Ch05.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import Markdown, display\n",
    "\n",
    "import os\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "\n",
    "from scipy.stats import fisher_exact\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "if os.path.exists('src'):\n",
    "    %load_ext autoreload\n",
    "    %autoreload 2\n",
    "\n",
    "    from src.chi2_test import chi2_test\n",
    "    from src.utils import combine_rows, combine_cols\n",
    "    from src.plot import barplot_ct\n",
    "else:\n",
    "    !git clone https://github.com/P-Mihail/medstats_ex.git\n",
    "\n",
    "    from medstats_ex.src.chi2_test import chi2_test  # type: ignore\n",
    "    from medstats_ex.src.utils import combine_rows, combine_cols  # type: ignore\n",
    "    from medstats_ex.src.plot import barplot_ct  # type: ignore"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Глава 5\n",
    "#### Анализ качественных признаков\n",
    "\n",
    "##### ТОЧНОСТЬ ОЦЕНКИ ДОЛЕЙ\n",
    "\n",
    "$\\sigma_{\\hat{p}} = \\cfrac{\\sigma}{\\sqrt n} = \\sqrt{\\cfrac{p(1-p)}{n}}$ - стандартная ошибка доли.\n",
    "\n",
    "$\\sigma = \\sqrt{p(1-p)}$ - стандартное отклонение, $n$ - объем выборки.\n",
    "\n",
    "заменив истинное значениедоли $p$ на её оценку $\\hat p$ получаем оценку стандартной ошибки доли: $S_{\\hat{p}} = \\sqrt{\\cfrac{\\hat{p}(1-\\hat{p})}{n}}$\n",
    "\n",
    "Из ЦПТ следует, что при достаточно большом объеме выборки $\\hat{p}$ приближенно подчиняется нормальному распределению, имеющему среднее $р$ и стандартное отклонение $\\sigma_{\\hat{p}}$. Однако при значениях $р$, близких к 0 или 1, и при малом объеме выборки это не так!\n",
    "\n",
    "Условие применомости оценки: $n\\hat{p} > 5$ и $n(1 - \\hat{p}) > 5$\n",
    "\n",
    "##### СРАВНЕНИЕ ДОЛЕЙ (качественный признак принимает два значения, число выборок тоже равно двум)\n",
    "\n",
    "z - критерий (аналогично t критерию Стьюденты)\n",
    "\n",
    "$z = \\cfrac{Разность\\ выборочных\\ долей}{Стандартная\\ ошибка\\ разности\\ выборочных\\ долей} = \\cfrac{\\hat{p_1}-\\hat{p_2}}{S_{\\hat{p_1}-\\hat{p_2}}} = \\cfrac{\\hat{p_1}-\\hat{p_2}}{\\sqrt{S^2_{\\hat{p_1}} + S^2_{\\hat{p_2}}}} = \\cfrac{\\hat{p_1}-\\hat{p_2}}{\\sqrt{\\cfrac{\\hat{p_1}(1-\\hat{p_1})}{n_1} + \\cfrac{\\hat{p_2}(1-\\hat{p_2})}{n_2}}}$\n",
    "\n",
    "Если нулевая гипотеза верна, то обе выборочные доли $\\hat{p_1}=\\cfrac{m_1}{n_1}$ и $\\hat{p_2}=\\cfrac{m_2}{n_2}$ — это две оценки одной и той же доли $\\hat{p}$, которую, следовательно, можно оценить как: $\\hat{p}=\\cfrac{m_1 + m_2}{n_1 + n_2}$\n",
    "\n",
    "Тогда: $z = \\cfrac{\\hat{p_1}-\\hat{p_2}}{\\sqrt{\\hat{p}(1-\\hat{p})(\\cfrac{1}{n_1} + \\cfrac{1}{n_2})}}$\n",
    "\n",
    "$z$ подчиняется стандартному нормальному распределению => критические значения $z$ для 5% уровня значимости = 1,96, для 1% — 2,58.\n",
    "\n",
    "**поправка Йеитса** называемая также **поправкой на непрерывность**\n",
    "\n",
    "Нормальное распределение служит лишь приближением для распределения $z$. При этом оценка $P$ оказывается заниженной, и нулевая гипотеза будет отвергаться слишком часто. Причина состоит в том, что $z$ принимает только дискретные значения, тогда как приближающее его нормальное распределение непрерывно.\n",
    "\n",
    "Поправка Йейтса слегка уменьшает значение $z$, уменьшая тем самым расхождение с нормальным распределением.\n",
    "\n",
    "$z = \\cfrac{|\\hat{p_1}-\\hat{p_2}| - \\cfrac{1}{2}(\\cfrac{1}{n_1} + \\cfrac{1}{n_2})}{\\sqrt{\\hat{p}(1-\\hat{p})(\\cfrac{1}{n_1} + \\cfrac{1}{n_2})}}$\n",
    "\n",
    "##### ТАБЛИЦЫ СОПРЯЖЕННОСТИ: КРИТЕРИЙ $\\chi^2$\n",
    "\n",
    "число значений признака и число выборок может быть большим двух.\n",
    "\n",
    "\n",
    "Критерий $\\chi^2$ не требует никаких предположений относительно параметров совокупности, из которой извлечены выборки (непараметрический критерий).\n",
    "\n",
    "$\\chi^2 = \\sum\\cfrac{(O - E)^2}{E}$, где $O$ - наблюдаемое число в клетке таблицы сопряженности, $E$ - ожидаемое число в той же клетке. Суммирование проводится по всем клеткам таблицы.\n",
    "\n",
    "\n",
