통계학_기말과제_2020131017_장윤태.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": "3.8.5"
    },
    "colab": {
      "name": "통계학 기말과제_2020131017_장윤태.ipynb",
      "provenance": [],
      "collapsed_sections": []
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2P9yIMlxOkCH"
      },
      "source": [
        "# 통계학 기말프로젝트\n",
        "### 2020131017 장윤태\n",
        "\n",
        "프로젝트 주제 : 국가별 GDP와 코로나 확진자 및 사망자 간의 연관성\n",
        "\n",
        "프로젝트 동기 : 코로나의 장기화 및 백신 보급이 이뤄지면서 국가 별 인프라 혹은 경제 수준이 국가 별 코로나 사망자 수에 영향을 미친다는 정보를 들었다. 이 정보 뿐만 아니라 그 외의 정보들을 통계학에서 배운 내용을 활용해 검증해보고자 한다.\n",
        "\n",
        "프로젝트 과정 : 국가별 GDP, 인구정보 및 국가별 코로나 확진자, 사망자, 백신 접종자 수에 대한 정보를 크롤링한 후 각각의 분포를 파악한다. 그 후 서로 간의 상관관계를 분석하여 연관성있는 항목을 찾는다."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#코랩에서 구글 드라이브에 접속할 수 있는 권한 마운트\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ED_FXBVQbL79",
        "outputId": "120e1b94-4e9f-49cc-aa53-5c1f0256604c"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hnqOK5hq_15p"
      },
      "source": [
        "# 1. 사용할 패키지 import"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jApMyTKZ_15w"
      },
      "source": [
        "#웹사이트로부터 html문서 요청 패키지 불러오기\n",
        "import requests\n",
        "\n",
        "#정규식 지정 패키지 불러오기\n",
        "import re\n",
        "\n",
        "#로컬 파일 접속 패키지 불러오기\n",
        "import os\n",
        "\n",
        "#html문서에서 추출 함수 불러오기\n",
        "from bs4 import BeautifulSoup\n",
        "\n",
        "#데이터를 처리할 프레임 패키지 불러오기\n",
        "import pandas as pd\n",
        "\n",
        "#수학 연산을 처리할 패키지 불러오기\n",
        "import numpy as np\n",
        "\n",
        "#그래프와 같은 그림 그리는 패키지 불러오기\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "#그래프와 같은 그림 그리는 패키지 불러오기\n",
        "import seaborn as sns \n",
        "\n",
        "#그림그리는 패키지에서 한글 폰트 사용 함수 불러오기\n",
        "import matplotlib.font_manager as fm\n",
        "\n",
        "#사이킷런의 통계 함수 패키지 불러오기\n",
        "from scipy import stats"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "snm-oS45_150"
      },
      "source": [
        "# 2. 웹사이트 html문서 불러오기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dtvMM3upVIGS"
      },
      "source": [
        "### 국가별 GDP 현황 html문서 불러오기"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BUC6ruv-VJ8L",
        "outputId": "2f6d460b-0ff3-4786-89cb-a2414a3ea6f0"
      },
      "source": [
        "#html문서를 담을 리스트 형성\n",
        "htmls = []\n",
        "\n",
        "#html문서 요청을 위한 사용자 header 지정\n",
        "headers = {'User-Agent' : 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.101 Safari/537.36'}\n",
        "\n",
        "#html문서를 가져올 url지정\n",
        "url = f'https://ko.wikipedia.org/wiki/%EB%AA%85%EB%AA%A9_%EA%B5%AD%EB%82%B4_%EC%B4%9D%EC%83%9D%EC%82%B0%EC%88%9C_%EB%82%98%EB%9D%BC_%EB%AA%A9%EB%A1%9D'\n",
        "\n",
        "#사용자 headers를 기반으로 국가별 gdp의 정보가 나와있는 html문서 불러오기\n",
        "response = requests.get(url, headers=headers)\n",
        "\n",
        "#불러온 html문서를 bs4의 html.parser방식으로 변환\n",
        "soup = BeautifulSoup(response.text, 'html.parser')\n",
        "\n",
        "#변환한 html문서를 htmls 리스트에 추가\n",
        "htmls.append(soup)\n",
        "\n",
        "#제대로 불러왔는지 title 표시\n",
        "print(soup.title)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<title>명목 국내 총생산순 나라 목록 - 위키백과, 우리 모두의 백과사전</title>\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CL_5zNGjXiOT"
      },
      "source": [
        "### 국가별 인구 현황 html문서 불러오기"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sH1dLkm6XjD6",
        "outputId": "dca5be59-f9e3-4835-e805-9f48ecb7b837"
      },
      "source": [
        "#html문서를 가져올 url지정\n",
        "url = f'https://ko.wikipedia.org/wiki/%EC%9D%B8%EA%B5%AC%EC%88%9C_%EB%82%98%EB%9D%BC_%EB%AA%A9%EB%A1%9D'\n",
        "\n",
        "#사용자 headers를 기반으로 국가별 gdp의 정보가 나와있는 html문서 불러오기\n",
        "response = requests.get(url, headers=headers)\n",
        "\n",
        "#불러온 html문서를 bs4의 html.parser방식으로 변환\n",
        "soup = BeautifulSoup(response.text, 'html.parser')\n",
        "\n",
        "#변환한 html문서를 htmls 리스트에 추가\n",
        "htmls.append(soup)\n",
        "\n",
        "#제대로 불러왔는지 title 표시\n",
        "print(soup.title)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<title>인구순 나라 목록 - 위키백과, 우리 모두의 백과사전</title>\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2WypVCPLWZQK"
      },
      "source": [
        "### 국가별 코로나 현황 html문서 불러오기"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1-mUbO7yWYf0",
        "outputId": "d725fed0-9546-45e2-82cf-08ac6a105f90"
      },
      "source": [
        "#html문서를 가져올 url지정\n",
        "url = f'https://www.worldometers.info/coronavirus/?'\n",
        "\n",
        "#사용자 headers를 기반으로 국가별 코로나 현황 정보가 나와있는 html문서 불러오기\n",
        "response = requests.get(url, headers=headers)\n",
        "\n",
        "#불러온 html문서를 bs4의 html.parser방식으로 변환\n",
        "soup = BeautifulSoup(response.text, 'html.parser')\n",
        "\n",
        "#변환한 html문서를 htmls 리스트에 추가\n",
        "htmls.append(soup)\n",
        "\n",
        "#제대로 불러왔는지 title 표시\n",
        "print(soup.title)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<title>COVID Live Update: 263,796,741 Cases and 5,242,874 Deaths from the Coronavirus - Worldometer</title>\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O6cIttLgXGZb"
      },
      "source": [
        "### 국가별 코로나 백신 접종 현황 html문서 불러오기"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "caMAQUTlXGAk",
        "outputId": "239adcc5-22d6-4744-836d-6ec9e590ce90"
      },
      "source": [
        "#html문서를 가져올 url지정\n",
        "url = f'https://www.google.com/search?q=%EC%A0%84%EC%84%B8%EA%B3%84+%EC%BD%94%EB%A1%9C%EB%82%9819+%EB%B0%B1%EC%8B%A0+%ED%86%B5%EA%B3%84&sxsrf=AOaemvImPB-a_UbZLJXdJBKnuALjzi-Xvg%3A1638428749846&ei=TXCoYeKLM-esmAXlybToBA&ved=0ahUKEwii5v7cxsT0AhVnFqYKHeUkDU0Q4dUDCA4&uact=5&oq=%EC%A0%84%EC%84%B8%EA%B3%84+%EC%BD%94%EB%A1%9C%EB%82%9819+%EB%B0%B1%EC%8B%A0+%ED%86%B5%EA%B3%84&gs_lcp=Cgdnd3Mtd2l6EAMyBQgAEM0CMgUIABDNAjIFCAAQzQI6BwgjELADECc6CQgAELADEAgQHjoHCCMQsAIQJzoICAAQCBANEB5KBAhBGAFQ5AFY8wpgpQxoAnAAeACAAXWIAcYGkgEDMy41mAEAoAEByAECwAEB&sclient=gws-wiz'\n",
        "\n",
        "#사용자 headers를 기반으로 국가별 코로나 현황 정보가 나와있는 html문서 불러오기\n",
        "response = requests.get(url, headers=headers)\n",
        "\n",
        "#불러온 html문서를 bs4의 html.parser방식으로 변환\n",
        "soup = BeautifulSoup(response.text, 'html.parser')\n",
        "\n",
        "#변환한 html문서를 htmls 리스트에 추가\n",
        "htmls.append(soup)\n",
        "\n",
        "#제대로 불러왔는지 title 표시\n",
        "print(soup.title)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<title>전세계 코로나19 백신 통계 - Google Search</title>\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kBx4lVjjX_jC"
      },
      "source": [
        "# 3. 데이터 전처리"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pMIw_zfbvhWx"
      },
      "source": [
        "### 국가별 gdp 데이터 전처리"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2Sygd-U4X2Mi",
        "outputId": "ccbe74c0-fe03-4cc3-f24e-5ce40ac28a53"
      },
      "source": [
        "#국가와 gdp를 넣을 빈 리스트 생성\n",
        "country = []\n",
        "gdp = []\n",
        "\n",
        "#국가와 gdp가 나와있는 태그만을 찾아 국가인 경우 국가명을 국가 리스트에 추가, gdp인 경우 gdp를 gdp 리스트에 추가\n",
        "for i in htmls[0].select('table > tbody > tr > td > span')[194:566]:\n",
        "  #gdp의 경우 text 추출 시 숫자가 존재하고 국가의 경우 text 추출 시 아무것도 없으므로 이를 통해 구분\n",
        "  if i.text:\n",
        "    gdp.append(int(''.join(i.text.split(','))))\n",
        "  else:\n",
        "    #국가명하고 (년도)가 나와있는 경우 존재해 국가명만 추출\n",
        "    if '(' in i.a['title']:\n",
        "      name = i.a['title'].split('(')[0]\n",
        "      #사이트별로 국가명의 띄어쓰기를 다르게하기에 공백을 제거함으로써 통일\n",
        "      name = ''.join(name.split(' '))\n",
        "      country.append(name)\n",
        "    else:\n",
        "      country.append(i.a['title'])\n",
        "\n",
        "#국가명은 없으나 gdp가 있어 제외\n",
        "gdp.remove(105422)\n",
        "\n",
        "#위 에러로 인해 아래 행의 국가명이 나오지 않아 추가 \n",
        "country.insert(60, '푸에르토리코')\n",
        "\n",
        "#국가명 리스트 길이와 gdp 리스트 길이가 일치하는지 확인함으로써 검증\n",
        "print(len(country))\n",
        "print(len(gdp))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "186\n",
            "186\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "bnevd48b6iNT",
        "outputId": "cf5b40cc-9c33-47f5-d98b-29dbda3bbca9"
      },
      "source": [
        "#국가 및 gdp 리스트를 데이터프레임화\n",
        "gdp_dict = {'국가명' : country, 'GDP' : gdp}\n",
        "gdp_df = pd.DataFrame(gdp_dict)\n",
        "\n",
        "#데이터프레임화가 됐는지 검증\n",
        "gdp_df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>GDP</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>미국</td>\n",
