTo/SFslPTzfRCUnG6Pu87UkMhoGg1ChKiIiInqR8g1ftra2aNSoEb7//nsAwL///gtJklClShUUK1ZMtQKJiIiIzInirSb++OMPTJ48Gc7OzjAYDIiPj8e0adN4jy8iIiKiZ6AYvr777jtERkbC3t4eQPbHDQUHBzN8ERERET0DxVtNWFpaysELAJycnGBpaWnSooiIiIjMleLMV/HixbF8+XI0b94cABAVFYXixYubvDAiIiIic6QYvsLCwvDVV18hMjISkiShXr16CA8PV6M2IiIiIrOjGL5Kly6NqVOnqlELAbC20iI5LUOxncEgIEkqFEREREQvlGL4InXVqVbGqHYbD1yFTm8wcTVERET0ovGzHYmIiIhUpBi+fvjhB6SlpalRCxEREZHZUwxfly5dQocOHTBmzBicOHFCjZqIiIiIzJbiNV+TJ0+GwWDA0aNHERkZiTlz5qBNmzbo0aMHSpQooUaNRERERGbDqGu+NBoNKlasCGdnZ2RmZiI6Ohq9e/fGnj17TF0fERERkVlRnPn6+eefERERgaSkJHTv3h0rVqxAiRIlkJKSgsDAQHh5ealRJxEREZFZUAxfhw4dQnBwMBo1aiQ/9vDhQ9jZ2SEoKMikxRERERGZG8XTjomJibmCFwD07t0bANC1a1fTVEVERERkpvKd+YqMjMTChQtx+/ZttG7dWn5cp9OhTBnjbgRKRERERLnlG746dOiA9u3bY/z48Rg2bJj8uEajgaOjoyrFEREREZmbfMPX+fPn4eLigo4dO+LGjRu5nrt27RqaNWtm8uKIiIiIzE2+4Wvz5s1wcXHBokWL8jwnSRLDFxEREdEzyDd8hYaGAgDWrFmjWjFERERE5i7f8NWrVy9IkpTviuvWrTNJQURERETmLN/wFRISomYd9AyEEDgaHWtUW2srLepU47tUiYiIXrZ8w5eNjQ1cXFxw+PBhNeuhpyAEUMKmqFFtk9MyTFwNERERGSPf8LVlyxZecE9ERET0guUbvsaNGweAF9wTERERvUiKHy90/PhxdO3aFfXq1UP9+vXx3nvv4dSpU2rURkRERGR2FD9Ye+rUqQgNDUWDBg0ghMDJkycxefJkREZGqlEfERERkVlRDF+lS5fOdX1XixYtULZsWZMWRURERGSu8g1fMTExAIA6depg+fLlaN68OTQaDQ4fPgwXFxfVCqT83X+ow5LIaKPaVna2RVNXZxNXREREREryDV9BQUGQJAlCCADA2rVr5eckScLw4cNNXx3lq3FNR0T9dduotqnpmbh6K8XEFREREZEx8g1fe/fuzXclXnD/8vXwfBOVnGyMus/XkshoGAxChaqIiIhIieI1X2lpadiyZQsSExMBADqdDps2bUJUVJTJiyMiIiIyN4q3mggJCcGlS5cQERGB+/fvY9++fZg8ebIKpRERERGZH8XwlZGRgalTp6JcuXIYM2YMVq9eje3btxvVeXh4ON577z0EBATgzz//fGKbL774An369Hm6qomIiIgKKcXwpdPpkJ6eDoPBgMTERJQsWVJ+J2RBjh07huvXr2P9+vUICwtDWFhYnjaXL1/G8ePHn61yIiIiokJIMXx17NgRGzZsQPfu3dGuXTu0b98eZcqUUez48OHD8PLyAgBUq1YNycnJSEtLy9Vm5syZGDFixDOWTkRERFT4KF5w37NnT/nrZs2a4d69e6hVq5Zix/Hx8XB1dZWX7e3tcffuXdjY2AAAIiIi0KRJE5QrV86oQkuVsoZWa2FU2+fl4GCrynael22JZJS0tVJsp9FIAArP63oa5viazBnHq3DheBUuHK/CQzF8xcXFYeXKlbh8+TIkSUL16tXh5OSE0qVLP9WGcu4XBgBJSUmIiIjAihUrcOfOHaPWT0xMf6rtPSsHB1vcvZuqyraelz5Dh+vJDxTbZWUZIEkoNK/LWIVprIjjVdhwvAoXjterp6AwrBi+QkJC0LRpUwQGBkIIgVOnTmH48OFYt25dges5OjoiPj5eXo6Li4ODgwMA4MiRI0hISEDv3r2RmZmJGzduIDw8HKGhoca+JgJQp5ry6V8A2HjgKnR6g4mrISIiImMohi8ACA4Olr92d3fH+++/r7hOixYtsGDBAgQEBCA6OhqOjo7yKUdfX1/4+voCAG7evIlx48YxeBEREdFrQTF81alTB9HR0fL1WxcuXECNGjUUO27QoAFcXV0REBAASZIwadIkREREwNbWFt7e3s9fOREREVEhlG/4cnd3lz/bcdWqVbCzs4MkSUhOTkaFChUwbtw4xc5HjRqVa7lmzZp52pQvXx5r1qx5htKJiIiICp98w9f333+vZh1EREREr4V8w1fOLSD0ej22bt2Kc+fOAQDq1asHf39/daojIiIiMjOK13xNnz4d9+7dQ9OmTSGEwPbt23HmzBlMmDBBjfqIiIiIzIpi+Prnn3+wdu1aeTkwMBC9evUyaVFERERE5sqoz3Y0GP53jyi9Xg+9Xm/SooiIiIjMleLMl7u7O7p164bGjRsDAI4ePYp27dqZvDAiIiIic6QYvgYPHozmzZvj7NmzkCQJU6dOhZubmxq1EREREZkdxfAVFhaG8ePHo169emrUQ0RERGTWFK/5srCwwOHDh5GRkQGDwSD/IyIiIqKnpzjz9dNPP2HVqlUQQsiPSZKECxcumLQwerGEEDgaHavYztpKa/QHdhMREdHTUwxfJ0+eVKMOMjEhgBI2RRXbJadlqFANERHR60sxfMXFxWHFihW4fPkyJElCjRo18P7776N06dJq1EdERERkVhSv+Ro5ciSsrKzQt29fBAYGQqPRYMSIEWrURkRERGR2FGe+hBAIDg6Wl9955x0EBQWZtCgiIiIic6U481WrVq1cF9dfvHgRNWrUMGlRREREROZKcebrwIEDWLduHUqWLAmDwYDk5GQ4OTlh586dkCQJ+/fvV6FMIiIiIvOgGL5WrlypQhlkavcf6rAkMlqxXWVnWzR1dVahIiIioteTYvgqV66cGnWQCTWu6Yiov24rtktNz8TVWykqVERERPT6UgxfVPj18HwTlZxsFO/ztSQyGgaDKLANERERPR/FC+6JiIiI6MXJd+Zr8+bNBa7YqVOnF14MERERkbnLN3wdOnQIAJCYmIiLFy+ibt260Ov1+PPPP1G/fn2GLyIiIqJnkG/4mj17NgBg+PDh2LNnD6ysrAAAaWlpmDBhgjrVEREREZkZxWu+bt26JQcvALCxscGtW7dMWhQRERGRuVJ8t+Nbb72FgIAA1K9fHxqNBmfPnkWlSpXUqI1eIGsrLZLTMgpsYzAISJJKBREREb2mFMNXeHg4/vjjD/z9998QQuCjjz5Cq1at1KiNXqA61coottl44Cp0eoMK1RAREb2+FMOXJEnQ6XSwtLREYGAgbty4AYnTI0RERETPRDF8zZ49G9evX8etW7cQGBiIrVu3IiEhARMnTlSjPlKZEAJHo2ONamttpTVqRo2IiIj+R/GC++PHj+Prr79G8eLFAQBDhgxBdLTyZwRS4SQEUMKmqFH/0h9mvexyiYiICh3F8FW0aPZH0uScatTr9dDr9aatioiIiMhMKZ52bNCgAcaNG4e4uDisWLECu3btQuPGjdWojYiIiMjsKIavESNGYMeOHbCyskJsbCz69euHtm3bqlEbERERkdlRDF9z5szBqFGj4OvrKz82fvx4hIWFmbQwIiIiInOUb/javXs3du3ahcOHDyMuLk5+PCsrC8eOHVOlOCIiIiJzk2/4atWqFezt7XHu3Dk0a9ZMflySJAwdOlSV4kh99x/qsCTSuHezVna2RVNXZxNXREREZF7yDV9WVlZo2LAhNm/eLL/jMcesWbMwZswYkxdH6mpc0xFRf902qm1qeiau3koxcUVERETmR/GarxMnTmDu3LlISkoCAGRmZqJkyZIMX2aoh+ebqORkgxI2RRXbLomMhsFg4A1ZiYiInpJi+Jo3bx4mTpyI8PBwhIWFYdu2bWjUqJEatdErTqPRGBXUACh+qDcREdHrQvEmqzY2NqhXrx4sLS3x1ltvITg4GCtWrFCjNiIiIiKzozjzlZWVhRMnTsDOzg4///wzqlWrhps3b6pRGxEREZHZUQxfU6ZMQXx8PD799FNMmzYN8fHxGDhwoBq1EREREZkdxfBVtWpVVK1aFQCwfPlykxdEREREZM4Uw9fmzZuxcuVKpKWlQQghP/7bb7+ZtDB69aWmZ/KeYERERE9JMXwtWrQI06dPh7Mzf3G+DqyttEa9M7Gysy2u3ko2qk/eE4yIiOh/jDrt2KRJEzVqoVeAsffiaurqjKPRsU9xTzCh2I6IiOh1oBi+AgIC0L9/f9StWxcWFhby4/yIISIiIqKnp3ifr1mzZsHJyQlCCGRlZcn/iIiIiOjpKc58OTg4YMaMGWrUQkRERGT2FMNXq1atEBERgfr160Or/V/zChUqmLQwMjeCnwNJREQEI8LXDz/8kOcxSZJ4qwl6KvwcSCIiomyK4Wvv3r1q1EFERET0Wsg3fC1duhQff/wxRo8eDUmS8jz/+eefm7QwevUZe0+w7NtM8FYTREREQAHhy8XFBQDQvHnzPM89KYzR68fY67I2HriKDJ3exNUQEREVDvmGr1atWgEArly5glGjRuV6bvz48ejUqZNpKyMiIiIyQ/mGr927d2PXrl04fPgw4uLi5Md1Oh1OnDihSnFkPu4/1Bn9OZAVHW2MaufslI5KZayfpywiIiLVFTjzZW9vj3PnzqFZs2by45IkYdiwYaoUR+ahcU1HRP1126i2qemZuBGXhnZGvDPy/kPd85ZGRESkunzDl5WVFRo2bIjNmzdDp9PBxsYG8fHxuHbtGipXrqxiiVTY9fB8E5WcbIz+HEgiIiJzpvjxQp9//jm2b9+OpKQkBAQEYO3atZg8ebIKpRERERGZH8Xwdf78eXTv3h3bt29H586dMW/ePFy/fl2N2oiIiIjMjmL4EiL7/kz79++Hp6cnACAzM9O0VRERERGZKcU73FepUgXt27dHqVKlUKtWLWzevBklSpQwqvPw8HCcPXsWkiQhNDQUbm5u8nNHjhzB3LlzodFoUKVKFYSFhUGjUcyC9BpITc806tqv6hVLwaW8ccciERHRq0IxfE2fPh1///03qlWrBgB48803MXv2bMWOjx07huvXr2P9+vW4cuUKQkNDsX79evn5zz77DKtXr4azszOGDx+OgwcPwt3d/TleCpmDGhVK4lJMkmK71PRM/H0jUYWKiIiIXqx8w9fy5cvRv39/WFhYoFatWvjrr79Qp04d1K5dG6GhoQgP9OfRUAAAEshJREFUDy+w48OHD8PLywsAUK1aNSQnJyMtLQ02Ntn3cIqIiJC/tre3R2Iif5ES4FG/HDzql1NstyQyGkIIHI2ONapfayut0XfkJyIiMqV8