    "Критическое значений $\\chi^2$ зависит от размера матрицы сопряженности, то есть от числа сравниваемых методов (строк таблицы) и числа возможных исходов (столбцов таблицы). Размер таблицы выражается числом степеней свободы $v$: $ν = (r – 1)(c – 1)$, где $r$ — число строк, а $с$ — число столбцов.\n",
    "\n",
    "**Критерии $\\chi^2$ для таблицы 2×2**\n",
    "\n",
    "для таблиц сопряженности размером 2×2 выполняется равенство: $\\chi^2 = z^2$ \n",
    "\n",
    "**поправка Йеитса** (только для таблиц 2х2)\n",
    "\n",
    "$\\chi^2 = \\sum\\cfrac{|O - E| - \\cfrac{1}{2}}{E}$\n",
    "\n",
    "**Общий порядок применения критерия $\\chi^2$**\n",
    "* Постройте по имеющимся данным таблицу сопряженности.\n",
    "* Подсчитайте число объектов в каждой строке и в каждом столбце и найдите, какую долю от общего числа объектов составляют эти величины.\n",
    "* Зная эти доли, подсчитайте с точностью до двух знаков после запятой ожидаемые числа — количество объектов, которое попало бы в каждую клетку таблицы, если бы связь между строками и столбцами отсутствовала.\n",
    "* Найдите величину, характеризующую различия наблюдаемых и ожидаемых значений. Если таблица сопряженности имеет размер 2×2, примените поправку Йеитса.\n",
    "* Вычислите число степеней свободы, выберите уровень значимости и по таблице, определите критическое значение $\\chi^2$. Сравните его с полученным для вашей таблицы.\n",
    "\n",
    "##### Применимость критерия $\\chi^2$:\n",
    "\n",
    "Применение критерия  $\\chi^2$ для таблицы 2х2 правомерно, если ожидаемое число в любой из клеток больше или равно 5.\n",
    "\n",
    "В общем случае критерии  $\\chi^2$ применим, если все ожидаемые числа не меньше 1 и доля клеток с ожидаемыми числами меньше 5 не превышает 20%. Если это не так, то можно попробовать объединить несколько строк или столбцов.\n",
    "\n",
    "##### Преобразование таблиц сопряженности\n",
    "\n",
    "Перед объединением в одну группу надо проверить гипотезу об отсутствии межгрупповых различий.\n",
    "\n",
    "Поскольку осуществляется несколько сравнений с использованием одних и тех же данных, необходимо использовать поправку Бонферрони, умножая уровень значимости на число сравнений!!!\n",
    "\n",
    "##### ТОЧНЫЙ КРИТЕРИЙ ФИШЕРА\n",
    "\n",
    "Когда число наблюдений невелико, невыполняется услование применимости критерия $\\chi^2$. В этом случае используют точный критерий Фишера. Он основан на переборе всех возможных вариантов заполнения таблицы сопряженности при данной численности групп, поэтому, чем она меньше, тем проще его применить.\n",
    "\n",
    "Нулевая гипотеза состоит в том, что между группами и исходом нет никакой связи. Тогда вероятность получить некоторую таблицу равна:\n",
    "\n",
    "$P = \\cfrac{\\cfrac{R_1!R_2!C_1!C_2!}{N!}}{O_{11}!O_{12}!O_{21}!O_{22}!}$, где $R_1$ и $R_2$ — суммы по строкам, $С_1$ и $С_2$ — суммы по столбцам. $O_{11}$, $O_{12}$, $O_{21}$ и $O_{22}$ — числа в клетках, $N$ — общее число наблюдений.\n",
    "\n",
    "Построив все остальные варианты заполнения таблицы, возможные при данных суммах по строкам и столбцам, по этой же формуле рассчитывают их вероятность. Вероятности, которые не превосходят вероятность исходной таблицы (включая саму эту вероятность), суммируют. Полученная сумма — это величина P для двустороннего варианта точного критерия Фишера.\n",
    "\n",
    "Для построения всех вариантов заполнения таблицы в одной из клеток перебираются все целые числа от нуля до максимально возможного, пересчитывая числа в остальных клетках так, чтобы суммы по строкам и столбцам оставались неизменными."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.1.**\n",
    "\n",
    "Т. Бишоп (Т. Bishop. High frequency neural modulation in dentistry. J. Am. Dent. Assoc., 112:176—177, 1986) изучил эффективность высокочастотной стимуляции нерва в качестве обезболивающего средства при удалении зуба. Все больные подключались к прибору, но в одних случаях он работал, в других был выключен. Ни стоматолог, ни больной не знали, включен ли прибор. Позволяют ли следующие данные считать высокочастотную стимуляцию нерва действенным анальгезируюшим средством?\n",
    "\n",
    "-|Прибор включен|Прибор выключен\n",
    "-|-|-\n",
    "Боли нет|24|3\n",
    "Боль есть|6|17"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "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>Боли нет</th>\n",
