CMSC320_Final_Project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CMSC320 Final Project",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/plato2000/covid19-datascience-tutorial/blob/master/CMSC320_Final_Project.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cCu2H5gJe7zs",
        "colab_type": "text"
      },
      "source": [
        "# CMSC320 Final Project: Data Science with COVID-19 Data\n",
        "## Parth Oza, Jagan Prem | CMSC320 | Spring 2020"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ROJLM7OC0_Ac",
        "colab_type": "text"
      },
      "source": [
        "## Introduction\n",
        "In this tutorial, we will go through the entire data science pipeline in Python, from data curation/parsing/management, exploratory data analysis, hypothesis testing, and machine learning.\n",
        "\n",
        "We will use COVID-19 testing data, as of 5/17/2020, in our analysis. We will look to answer the question, *how many deaths would the United States have had as of today if we had more testing than we did now?* Answering this question is critically important to determine if countries could have done more to reduce deaths due to COVID-19. If we would have had fewer deaths with more testing, the governments of countries which had lower testing are partially to blame for their increased deaths.\n",
        "\n",
        "Data science is crucial in many similar applications to this. While some might believe it to be obvious that more testing would result in fewer deaths, doing a statistical analysis, as we will do in this tutorial, is very important to scientifically prove a correlation rather than relying on what people believe to be obvious. After all, different people may consider different things to be obvious, so proving statements with statistics is crucial.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r_LBbzNa1hD4",
        "colab_type": "text"
      },
      "source": [
        "## Initial Setup\n",
        "\n",
        "We will use some Python libraries to do our data analysis. Specifically, we will use `pandas` for data manipulation and analysis, `numpy` for math, `plotnine` for plotting (replicating R `ggplot`'s functionality), and `sklearn` for hypothesis testing and machine learning.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bZ5jbEcyfO2H",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from plotnine import *\n",
        "from scipy import stats\n",
        "from scipy.linalg import sqrtm\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BEPU9PdQ1_Dv",
        "colab_type": "text"
      },
      "source": [
        "## Data Curation, Parsing, and Management\n",
        "\n",
        "The first step to any data science project is to figure out what data we're going to use. In this case, we were able to find a well-formatted and up to date dataset simply by Googling \"covid 19 data csv\". From here, we can get our link to the CSV and feed it to [`pandas`' `read_csv` function.](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_csv.html)\n",
        "\n",
        "Our data was found from https://ourworldindata.org/coronavirus-source-data, which tells us they source their data from the European CDC (which collects data from around the world). From this site, we get the link for the complete dataset as a CSV file."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QQkA-ei_lTBp",
        "colab_type": "code",
        "outputId": "014b6d9e-a8b6-4617-a09b-4e1314a2bd74",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 609
        }
      },
      "source": [
        "df = pd.read_csv(\"https://covid.ourworldindata.org/data/owid-covid-data.csv\", parse_dates=[2])\n",
        "df = df.dropna(subset=['location', 'date', 'new_tests', 'total_tests_per_thousand', 'population_density', 'gdp_per_capita', 'total_deaths_per_million', 'male_smokers', 'female_smokers'])\n",
        "df"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>iso_code</th>\n",
              "      <th>location</th>\n",
              "      <th>date</th>\n",
              "      <th>total_cases</th>\n",
              "      <th>new_cases</th>\n",
              "      <th>total_deaths</th>\n",
              "      <th>new_deaths</th>\n",
              "      <th>total_cases_per_million</th>\n",
              "      <th>new_cases_per_million</th>\n",
              "      <th>total_deaths_per_million</th>\n",
              "      <th>new_deaths_per_million</th>\n",
              "      <th>total_tests</th>\n",
              "      <th>new_tests</th>\n",
              "      <th>total_tests_per_thousand</th>\n",
              "      <th>new_tests_per_thousand</th>\n",
              "      <th>tests_units</th>\n",
              "      <th>population</th>\n",
              "      <th>population_density</th>\n",
              "      <th>median_age</th>\n",
              "      <th>aged_65_older</th>\n",
              "      <th>aged_70_older</th>\n",
              "      <th>gdp_per_capita</th>\n",
              "      <th>extreme_poverty</th>\n",
              "      <th>cvd_death_rate</th>\n",
              "      <th>diabetes_prevalence</th>\n",
              "      <th>female_smokers</th>\n",
              "      <th>male_smokers</th>\n",
              "      <th>handwashing_facilities</th>\n",
              "      <th>hospital_beds_per_100k</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>602</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>1795</td>\n",
              "      <td>80</td>\n",
              "      <td>65</td>\n",
              "      <td>5</td>\n",
              "      <td>39.716</td>\n",
              "      <td>1.770</td>\n",
              "      <td>1.438</td>\n",
              "      <td>0.111</td>\n",
              "      <td>14850.0</td>\n",
              "      <td>1520.0</td>\n",
              "      <td>0.329</td>\n",
              "      <td>0.034</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>603</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-10</td>\n",
              "      <td>1894</td>\n",
              "      <td>99</td>\n",
              "      <td>79</td>\n",