w9f+/fvRv39/eXn27NlYvXo1ACAmJkax4/j4eLi6usrL9vb2uHv3rhy4cv6Pi4vDoUOHEBwcXGB/pUpZQ6u1UNzui+DgYKvKdl4nzk7pRt+XK1Onh7N98QLbaDQSAAmVypcyqs+E1Icc11cAx6Bw4XgVLhyvwuP/2rv3oCqrf4/jH2C7RWL/RGxvUPRUmknHS0WlKZqXLn84TTOl/jC7OdM4pmOXmbQYdULHO/qHZY46Fl00hULG+TWlljOaFqjlkBesU+YMASECRxECRGCfPxw5YV6eLT43eL/+crMf17NgCXxcaz3fddXwdWmj/ZVe38jZjpe3J0mVlZV6+eWXlZaWpm7drv1L9MyZ2pDveSP8fp/Ky6stuVdHEkol+gMFp3Q2/Nr/xpqbgwoPD9PZs8b+XVTXnGdcbcb3lrswXu7CeDnPtcLwVXe4XytgXSlIXS4QCKiioqLl9enTp+X3+1te19TUaOrUqXr99dc1YsSI67YHAADQHhh+vPDvYczIzFdycrJ27twpSSooKFAgEGhZapSkZcuW6cUXX9TDDz8cSn8BAABc7arLjvn5+Ro9enTL68rKSo0ePVrBYNDQ5vikpCQNGDBAkyZNUlhYmNLS0pSTkyOfz6cRI0Zo27ZtKiwsVHZ2tiTpiSeeUEpKSts/IwAAAAe7avjasWNHmxufNWtWq9eJiYktfz527Fib20f7FBXpUVXN+Wte09wcNLT8DQCA01w1fCUkXP9xf8AMRkpCZH97Uv9bXW/4IO7b430aOiC+rV0DAKDNrltkFXCiBxMDyjVY46u6tkEn/6yiJhgAwBEIX3Clf4+9UwPv8iu8qfm6116aHesa3dlQ29db8gQAoC04TBEAAMBCzHwBl2loamaJEgBgGsIXcBl/1y6Gr2WJEgAQKsIXOoTq2gbDT0b27x1j6HBvAABuBOELrnVLZCedKjt33etuj/fp5J9Vhtqsrm3Q/xSdJXwBAExD+IJrPXB3nMoNHNg9dEC8DhScMvS0o9HZsUvYHwYACBXhC2gD9ocBAEJF+AIuw/4wAICZqPMF/E3/3jHyRXkNXXtpfxgAAKFg5gv4mzH3JRieyQp1fxgAABIzXwAAAJZi5gsdQlSkx/CG94amZsMb6UPZH/ZfgWhD1/FUJAC0b4QvdAihhBmjpSP6944xvOerurZBf5yu0TgD5S54KhIA2jfCF3CDzNofRu0wAGjfCF+Aw1A7DADaN8IXcBmz9oeZgVkyAHAfwhdwGTP2h0nGN+eHUriVWTIAcB/CF2ABo5vzzTzYm1kyAHAGwhfQBkaXKJP63aqBfWKvO1NlZuFWZskAwBkIX0AbmLFEydmSANC+Eb4ABwmldti5vxr0wy+nDV8fSlBjiRIAzEP4AhwklNphu/NLQiryGspeMpYoAcA8hC/ApUIt8mrWcqbRWbL4uFrddmuUoTYBoD0jfAEWMbo534zaYaEehWTGLNmZ6nqdKjtn6FqWMgG0Z4QvwCJGw0QotcOMMusopFDEx96is+Fhhq5lKRNAe0b4AhzGCRX27X7ikg3/ANozwhfgMGZV2DfKzCVKo9jwD6A9I3wBLmbGLJlZG/nvufNWDfvvOEPthoJZMgBuQ/gCXMwts2Tn/mrQvsN/6vCJCsPtUhYDQHtF+AI6CDtnyXbnl+i3kio1Nwevey3nWwJo78KCweD1fxo6QHl5tSX38ft9lt0LbcNYmedAwSl1je58U9uMiYnS2bO1173u0lKmL8prqF2zjlgqr6qTNyL8prbppkDH95e7MF7O4/f7rvoeM18A/sHOJy6dsOFfCm050yiWPQFIhC8AV2DnXjKnVO43A8ueACTCF4A2MjpLVt8cVKTBIqtGOWWWzKhQZtPKq+pMeUiCUAfYj/AFoE2M/iIvrKg1fLyQ3WUxJPtnycxY9pRY+gScgPAFwBIP3B2ncoMHa9tdPPbcXw364ZfThq8PpQ92BjqJg9ABJyB8AXAcu4vH7s4vuenBK9RAZ1ZQ4yB0wH6ELwCOY3fx2FCCmlGhBDon7E8L5SD0UPanEdQAwhcAWMLM/WmhMGNGjVMGgNAQvgC4mp01ycwSyv60UDhhRo1yGwDhC4DLhfLL+ejvFa4IamYse0rOOAidWTKA8AWgAzErqBll98ybEw5CD0Uos2RGMZsGJyB8AcAVmPEL2oyHA0JhxkHoZj7FyRFPMMsvhWeUeFs32+5P+AIAi7hlf9qY+xL01Jh+hg5Cd9tTnIAThAWDwev/18YBrDqtnZPh3YOxchfGKzRHf69QbX3jTW/XaKiLiYkyFL5CcWnPmS/Ka+h6M5Yzy6vq5I0Iv6ltOoGvaxdVV9XZ3Q3X6OyN0L39/Kbew+/3XfU9Zr4AwIHM2pdk59KnE87idMPTrjcixhep8KZmu7vhGnYvPxO+AKADsfMgdCfUOnPCEU8A4QsAOhA7D0IPhRm1zthzBqcgfAEA/iGUg9DNqJ9mRq0zM04MAG4E4QsA0CZ2n8UZCrOObbJbeHiYodIguOj2eJ+GDoi37f6ELwCAZewst2HWsU1wl+raBp3809iSulkIXwAAy9g5S2bWsU1OYEZpkPZq3X8KbJ8lJHwBABwplFkyo+w+4gmQCF8AAIdqj0c8ARLhCwDQgZgxm+YEzRHhqm6Hn5cZmpuDCru5JexCRvgCAHQYZp0cYDeO7zKusKxGlefqbe0D4QsAAHQY/x57p34pPGNrH0w9XXTJkiVKSUnRpEmTdOTIkVbv5ebmasKECUpJSdGaNWvM7AYAAIBjmBa+Dh48qMLCQmVlZWnx4sVavHhxq/cXLVqk1atXa8uWLfr+++914sQJs7oCAADgGKaFr7y8PD366KOSpL59+6qqqko1NTWSpKKiInXt2lU9evRQeHi4Ro0apby8PLO6AgAA4Bim7fmqqKjQgAEDWl7HxsaqvLxc0dHRKi8vV2xsbKv3ioqKrtlet25R8ngizOpuK36/z5L7oO0YK3dhvNyF8XIXxsu4U1Xnbf16WbbhPhhsWzXZM2esqdzLEyPuwVi5C+PlLoyXuzBeoYnv2tn0r9e1wp1py46BQEAVFRUtr0+fPi2/33/F98rKyhQIBMzqCgAAgGOYFr6Sk5O1c+dOSVJBQYECgYCio6MlSb169VJNTY2Ki4vV2Nio3bt3Kzk52ayuAAAAOIZpy45JSUkaMGCAJk2apLCwMKWlpSknJ0c+n0+PPfaY5s+frzfeeEOSNG7cON1xxx1mdQUAAMAxwoJt3YxlEavWslk3dw/Gyl0YL3dhvNyF8XIeW/Z8AQAA4J8IXwAAABYifAEAAFiI8AUAAGAhwhcAAICFCF8AAAAWInwBAABYiPAFAABgIcIXAACAhQhfAAAAFiJ8AQAAWMg1ZzsCAAC0B8x8AQAAWIjwBQAAYCHCFwAAgIUIXwAAABYifAEAAFiI8AUAAGAhwhcAAICFPHZ3wCmWLFmiw4cPKywsTHPmzNHgwYPt7hKu4Ndff9WMGTM0ZcoUPffccyotLdWbb76ppqYm+f1+rVixQl6v1+5uQlJ6eroOHTqkxsZGTZs2TYMGDWKsHKqurk6pqamqrKzU+fPnNWPGDCUmJjJeDldfX68nnnhCM2bM0LBhwxgvF2HmS9LBgwdVWFiorKwsLV68WIsXL7a7S7iC2tpaLVy4UMOGDWv52LvvvqvJkydr8+bNuu2225SdnW1jD3HJ/v379dtvvykrK0vvv/++lixZwlg52O7duzVw4EBt2rRJq1at0rJlyxgvF1i7dq26du0qiZ+FbkP4kpSXl6dHH31UktS3b19VVVWppqbG5l7hcl6vVxs2bFAgEGj52IEDB/TII49IksaMGaO8vDy7uoe/efDBB/XOO+9Ikv71r3+prq6OsXKwcePGaerUqZKk0tJSxcXFMV4O9/vvv+vEiRMaPXq0JH4Wug3hS1JFRYW6devW8jo2Nlbl5eU29ghX4vF4FBkZ2epjdXV1LVPr3bt3Z9wcIiIiQlFRUZKk7OxsPfzww4yVC0yaNEmzZs3SnDlzGC+HW758uVJTU1teM17uwp6vK+C4S3di3Jxn165dys7OVkZGhh5//PGWjzNWzpSZmamff/5Zs2fPbjVGjJezbNu2Tffee6969+59xfcZL+cjfEkKBAKqqKhoeX369Gn5/X4bewSjoqKiVF9fr8jISJWVlbVakoS99u3bp3Xr1un999+Xz+djrBzs2LFj6t69u3r06KG7775bTU1NuuWWWxgvh9qzZ4+Kioq0Z88enTp1Sl6vl+8vl2HZUVJycrJ27twpSSooKFAgEFB0dLTNvYIRw4cPbxm7r7/+WiNHjrS5R5Ck6upqpaena/369YqJiZHEWDnZjz/+qIyMDEkXt2HU1tYyXg62atUqbd26VZ999pkmTpyoGTNmMF4uExZkflKStHLlSv34448KCwtTWlqaEhMT7e4SLnPs2DEtX75cJSUl8ng8iouL08qVK5Wamqrz58+rZ8+eWrp0qTp16mR3Vzu8rKwsrV69WnfccUfLx5YtW6Z58+YxVg5UX1+vuXPnqrS0VPX19Zo5c6YGDhyot956i/FyuNWrVyshIUEjRoxgvFyE8AUAAGAhlh0BAAAsRPgCAACwEOELAADAQoQvAAAACxG+AAAALESRVQCWSk9P19GjR3X+/HkdP35c9913nyRp/Pjx+vbbb5Wamqq4uLibes/33ntPXbp0Ub9+/VRQUKDp06ff1Pbbqri4WLNnz9aGDRuoMQh0AJSaAGCL4uJiTZ48WXv37jX1PkeOHNHSpUu1ZcsWU+/TVllZWTp+/LgWLFhgd1cAmIyZLwCOMXbsWH344Yc6dOiQ9u3bp2AwqOPHj+vJJ5/UhQsXdODAAQWDQX344YeKiorSV199pU2bNikYDCo2NlaLFi1St27dWrW5du1aTZkyRZKUk5Oj3NxcrVy5UmPHjtULL7ygvXv3qri4WAsWLNCwYcNa/t7333+vdevWaePGjZKkw4cPa+HChcrOztbGjRu1fft2NTU1qU+fPkpLS1NkZKTeeecd5eXlSZLi4+O1YsUKderUSUlJSZowYYKam5s1depUzZo1S9LF4qYpKSmaMGGCnn76aa1evVqvvfaaYmNjLfhqA7ALe74AONKxY8eUnp6ujIwMrVmzRsOHD1dmZqa8Xq9yc3NVWlqqdevW6aOPPtKWLVs0ZMgQrV+/vlUbTU1N2r9/v5KTk694j86dOysjI0PTp0/XJ5980uq94cOHq6ysTEVFRZKk7du3a+LEiTpy5Ii++eYbffrpp8rKypLP59Pnn3+uxsZGdenSRZs3b1ZmZqaqq6v13XffSZJqa2s1atQozZs3T9u3b1efPn20ceNGbdq0SfX19ZLUEtIuhTcA7RczXwAcaeDAgfJ6vYqPj1dzc7Puv/9+SVJcXJyqq6uVn5+v8vJyvfTSS5KkhoYG9erVq1UbZ8+eVadOna66j2rIkCGSpJ49e6qqqqrVe2FhYZowYYK2bdummTNnau/evZo5c6a2bNmiP/74Qy+88IKki8HK4/HI4/EoPDxckydPlsfj0cmTJ3XmzBlJUjAYVFJSkiRp5MiR2rx5s1JTUzVq1CilpKS03DMhIUElJSVt/dIBcDjCFwBHioiIaPXa4/n/H1fBYFBer1eDBw/+x2xXKC5v83Ljx4/Xc889pxEjRuiee+5RdHS0vF6vxo4dq7fffrvVtYcOHdLWrVu1detWRUVF6dVXX231/qVz9vr27asvv/xSP/zwg3bs2KGPP/5YmZmZN/w5AHAflh0BuNKgQYN05MgRlZeXS7q4LLhr165W18TExOjChQuqqam5oXt0795d/fv3V3p6usaPHy9JSkpK0t69e/XXX39Jkj799FPl5+ersrJSCQkJioqKUklJiX766Sc1NDT8o80vvvhCR48e1fDhw5WWlqbS0lI1NjZKkkpKSpSQkHBDfQXgHsx8AXCluLg4zZ07V9OmTVOXLl0UGRmp5cuXt7omIiJCDz30kHJzc/X444/f0H2eeuopLVu2TA888ICki6Hv2Wef1fPPP6/OnTsrEAjo6aef1l133aWMjAw988wz6tevn1555RWtWbNGQ4cObdXenXfeqbS0NHm9XgWDQU2dOlUej0eNjY3Kz8/X/Pnzb6ifANyDUhMA2rW2lppYsGCBEhMTW+3NMsNnn32mgoICSk0AHQDLjgDatcGDBys5OVkffPBBSH+vrKxMEydOVG1trSZOnGhS7y4qLi5WTk6OZs+ebep9ADgDM18AAAAWYuYLAADAQoQvAAAACxG+AAAALET4AgAAsBDhCwAAwEL/B/+7P5ArrQaPAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pdrG0N6nuTwL",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a8bb1b3c-412d-44e5-abf1-525428446be8"
      },
      "source": [
        "print(f'The median number of years of government survival is {kmf.median_survival_time_}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The median number of years of government survival is 4.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "id": "NQHPG2x6uTwQ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "c42631e6-314b-4488-8309-89a699eea278"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(10,7))\n",
        "for r in df['democracy'].unique():\n",
        "    ix = df['democracy'] == r\n",
        "    kmf.fit(df['duration'].loc[ix], df['observed'].loc[ix], label=r)\n",
        "    kmf.plot(ax=ax)\n",
        "plt.title('Estimated probability of government survival vs number of years')\n",
        "plt.xlabel('Time (in years)')\n",
        "plt.ylabel('Estimated probability of government survival')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PBKPtOFKuTwV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3aca95f8-dea3-4165-ed55-9cef803d5fbc"
      },
      "source": [
        "for r in df['democracy'].unique():\n",
        "    ix = df['democracy'] == r\n",
        "    kmf.fit(df['duration'].loc[ix], df['observed'].loc[ix], label=r)\n",
        "    print(f'The median number of years for a {r} is {kmf.median_survival_time_}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The median number of years for a Non-democracy is 6.0\n",
            "The median number of years for a Democracy is 3.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZhK-X5-fuTwc"
      },
      "source": [
        "How can we tell if these survival functions are different?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uL59D8GAuTwc"
      },
      "source": [
        "# Log-rank test (not recommended but still very common statistical test)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "id": "LDq8D5ovuTwe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0384b506-52dd-45b3-a941-bc969ced78b7"
      },
      "source": [
        "df['democracy'].unique()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Non-democracy', 'Democracy'], dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kaRMHNtSuTwi"
      },
      "source": [
        "from lifelines.statistics import logrank_test"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1qrrTDDWuTwn"
      },
      "source": [
        "ix = df['democracy'] == 'Democracy'\n",
        "T_democracy, E_democracy = df.loc[ix, 'duration'], df.loc[ix, 'observed']\n",
        "T_non_democracy, E_non_democracy = df.loc[~ix, 'duration'], df.loc[~ix, 'observed']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "Qw20wiiRuTws",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "3b3839be-bdcc-409a-f982-4095da270066"
      },
      "source": [
        "results = logrank_test(T_democracy, T_non_democracy, event_observed_A=E_democracy, event_observed_B=E_non_democracy)\n",
        "results.print_summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<lifelines.StatisticalResult: logrank_test>\n",
              "               t_0 = -1\n",
              " null_distribution = chi squared\n",
              "degrees_of_freedom = 1\n",
              "         test_name = logrank_test\n",
              "\n",
              "---\n",
              " test_statistic      p  -log2(p)\n",
              "         260.47 <0.005    192.23"
            ],
            "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",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>t_0</th>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>null_distribution</th>\n",
              "      <td>chi squared</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>degrees_of_freedom</th>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>test_name</th>\n",
              "      <td>logrank_test</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div><table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>test_statistic</th>\n",
              "      <th>p</th>\n",
              "      <th>-log2(p)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>260.47</td>\n",
              "      <td>&lt;0.005</td>\n",
              "      <td>192.23</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/latex": "\\begin{tabular}{lrrr}\n{} & {test_statistic} & {p} & {-log2(p)} \\\\\n0 & 260.47 & 0.00 & 192.23 \\\\\n\\end{tabular}\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1Oms68oOuTww",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8bcfdafc-68fe-4a84-8703-e98e061aa63c"