              "      <td>21427700</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>중화인민공화국</td>\n",
              "      <td>14342903</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>일본</td>\n",
              "      <td>5081770</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>독일</td>\n",
              "      <td>3845630</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>인도</td>\n",
              "      <td>2875142</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       국가명       GDP\n",
              "0       미국  21427700\n",
              "1  중화인민공화국  14342903\n",
              "2       일본   5081770\n",
              "3       독일   3845630\n",
              "4       인도   2875142"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SKih_CFkaR9l",
        "outputId": "dd36d8e7-3681-4f8d-ec97-c3e6df736833"
      },
      "source": [
        "#공백 데이터가 존재하는지 검증\n",
        "gdp_df.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 186 entries, 0 to 185\n",
            "Data columns (total 2 columns):\n",
            " #   Column  Non-Null Count  Dtype \n",
            "---  ------  --------------  ----- \n",
            " 0   국가명     186 non-null    object\n",
            " 1   GDP     186 non-null    int64 \n",
            "dtypes: int64(1), object(1)\n",
            "memory usage: 3.0+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t2QQWzAB7J47"
      },
      "source": [
        "### 국가별 인구 데이터 전처리"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W6hO7Ziuv9CZ",
        "outputId": "c3aa1759-7e1a-4c76-ff54-dae6b920fc2e"
      },
      "source": [
        "#국가와 인구수를 넣을 빈 리스트 생성\n",
        "country = []\n",
        "population = []\n",
        "\n",
        "#국가와 인구수가 tr 태그로 묶여져 있기에 해당 태그를 모두 찾은 후 하나씩 정보 추출\n",
        "for i in htmls[1].select('table > tbody > tr')[:240]:\n",
        "  #tr 태그 내에서 td 태그 중 인구수에 해당되는게 인덱스 1로 이를 추출\n",
        "  popu = i('td')[1].text\n",
        "  #추출한 인구수를 ,제거 및 정수형으로 변환\n",
        "  popu = int(''.join(popu.split(',')))\n",
        "  #인구수를 인구 리스트에 추가\n",
        "  population.append(popu)\n",
        "  #tr 태그 내에서 td 태그 중 국가명에 해당되는게 인덱스 0이며 a 태그 중 인덱스 1의 text로 이를 추출\n",
        "  name = i('td')[0]('a')[1].text\n",
        "  #사이트별로 국가명의 띄어쓰기를 다르게하기에 공백을 제거함으로써 통일\n",
        "  name = ''.join(name.split(' '))\n",
        "  country.append(name)\n",
        "\n",
        "#국가명 리스트 길이와 인구수 리스트 길이가 일치하는지 확인함으로써 검증\n",
        "print(len(country))\n",
        "print(len(population))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "240\n",
            "240\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "3TE-wewRvtT7",
        "outputId": "a1063d63-5b18-487a-a78c-2efc36d00f4c"
      },
      "source": [
        "#국가 및 인구수 리스트를 데이터프레임화\n",
        "popu_dict = {'국가명' : country, '인구수' : population}\n",
        "popu_df = pd.DataFrame(popu_dict)\n",
        "\n",
        "#데이터프레임화가 됐는지 검증\n",
        "popu_df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>인구수</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>중국</td>\n",
              "      <td>1411718724</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>인도</td>\n",
              "      <td>1379619774</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>미국</td>\n",
              "      <td>332915074</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>인도네시아</td>\n",
              "      <td>276361788</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>파키스탄</td>\n",
              "      <td>225199929</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     국가명         인구수\n",
              "0     중국  1411718724\n",
              "1     인도  1379619774\n",
              "2     미국   332915074\n",
              "3  인도네시아   276361788\n",
              "4   파키스탄   225199929"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FWpgHVsUZhe8",
        "outputId": "7c28d861-7a29-4930-ee4f-b276670a0363"
      },
      "source": [
        "#공백데이터가 존재하는지 검증\n",
        "popu_df.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 240 entries, 0 to 239\n",
            "Data columns (total 2 columns):\n",
            " #   Column  Non-Null Count  Dtype \n",
            "---  ------  --------------  ----- \n",
            " 0   국가명     240 non-null    object\n",
            " 1   인구수     240 non-null    int64 \n",
            "dtypes: int64(1), object(1)\n",
            "memory usage: 3.9+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5qt2v-CTANpq"
      },
      "source": [
        "### 국가별 코로나 현황 데이터 전처리"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0zMh4vqWE5-I"
      },
      "source": [
        "#코로나 현황 사이트는 국가명이 영어로 표기되기에 파파고 api를 이용해 한국어로 번역하는 함수 생성\n",
        "def get_translate_etok(text):\n",
        "  #파파고 api를 사용할 사용자 id와 코드 입력\n",
        "  client_id = \"j3f8zrLyKM0fhEKu0H0W\"\n",
        "  client_secret = \"oJffUyHgQa\"\n",
        "  #번역할 데이터와 언어 지정\n",
        "  data = {'text' : text,\n",
        "          'source' : 'en',\n",
        "          'target': 'ko'}\n",
        "  #파파고 api를 사용할 url 지정\n",
        "  url = \"https://openapi.naver.com/v1/papago/n2mt\"\n",
        "  #사용자 id 및 코드를 헤더로 입력\n",
        "  header = {\"X-Naver-Client-Id\":client_id,\n",
        "            \"X-Naver-Client-Secret\":client_secret}\n",
        "  #requests 함수를 통해 파파고 url에 헤더와 번역할 내용 전송\n",
        "  response = requests.post(url, headers=header, data=data)\n",
        "  #응답코드 저장\n",
        "  rescode = response.status_code\n",
        "  #에러가 뜨는 경우 에러 코드를 출력하도록 조건문 생성\n",
        "  if(rescode==200):\n",
        "      send_data = response.json()\n",
        "      trans_data = (send_data['message']['result']['translatedText'])\n",
        "      return trans_data\n",
        "  else:\n",
        "      print(\"Error Code:\" , rescode)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RyF-Hxek8Aii",
        "outputId": "449d8aa2-8eb5-4670-d116-95e0ea455247"
      },
      "source": [
        "#국가, 확진자수, 사망자수를 넣을 빈 리스트 생성\n",
        "country = []\n",
        "confirm = []\n",
        "death = []\n",
        "\n",
        "#국가와 인구수가 tr 태그로 묶여져 있기에 해당 태그를 모두 찾은 후 하나씩 정보 추출\n",
        "for i in htmls[2].select('table > tbody > tr')[:231]:\n",
        "  #tr 태그 내에서 td 태그 중 국가명에 해당되는게 인덱스 0이며 a 태그 중 인덱스 1의 text로 이를 추출\n",
        "  name = i('td')[1].text\n",
        "  #이 사이트는 국가명 양쪽으로 공백이 있기에 공백제거\n",
        "  name = name.strip()\n",
        "  #이 사이트는 국가명이 영어이기에 파파고 api를 활용해 한국어로 번역\n",
        "  name = get_translate_etok(name)\n",
        "  #사이트별로 국가명의 띄어쓰기를 다르게하기에 공백을 제거함으로써 통일(국가명이 공백인 경우가 존재하여 이는 continue 처리)\n",
        "  try:\n",
        "    name = ''.join(name.split(' '))\n",
        "  except:\n",
        "    continue\n",
        "  #파파고 번역으로 인해 국가명 뒤에 .이 추가되는 경우 존재\n",
        "  if name[-1] == '.':\n",
        "    name = name[:-1]\n",
        "  #국가명을 국가 리스트에 추가\n",
        "  country.append(name)\n",
        "  #tr 태그 내에서 td 태그 중 확진자수에 해당되는게 인덱스 2로 이를 추출\n",
        "  conf = i('td')[2].text\n",
        "  #추출한 확진자수를 ,제거 및 정수형으로 변환\n",
        "  conf = int(''.join(conf.split(',')))\n",
        "  #확진자수를 확진자수 리스트에 추가\n",
        "  confirm.append(conf)\n",
        "  #tr 태그 내에서 td 태그 중 사망자수에 해당되는게 인덱스 4로 이를 추출\n",
        "  die = i('td')[4].text\n",
        "  #사망자가 없는 국가의 경우 0 값으로 추가\n",
        "  if die == ' ':\n",
        "    death.append(0)\n",
        "  else:\n",
        "    #추출한 사망자수를 ,제거 및 정수형으로 변환\n",
        "    die = int(''.join(die.split(',')))\n",
        "    #사망자수를 사망자수 리스트에 추가\n",
        "    death.append(die)\n",
        "\n",
        "#국가명 리스트 길이와 인구 리스트 길이가 일치하는지 확인함으로써 검증\n",
        "print(len(country))\n",
        "print(len(confirm))\n",
        "print(len(death))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Error Code: 400\n",
            "230\n",
            "230\n",
            "230\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "BkK3yfVIAf4B",
        "outputId": "76447a1a-2e3d-4065-86a5-06d9b422959c"
      },
      "source": [
        "#국가, 확진자수 및 사망자수 리스트를 데이터프레임화\n",
        "corona_dict = {'국가명' : country, '확진자수' : confirm, '사망자수' : death}\n",
        "corona_df = pd.DataFrame(corona_dict)\n",
        "\n",
        "#데이터프레임화가 됐는지 검증\n",
        "corona_df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>확진자수</th>\n",
              "      <th>사망자수</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>북아메리카</td>\n",
              "      <td>59282891</td>\n",
              "      <td>1196385</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>아시아</td>\n",
              "      <td>82226824</td>\n",
              "      <td>1215571</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>남아메리카</td>\n",