              "      <td>14</td>\n",
              "      <td>41.907</td>\n",
              "      <td>2.190</td>\n",
              "      <td>1.748</td>\n",
              "      <td>0.310</td>\n",
              "      <td>16379.0</td>\n",
              "      <td>1529.0</td>\n",
              "      <td>0.362</td>\n",
              "      <td>0.034</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>604</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-11</td>\n",
              "      <td>1975</td>\n",
              "      <td>81</td>\n",
              "      <td>82</td>\n",
              "      <td>3</td>\n",
              "      <td>43.699</td>\n",
              "      <td>1.792</td>\n",
              "      <td>1.814</td>\n",
              "      <td>0.066</td>\n",
              "      <td>18027.0</td>\n",
              "      <td>1648.0</td>\n",
              "      <td>0.399</td>\n",
              "      <td>0.036</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>607</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-14</td>\n",
              "      <td>2272</td>\n",
              "      <td>69</td>\n",
              "      <td>98</td>\n",
              "      <td>3</td>\n",
              "      <td>50.270</td>\n",
              "      <td>1.527</td>\n",
              "      <td>2.168</td>\n",
              "      <td>0.066</td>\n",
              "      <td>22805.0</td>\n",
              "      <td>3047.0</td>\n",
              "      <td>0.505</td>\n",
              "      <td>0.067</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>608</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-15</td>\n",
              "      <td>2432</td>\n",
              "      <td>160</td>\n",
              "      <td>105</td>\n",
              "      <td>7</td>\n",
              "      <td>53.810</td>\n",
              "      <td>3.540</td>\n",
              "      <td>2.323</td>\n",
              "      <td>0.155</td>\n",
              "      <td>24374.0</td>\n",
              "      <td>1569.0</td>\n",
              "      <td>0.539</td>\n",
              "      <td>0.035</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17994</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-10</td>\n",
              "      <td>36</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.422</td>\n",
              "      <td>0.067</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>9872.0</td>\n",
              "      <td>279.0</td>\n",
              "      <td>0.664</td>\n",
              "      <td>0.019</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17995</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-11</td>\n",
              "      <td>36</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.422</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>10598.0</td>\n",
              "      <td>726.0</td>\n",
              "      <td>0.713</td>\n",
              "      <td>0.049</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17996</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-12</td>\n",
              "      <td>37</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.489</td>\n",
              "      <td>0.067</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>11058.0</td>\n",
              "      <td>460.0</td>\n",
              "      <td>0.744</td>\n",
              "      <td>0.031</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17997</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-13</td>\n",
              "      <td>37</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.489</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>11191.0</td>\n",
              "      <td>133.0</td>\n",
              "      <td>0.753</td>\n",
              "      <td>0.009</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18000</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-16</td>\n",
              "      <td>42</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.826</td>\n",
              "      <td>0.336</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>12275.0</td>\n",
              "      <td>874.0</td>\n",
              "      <td>0.826</td>\n",
              "      <td>0.059</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3852 rows × 29 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      iso_code   location  ... handwashing_facilities  hospital_beds_per_100k\n",
              "602        ARG  Argentina  ...                    NaN                     5.0\n",
              "603        ARG  Argentina  ...                    NaN                     5.0\n",
              "604        ARG  Argentina  ...                    NaN                     5.0\n",
              "607        ARG  Argentina  ...                    NaN                     5.0\n",
              "608        ARG  Argentina  ...                    NaN                     5.0\n",
              "...        ...        ...  ...                    ...                     ...\n",
              "17994      ZWE   Zimbabwe  ...                 36.791                     1.7\n",
              "17995      ZWE   Zimbabwe  ...                 36.791                     1.7\n",
              "17996      ZWE   Zimbabwe  ...                 36.791                     1.7\n",
              "17997      ZWE   Zimbabwe  ...                 36.791                     1.7\n",
              "18000      ZWE   Zimbabwe  ...                 36.791                     1.7\n",
              "\n",
              "[3852 rows x 29 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 143
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H3RbKaoM23qk",
        "colab_type": "text"
      },
      "source": [
        "The `read_csv` function is quite useful to be able to automatically convert a CSV file into a a `DataFrame`. Looking at the documentation linked above, we can see `read_csv` has many options it can take to make reading poorly formatted CSV files more easy. Fortunately, in this case, the data is already tidied for us and is easy to parse.\n",
        "\n",