      },
      "source": [
        "print(results.p_value)\n",
        "print(results.test_statistic)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.3557143218482446e-58\n",
            "260.46953907795944\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qTX47C6-uTw1"
      },
      "source": [
        "# Univariate Cox regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "meyGKaYcuTw4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 474
        },
        "outputId": "a5d96704-cea4-49f0-f4d9-2c62b5fa1a8a"
      },
      "source": [
        "from lifelines import CoxPHFitter\n",
        "\n",
        "cph = CoxPHFitter()\n",
        "df_Uni_Cox = df.copy()\n",
        "df_Uni_Cox['indicator'] = df_Uni_Cox['democracy'] == 'Democracy'\n",
        "cph.fit(df_Uni_Cox[['indicator', 'duration', 'observed']], 'duration', 'observed')\n",
        "cph.print_summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<lifelines.CoxPHFitter: fitted with 1808 total observations, 340 right-censored observations>\n",
              "             duration col = 'duration'\n",
              "                event col = 'observed'\n",
              "      baseline estimation = breslow\n",
              "   number of observations = 1808\n",
              "number of events observed = 1468\n",
              "   partial log-likelihood = -9614.27\n",
              "         time fit was run = 2023-02-08 10:26:23 UTC\n",
              "\n",
              "---\n",
              "            coef  exp(coef)   se(coef)   coef lower 95%   coef upper 95%  exp(coef) lower 95%  exp(coef) upper 95%\n",
              "covariate                                                                                                         \n",
              "indicator   0.96       2.62       0.06             0.84             1.09                 2.32                 2.96\n",
              "\n",
              "            cmp to     z      p   -log2(p)\n",
              "covariate                                 \n",
              "indicator     0.00 15.40 <0.005     175.43\n",
              "---\n",
              "Concordance = 0.59\n",
              "Partial AIC = 19230.53\n",
              "log-likelihood ratio test = 264.03 on 1 df\n",
              "-log2(p) of ll-ratio test = 194.81"
            ],
            "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",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>model</th>\n",
              "      <td>lifelines.CoxPHFitter</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>duration col</th>\n",
              "      <td>'duration'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>event col</th>\n",
              "      <td>'observed'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>baseline estimation</th>\n",
              "      <td>breslow</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number of observations</th>\n",
              "      <td>1808</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number of events observed</th>\n",
              "      <td>1468</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>partial log-likelihood</th>\n",
              "      <td>-9614.27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>time fit was run</th>\n",
              "      <td>2023-02-08 10:26:23 UTC</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div><table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th style=\"min-width: 12px;\"></th>\n",
              "      <th style=\"min-width: 12px;\">coef</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef)</th>\n",
              "      <th style=\"min-width: 12px;\">se(coef)</th>\n",
              "      <th style=\"min-width: 12px;\">coef lower 95%</th>\n",
              "      <th style=\"min-width: 12px;\">coef upper 95%</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef) lower 95%</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef) upper 95%</th>\n",
              "      <th style=\"min-width: 12px;\">cmp to</th>\n",
              "      <th style=\"min-width: 12px;\">z</th>\n",
              "      <th style=\"min-width: 12px;\">p</th>\n",
              "      <th style=\"min-width: 12px;\">-log2(p)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>indicator</th>\n",
              "      <td>0.96</td>\n",
              "      <td>2.62</td>\n",
              "      <td>0.06</td>\n",
              "      <td>0.84</td>\n",
              "      <td>1.09</td>\n",
              "      <td>2.32</td>\n",
              "      <td>2.96</td>\n",
              "      <td>0.00</td>\n",
              "      <td>15.40</td>\n",
              "      <td>&lt;0.005</td>\n",
              "      <td>175.43</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table><br><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",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Concordance</th>\n",
              "      <td>0.59</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Partial AIC</th>\n",
              "      <td>19230.53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>log-likelihood ratio test</th>\n",
              "      <td>264.03 on 1 df</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>-log2(p) of ll-ratio test</th>\n",