       "      <th>Боль есть</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Прибор включен</th>\n",
       "      <td>24</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Прибор выключен</th>\n",
       "      <td>3</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Боли нет  Боль есть\n",
       "Прибор включен         24          6\n",
       "Прибор выключен         3         17"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 17.878, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[24, 6], [3, 17]], columns=[\"Боли нет\", \"Боль есть\"], index=[\"Прибор включен\", \"Прибор выключен\"])\n",
    "\n",
    "display(df)\n",
    "\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\")\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Да, позволяют: $\\chi^2$= 17,878; ν = 1; Р = 2.35e-05 < 0,001."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.2.**\n",
    "\n",
    "Синдром внезапной детской смерти — основная причина смерти детей в возрасте от 1 недели до 1 года. Обычно смерть наступает на фоне полного здоровья незаметно, во сне, поэтому\n",
    "определение факторов риска имеет первостепенное значение.\n",
    "\n",
    "Считается, что синдром внезапной детской смерти чаще случается у недоношенных детей, негров, а также в семьях с низкими доходами. Н. Левак и соавт. (N. Lewak et al. Sudden infant death syndrome risk factors: prospective data review. Clin. Pediatr., 18:404—411, 1979) решили уточнить эти данные. Исследователи собрали сведения о 19047 детях, родившихся в одном из роддомов Окленда, штат Калифорния, с 1960 по 1967 г. Судьбу детей проследили до 1 года. Данных о 48 детях получить не удалось.\n",
    "\n",
    "От синдрома внезапной детской смерти умерли 44 ребенка. Данные о предполагаемых факторах риска представлены в таблице. Найдите признаки, связанные с риском синдрома внезапной детской смерти.\n",
    "\n",
    "\n",
    "Фактор| |+|–\n",
    "-|-|-|-\n",
    "Возраст матери|До 25 лет|29|7301\n",
    "| |25 лет и старше|15|11241\n",
    "Время от окончания предыдущей беременности|Менее 1 года|23|4694\n",
    "||Более 1 года|11|7339\n",
    "Планировалась ли беременность|Нет|23|7654\n",
    "||Да|5|4253\n",
    "Повторная беременность|Нет|36|12987\n",
    "||Да|8|4999\n",
    "Курение во время беременности|Да|24|5228\n",
    "||Нет|10|9595\n",
    "Посещения врача во время беременности|Менее 11 раз|31|10512\n",
    "||11 раз или более|11|8154\n",
    "Самый низкий гемоглобин во время беременности|Менее 12 мг%|26|12613\n",
    "||12 мг% и более|7|2678\n",
    "Раса|Белые|31|12240\n",
    "||Негры|9|4323\n",
    "||Другие|4|2153"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "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></th>\n",
       "      <th>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Возраст матери</th>\n",
       "      <th>До 25 лет</th>\n",
       "      <td>29</td>\n",
       "      <td>7301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25 лет и старше</th>\n",
       "      <td>15</td>\n",
       "      <td>11241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Время от окончания предыдущей беременности</th>\n",
       "      <th>Менее 1 года</th>\n",
       "      <td>23</td>\n",
       "      <td>4694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Более 1 года</th>\n",
       "      <td>11</td>\n",
       "      <td>7339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Планировалась ли беременность</th>\n",
       "      <th>Нет</th>\n",
       "      <td>23</td>\n",
       "      <td>7654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Да</th>\n",
       "      <td>5</td>\n",
       "      <td>4253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Повторная беременность</th>\n",
       "      <th>Нет</th>\n",
       "      <td>36</td>\n",
       "      <td>12987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Да</th>\n",
       "      <td>8</td>\n",
       "      <td>4999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Курение во время беременности</th>\n",
       "      <th>Да</th>\n",
       "      <td>24</td>\n",
       "      <td>5228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Нет</th>\n",