        "If the data had been in an [HTML table](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_html.html) or [Excel document](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_excel.html) or [JSON file](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_json.html), `pandas` has functions to make reading those formats simple as well.\n",
        "\n",
        "Looking at the `read_csv` function documentation, we can see that it has an optional parameter `parse_dates`. This allows us to automatically convert the date column in our dataset into a datetime object rather than a string. \n",
        "\n",
        "We know that we're planning on analyzing data from this dataset regarding testing and death rate. We also know that not all countries report their death rates and testing, but report some other metrics. As a result, our dataframe has some `NaN` values (meaning no data) for some countries' data in those columns. To make sure that this doesn't impact our data, we can run `df.dropna` on  the columns we care about, which will remove the rows that have a `NaN` value in those columns.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ent9CAWxsud0",
        "colab_type": "code",
        "outputId": "e0220208-c7c8-40fc-d6f5-b317ba5ece38",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        }
      },
      "source": [
        "df.location.unique()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Argentina', 'Australia', 'Austria', 'Belgium', 'Bangladesh',\n",
              "       'Bulgaria', 'Bahrain', 'Belarus', 'Canada', 'Switzerland', 'Chile',\n",
              "       'Colombia', 'Costa Rica', 'Czech Republic', 'Denmark', 'Estonia',\n",
              "       'Ethiopia', 'Finland', 'France', 'United Kingdom', 'Ghana',\n",
              "       'Greece', 'Croatia', 'Hungary', 'Indonesia', 'India', 'Iran',\n",
              "       'Iceland', 'Israel', 'Italy', 'Kazakhstan', 'Kenya', 'South Korea',\n",
              "       'Lithuania', 'Luxembourg', 'Latvia', 'Morocco', 'Mexico',\n",
              "       'Myanmar', 'Malaysia', 'Nigeria', 'Norway', 'Nepal', 'New Zealand',\n",
              "       'Pakistan', 'Panama', 'Philippines', 'Poland', 'Portugal',\n",
              "       'Paraguay', 'Qatar', 'Romania', 'Russia', 'Rwanda', 'Saudi Arabia',\n",
              "       'Senegal', 'El Salvador', 'Serbia', 'Slovakia', 'Slovenia',\n",
              "       'Tunisia', 'Turkey', 'Uganda', 'Ukraine', 'Uruguay',\n",
              "       'United States', 'Vietnam', 'South Africa', 'Zimbabwe'],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 144
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8Wqm5VeLxtA7",
        "colab_type": "code",
        "outputId": "0c4e8997-a8cd-4ee7-9280-5b0524bcb97c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(df.location.unique())"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "69"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 145
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mhzi_K0Z5Ytb",
        "colab_type": "text"
      },
      "source": [
        "We can see that after dropping the data that would leave our dataset incomplete, we have data from 71 countries remaining.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iKLh2Q0-6vJa",
        "colab_type": "text"
      },
      "source": [
        "## Exploratory Data Analysis\n",
        "\n",
        "First, we will look at testing over time to make sure that there is enough difference in testing levels to actually be able to do our final analysis. To do this, we'll plot total tests per thousand against date.\n",
        "\n",
        "We choose to plot tests per thousand instead of raw test counts because otherwise, large countries with high populations would show as having tested lots of people even if only small percentages of the population were tested.\n",
        "\n",
        "Since 71 countries is a lot of data to show on one chart, we picked a few countries that (based on prior knowledge) represent different geographic areas, average wealths, and government responses, in order to get a somewhat representative group of countries.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XfwZHUX5xUw4",
        "colab_type": "code",
        "outputId": "e9aeda02-810e-48e9-87ba-4100c175db85",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 566
        }
      },
      "source": [
        "countries_to_analyze = [\"United States\", \"South Korea\", \"Norway\", \"India\", \"Argentina\", \"Denmark\", \"Luxembourg\", \"Rwanda\", \"New Zealand\", \"Australia\"]\n",
        "\n",
        "(ggplot(df.loc[df[\"location\"].isin(countries_to_analyze)])\n",
        "    + aes(x=\"date\", y=\"total_tests_per_thousand\", color=\"location\")\n",
        "    # + geom_point()\n",
        "    + geom_smooth(span=0.1)\n",
        "    + xlab(\"Date\")\n",
        "    + ylab(\"Total Tests per Thousand\")\n",
        "    + ggtitle(\"Total testing over time for a selection of countries\")\n",
        "    + theme(axis_text_x=element_text(angle=90, vjust=1))\n",
        ").draw();"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/plotnine/stats/smoothers.py:168: PlotnineWarning: Confidence intervals are not yet implementedfor lowess smoothings.\n",
            "  \"for lowess smoothings.\", PlotnineWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "guljFAxVGVrc",
        "colab_type": "text"
      },
      "source": [
        "We've used `ggplot` to make our graph, which is included in the `plotnine` Python library. `ggplot` is originally a plotting tool in R, and `plotnine` was created to help R programmers easily transition to Python. As a result, almost everything in `ggplot` for Python is nearly identical to the R version. [The plotnine documentation ](https://plotnine.readthedocs.io/en/stable/) has plenty of information on how to use the basics of `ggplot`, but if you want more advanced examples, you can look at the[ R documentation](https://ggplot2.tidyverse.org/reference/index.html). There are some differences in syntax, but it's easy to adjust to these by pattern matching against the `plotnine` documentation (for example, the R library uses dot notation for some functions which in Python use underscores, such as element.text vs element_text).\n",