              "      <td>39007037</td>\n",
              "      <td>1182433</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>유럽</td>\n",
              "      <td>74155132</td>\n",
              "      <td>1420355</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>아프리카</td>\n",
              "      <td>8748400</td>\n",
              "      <td>223815</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     국가명      확진자수     사망자수\n",
              "0  북아메리카  59282891  1196385\n",
              "1    아시아  82226824  1215571\n",
              "2  남아메리카  39007037  1182433\n",
              "3     유럽  74155132  1420355\n",
              "4   아프리카   8748400   223815"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5DwO1FDRZWhm",
        "outputId": "b3bc8db7-991b-4538-ccd3-359d1d92de75"
      },
      "source": [
        "#공백 데이터가 존재하는지 검증\n",
        "corona_df.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 230 entries, 0 to 229\n",
            "Data columns (total 3 columns):\n",
            " #   Column  Non-Null Count  Dtype \n",
            "---  ------  --------------  ----- \n",
            " 0   국가명     230 non-null    object\n",
            " 1   확진자수    230 non-null    int64 \n",
            " 2   사망자수    230 non-null    int64 \n",
            "dtypes: int64(2), object(1)\n",
            "memory usage: 5.5+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7ggfoLCtMO_z"
      },
      "source": [
        "### 국가별 코로나 백신 현황 데이터 전처리"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W7b9Ki8XMQXU",
        "outputId": "c9f1ab59-128b-4cc4-d404-1192d3c52e8e"
      },
      "source": [
        "#국가와 백신 접종비율을 넣을 빈 리스트 생성\n",
        "country = []\n",
        "vaccine = []\n",
        "\n",
        "#국가와 접종비율이 tr 태그로 묶여져 있기에 해당 태그를 모두 찾은 후 하나씩 정보 추출\n",
        "for i in htmls[3].select('table > tbody > tr')[:191]:\n",
        "  #tr 태그 내에서 td 태그 중 국가명에 해당되는게 인덱스 0이며 span 태그 중 첫 번째의 text로 이를 추출\n",
        "  name = i('td')[0]('span')[0].text\n",
        "  #국가명하고 (세부내용)이 나와있는 경우 존재해 국가명만 추출\n",
        "  if '(' in name:\n",
        "    name = name.split('(')[0]\n",
        "  else:\n",
        "    pass\n",
        "  #이 사이트는 국가명이 영어이기에 파파고 api를 활용해 한국어로 번역\n",
        "  name = get_translate_etok(name)\n",
        "  #사이트별로 국가명의 띄어쓰기를 다르게하기에 공백을 제거함으로써 통일\n",
        "  name = ''.join(name.split(' '))\n",
        "  #파파고 번역으로 인해 국가명 뒤에 .이 추가되는 경우 존재\n",
        "  if name[-1] == '.':\n",
        "    name = name[:-1]\n",
        "  #국가명을 국가 리스트에 추가\n",
        "  country.append(name)\n",
        "  #tr 태그 내에서 td 태그 중 접종비율에 해당되는게 인덱스 3, span 태그 중 첫 번재의 text로 이를 추출(중국만 공백으로 처리되어 있어 이는 따로 처리)\n",
        "  try:\n",
        "    vac = i('td')[3]('span')[0].text[:-1]\n",
        "  except:\n",
        "    vaccine.append(75.8)\n",
        "    continue\n",
        "  #추출한 접종비율은 % 형태로 실수형으로 변환\n",
        "  vac = float(vac)\n",
        "  #접종비율을 백신 리스트에 추가\n",
        "  vaccine.append(vac)\n",
        "\n",
        "#국가명 리스트 길이와 인구수 리스트 길이가 일치하는지 확인함으로써 검증\n",
        "print(len(country))\n",
        "print(len(vaccine))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "191\n",
            "191\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sVi2cW-tMTOy",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "2e257e33-7c81-42b8-fe0a-989452b14ae5"
      },
      "source": [
        "#국가 및 접종비율 리스트를 데이터프레임화\n",
        "vac_dict = {'국가명' : country, '접종비율' : vaccine}\n",
        "vac_df = pd.DataFrame(vac_dict)\n",
        "\n",
        "#데이터프레임화가 됐는지 검증\n",
        "vac_df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>접종비율</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>중국</td>\n",
              "      <td>75.8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>인도</td>\n",
              "      <td>32.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>미국</td>\n",
              "      <td>59.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>브라질</td>\n",
              "      <td>63.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>인도네시아</td>\n",
              "      <td>34.5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     국가명  접종비율\n",
              "0     중국  75.8\n",
              "1     인도  32.5\n",
              "2     미국  59.4\n",
              "3    브라질  63.0\n",
              "4  인도네시아  34.5"
            ]
          },
          "metadata": {},
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GDbBS8W7aNI8",
        "outputId": "92dcd600-584b-4028-cd4c-d9e969b9f20d"
      },
      "source": [
        "#공백 데이터가 존재하는지 검증\n",
        "vac_df.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 191 entries, 0 to 190\n",
            "Data columns (total 2 columns):\n",
            " #   Column  Non-Null Count  Dtype  \n",
            "---  ------  --------------  -----  \n",
            " 0   국가명     191 non-null    object \n",
            " 1   접종비율    191 non-null    float64\n",
            "dtypes: float64(1), object(1)\n",
            "memory usage: 3.1+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gZntVwZaTFyY"
      },
      "source": [
        "### 생성한 데이터프레임 파일로 생성\n",
        "\n",
        "검증하고하는 주제가 국가 gdp에 따른 코로나 확진자수 및 백신 접종자수로 코로나 확진자 데이터가 없는 국가, 백신 접종자 데이터가 없는 국가를 기준으로 데이터프레임 병합한다."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kt-8OMrKPfdn",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "52c2b2b2-3c7b-4432-90cd-dca7f2c18c47"
      },
      "source": [
        "#국가별 gdp 및 인구수 데이터프레임 병합\n",
        "df = pd.merge(gdp_df, popu_df, how = 'inner', on = '국가명')\n",
        "df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>GDP</th>\n",
              "      <th>인구수</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>미국</td>\n",
              "      <td>21427700</td>\n",
              "      <td>332915074</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>일본</td>\n",
              "      <td>5081770</td>\n",
              "      <td>125960000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>독일</td>\n",
              "      <td>3845630</td>\n",
              "      <td>83783942</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>인도</td>\n",
              "      <td>2875142</td>\n",
              "      <td>1379619774</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>영국</td>\n",
              "      <td>2827113</td>\n",
              "      <td>67886011</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  국가명       GDP         인구수\n",
              "0  미국  21427700   332915074\n",
              "1  일본   5081770   125960000\n",
              "2  독일   3845630    83783942\n",
              "3  인도   2875142  1379619774\n",
              "4  영국   2827113    67886011"
            ]
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JfDHXxc4YS_4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "f64ea454-eb8c-473a-9582-a13e1bee7d36"
      },
      "source": [
        "#최종 데이터프레임 및 코로나 현황 데이터프레임 병합\n",
        "df = pd.merge(df, corona_df, how = 'inner', on = '국가명')\n",
        "df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>GDP</th>\n",
              "      <th>인구수</th>\n",
              "      <th>확진자수</th>\n",
              "      <th>사망자수</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>미국</td>\n",
              "      <td>21427700</td>\n",
              "      <td>332915074</td>\n",
              "      <td>49573122</td>\n",
              "      <td>805004</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>일본</td>\n",
              "      <td>5081770</td>\n",
              "      <td>125960000</td>\n",
              "      <td>1727430</td>\n",
              "      <td>18362</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>독일</td>\n",
              "      <td>3845630</td>\n",
              "      <td>83783942</td>\n",
              "      <td>5953310</td>\n",
              "      <td>102552</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>인도</td>\n",
              "      <td>2875142</td>\n",
              "      <td>1379619774</td>\n",
              "      <td>34606541</td>\n",
              "      <td>469724</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>영국</td>\n",
              "      <td>2827113</td>\n",
              "      <td>67886011</td>\n",
              "      <td>10276007</td>\n",
              "      <td>145140</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  국가명       GDP         인구수      확진자수    사망자수\n",
              "0  미국  21427700   332915074  49573122  805004\n",
              "1  일본   5081770   125960000   1727430   18362\n",
              "2  독일   3845630    83783942   5953310  102552\n",
              "3  인도   2875142  1379619774  34606541  469724\n",