       "      <td>10</td>\n",
       "      <td>9595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Посещения врача во время беременности</th>\n",
       "      <th>Менее 11 раз</th>\n",
       "      <td>31</td>\n",
       "      <td>10512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11 раз или более</th>\n",
       "      <td>11</td>\n",
       "      <td>8154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Самый низкий гемоглобин во время беременности</th>\n",
       "      <th>Менее 12 мг%</th>\n",
       "      <td>26</td>\n",
       "      <td>12613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12 мг% и более</th>\n",
       "      <td>7</td>\n",
       "      <td>2678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Раса</th>\n",
       "      <th>Белые</th>\n",
       "      <td>31</td>\n",
       "      <td>12240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Негры</th>\n",
       "      <td>9</td>\n",
       "      <td>4323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Другие</th>\n",
       "      <td>4</td>\n",
       "      <td>2153</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                 +      -\n",
       "Возраст матери                                До 25 лет         29   7301\n",
       "                                              25 лет и старше   15  11241\n",
       "Время от окончания предыдущей беременности    Менее 1 года      23   4694\n",
       "                                              Более 1 года      11   7339\n",
       "Планировалась ли беременность                 Нет               23   7654\n",
       "                                              Да                 5   4253\n",
       "Повторная беременность                        Нет               36  12987\n",
       "                                              Да                 8   4999\n",
       "Курение во время беременности                 Да                24   5228\n",
       "                                              Нет               10   9595\n",
       "Посещения врача во время беременности         Менее 11 раз      31  10512\n",
       "                                              11 раз или более  11   8154\n",
       "Самый низкий гемоглобин во время беременности Менее 12 мг%      26  12613\n",
       "                                              12 мг% и более     7   2678\n",
       "Раса                                          Белые             31  12240\n",
       "                                              Негры              9   4323\n",
       "                                              Другие             4   2153"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[29, 7301], [15, 11241], [23, 4694], [11, 7339], [23, 7654],\n",
    "                        [5, 4253], [36, 12987], [8, 4999], [24, 5228], [10, 9595],\n",
    "                        [31, 10512], [11, 8154], [26, 12613], [7, 2678], [31, 12240],\n",
    "                        [9, 4323], [4, 2153]], \n",
    "                 columns=[\"+\", \"-\"], \n",
    "                 index=pd.MultiIndex.from_arrays([[\"Возраст матери\", \"Возраст матери\", \"Время от окончания предыдущей беременности\", \n",
    "                     \"Время от окончания предыдущей беременности\", \"Планировалась ли беременность\", \"Планировалась ли беременность\", \n",
    "                     \"Повторная беременность\", \"Повторная беременность\", \"Курение во время беременности\", \"Курение во время беременности\", \n",
    "                     \"Посещения врача во время беременности\", \"Посещения врача во время беременности\", \"Самый низкий гемоглобин во время беременности\", \n",
    "                     \"Самый низкий гемоглобин во время беременности\", \"Раса\", \"Раса\", \"Раса\"], [\"До 25 лет\", \"25 лет и старше\", \"Менее 1 года\", \n",
    "                     \"Более 1 года\", \"Нет\", \"Да\", \"Нет\", \"Да\", \"Да\", \"Нет\", \"Менее 11 раз\", \"11 раз или более\", \"Менее 12 мг%\", \"12 мг% и более\", \n",
    "                     \"Белые\", \"Негры\", \"Другие\"]]))\n",
    "\n",
    "display(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "##### 1. Возраст матери:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>До 25 лет</th>\n",