        "\n",
        "`ggplot` plots are initialized with some data. In this case, we pass in our dataframe as the data to be plotted. Specifically, we pass in the subset of rows whose \"location\" column is in our list of countries to plot.\n",
        "\n",
        "Then, we can add aesthetic mappings, layers, and more to our plot. The base aesthetic mapping is for the plot, but there can be additional aesthetic mappings for each layer we add. The base aesthetic mapping can tell the plot what to plot on each axis and how to determine the color of what is plotted. Then, we add in `geom_point` or `geom_smooth` or both, which actually place points or a curve on our plot. `geom_smooth` is able to take arguments for how it determines the smoothing and how much it smooths (this is what `span` is). In this case, we didn't plot each point because the chart would look significantly more noisy with points, and the lines convey the same information.\n",
        "\n",
        "Then, we add in guides, which allow us to label the axes and the plot. Finally, we add a theme, which allows us to style the non-data elements of the plot. In particular, we are able to rotate the date labels 90 degrees to make them not overlap with each other.\n",
        "\n",
        "### Plot Analysis\n",
        "\n",
        "This plot shows us that different countries ramped up their testing at different times, and currently have wildly different levels of tests per capita performed. This is enough information for us to move on to our hypothesis testing, as there is clearly enough difference between countries here."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R07F2McKV30q",
        "colab_type": "text"
      },
      "source": [
        "### More Exploratory Data Analysis\n",
        "\n",
        "Even though we could theoretically move on to our hypothesis testing now, we can do some more analysis on our data to learn more about COVID-19 deaths and how they relate to testing.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5N0limZPG8_b",
        "colab_type": "code",
        "outputId": "b9723516-6e5a-42ef-c447-fbb83d348e78",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 502
        }
      },
      "source": [
        "agg_data = df.groupby(\"location\").agg({\"total_tests_per_thousand\": \"max\", \"total_deaths_per_million\": \"max\"})\n",
        "(ggplot(agg_data)\n",
        "    + aes(x=\"total_deaths_per_million\", y=\"total_tests_per_thousand\")\n",
        "    + geom_point(alpha=0.2)\n",
        "    + ggtitle(\"Deaths per capita vs. tests per capita for all countries in dataset\")\n",
        "    + theme(axis_text_x=element_text(angle=90, vjust=1))\n",
        "    + scale_color_discrete(guide=False))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<ggplot: (8763678924741)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 147
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UW96oiqMW8Cp",
        "colab_type": "text"
      },
      "source": [
        "We want to show how different countries' testing corresponds to their deaths. We don't care about time in this plot, and we only want one point per country (we want to show the latest total data for each country on our plot). We can do this by grouping and aggregating our data. Using `pandas`, there are two functions we use: `groupby` and `agg`. These functions are used in conjunction with each other to produce a condensed dataframe. Here, we group rows into countries, so  that each country is now one row, and aggregate our total tests per thousand and total deaths per million for each of those rows. We select the max in total tests and total deaths for each country, since we're looking for the most recent data, and total tests and total deaths cannot go down over time. \n",
        "\n",
        "Then, we simply plot our data the same way we did before. Here, we use an alpha value on each of our points since there is a lot of clustering. Making each point translucent allows us to see where there are large clumps of points.\n",
        "\n",
        "Our plot shows that there are a lot of countries in our dataset which have both low testing and low deaths reported. This intuitively makes sense, because either they truly have low deaths due to limited exposure to COVID-19, and therefore don't need tests, or do have cases of COVID-19 but can't attribute deaths to COVID-19 due to lack of testing to determine if people died of it.\n",
        "\n",
        "Outside of this clump, though, the trend is still interesting. It appears that there is something of an inverse relationship between tests and deaths, with some countries having very high testing and very low deaths, and other countries having little testing but more deaths. It appears that there is a floor to testing -- if there are too few tests, the death data may be unreliable, but once there is some threshold of tests (in this plot, it appears to be around 25 tests per thousand), we can see that more tests are correlated with less deaths.\n",
        "\n",
        "Knowing this, we can use more data in our hypothesis testing to account for countries in that first clump. For example, we can add data about GDP per capita or population density and see if adding those variables allows us to get a better fitted regression that doesn't require us to simply ignore a large clump of countries.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fVrQHoD7Z3tO",
        "colab_type": "text"
      },
      "source": [
        "## Hypothesis Testing\n",
        "\n",
        "The first step in our hypothesis testing is to determine our hypotheses (the null and alternative hypotheses):\n",
        "\n",
        "$H_0$ is that given a regression model correlating total tests per capita, population density, GDP per capita, male and female smoking rate, and time since first case with death rate, all of the individual regression coefficients will be zero.\n",
        "\n",