              "4  영국   2827113    67886011  10276007  145140"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9eY6sqrjYJ7s",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "cbd69023-9f30-4bac-ff6a-164fd76ff372"
      },
      "source": [
        "#최종 데이터프레임 및 코로나 백신 데이터프레임 병합\n",
        "df = pd.merge(df, vac_df, how = 'inner', on = '국가명')\n",
        "df.head()"
      ],
      "execution_count": null,
      "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>국가명</th>\n",
              "      <th>GDP</th>\n",
              "      <th>인구수</th>\n",
              "      <th>확진자수</th>\n",
              "      <th>사망자수</th>\n",
              "      <th>접종비율</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>미국</td>\n",
              "      <td>21427700</td>\n",
              "      <td>332915074</td>\n",
              "      <td>49573122</td>\n",
              "      <td>805004</td>\n",
              "      <td>59.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>일본</td>\n",
              "      <td>5081770</td>\n",
              "      <td>125960000</td>\n",
              "      <td>1727430</td>\n",
              "      <td>18362</td>\n",
              "      <td>77.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>독일</td>\n",
              "      <td>3845630</td>\n",
              "      <td>83783942</td>\n",
              "      <td>5953310</td>\n",
              "      <td>102552</td>\n",
              "      <td>68.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>인도</td>\n",
              "      <td>2875142</td>\n",
              "      <td>1379619774</td>\n",
              "      <td>34606541</td>\n",
              "      <td>469724</td>\n",
              "      <td>32.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>영국</td>\n",
              "      <td>2827113</td>\n",
              "      <td>67886011</td>\n",
              "      <td>10276007</td>\n",
              "      <td>145140</td>\n",
              "      <td>69.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  국가명       GDP         인구수      확진자수    사망자수  접종비율\n",
              "0  미국  21427700   332915074  49573122  805004  59.4\n",
              "1  일본   5081770   125960000   1727430   18362  77.4\n",
              "2  독일   3845630    83783942   5953310  102552  68.5\n",
              "3  인도   2875142  1379619774  34606541  469724  32.5\n",
              "4  영국   2827113    67886011  10276007  145140  69.0"
            ]
          },
          "metadata": {},
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 추가 데이터 전처리"
      ],
      "metadata": {
        "id": "CVJvsxWFamNu"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "객관적인 비교를 위해 각 국가별 GDP, 확진자수, 사망자수를 인구수로 나눠 접종비율과 같이 비율로 나타낸다."
      ],
      "metadata": {
        "id": "Hbi87wFlaqrc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 칼럼 생성(사이트에 백만달러가 단위이기에 GDP에 백만을 곱한다.)\n",
        "df['1인당 GDP'] = df['GDP'] * 1000000 / df['인구수']\n",
        "\n",
        "#확진비율 칼럼 생성\n",
        "df['확진비율'] = df['확진자수'] / df['인구수'] * 100\n",
        "\n",
        "#사망비율 칼럼 생성\n",
        "df['사망비율'] = df['사망자수'] / df['인구수'] * 100\n",
        "\n",
        "#칼럼이 제대로 생성됐는지 확인\n",
        "df.head()"
      ],
      "metadata": {
        "id": "nJJPBDcQc0-L"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L1oJj8bVapt-"
      },
      "source": [
        "#생성한 최종 데이터프레임 파일로 생성\n",
        "df[['국가명', '1인당 GDP', '접종비율', '확진비율', '사망비율']].to_excel('/content/drive/MyDrive/통계학 기말과제 국가별 gdp 및 코로나 데이터.xlsx', index = False)"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#4. 데이터 분포 확인"
      ],
      "metadata": {
        "id": "Xq670J9QcEj8"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C2kb5qQVb15L",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "d74726c4-1ac3-449b-cc04-d5089ff06b5c"
      },
      "source": [
        "#최종 데이터프레임 파일 로드\n",
        "df = pd.read_excel('/content/drive/MyDrive/통계학 기말과제 국가별 gdp 및 코로나 데이터.xlsx')\n",
        "\n",
        "#제대로 로드됐는지 확인\n",
        "df.head()"
      ],
      "execution_count": 3,
      "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>국가명</th>\n",
              "      <th>1인당 GDP</th>\n",
              "      <th>접종비율</th>\n",
              "      <th>확진비율</th>\n",
              "      <th>사망비율</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>미국</td>\n",
              "      <td>64363.862358</td>\n",
              "      <td>59.4</td>\n",
              "      <td>14.890621</td>\n",
              "      <td>0.241805</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>일본</td>\n",
              "      <td>40344.315656</td>\n",
              "      <td>77.4</td>\n",
              "      <td>1.371412</td>\n",
              "      <td>0.014578</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>독일</td>\n",
              "      <td>45899.368163</td>\n",
              "      <td>68.5</td>\n",
              "      <td>7.105550</td>\n",
              "      <td>0.122401</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>인도</td>\n",
              "      <td>2084.010431</td>\n",
              "      <td>32.5</td>\n",
              "      <td>2.508411</td>\n",
              "      <td>0.034047</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>영국</td>\n",
              "      <td>41645.001059</td>\n",
              "      <td>69.0</td>\n",
              "      <td>15.137150</td>\n",
              "      <td>0.213800</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  국가명       1인당 GDP  접종비율       확진비율      사망비율\n",
              "0  미국  64363.862358  59.4  14.890621  0.241805\n",
              "1  일본  40344.315656  77.4   1.371412  0.014578\n",
              "2  독일  45899.368163  68.5   7.105550  0.122401\n",
              "3  인도   2084.010431  32.5   2.508411  0.034047\n",
              "4  영국  41645.001059  69.0  15.137150  0.213800"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#데이터들의 기초 통계량인 평균, 표준편차, 4분위수 확인\n",
        "df.describe()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "ecwo1rsMQV29",
        "outputId": "7198c07e-69da-4338-a0fd-e9a7c6611a53"
      },
      "execution_count": 9,
      "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>1인당 GDP</th>\n",
              "      <th>접종비율</th>\n",
              "      <th>확진비율</th>\n",
              "      <th>사망비율</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>151.000000</td>\n",
              "      <td>151.000000</td>\n",
              "      <td>151.000000</td>\n",
              "      <td>151.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>15329.447487</td>\n",
              "      <td>41.173510</td>\n",
              "      <td>6.204500</td>\n",
              "      <td>0.098506</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>20476.726822</td>\n",
              "      <td>26.969792</td>\n",
              "      <td>6.048078</td>\n",
              "      <td>0.104902</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>426.549775</td>\n",
              "      <td>0.400000</td>\n",
              "      <td>0.000997</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>2097.619013</td>\n",
              "      <td>14.800000</td>\n",
              "      <td>0.574304</td>\n",
              "      <td>0.013179</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>6109.290207</td>\n",
              "      <td>42.600000</td>\n",
              "      <td>5.032332</td>\n",
              "      <td>0.066322</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>19483.656226</td>\n",
              "      <td>64.900000</td>\n",
              "      <td>10.120234</td>\n",
              "      <td>0.162021</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>115826.184977</td>\n",
              "      <td>89.300000</td>\n",
              "      <td>25.324772</td>\n",
              "      <td>0.619089</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "             1인당 GDP        접종비율        확진비율        사망비율\n",
              "count     151.000000  151.000000  151.000000  151.000000\n",
              "mean    15329.447487   41.173510    6.204500    0.098506\n",
              "std     20476.726822   26.969792    6.048078    0.104902\n",
              "min       426.549775    0.400000    0.000997    0.000000\n",
              "25%      2097.619013   14.800000    0.574304    0.013179\n",
              "50%      6109.290207   42.600000    5.032332    0.066322\n",
              "75%     19483.656226   64.900000   10.120234    0.162021\n",
              "max    115826.184977   89.300000   25.324772    0.619089"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "기초통계량만 봤을 때는 평균, 분산, 4분위수 모두 다르기에 네 집단 간에 유사성이 없는 것으로 보인다."