       "      <td>29</td>\n",
       "      <td>7301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25 лет и старше</th>\n",
       "      <td>15</td>\n",
       "      <td>11241</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  +      -\n",
       "До 25 лет        29   7301\n",
       "25 лет и старше  15  11241"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 11.852, v=1, P=0.001\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 2. Время от окончания предыдущей беременности:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 1 года</th>\n",
       "      <td>23</td>\n",
       "      <td>4694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Более 1 года</th>\n",
       "      <td>11</td>\n",
       "      <td>7339</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               +     -\n",
       "Менее 1 года  23  4694\n",
       "Более 1 года  11  7339"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 10.506, v=1, P=0.001\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 3. Курение во время беременности:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Да</th>\n",
       "      <td>24</td>\n",
       "      <td>5228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Нет</th>\n",
       "      <td>10</td>\n",
       "      <td>9595</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      +     -\n",
       "Да   24  5228\n",
       "Нет  10  9595"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 17.002, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 4. Планировалась ли беременность:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Нет</th>\n",
       "      <td>23</td>\n",
       "      <td>7654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Да</th>\n",
       "      <td>5</td>\n",
       "      <td>4253</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      +     -\n",
       "Нет  23  7654\n",
       "Да    5  4253"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 3.144, v=1, P=0.076\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 5. Повторная беременность:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Нет</th>\n",
       "      <td>36</td>\n",
       "      <td>12987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Да</th>\n",
       "      <td>8</td>\n",
       "      <td>4999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      +      -\n",
       "Нет  36  12987\n",
       "Да    8   4999"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 1.571, v=1, P=0.210\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 6. Посещения врача во время беременности:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 11 раз</th>\n",
       "      <td>31</td>\n",
       "      <td>10512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11 раз или более</th>\n",
       "      <td>11</td>\n",
       "      <td>8154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   +      -\n",
       "Менее 11 раз      31  10512\n",
       "11 раз или более  11   8154"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 4.527, v=1, P=0.033\n",
      "Пограничный случай, есть основания задуматься над наличием различий.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 7. Раса:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Белые</th>\n",
       "      <td>31</td>\n",
       "      <td>12240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Негры</th>\n",
       "      <td>9</td>\n",
       "      <td>4323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Другие</th>\n",
       "      <td>4</td>\n",
       "      <td>2153</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         +      -\n",
       "Белые   31  12240\n",
       "Негры    9   4323\n",
       "Другие   4   2153"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 0.527, v=2, P=0.769\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "##### 8. Самый низкий гемоглобин во время беременности:"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>+</th>\n",
       "      <th>-</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 12 мг%</th>\n",
       "      <td>26</td>\n",
       "      <td>12613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12 мг% и более</th>\n",
       "      <td>7</td>\n",