        "$H_1$ is that given a regression model correlating total tests per capita, population density, GDP per capita, male and female smoking rate, and time since first case with death rate, at least one of the individual regression coefficients will be significantly different from zero.\n",
        "\n",
        "From here, we know what data we need to test our hypothesis. The only data that we need that isn't in our raw data table is the time since first case for each country. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mMfIZNlOoJni",
        "colab_type": "code",
        "outputId": "c2fa2128-cc74-4c44-f040-6daf56206929",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 450
        }
      },
      "source": [
        "agg_data = df.groupby(\"location\").agg({\"total_deaths_per_million\": \"max\", \"total_cases_per_million\": \"max\"})\n",
        "agg_data[\"first_case_date\"] = pd.to_datetime(\"today\")\n",
        "\n",
        "for _, row in df.iterrows():\n",
        "    if row[\"total_cases\"] > 0:\n",
        "        if agg_data.loc[row[\"location\"]][\"first_case_date\"] > row[\"date\"]:\n",
        "            agg_data.loc[row[\"location\"], [\"first_case_date\"]] = row[\"date\"]\n",
        "agg_data"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>total_deaths_per_million</th>\n",
              "      <th>total_cases_per_million</th>\n",
              "      <th>first_case_date</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>location</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Argentina</th>\n",
              "      <td>8.098</td>\n",
              "      <td>172.405</td>\n",
              "      <td>2020-04-09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Australia</th>\n",
              "      <td>3.843</td>\n",
              "      <td>276.276</td>\n",
              "      <td>2020-04-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Austria</th>\n",
              "      <td>69.839</td>\n",
              "      <td>1793.613</td>\n",
              "      <td>2020-02-26</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Bahrain</th>\n",
              "      <td>7.052</td>\n",
              "      <td>4087.958</td>\n",
              "      <td>2020-03-10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Bangladesh</th>\n",
              "      <td>1.992</td>\n",
              "      <td>135.212</td>\n",
              "      <td>2020-03-09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>United Kingdom</th>\n",
              "      <td>510.208</td>\n",
              "      <td>3589.768</td>\n",
              "      <td>2020-01-31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>United States</th>\n",
              "      <td>268.137</td>\n",
              "      <td>4434.659</td>\n",
              "      <td>2020-03-08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Uruguay</th>\n",
              "      <td>5.758</td>\n",
              "      <td>211.300</td>\n",
              "      <td>2020-03-29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Vietnam</th>\n",
              "      <td>0.000</td>\n",
              "      <td>2.774</td>\n",
              "      <td>2020-03-12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Zimbabwe</th>\n",
              "      <td>0.269</td>\n",
              "      <td>2.826</td>\n",
              "      <td>2020-05-07</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>69 rows × 3 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                total_deaths_per_million  ...  first_case_date\n",
              "location                                  ...                 \n",
              "Argentina                          8.098  ...       2020-04-09\n",
              "Australia                          3.843  ...       2020-04-06\n",
              "Austria                           69.839  ...       2020-02-26\n",
              "Bahrain                            7.052  ...       2020-03-10\n",
              "Bangladesh                         1.992  ...       2020-03-09\n",
              "...                                  ...  ...              ...\n",
              "United Kingdom                   510.208  ...       2020-01-31\n",
              "United States                    268.137  ...       2020-03-08\n",
              "Uruguay                            5.758  ...       2020-03-29\n",
              "Vietnam                            0.000  ...       2020-03-12\n",
              "Zimbabwe                           0.269  ...       2020-05-07\n",
              "\n",
              "[69 rows x 3 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 148
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ho2BncUFIUnb",
        "colab_type": "text"
      },
      "source": [
        "To get the data for time since first case, we need to find out when the first case date was for each country. We can then later subtract this date from the date of each data point.\n",
        "\n",
        "We do our aggregation by country the same way we did it before, then simply iterate through each row in our original dataframe to find the minimum date with more than 0 cases for each country.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6BVdGlGAK3Qq",
        "colab_type": "code",
        "outputId": "2f27d03a-b356-4cfb-a5f5-144b423bd068",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 609
        }
      },
      "source": [
        "df = df.join(agg_data[[\"first_case_date\"]], on=\"location\")\n",
        "df[\"since_case_date\"] = (df[\"date\"] - df[\"first_case_date\"]).astype(int)\n",
        "df"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>iso_code</th>\n",
              "      <th>location</th>\n",
              "      <th>date</th>\n",
              "      <th>total_cases</th>\n",
              "      <th>new_cases</th>\n",
              "      <th>total_deaths</th>\n",
              "      <th>new_deaths</th>\n",
              "      <th>total_cases_per_million</th>\n",
              "      <th>new_cases_per_million</th>\n",
              "      <th>total_deaths_per_million</th>\n",
              "      <th>new_deaths_per_million</th>\n",
              "      <th>total_tests</th>\n",
              "      <th>new_tests</th>\n",