              "      <td>194.81</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/latex": "\\begin{tabular}{lrrrrrrrrrrr}\n{} & {coef} & {exp(coef)} & {se(coef)} & {coef lower 95%} & {coef upper 95%} & {exp(coef) lower 95%} & {exp(coef) upper 95%} & {cmp to} & {z} & {p} & {-log2(p)} \\\\\n{covariate} & {} & {} & {} & {} & {} & {} & {} & {} & {} & {} & {} \\\\\nindicator & 0.96 & 2.62 & 0.06 & 0.84 & 1.09 & 2.32 & 2.96 & 0.00 & 15.40 & 0.00 & 175.43 \\\\\n\\end{tabular}\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:17.131185Z",
          "start_time": "2020-01-09T22:37:16.573714Z"
        },
        "id": "USPOd2fOuTw8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "7c02f5fa-61f4-4d87-a4c0-21e96c70bba9"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(10,7))\n",
        "\n",
        "for r in df['regime'].unique():\n",
        "    ix = df['regime'] == r\n",
        "    kmf.fit(df['duration'].loc[ix], df['observed'].loc[ix], label=r)\n",
        "    kmf.survival_function_.plot(ax=ax)\n",
        "plt.title('Estimated probability of government survival vs number of years')\n",
        "plt.xlabel('Time (in years)')\n",
        "plt.ylabel('Estimated probability of government survival')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:17.690930Z",
          "start_time": "2020-01-09T22:37:17.134877Z"
        },
        "scrolled": false,
        "id": "ydmaYBqnuTxA",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "acd75c18-71f0-4fe2-d06b-bb1c061da7aa"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(10,7))\n",
        "for r in df['un_continent_name'].unique():\n",
        "    ix = df['un_continent_name'] == r\n",
        "    kmf.fit(df['duration'].loc[ix], df['observed'].loc[ix], label=r)\n",
        "    kmf.plot(ax=ax)\n",
        "plt.title('Estimated probability of government survival vs number of years')\n",
        "plt.xlabel('Time (in years)')\n",
        "plt.ylabel('Estimated probability of government survival')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:17.723699Z",
          "start_time": "2020-01-09T22:37:17.694880Z"
        },
        "scrolled": false,
        "id": "NXIl1sBfuTxF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "15aea511-f39e-4c5d-95c0-688ab09c1608"
      },
      "source": [
        "df.query('ctryname == \"United States of America\"').tail(3)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                      ctryname    un_region_name un_continent_name  \\\n",
              "1721  United States of America  Northern America          Americas   \n",
              "1722  United States of America  Northern America          Americas   \n",
              "1723  United States of America  Northern America          Americas   \n",
              "\n",
              "               ehead  democracy            regime  start_year  duration  \\\n",
              "1721     George Bush  Democracy  Presidential Dem        1989         4   \n",
              "1722    Bill Clinton  Democracy  Presidential Dem        1993         8   \n",
              "1723  George W. Bush  Democracy  Presidential Dem        2001         8   \n",
              "\n",
              "      observed  \n",
              "1721         1  \n",
              "1722         1  \n",
              "1723         0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-2505bf06-ec83-4c17-8878-4bd68652c658\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ctryname</th>\n",
              "      <th>un_region_name</th>\n",
              "      <th>un_continent_name</th>\n",
              "      <th>ehead</th>\n",
              "      <th>democracy</th>\n",
              "      <th>regime</th>\n",
              "      <th>start_year</th>\n",
              "      <th>duration</th>\n",
              "      <th>observed</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1721</th>\n",
              "      <td>United States of America</td>\n",
              "      <td>Northern America</td>\n",
              "      <td>Americas</td>\n",
              "      <td>George Bush</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Presidential Dem</td>\n",
              "      <td>1989</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1722</th>\n",
              "      <td>United States of America</td>\n",
              "      <td>Northern America</td>\n",
              "      <td>Americas</td>\n",
              "      <td>Bill Clinton</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Presidential Dem</td>\n",
              "      <td>1993</td>\n",
              "      <td>8</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1723</th>\n",
              "      <td>United States of America</td>\n",
              "      <td>Northern America</td>\n",
              "      <td>Americas</td>\n",
              "      <td>George W. Bush</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Presidential Dem</td>\n",
              "      <td>2001</td>\n",
              "      <td>8</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2505bf06-ec83-4c17-8878-4bd68652c658')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2505bf06-ec83-4c17-8878-4bd68652c658 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2505bf06-ec83-4c17-8878-4bd68652c658');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:17.748036Z",
          "start_time": "2020-01-09T22:37:17.725349Z"
        },
        "scrolled": false,