      ],
      "metadata": {
        "id": "C0u4JpMvfSv_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "단일 데이터들이 정규분포를 띠는지 확인 <br>\n",
        "box plot, histogram, skewtest, kurtosistest, normaltest, jarqueberatest, shapirotest"
      ],
      "metadata": {
        "id": "saVymMhh2VRJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#한글 폰트 다운로드\n",
        "!sudo apt-get install -y fonts-nanum\n",
        "!sudo fc-cache -fv\n",
        "!rm ~/.cache/matplotlib -rf\n",
        "\n",
        "#나눔고딕으로 matplotlib 폰트 설정\n",
        "plt.rc('font',family='NanumGothic')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8cbp8Ged9Aql",
        "outputId": "9af208b8-deb3-4664-8944-5a2147b7975b"
      },
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "The following NEW packages will be installed:\n",
            "  fonts-nanum\n",
            "0 upgraded, 1 newly installed, 0 to remove and 37 not upgraded.\n",
            "Need to get 9,604 kB of archives.\n",
            "After this operation, 29.5 MB of additional disk space will be used.\n",
            "Get:1 http://archive.ubuntu.com/ubuntu bionic/universe amd64 fonts-nanum all 20170925-1 [9,604 kB]\n",
            "Fetched 9,604 kB in 1s (7,364 kB/s)\n",
            "debconf: unable to initialize frontend: Dialog\n",
            "debconf: (No usable dialog-like program is installed, so the dialog based frontend cannot be used. at /usr/share/perl5/Debconf/FrontEnd/Dialog.pm line 76, <> line 1.)\n",
            "debconf: falling back to frontend: Readline\n",
            "debconf: unable to initialize frontend: Readline\n",
            "debconf: (This frontend requires a controlling tty.)\n",
            "debconf: falling back to frontend: Teletype\n",
            "dpkg-preconfigure: unable to re-open stdin: \n",
            "Selecting previously unselected package fonts-nanum.\n",
            "(Reading database ... 155222 files and directories currently installed.)\n",
            "Preparing to unpack .../fonts-nanum_20170925-1_all.deb ...\n",
            "Unpacking fonts-nanum (20170925-1) ...\n",
            "Setting up fonts-nanum (20170925-1) ...\n",
            "Processing triggers for fontconfig (2.12.6-0ubuntu2) ...\n",
            "/usr/share/fonts: caching, new cache contents: 0 fonts, 1 dirs\n",
            "/usr/share/fonts/truetype: caching, new cache contents: 0 fonts, 3 dirs\n",
            "/usr/share/fonts/truetype/humor-sans: caching, new cache contents: 1 fonts, 0 dirs\n",
            "/usr/share/fonts/truetype/liberation: caching, new cache contents: 16 fonts, 0 dirs\n",
            "/usr/share/fonts/truetype/nanum: caching, new cache contents: 10 fonts, 0 dirs\n",
            "/usr/local/share/fonts: caching, new cache contents: 0 fonts, 0 dirs\n",
            "/root/.local/share/fonts: skipping, no such directory\n",
            "/root/.fonts: skipping, no such directory\n",
            "/var/cache/fontconfig: cleaning cache directory\n",
            "/root/.cache/fontconfig: not cleaning non-existent cache directory\n",
            "/root/.fontconfig: not cleaning non-existent cache directory\n",
            "fc-cache: succeeded\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#전체 데이터의 박스플랏 생성\n",
        "#전체 plot 사이즈 지정\n",
        "plt.figure(figsize=(15,10))\n",
        "\n",
        "#첫 번째 plot 지정\n",
        "plt.subplot(2, 2, 1)\n",
        "df.boxplot('1인당 GDP')\n",
        "\n",
        "#두 번째 plot 지정\n",
        "plt.subplot(2, 2, 2)\n",
        "df.boxplot('확진비율')\n",
        "\n",
        "#세 번째 plot 지정\n",
        "plt.subplot(2, 2, 3)\n",
        "df.boxplot('사망비율')\n",
        "\n",
        "#네 번째 plot 지정\n",
        "plt.subplot(2, 2, 4)\n",
        "df.boxplot('접종비율')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 614
        },
        "id": "2XQqFW0z7UrO",
        "outputId": "ccf434f2-8639-4839-a430-42bf0fcbd3ac"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f688b87fd90>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 4 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#전체 데이터의 히스토그램 생성\n",
        "df.hist(figsize = (15, 10), bins = 50)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 701
        },
        "id": "LF_ABgLWVO1a",
        "outputId": "00baab99-3794-4bfe-ea00-c4177798e490"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7f6887ba2b90>,\n",
              "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f6887bc8810>],\n",
              "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7f6887b46510>,\n",
              "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f6887aeba10>]],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 39
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 4 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "4개 데이터 집단의 분포를 박스플랏과 히스토그램으로 봤을 때 정규분포의 형태가 아니나 다른 분포로의 유사성은 나타나는 것으로 보인다."
      ],
      "metadata": {
        "id": "rBK90q1VaBLf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 데이터의 skew test\n",
        "_, pval = stats.skewtest(np.array(df['1인당 GDP']))\n",
        "print(f\"1인당 GDP의 pval : {pval}\")\n",
        "\n",
        "#접종비율 데이터의 skew test\n",
        "_, pval = stats.skewtest(np.array(df['접종비율']))\n",
        "print(f\"접종비율의 pval : {pval}\")\n",
        "\n",
        "#확진비율의 skew test\n",
        "_, pval = stats.skewtest(np.array(df['확진비율']))\n",
        "print(f\"확진비율의 pval : {pval}\")\n",
        "\n",
        "#사망비율의 skew test\n",
        "_, pval = stats.skewtest(np.array(df['사망비율']))\n",
        "print(f\"사망비율의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FkEysoEyRA7c",
        "outputId": "bb25a671-3430-4ad1-ffe2-5073f8327097"
      },
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP의 pval : 3.875179895588504e-13\n",
            "접종비율의 pval : 0.7079728025990508\n",
            "확진비율의 pval : 6.195141613222321e-06\n",
            "사망비율의 pval : 1.4182926890073933e-09\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "접종비율 집단은 pval가 0.5를 초과하므로 중간값과 평균값의 차이가 없다는 귀무가설을 채택하나 그 외 집단은 0.5 미만으로 중간값과 평균값의 차이가 존재한다는 대립가설을 채택한다."
      ],
      "metadata": {
        "id": "tKJi2DyMagF5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 데이터의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(df['1인당 GDP']))\n",
        "print(f\"1인당 GDP의 pval : {pval}\")\n",
        "\n",
        "#접종비율 데이터의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(df['접종비율']))\n",
        "print(f\"접종비율의 pval : {pval}\")\n",
        "\n",
        "#확진비율의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(df['확진비율']))\n",
        "print(f\"확진비율의 pval : {pval}\")\n",
        "\n",
        "#사망비율의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(df['사망비율']))\n",
        "print(f\"사망비율의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hsLhJ1Ytaxo_",
        "outputId": "69116e88-00e9-4749-defb-5b1c9921b67e"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP의 pval : 2.536843007844659e-06\n",
            "접종비율의 pval : 4.88338677460066e-38\n",
            "확진비율의 pval : 0.1589629872071303\n",
            "사망비율의 pval : 3.402924076213332e-05\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "확진비율 집단은 pval가 0.5를 초과하므로 K값이 정규분포의 K값인 3과 차이가 없다는 귀무가설을 채택하나 그 외 집단은 0.5 미만으로 차이가 존재한다는 대립가설을 채택한다."