              "      <th>total_tests_per_thousand</th>\n",
              "      <th>new_tests_per_thousand</th>\n",
              "      <th>tests_units</th>\n",
              "      <th>population</th>\n",
              "      <th>population_density</th>\n",
              "      <th>median_age</th>\n",
              "      <th>aged_65_older</th>\n",
              "      <th>aged_70_older</th>\n",
              "      <th>gdp_per_capita</th>\n",
              "      <th>extreme_poverty</th>\n",
              "      <th>cvd_death_rate</th>\n",
              "      <th>diabetes_prevalence</th>\n",
              "      <th>female_smokers</th>\n",
              "      <th>male_smokers</th>\n",
              "      <th>handwashing_facilities</th>\n",
              "      <th>hospital_beds_per_100k</th>\n",
              "      <th>first_case_date</th>\n",
              "      <th>since_case_date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>602</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>1795</td>\n",
              "      <td>80</td>\n",
              "      <td>65</td>\n",
              "      <td>5</td>\n",
              "      <td>39.716</td>\n",
              "      <td>1.770</td>\n",
              "      <td>1.438</td>\n",
              "      <td>0.111</td>\n",
              "      <td>14850.0</td>\n",
              "      <td>1520.0</td>\n",
              "      <td>0.329</td>\n",
              "      <td>0.034</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>603</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-10</td>\n",
              "      <td>1894</td>\n",
              "      <td>99</td>\n",
              "      <td>79</td>\n",
              "      <td>14</td>\n",
              "      <td>41.907</td>\n",
              "      <td>2.190</td>\n",
              "      <td>1.748</td>\n",
              "      <td>0.310</td>\n",
              "      <td>16379.0</td>\n",
              "      <td>1529.0</td>\n",
              "      <td>0.362</td>\n",
              "      <td>0.034</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>86400000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>604</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-11</td>\n",
              "      <td>1975</td>\n",
              "      <td>81</td>\n",
              "      <td>82</td>\n",
              "      <td>3</td>\n",
              "      <td>43.699</td>\n",
              "      <td>1.792</td>\n",
              "      <td>1.814</td>\n",
              "      <td>0.066</td>\n",
              "      <td>18027.0</td>\n",
              "      <td>1648.0</td>\n",
              "      <td>0.399</td>\n",
              "      <td>0.036</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>172800000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>607</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-14</td>\n",
              "      <td>2272</td>\n",
              "      <td>69</td>\n",
              "      <td>98</td>\n",
              "      <td>3</td>\n",
              "      <td>50.270</td>\n",
              "      <td>1.527</td>\n",
              "      <td>2.168</td>\n",
              "      <td>0.066</td>\n",
              "      <td>22805.0</td>\n",
              "      <td>3047.0</td>\n",
              "      <td>0.505</td>\n",
              "      <td>0.067</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>432000000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>608</th>\n",
              "      <td>ARG</td>\n",
              "      <td>Argentina</td>\n",
              "      <td>2020-04-15</td>\n",
              "      <td>2432</td>\n",
              "      <td>160</td>\n",
              "      <td>105</td>\n",
              "      <td>7</td>\n",
              "      <td>53.810</td>\n",
              "      <td>3.540</td>\n",
              "      <td>2.323</td>\n",
              "      <td>0.155</td>\n",
              "      <td>24374.0</td>\n",
              "      <td>1569.0</td>\n",
              "      <td>0.539</td>\n",
              "      <td>0.035</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>45195777.0</td>\n",
              "      <td>16.177</td>\n",
              "      <td>31.9</td>\n",
              "      <td>11.198</td>\n",
              "      <td>7.441</td>\n",
              "      <td>18933.907</td>\n",
              "      <td>0.6</td>\n",
              "      <td>191.032</td>\n",
              "      <td>5.50</td>\n",
              "      <td>16.2</td>\n",
              "      <td>27.7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020-04-09</td>\n",
              "      <td>518400000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17994</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-10</td>\n",
              "      <td>36</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.422</td>\n",
              "      <td>0.067</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>9872.0</td>\n",
              "      <td>279.0</td>\n",
              "      <td>0.664</td>\n",
              "      <td>0.019</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "      <td>2020-05-07</td>\n",
              "      <td>259200000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17995</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-11</td>\n",
              "      <td>36</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.422</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>10598.0</td>\n",
              "      <td>726.0</td>\n",
              "      <td>0.713</td>\n",
              "      <td>0.049</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "      <td>2020-05-07</td>\n",
              "      <td>345600000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17996</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-12</td>\n",
              "      <td>37</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.489</td>\n",
              "      <td>0.067</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>11058.0</td>\n",
              "      <td>460.0</td>\n",
              "      <td>0.744</td>\n",