        "id": "Qzy7VJ_luTxI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "0b6e4490-bf17-4f96-a3b2-8677a824d05e"
      },
      "source": [
        "df.query('ctryname == \"United Kingdom\"').tail(3)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            ctryname   un_region_name un_continent_name         ehead  \\\n",
              "1710  United Kingdom  Northern Europe            Europe    John Major   \n",
              "1711  United Kingdom  Northern Europe            Europe    Tony Blair   \n",
              "1712  United Kingdom  Northern Europe            Europe  Gordon Brown   \n",
              "\n",
              "      democracy             regime  start_year  duration  observed  \n",
              "1710  Democracy  Parliamentary Dem        1990         7         1  \n",
              "1711  Democracy  Parliamentary Dem        1997        10         1  \n",
              "1712  Democracy  Parliamentary Dem        2007         2         0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-332a83ae-14e0-4854-b455-45d1e30a8c23\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ctryname</th>\n",
              "      <th>un_region_name</th>\n",
              "      <th>un_continent_name</th>\n",
              "      <th>ehead</th>\n",
              "      <th>democracy</th>\n",
              "      <th>regime</th>\n",
              "      <th>start_year</th>\n",
              "      <th>duration</th>\n",
              "      <th>observed</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1710</th>\n",
              "      <td>United Kingdom</td>\n",
              "      <td>Northern Europe</td>\n",
              "      <td>Europe</td>\n",
              "      <td>John Major</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Parliamentary Dem</td>\n",
              "      <td>1990</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1711</th>\n",
              "      <td>United Kingdom</td>\n",
              "      <td>Northern Europe</td>\n",
              "      <td>Europe</td>\n",
              "      <td>Tony Blair</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Parliamentary Dem</td>\n",
              "      <td>1997</td>\n",
              "      <td>10</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1712</th>\n",
              "      <td>United Kingdom</td>\n",
              "      <td>Northern Europe</td>\n",
              "      <td>Europe</td>\n",
              "      <td>Gordon Brown</td>\n",
              "      <td>Democracy</td>\n",
              "      <td>Parliamentary Dem</td>\n",
              "      <td>2007</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-332a83ae-14e0-4854-b455-45d1e30a8c23')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-332a83ae-14e0-4854-b455-45d1e30a8c23 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-332a83ae-14e0-4854-b455-45d1e30a8c23');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "ExecuteTime": {
          "end_time": "2020-01-09T22:37:18.129226Z",
          "start_time": "2020-01-09T22:37:17.749639Z"
        },
        "id": "010659VeuTxM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "dc9cbd4f-61d3-413e-d88c-d826ec852725"
      },
      "source": [
        "ix_US = df['ctryname'] == 'United States of America'\n",
        "ix_UK = df['ctryname'] == 'United Kingdom'\n",
        "\n",
        "kmf_US = KaplanMeierFitter()\n",
        "kmf_US.fit(df['duration'].loc[ix_US], df['observed'].loc[ix_US], label='USA')\n",
        "\n",
        "kmf_UK = KaplanMeierFitter()\n",
        "kmf_UK.fit(df['duration'].loc[ix_UK], df['observed'].loc[ix_UK], label='UK')\n",
        "\n",
        "plt.figure(figsize=(10,7))\n",
        "ax = plt.subplot(111)\n",
        "kmf_US.plot(ax=ax)\n",
        "kmf_UK.plot(ax=ax)\n",
        "plt.title('Estimated probability of government survival vs number of years')\n",
        "plt.xlabel('Time (in years)')\n",
        "plt.ylabel('Estimated probability of government survival')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "id": "nCPi7YY1uTxQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4d73accc-de24-4f0a-c01e-5d484efa837a"
      },
      "source": [
        "df.columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['ctryname', 'un_region_name', 'un_continent_name', 'ehead', 'democracy',\n",
              "       'regime', 'start_year', 'duration', 'observed'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "xphf-zpauTxU"
      },
      "source": [
        "df = df.drop(columns=['ctryname', 'un_region_name', 'ehead', 'regime', 'start_year'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NOuRvPJXuTxZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "597ec472-bf1c-4f65-d60c-546309341f36"
      },
      "source": [
        "df.columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['un_continent_name', 'democracy', 'duration', 'observed'], dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tzcg35DtuTxd"
      },
      "source": [
        "# df_hazard = pd.get_dummies(df, drop_first=True, columns=df.columns.drop(['duration', 'observed']))\n",
        "df_hazard = pd.get_dummies(df, columns=df.columns.drop(['duration', 'observed']))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MGxRvM20uTxg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "bc185949-7546-48cf-f3dd-3fd76417dfb4"
      },
      "source": [