      ],
      "metadata": {
        "id": "5VDARc_iaxv5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 데이터의 normal test\n",
        "_, pval = stats.kurtosistest(np.array(df['1인당 GDP']))\n",
        "print(f\"1인당 GDP의 pval : {pval}\")\n",
        "\n",
        "#접종비율 데이터의 normal test\n",
        "_, pval = stats.normaltest(np.array(df['접종비율']))\n",
        "print(f\"접종비율의 pval : {pval}\")\n",
        "\n",
        "#확진비율의 normal test\n",
        "_, pval = stats.normaltest(np.array(df['확진비율']))\n",
        "print(f\"확진비율의 pval : {pval}\")\n",
        "\n",
        "#사망비율의 normal test\n",
        "_, pval = stats.normaltest(np.array(df['사망비율']))\n",
        "print(f\"사망비율의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ku3zY5l7ax2A",
        "outputId": "2d94e647-ed9b-4c44-cdd2-8dbd94091d4d"
      },
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP의 pval : 2.536843007844659e-06\n",
            "접종비율의 pval : 7.400572138510754e-37\n",
            "확진비율의 pval : 1.3599146574895648e-05\n",
            "사망비율의 pval : 2.054603195528273e-12\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "네 집단 모두 pval가 0.05 미만이므로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택한다."
      ],
      "metadata": {
        "id": "c8MAPTLMax7P"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 데이터의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(df['1인당 GDP']))\n",
        "print(f\"1인당 GDP의 pval : {pval}\")\n",
        "\n",
        "#접종비율 데이터의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(df['접종비율']))\n",
        "print(f\"접종비율의 pval : {pval}\")\n",
        "\n",
        "#확진비율의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(df['확진비율']))\n",
        "print(f\"확진비율의 pval : {pval}\")\n",
        "\n",
        "#사망비율의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(df['사망비율']))\n",
        "print(f\"사망비율의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lQhfKlhHayAj",
        "outputId": "4f27344a-5441-4a7a-a12b-2857533d4580"
      },
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP의 pval : 0.0\n",
            "접종비율의 pval : 0.0027768774093495097\n",
            "확진비율의 pval : 1.1691854426842951e-06\n",
            "사망비율의 pval : 0.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "네 집단 모두 pval가 0.05 미만이므로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택한다."
      ],
      "metadata": {
        "id": "Fn1arLObayFy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP 데이터의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(df['1인당 GDP']))\n",
        "print(f\"1인당 GDP의 pval : {pval}\")\n",
        "\n",
        "#접종비율 데이터의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(df['접종비율']))\n",
        "print(f\"접종비율의 pval : {pval}\")\n",
        "\n",
        "#확진비율의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(df['확진비율']))\n",
        "print(f\"확진비율의 pval : {pval}\")\n",
        "\n",
        "#사망비율의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(df['사망비율']))\n",
        "print(f\"사망비율의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "83P_PNGxbdzw",
        "outputId": "ac864fed-5003-4799-ad00-e44ee03edd89"
      },
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP의 pval : 1.3368788769150403e-15\n",
            "접종비율의 pval : 6.998612889219658e-07\n",
            "확진비율의 pval : 1.1603558114359203e-09\n",
            "사망비율의 pval : 1.630544137209977e-11\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "네 집단 모두 pval가 0.05 미만이므로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택한다. <br>\n",
        "따라서 네 집단은 정규분포를 따르지 않는다는 사실을 확인하였다."
      ],
      "metadata": {
        "id": "8-KoQxNCbd3W"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#5. 데이터 상관관계 확인"
      ],
      "metadata": {
        "id": "XCSWVpLrcI0Y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#전체 데이터간 산점도 생성\n",
        "#전체 plot 사이즈 지정\n",
        "plt.figure(figsize=(15, 10))\n",
        "\n",
        "#첫 번째 plot 지정\n",
        "plt.subplot(2, 3, 1)\n",
        "plt.title('1인당 GDP와 접종비율')\n",
        "plt.scatter(df['1인당 GDP'], df['접종비율'])\n",
        "\n",
        "#두 번째 plot 지정\n",
        "plt.subplot(2, 3, 2)\n",
        "plt.title('1인당 GDP와 확진비율')\n",
        "plt.scatter(df['1인당 GDP'], df['확진비율'])\n",
        "\n",
        "#세 번째 plot 지정\n",
        "plt.subplot(2, 3, 3)\n",
        "plt.title('1인당 GDP와 사망비율')\n",
        "plt.scatter(df['1인당 GDP'], df['사망비율'])\n",
        "\n",
        "#네 번째 plot 지정\n",
        "plt.subplot(2, 3, 4)\n",
        "plt.title('접종비율과 확진비율')\n",
        "plt.scatter(df['접종비율'], df['확진비율'])\n",
        "\n",
        "#다섯 번째 plot 지정\n",
        "plt.subplot(2, 3, 5)\n",
        "plt.title('접종비율과 사망비율')\n",
        "plt.scatter(df['접종비율'], df['사망비율'])\n",
        "\n",
        "#여섯 번째 plot 지정\n",
        "plt.subplot(2, 3, 6)\n",
        "plt.title('확진비율과 사망비율')\n",
        "plt.scatter(df['확진비율'], df['사망비율'])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 623
        },
        "id": "TMr4Q50-d3-m",
        "outputId": "ac6a040d-ad5a-420e-bb90-a3558a5ae4ec"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7fd516d95f10>"
            ]
          },
          "metadata": {},
          "execution_count": 8
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 6 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "일직선 형태로 뚜렷한 데이터 분포가 나타나지 않으나 GDP와 다른 데이터 간에는 가파른 일직선에 몰려있고 그 외에 부채꼴 모양으로 데이터 분포가 약간 퍼져있으며 접종비율, 확진비율, 사망비율 간에는 아랫쪽 방향의 부채꼴 모양의 분포가 나타나는 것으로 보아 서로 간에 어느 정도의 상관관계가 있는 것으로 여겨진다."
      ],
      "metadata": {
        "id": "cF1zYCCVfs2p"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "데이터 간 구체적인 상관관계 확인 <br>\n",
        "피어슨 상관계수 표 활용"
      ],
      "metadata": {
        "id": "MQBvQngoGO2y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#데이터들간 상관계수 계산\n",
        "corr = df.corr(method = 'pearson')\n",
        "\n",
        "#피어슨 상관계수 표 생성\n",
        "plt.rcParams[\"figure.figsize\"] = (15,10)\n",
        "sns.heatmap(corr,\n",
        "           annot = True,\n",
        "           cmap = 'Greens',\n",
        "           vmin = 0, vmax=1, \n",
        "          )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 613
        },
        "id": "8OnOA8-zcKNW",
        "outputId": "11c1eb28-e6f6-46e3-f03e-f3cf5e956f54"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd516c09950>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "피어슨 상관계수 표를 통해 상관계수가 0.6을 넘는 (1인당 GDP, 접종비율), (확진비율, 사망비율) 간에는 확실한 종속 관계가, 그 외의 데이터들 간에는 약한 종속 관계가 있다고 볼 수 있다. <br>\n",
        "이제 통계 테스트를 이용해 구체적으로 분석이 이뤄지며 전체 데이터는 연속형 데이터이고 nonnormality하므로 test 시행시 numeric data의 nonparametric의 test를 사용한다."
      ],
      "metadata": {
        "id": "-vxEatXQ-OUn"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "데이터 간의 상관관계 확인 <br>\n",
        "spearman's R test 실시"
      ],
      "metadata": {
        "id": "sfQ9Zy1W2Wz4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP와 접종비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['1인당 GDP'], df['접종비율'])\n",
        "print(f\"1인당 GDP와 접종비율 간의 pval : {pval}\")\n",
        "\n",
        "#1인당 GDP와 확진비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['1인당 GDP'], df['확진비율'])\n",
        "print(f\"1인당 GDP와 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#1인당 GDP와 사망비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['1인당 GDP'], df['사망비율'])\n",
        "print(f\"1인당 GDP와 사망비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 확진비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['확진비율'], df['접종비율'])\n",
        "print(f\"접종비율과 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 사망비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['접종비율'], df['사망비율'])\n",
        "print(f\"접종비율과 사망비율 간의 pval : {pval}\")\n",
        "\n",
        "#확진비율과 사망비율 간의 spearmanr test\n",
        "_, pval = stats.spearmanr(df['확진비율'], df['사망비율'])\n",
        "print(f\"확진비율과 사망비율 간의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FUexehb7dgX-",
        "outputId": "45bea2c5-6df0-4e2a-f528-885be09c8046"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP와 접종비율 간의 pval : 2.597240221890674e-36\n",
            "1인당 GDP와 확진비율 간의 pval : 1.287212904358471e-29\n",
            "1인당 GDP와 사망비율 간의 pval : 3.70444169183012e-14\n",
            "접종비율과 확진비율 간의 pval : 1.5729070785450133e-18\n",
            "접종비율과 사망비율 간의 pval : 8.731466185236383e-09\n",
            "확진비율과 사망비율 간의 pval : 1.7828246549734355e-43\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단 데이터들 간에 모두 pval가 0.05 미만이므로 두 집단 간에 상관관계가 있다는 대립가설을 채택한다. <br>\n",
        "전체 데이터들 간에 상관관계가 있다는 놀라운 결과가 나왔는데 이는 코로나 발병으로 인해 확진, 사망, 접종의 전체적인 흐름이 같이 증가하기 때문에 기본적인 상관관계가 존재하고 이것이 반영된 것으로 보인다. <br>\n",
        "그래도 피어슨 상관계수 표에서 알 수 있듯이 우리가 예상했던 접종비율과 1인당 GDP, 접종비율과 사망비율 간에는 뚜렷한 연관성이 있음을 알 수 있다. <br>\n",
        "앞으로는 데이터들이 같은 분포를 띠는지 test가 진행될 예정이며 데이터들간에는 종속관계가 있으므로 paired sample의 test를 사용한다."