              "      <td>0.031</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "      <td>2020-05-07</td>\n",
              "      <td>432000000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17997</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-13</td>\n",
              "      <td>37</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.489</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>11191.0</td>\n",
              "      <td>133.0</td>\n",
              "      <td>0.753</td>\n",
              "      <td>0.009</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "      <td>2020-05-07</td>\n",
              "      <td>518400000000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18000</th>\n",
              "      <td>ZWE</td>\n",
              "      <td>Zimbabwe</td>\n",
              "      <td>2020-05-16</td>\n",
              "      <td>42</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>2.826</td>\n",
              "      <td>0.336</td>\n",
              "      <td>0.269</td>\n",
              "      <td>0.000</td>\n",
              "      <td>12275.0</td>\n",
              "      <td>874.0</td>\n",
              "      <td>0.826</td>\n",
              "      <td>0.059</td>\n",
              "      <td>tests performed</td>\n",
              "      <td>14862927.0</td>\n",
              "      <td>42.729</td>\n",
              "      <td>19.6</td>\n",
              "      <td>2.822</td>\n",
              "      <td>1.882</td>\n",
              "      <td>1899.775</td>\n",
              "      <td>21.4</td>\n",
              "      <td>307.846</td>\n",
              "      <td>1.82</td>\n",
              "      <td>1.6</td>\n",
              "      <td>30.7</td>\n",
              "      <td>36.791</td>\n",
              "      <td>1.7</td>\n",
              "      <td>2020-05-07</td>\n",
              "      <td>777600000000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3852 rows × 31 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      iso_code   location  ... first_case_date  since_case_date\n",
              "602        ARG  Argentina  ...      2020-04-09                0\n",
              "603        ARG  Argentina  ...      2020-04-09   86400000000000\n",
              "604        ARG  Argentina  ...      2020-04-09  172800000000000\n",
              "607        ARG  Argentina  ...      2020-04-09  432000000000000\n",
              "608        ARG  Argentina  ...      2020-04-09  518400000000000\n",
              "...        ...        ...  ...             ...              ...\n",
              "17994      ZWE   Zimbabwe  ...      2020-05-07  259200000000000\n",
              "17995      ZWE   Zimbabwe  ...      2020-05-07  345600000000000\n",
              "17996      ZWE   Zimbabwe  ...      2020-05-07  432000000000000\n",
              "17997      ZWE   Zimbabwe  ...      2020-05-07  518400000000000\n",
              "18000      ZWE   Zimbabwe  ...      2020-05-07  777600000000000\n",
              "\n",
              "[3852 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 149
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uqT6FBjVLSES",
        "colab_type": "text"
      },
      "source": [
        "To get our dates for each country into our original dataframe, we use `join`. The default `join` in `pandas` is a left join, which keeps everything in the \"left\" table and adds in things from the \"right\" table. In this case, the \"left\" table is `df` and the right table is `agg_data`. This will put the first case date into each row based on its country, since we're joining on location.\n",
        "\n",
        "Then, we can subtract the the first case date from the date column to get a measure of how much time there has been since the first case date for each data point.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WogGFtnxW7Uy",
        "colab_type": "code",
        "outputId": "be6bc16e-fed9-488c-ce86-825fd87d6660",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "X = df[['total_tests_per_thousand', 'population_density', 'gdp_per_capita', 'since_case_date', 'male_smokers', 'female_smokers']]\n",
        "y = df['total_deaths_per_million']\n",
        "poly = PolynomialFeatures(degree=4)\n",
        "poly_vars = poly.fit_transform(X)\n",
        "poly_var_train, poly_var_test, y_train, y_test = train_test_split(poly_vars, y, test_size = 0.3, random_state=4)\n",
        "\n",
        "lm = LinearRegression(normalize=True, fit_intercept=True).fit(poly_var_train, y_train)\n",
        "lm.score(poly_var_test, y_test)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9893102516684293"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 150
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Dr1O2WlJXObu",
        "colab_type": "text"
      },
      "source": [
        "We then fit our data, containing all the relevant columns, with a polynomial model using the `LinearRegression` class from `sklearn` in conjuction with `PolynomialFeatures`. This allows us to determine a polynomial curve that fits the relationship between the input and output columns. Additionally, we perform train-test split so that we can verify our results against data unused in our training operation.\n",
        "\n",
        "The output produced above corresponds to the $R^2$ value used to score our linear regression model. This indicates that 95.57% of the variance in the total deaths per million may be explained by our selected variables, which is quite a strong relationship."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D7q6SNxSQDyT",
        "colab_type": "code",
        "outputId": "891bae98-083d-4ed9-8977-0de8554e8971",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 453
        }
      },
      "source": [
        "params = np.append(lm.intercept_, lm.coef_)\n",
        "predictions = lm.predict(poly_var_test)\n",
        "\n",
        "newX = pd.DataFrame({\"Constant\":np.ones(len(poly_var_test))}).join(pd.DataFrame(poly_var_test))\n",
        "MSE = (sum((y_test-predictions)**2))/(len(newX)-len(newX.columns))\n",
        "\n",
        "var_b = MSE*(np.linalg.inv(np.dot(newX.T,newX)).diagonal())\n",
        "sd_b = np.sqrt(var_b)\n",