        "df_hazard.columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['duration', 'observed', 'un_continent_name_Africa',\n",
              "       'un_continent_name_Americas', 'un_continent_name_Asia',\n",
              "       'un_continent_name_Europe', 'un_continent_name_Oceania',\n",
              "       'democracy_Democracy', 'democracy_Non-democracy'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i9b7PSd2uTxl"
      },
      "source": [
        "df_hazard = df_hazard.drop(columns=['un_continent_name_Americas', 'democracy_Democracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3-O2MTLPuTxo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c06e3a8d-ceff-4800-b578-88cf82a1ac2e"
      },
      "source": [
        "df_hazard.columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['duration', 'observed', 'un_continent_name_Africa',\n",
              "       'un_continent_name_Asia', 'un_continent_name_Europe',\n",
              "       'un_continent_name_Oceania', 'democracy_Non-democracy'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VaQCN1ChuTxs"
      },
      "source": [
        "# Multivariate Cox regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IUKzqcFvuTxs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 663
        },
        "outputId": "5f3bbaad-7937-4625-d53d-c47d0906f140"
      },
      "source": [
        "from lifelines import CoxPHFitter\n",
        "\n",
        "cph = CoxPHFitter(penalizer=0.1)\n",
        "cph.fit(df_hazard, 'duration', 'observed')\n",
        "cph.print_summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<lifelines.CoxPHFitter: fitted with 1808 total observations, 340 right-censored observations>\n",
              "             duration col = 'duration'\n",
              "                event col = 'observed'\n",
              "                penalizer = 0.1\n",
              "                 l1 ratio = 0.0\n",
              "      baseline estimation = breslow\n",
              "   number of observations = 1808\n",
              "number of events observed = 1468\n",
              "   partial log-likelihood = -9613.12\n",
              "         time fit was run = 2023-02-08 10:28:32 UTC\n",
              "\n",
              "---\n",
              "                            coef  exp(coef)   se(coef)   coef lower 95%   coef upper 95%  exp(coef) lower 95%  exp(coef) upper 95%\n",
              "covariate                                                                                                                         \n",
              "un_continent_name_Africa   -0.20       0.82       0.08            -0.36            -0.04                 0.70                 0.96\n",
              "un_continent_name_Asia     -0.06       0.95       0.07            -0.20             0.09                 0.82                 1.09\n",
              "un_continent_name_Europe    0.23       1.26       0.06             0.11             0.35                 1.12                 1.43\n",
              "un_continent_name_Oceania  -0.12       0.89       0.11            -0.33             0.10                 0.72                 1.10\n",
              "democracy_Non-democracy    -0.72       0.49       0.06            -0.84            -0.60                 0.43                 0.55\n",
              "\n",
              "                            cmp to      z      p   -log2(p)\n",
              "covariate                                                  \n",
              "un_continent_name_Africa      0.00  -2.48   0.01       6.25\n",
              "un_continent_name_Asia        0.00  -0.77   0.44       1.19\n",
              "un_continent_name_Europe      0.00   3.79 <0.005      12.70\n",
              "un_continent_name_Oceania     0.00  -1.07   0.29       1.80\n",
              "democracy_Non-democracy       0.00 -11.48 <0.005      98.97\n",
              "---\n",
              "Concordance = 0.62\n",
              "Partial AIC = 19236.25\n",
              "log-likelihood ratio test = 266.31 on 5 df\n",
              "-log2(p) of ll-ratio test = 181.91"
            ],
            "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",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>model</th>\n",
              "      <td>lifelines.CoxPHFitter</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>duration col</th>\n",
              "      <td>'duration'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>event col</th>\n",
              "      <td>'observed'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>penalizer</th>\n",
              "      <td>0.1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>l1 ratio</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>baseline estimation</th>\n",
              "      <td>breslow</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number of observations</th>\n",
              "      <td>1808</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>number of events observed</th>\n",
              "      <td>1468</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>partial log-likelihood</th>\n",
              "      <td>-9613.12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>time fit was run</th>\n",
              "      <td>2023-02-08 10:28:32 UTC</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div><table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th style=\"min-width: 12px;\"></th>\n",