      ],
      "metadata": {
        "id": "8gNpx_bsio5W"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "전체 데이터들 간에 동일 분포인지 확인 <br>\n",
        "Paired Wilcoxen Signed-Rank test 실시"
      ],
      "metadata": {
        "id": "HejFbXv4GV8R"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP와 접종비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['1인당 GDP'], df['접종비율'], zero_method=\"pratt\")\n",
        "print(f\"1인당 GDP와 접종비율 간의 pval : {pval}\")\n",
        "\n",
        "#1인당 GDP와 확진비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['1인당 GDP'], df['확진비율'], zero_method=\"pratt\")\n",
        "print(f\"1인당 GDP와 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#1인당 GDP와 사망비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['1인당 GDP'], df['사망비율'], zero_method=\"pratt\")\n",
        "print(f\"1인당 GDP와 사망비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 확진비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['확진비율'], df['접종비율'], zero_method=\"pratt\")\n",
        "print(f\"접종비율과 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 사망비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['접종비율'], df['사망비율'], zero_method=\"pratt\")\n",
        "print(f\"접종비율과 사망비율 간의 pval : {pval}\")\n",
        "\n",
        "#확진비율과 사망비율 간의 wilcoxon test\n",
        "_, pval = stats.wilcoxon(df['확진비율'], df['사망비율'], zero_method=\"pratt\")\n",
        "print(f\"확진비율과 사망비율 간의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZGvSKAXZGjXJ",
        "outputId": "58ff9d6f-a521-4d99-8580-ac7901a135fe"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP와 접종비율 간의 pval : 1.5753513023728927e-26\n",
            "1인당 GDP와 확진비율 간의 pval : 1.5753513023728927e-26\n",
            "1인당 GDP와 사망비율 간의 pval : 1.5753513023728927e-26\n",
            "접종비율과 확진비율 간의 pval : 1.5753513023728927e-26\n",
            "접종비율과 사망비율 간의 pval : 1.5753513023728927e-26\n",
            "확진비율과 사망비율 간의 pval : 1.5753513023728927e-26\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단 데이터들 간에는 모두 pval이 0.05 미만으로 두 집단은 동일하지 않은 분포를 띤다는 대립가설을 채택한다."
      ],
      "metadata": {
        "id": "5mzD2mv8GuPr"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "상관계수가 0.6보다 낮았던 데이터들간에는 독립이라고 가정하고 추가적으로 동일 분포인지 확인 <br>\n",
        "Mann-Whitney test 실시"
      ],
      "metadata": {
        "id": "zLiV5cfT3lPp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#1인당 GDP와 확진비율 간의 mannwhitneyu test\n",
        "_, pval = stats.mannwhitneyu(df['1인당 GDP'], df['확진비율'])\n",
        "print(f\"1인당 GDP와 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#1인당 GDP와 사망비율 간의 mannwhitneyu test\n",
        "_, pval = stats.mannwhitneyu(df['1인당 GDP'], df['사망비율'])\n",
        "print(f\"1인당 GDP와 사망비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 확진비율 간의 mannwhitneyu test\n",
        "_, pval = stats.mannwhitneyu(df['확진비율'], df['접종비율'])\n",
        "print(f\"접종비율과 확진비율 간의 pval : {pval}\")\n",
        "\n",
        "#접종비율과 사망비율 간의 mannwhitneyu test\n",
        "_, pval = stats.mannwhitneyu(df['접종비율'], df['사망비율'])\n",
        "print(f\"접종비율과 사망비율 간의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sNkKvzSHGju5",
        "outputId": "a6f2f1cc-48cf-4fad-8973-e18a1dbbdd36"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1인당 GDP와 확진비율 간의 pval : 2.540870257033502e-51\n",
            "1인당 GDP와 사망비율 간의 pval : 2.54080751191977e-51\n",
            "접종비율과 확진비율 간의 pval : 2.9117640754885746e-30\n",
            "접종비율과 사망비율 간의 pval : 2.6959306833574575e-51\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단 데이터들 간에는 모두 pval이 0.05 미만으로 두 집단은 동일하지 않은 분포를 띤다는 대립가설을 채택한다. <br>\n",
        "모든 데이터들 간에는 유의미한 상관관계가 있으나 동일한 분포가 나타나지 않음을 알 수 있다. 이는 데이터들 간의 단위가 다르기에 나타난 것으로 보이며 각 데이터마다 히스토그램에서 봤듯이 표준화시킨다면 동일한 분포가 나올 것으로 예상이 된다."
      ],
      "metadata": {
        "id": "zapVVxAzGkIR"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#6. (추가) 표본 및 표본평균의 분포 확인"
      ],
      "metadata": {
        "id": "PY49qfVM3s6E"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "추가로 표본과 표본평균들의 분포를 확인해보고자 그나마 균일하게 분포돼 있는 접종비율에서 임의로 표본을 추출하여 활용한다."
      ],
      "metadata": {
        "id": "wbLbbLfG383R"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#접종비율에서 1000개의 표본 샘플 추출\n",
        "sample = pd.DataFrame({\n",
        "    'sample_mean_05': [df['접종비율'].sample(5).mean() for _ in range(1000)],\n",
        "    'sample_mean_20': [df['접종비율'].sample(20).mean() for _ in range(1000)],\n",
        "})\n",
        "\n"
      ],
      "metadata": {
        "id": "BKul1Oc63yck"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#데이터들의 기초 통계량인 평균, 표준편차, 4분위수 확인\n",
        "sample.describe()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "LFzJZDgPhG5t",
        "outputId": "32ea42f5-435d-46e9-828f-50aaf1eac83f"
      },
      "execution_count": 47,
      "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>sample_mean_05</th>\n",
              "      <th>sample_mean_20</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>41.389980</td>\n",
              "      <td>41.150065</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>12.414809</td>\n",
              "      <td>5.618035</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>6.420000</td>\n",
              "      <td>25.150000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>32.705000</td>\n",
              "      <td>37.411250</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>41.780000</td>\n",
              "      <td>41.255000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>50.520000</td>\n",
              "      <td>45.166250</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>72.580000</td>\n",
              "      <td>57.500000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       sample_mean_05  sample_mean_20\n",
              "count     1000.000000     1000.000000\n",
              "mean        41.389980       41.150065\n",
              "std         12.414809        5.618035\n",
              "min          6.420000       25.150000\n",
              "25%         32.705000       37.411250\n",
              "50%         41.780000       41.255000\n",
              "75%         50.520000       45.166250\n",
              "max         72.580000       57.500000"
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "추출한 20개 샘플 두 집단이 동일한 분포를 띠는지 확인 <br>\n",
        "Mann-Whitney test 실시"
      ],
      "metadata": {
        "id": "_9aajqnIG7Xi"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#접종비율에서 20개 샘플 두 집단을 추출\n",
        "sample_20 = pd.DataFrame({\n",
        "    'vaccine1': df['접종비율'].sample(20),\n",
        "    'vaccine2': df['접종비율'].sample(20)\n",
        "})"
      ],
      "metadata": {
        "id": "-BxF2PqOhJR-"
      },
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#nonnormality한 모집단에서 추출한 두 표본집단 간의 mannwhitneyu test 실시\n",
        "_, pval = stats.mannwhitneyu(sample_20['vaccine1'], sample_20['vaccine2'])\n",
        "print(f\"nonnormality한 모집단에서 추출한 두 표본집단 간의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m1hiusjKHCmp",
        "outputId": "7fc15d85-b81f-433a-f27f-687f7819796a"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "nonnormality한 모집단에서 추출한 두 표본집단 간의 pval : 0.40494651883434646\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단의 데이터 간에는 pval이 0.05를 초과하으로 두 집단은 동일한 분포라는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "UGAtVQBPbj8u"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "추출한 표본 평균들이 정규분포를 띠는지 확인 <br>\n",
        "box plot, histogram, skewtest, kurtosistest, normaltest, jarqueberatest, shapirotest"
      ],
      "metadata": {
        "id": "T_KtV2lgHSto"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#전체 데이터의 박스플랏 생성\n",
        "#전체 plot 사이즈 지정\n",
        "plt.figure(figsize=(15,10))\n",
        "\n",
        "#첫 번째 plot 지정\n",
        "plt.subplot(1, 2, 1)\n",
        "sample.boxplot('sample_mean_05')\n",
        "\n",
        "#두 번째 plot 지정\n",
        "plt.subplot(1, 2, 2)\n",
        "sample.boxplot('sample_mean_20')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 610
        },
        "id": "AJ5UTIIyTvKb",
        "outputId": "a04c1b27-c11e-4a70-bf95-bec015ff33e3"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fa3e0e53b50>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#표본평균 데이터의 히스토그램 생성\n",
        "sample.hist(figsize = (15, 10), bins = 50)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 662
        },
        "id": "8eqJkB2eReSD",
        "outputId": "3995c39a-74a1-4a5a-a93b-45a3ddfcceba"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7fa3e0dd9c50>,\n",
              "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fa3e0d008d0>]],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "임의로 5개, 20개씩 추출한 샘플 평균들을 박스플랏과 히스토그램으로 봤을 때 정규분포의 형태인 것처럼 보인다."