        "ts_b = params/ sd_b\n",
        "\n",
        "p_values =[2*(1-stats.t.cdf(np.abs(i),(len(newX)-1))) for i in ts_b]\n",
        "\n",
        "sd_b = np.round(sd_b,3)\n",
        "ts_b = np.round(ts_b,3)\n",
        "p_values = np.round(p_values,3)\n",
        "params = np.round(params,4)\n",
        "\n",
        "myDF3 = pd.DataFrame()\n",
        "myDF3[\"Coefficients\"], myDF3[\"Standard Errors\"], myDF3[\"t statistics\"], myDF3[\"p-values\"] = [params,sd_b,ts_b,p_values]\n",
        "myDF3"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:8: RuntimeWarning: invalid value encountered in sqrt\n",
            "  \n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Coefficients</th>\n",
              "      <th>Standard Errors</th>\n",
              "      <th>t statistics</th>\n",
              "      <th>p-values</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>3.587018e+10</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-3.000000e-04</td>\n",
              "      <td>1.854319e+10</td>\n",
              "      <td>-0.000000e+00</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.082425e+02</td>\n",
              "      <td>2.995400e+01</td>\n",
              "      <td>3.614000e+00</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-3.350518e+08</td>\n",
              "      <td>6.763000e+00</td>\n",
              "      <td>-4.954035e+07</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-6.812252e+06</td>\n",
              "      <td>4.500000e-02</td>\n",
              "      <td>-1.528722e+08</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206</th>\n",
              "      <td>2.929992e+04</td>\n",
              "      <td>1.000000e-03</td>\n",
              "      <td>5.183677e+07</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>207</th>\n",
              "      <td>-7.347636e+04</td>\n",
              "      <td>3.000000e-03</td>\n",
              "      <td>-2.807108e+07</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>208</th>\n",
              "      <td>3.478106e+05</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>209</th>\n",
              "      <td>-2.449073e+05</td>\n",
              "      <td>4.000000e-03</td>\n",
              "      <td>-6.974031e+07</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>210</th>\n",
              "      <td>-1.395557e+05</td>\n",
              "      <td>2.000000e-03</td>\n",
              "      <td>-8.345026e+07</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>211 rows × 4 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Coefficients  Standard Errors  t statistics  p-values\n",
              "0    3.587018e+10              NaN           NaN       NaN\n",
              "1   -3.000000e-04     1.854319e+10 -0.000000e+00       1.0\n",
              "2    1.082425e+02     2.995400e+01  3.614000e+00       0.0\n",
              "3   -3.350518e+08     6.763000e+00 -4.954035e+07       0.0\n",
              "4   -6.812252e+06     4.500000e-02 -1.528722e+08       0.0\n",
              "..            ...              ...           ...       ...\n",
              "206  2.929992e+04     1.000000e-03  5.183677e+07       0.0\n",
              "207 -7.347636e+04     3.000000e-03 -2.807108e+07       0.0\n",
              "208  3.478106e+05              NaN           NaN       NaN\n",
              "209 -2.449073e+05     4.000000e-03 -6.974031e+07       0.0\n",
              "210 -1.395557e+05     2.000000e-03 -8.345026e+07       0.0\n",
              "\n",
              "[211 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 151
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N79EqbboY0PZ",
        "colab_type": "text"
      },
      "source": [
        "As may be observed above, there are a number of p-values significantly below 0.05, indicating that there are coefficients significantly different than 0. Thus, we may reject $H_0$ in favor of $H_1$.\n",
        "\n",
        "Let us now use our model to make a prediction about how the United States would be faring in terms of deaths per million if we had a 50% increase in testing capacity:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HWeNNULdkZ00",
        "colab_type": "code",
        "outputId": "d94a0247-1dd8-4abf-c546-de7fc932e1bd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "row = df.loc[df['location'] == 'United States'].iloc[-1]\n",
        "x = np.hstack((1.5 * df.loc[df['location'] == 'United States'].iloc[-1]['total_tests_per_thousand'], row[['population_density', 'gdp_per_capita', 'since_case_date', 'male_smokers', 'female_smokers']]))[None]\n",
        "deaths = lm.predict(poly.fit_transform(x))\n",
        "deaths[0]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "117.91931915283203"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 152
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZEKL8WkPtbTR",
        "colab_type": "code",
        "outputId": "673cc962-b167-45a5-8536-846272478513",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "row[\"total_deaths_per_million\"]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "268.137"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 153
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I5yVaLIYtWGD",
        "colab_type": "text"
      },
      "source": [
        "We can see that this is less than half the number of deaths  that we currently have. We can also look at how many deaths we would have if we had half the testing that we currently have:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pXwrC9uRtVqJ",
        "colab_type": "code",
        "outputId": "52f2c0bf-522c-4ee7-99aa-a55f6eca73e5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "x = np.hstack((0.5 * df.loc[df['location'] == 'United States'].iloc[-1]['total_tests_per_thousand'], row[['population_density', 'gdp_per_capita', 'since_case_date', 'male_smokers', 'female_smokers']]))[None]\n",
        "deaths = lm.predict(poly.fit_transform(x))\n",
        "deaths[0]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "344.7753372192383"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 154
        }
      ]
    }
  ]
}