              "      <th style=\"min-width: 12px;\">coef</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef)</th>\n",
              "      <th style=\"min-width: 12px;\">se(coef)</th>\n",
              "      <th style=\"min-width: 12px;\">coef lower 95%</th>\n",
              "      <th style=\"min-width: 12px;\">coef upper 95%</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef) lower 95%</th>\n",
              "      <th style=\"min-width: 12px;\">exp(coef) upper 95%</th>\n",
              "      <th style=\"min-width: 12px;\">cmp to</th>\n",
              "      <th style=\"min-width: 12px;\">z</th>\n",
              "      <th style=\"min-width: 12px;\">p</th>\n",
              "      <th style=\"min-width: 12px;\">-log2(p)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>un_continent_name_Africa</th>\n",
              "      <td>-0.20</td>\n",
              "      <td>0.82</td>\n",
              "      <td>0.08</td>\n",
              "      <td>-0.36</td>\n",
              "      <td>-0.04</td>\n",
              "      <td>0.70</td>\n",
              "      <td>0.96</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-2.48</td>\n",
              "      <td>0.01</td>\n",
              "      <td>6.25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>un_continent_name_Asia</th>\n",
              "      <td>-0.06</td>\n",
              "      <td>0.95</td>\n",
              "      <td>0.07</td>\n",
              "      <td>-0.20</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.82</td>\n",
              "      <td>1.09</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.77</td>\n",
              "      <td>0.44</td>\n",
              "      <td>1.19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>un_continent_name_Europe</th>\n",
              "      <td>0.23</td>\n",
              "      <td>1.26</td>\n",
              "      <td>0.06</td>\n",
              "      <td>0.11</td>\n",
              "      <td>0.35</td>\n",
              "      <td>1.12</td>\n",
              "      <td>1.43</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.79</td>\n",
              "      <td>&lt;0.005</td>\n",
              "      <td>12.70</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>un_continent_name_Oceania</th>\n",
              "      <td>-0.12</td>\n",
              "      <td>0.89</td>\n",
              "      <td>0.11</td>\n",
              "      <td>-0.33</td>\n",
              "      <td>0.10</td>\n",
              "      <td>0.72</td>\n",
              "      <td>1.10</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-1.07</td>\n",
              "      <td>0.29</td>\n",
              "      <td>1.80</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>democracy_Non-democracy</th>\n",
              "      <td>-0.72</td>\n",
              "      <td>0.49</td>\n",
              "      <td>0.06</td>\n",
              "      <td>-0.84</td>\n",
              "      <td>-0.60</td>\n",
              "      <td>0.43</td>\n",
              "      <td>0.55</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-11.48</td>\n",
              "      <td>&lt;0.005</td>\n",
              "      <td>98.97</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table><br><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",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Concordance</th>\n",
              "      <td>0.62</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Partial AIC</th>\n",
              "      <td>19236.25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>log-likelihood ratio test</th>\n",
              "      <td>266.31 on 5 df</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>-log2(p) of ll-ratio test</th>\n",
              "      <td>181.91</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/latex": "\\begin{tabular}{lrrrrrrrrrrr}\n{} & {coef} & {exp(coef)} & {se(coef)} & {coef lower 95%} & {coef upper 95%} & {exp(coef) lower 95%} & {exp(coef) upper 95%} & {cmp to} & {z} & {p} & {-log2(p)} \\\\\n{covariate} & {} & {} & {} & {} & {} & {} & {} & {} & {} & {} & {} \\\\\nun_continent_name_Africa & -0.20 & 0.82 & 0.08 & -0.36 & -0.04 & 0.70 & 0.96 & 0.00 & -2.48 & 0.01 & 6.25 \\\\\nun_continent_name_Asia & -0.06 & 0.95 & 0.07 & -0.20 & 0.09 & 0.82 & 1.09 & 0.00 & -0.77 & 0.44 & 1.19 \\\\\nun_continent_name_Europe & 0.23 & 1.26 & 0.06 & 0.11 & 0.35 & 1.12 & 1.43 & 0.00 & 3.79 & 0.00 & 12.70 \\\\\nun_continent_name_Oceania & -0.12 & 0.89 & 0.11 & -0.33 & 0.10 & 0.72 & 1.10 & 0.00 & -1.07 & 0.29 & 1.80 \\\\\ndemocracy_Non-democracy & -0.72 & 0.49 & 0.06 & -0.84 & -0.60 & 0.43 & 0.55 & 0.00 & -11.48 & 0.00 & 98.97 \\\\\n\\end{tabular}\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iYkJ1coruTxx",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 457
        },
        "outputId": "b4047e03-7054-4b69-8054-613f14822a4c"
      },
      "source": [
        "fig_coef, ax_coef = plt.subplots(figsize=(12,7))\n",
        "ax_coef.set_title('Survival Regression: Coefficients and Confident Intervals')\n",
        "cph.plot(ax=ax_coef)\n",
        "plt.savefig(\"Survival-Regression.png\", bbox_inches='tight', format='png', dvi=700)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H3hWrJk8uTx1"
      },
      "source": [
        "from google.colab import files\n",
        "files.download(\"Survival-Regression.png\")"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}