      ],
      "metadata": {
        "id": "sr5ke1FwW_zL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#5개 샘플 평균들의 skew test\n",
        "_, pval = stats.skewtest(np.array(sample['sample_mean_05']))\n",
        "print(f\"sample_mean_05's pval : {pval}\")\n",
        "\n",
        "#20개 샘플 평균들의 skew test\n",
        "_, pval = stats.skewtest(np.array(sample['sample_mean_20']))\n",
        "print(f\"sample_mean_20's pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K0Y_eic7SYkG",
        "outputId": "bbe623a0-04f7-4dfc-e9ef-07a9f1829129"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sample_mean_05's pval : 0.32164052811659705\n",
            "sample_mean_20's pval : 0.8685217380679652\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단은 모두 pval가 0.05를 초과하므로 중간값과 평균값의 차이가 없다는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "7xme5S9kXbxr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#5개 샘플 평균들의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(sample['sample_mean_05']))\n",
        "print(f\"sample_mean_05's pval : {pval}\")\n",
        "\n",
        "#20개 샘플 평균들의 kurtosis test\n",
        "_, pval = stats.kurtosistest(np.array(sample['sample_mean_20']))\n",
        "print(f\"sample_mean_20's pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cHEXacREWVE3",
        "outputId": "13e1ef08-e8a1-4683-b1b7-f3443c95a02a"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sample_mean_05's pval : 0.0002472157609939327\n",
            "sample_mean_20's pval : 0.09413913859044315\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본개수가 5개인 표본평균 집단은 pval가 0.05 미만으로 K값이 정규분포의 K값인 3과 차이가 존재한다는 대립가설을 채택하나 표본개수가 20개인 표본평균 집단은 pval가 0.05를 초과하므로 정규분포의 K값과 차이가 없다는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "kyp52EnFYevy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#5개 샘플 평균들의 normal test\n",
        "_, pval = stats.normaltest(np.array(sample['sample_mean_05']))\n",
        "print(f\"sample_mean_05's pval : {pval}\")\n",
        "\n",
        "#20개 샘플 평균들의 normal test\n",
        "_, pval = stats.normaltest(np.array(sample['sample_mean_20']))\n",
        "print(f\"sample_mean_20's pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nZlshF4QYHHA",
        "outputId": "c9ee0de0-0cbe-412e-dc75-0155c3fa3ee3"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sample_mean_05's pval : 0.000740851514297894\n",
            "sample_mean_20's pval : 0.24298235965301704\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본개수가 5개인 표본평균 집단은 pval가 0.05 미만으로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택하나 표본개수가 20개인 표본평균 집단은 pval가 0.05를 초과하므로 데이터의 분포가 정규분포라는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "Y5gNl4ibY0SI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#5개 샘플 평균들의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(sample['sample_mean_05']))\n",
        "print(f\"sample_mean_05's pval : {pval}\")\n",
        "\n",
        "#20개 샘플 평균들의 jarque_bera test\n",
        "_, pval = stats.jarque_bera(np.array(sample['sample_mean_20']))\n",
        "print(f\"sample_mean_20's pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mUftiSCjZqmg",
        "outputId": "472f42ea-9715-4798-ae86-3a411b735da1"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sample_mean_05's pval : 0.0094973481121301\n",
            "sample_mean_20's pval : 0.29347054503761294\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본개수가 5개인 표본평균 집단은 pval가 0.05 미만으로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택하나 표본개수가 20개인 표본평균 집단은 pval가 0.05를 초과하므로 데이터의 분포가 정규분포라는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "H8cDVqygZq6M"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#5개 샘플 평균들의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(sample['sample_mean_05']))\n",
        "print(f\"sample_mean_05's pval : {pval}\")\n",
        "\n",
        "#20개 샘플 평균들의 shapiro test\n",
        "_, pval = stats.shapiro(np.array(sample['sample_mean_20']))\n",
        "print(f\"sample_mean_20's pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gp0O_gOAYzf1",
        "outputId": "6e23d6a5-89f7-43aa-c9cc-4aa7c7812ffc"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sample_mean_05's pval : 0.008294115774333477\n",
            "sample_mean_20's pval : 0.37569934129714966\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본개수가 5개인 표본평균 집단은 pval가 0.05 미만으로 데이터의 분포가 정규분포가 아니라는 대립가설을 채택하나 표본개수가 20개인 표본평균 집단은 pval가 0.05를 초과하므로 데이터의 분포가 정규분포라는 귀무가설을 채택한다.\n",
        "<br> 따라서 표본개수가 클수록 정규분포를 띠는 중심극한정리를 확인하였다.\n"
      ],
      "metadata": {
        "id": "hhTQSmPaZKEB"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "추출한 20개 표본평균 샘플 두 집단이 동일한 분포를 띠는지 확인 <br>\n",
        "Two-sample Kolmogorov-Smirnov test 실시"
      ],
      "metadata": {
        "id": "qIaXkh6xgCZ2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#20개 샘플의 표본평균 집단에서 20개 샘플 두 집단을 추출\n",
        "sample_mean_20 = pd.DataFrame({\n",
        "    'vaccine1': sample['sample_mean_20'].sample(20),\n",
        "    'vaccine2': sample['sample_mean_20'].sample(20)\n",
        "})"
      ],
      "metadata": {
        "id": "buWA8VHNge1u"
      },
      "execution_count": 63,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#normality한 표본평균 모집단에서 추출한 두 표본평균 집단 간의 ks_2samp test 실시\n",
        "_, pval = stats.ks_2samp(sample_mean_20['vaccine1'], sample_mean_20['vaccine2'])\n",
        "print(f\"normality한 모집단에서 추출한 두 표본집단 간의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ngrZ9OuMZTWJ",
        "outputId": "53697482-2912-4d84-fd4e-aae64a341396"
      },
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "normality한 모집단에서 추출한 두 표본집단 간의 pval : 0.9188052214121167\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "두 집단의 데이터 간에는 pval이 0.05를 초과하으로 두 집단은 동일한 분포라는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "hUc8EfWJZSxZ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "추출한 20개 표본평균 샘플의 평균이 표본평균 집단의 평균과 동일한지 확인(표본평균 집단의 표준편차를 모른다고 가정) <br>\n",
        "one sample T test 실시(표본평균 집단의 평균은 위에서 확인한 기초통계량 참조)"
      ],
      "metadata": {
        "id": "e7MUgqmVguWF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#normality한 표본평균 모집단에서 추출한 표본평균 샘플의 t test 실시\n",
        "_, pval = stats.ttest_1samp(sample['sample_mean_20'].sample(20), 41.150065)\n",
        "print(f\"모집단에서 추출한 표본집단의 pval : {pval}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V_BvmpjAgs7q",
        "outputId": "3a97e578-1dab-4520-a9b3-5480973331c6"
      },
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "모집단에서 추출한 표본집단의 pval : 0.942838012657401\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본집단의 pval이 0.05를 초과하으로 표본집단의 평균이 모집단의 평균과 동일하다는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "J9jqPrSzi9Bz"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "추출한 20개 표본평균 샘플의 표준편차가 표본평균 집단의 표준편차와 동일한지 확인 <br>\n",
        "Chi-Square test 실시(표본평균 집단의 표준편차는 위에서 확인한 기초통계량 참조)"
      ],
      "metadata": {
        "id": "qPpQwfAwjt20"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#normality한 표본평균 모집단에서 임의로 20개 샘플 추출\n",
        "a = sample['sample_mean_20'].sample(20)\n",
        "sigma_0 = 5.618035"
      ],
      "metadata": {
        "id": "lSPLrue_iF3N"
      },
      "execution_count": 67,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#샘플의 통계량 추출\n",
        "stat = (a.shape[0] - 1) * a.var(ddof=1) / sigma_0**2\n",
        "stat"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FhrliieykJd9",
        "outputId": "831371f7-1712-4c9d-e16c-0b407dba34de"
      },
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "17.817504188479642"
            ]
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#normality한 표본평균 모집단에서 추출한 표본평균 샘플의 Chi-Square test 실시\n",
        "pval = 2 * min(\n",
        "    stats.chi2(a.shape[0]-1).cdf(stat), \n",
        "    stats.chi2(a.shape[0]-1).sf(stat)\n",
        ")\n",
        "pval"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tIrmemMvkc_l",
        "outputId": "2cfec079-d8e0-41a4-e1ce-d7d3fa25eb63"
      },
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9306841664739159"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "표본집단의 pval이 0.05를 초과하으로 표본집단의 표준편차가 모집단의 표준편차와 동일하다는 귀무가설을 채택한다."
      ],
      "metadata": {
        "id": "Y39jkSeKkrUN"
      }
    }
  ]
}