       "      <td>2678</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 +      -\n",
       "Менее 12 мг%    26  12613\n",
       "12 мг% и более   7   2678"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 0.108, v=1, P=0.742\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i, (factor, new_df) in enumerate(df.groupby(level=0), start=1):\n",
    "    display(Markdown(f\"##### {i}. {factor}:\"))\n",
    "    display(new_df.droplevel(0))\n",
    "    \n",
    "    chi2_test(new_df, silent=False)\n",
    "\n",
    "    display(Markdown(\"---\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Статистически значимые различия (факторы риска):\n",
    "* Возраст матери до 25 лет\n",
    "* Время от окончания предыдущей беременности менее 1 года\n",
    "* Курение во время беременности\n",
    "\n",
    "Так же стоит обратить внимание на пограничный случай:\n",
    "* Посещения врача во время беременности менее 11 раз"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.3.**\n",
    "\n",
    "Могло ли повлиять отсутствие данных о 48 детях на результаты исследования? Если да, то как?\n",
    "\n",
    "Зависит от причин по которой были утрачены данные об этих детях. Если предполагать, что утрата случайна и они пренадлежат той же ГС, то эта утрата не могла сильно повлиять на результат. Однако в случае если была утрата из группы \"+\" или явная подтасовка результатов, тогда результату нельзя доверять."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.4.**\n",
    "\n",
    "Р. Феннел и соавт. (R. Fennell et al. Urinary tract infections in children effect of short course antibiotic therapy on recurrence rate in children with previous infections. Clin. Pediatr., 19:121—124, 1980) сравнили эффективность трех антибиотиков при рецидивирующей инфекции мочевых путей у девочек 3—16 лет. После короткого курса одного из антибактериальных препаратов (назначенного случайным образом) в течение года делали повторные посевы мочи. При выявлении бактериурии констатировали рецидив. Были получены следующие результаты.\n",
    "\n",
    "||Рецидив \"+\"|Рецидив \"-\"\n",
    "-|-|-\n",
    "Ампициллин|20|7\n",
    "Триметоприм/сульфаметоксазол|24|21\n",
    "Цефалексин|14|2\n",
    "\n",
    "Есть ли основания говорить о разной эффективности препаратов? Если да, то какой лучше?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "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>Рецидив \"+\"</th>\n",
       "      <th>Рецидив \"-\"</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Ампициллин</th>\n",
       "      <td>20</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Триметоприм/сульфаметоксазол</th>\n",
       "      <td>24</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Цефалексин</th>\n",
       "      <td>14</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Рецидив \"+\"  Рецидив \"-\"\n",
       "Ампициллин                             20            7\n",
       "Триметоприм/сульфаметоксазол           24           21\n",
       "Цефалексин                             14            2"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 7.288, v=2, P=0.026\n",
      "Пограничный случай, есть основания задуматься над наличием различий.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[20, 7], [24, 21], [14, 2]], \n",
    "                  index=[\"Ампициллин\", \"Триметоприм/сульфаметоксазол\", \"Цефалексин\"],\n",
    "                  columns=[\"Рецидив \\\"+\\\"\", \"Рецидив \\\"-\\\"\"])\n",
    "\n",
    "display(df)\n",
    "\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\")\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Исходя из графиков, можно предположить, о возможноти сгруппировать \"Ампицилин\" и \"Цефаликсин\" в случае отсутствия статистических различий между ними и сравнить полученную группу с группой \"Триметоприм/сульфаметоксазол\".\n",
    "Поскольку надо провести 2 сравнения, то стоит помнить про поправку Бонферрони для множественных сравнения."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Не выполнено условие применимости для матрицы 2х2, в ожидаемой матрице существует элемент < 5\n",
      "Критерий CHI2 не применим для этих данных.\n",
      "Применение точного критерия Фишера.\n",
      "P=0.446\n",
      "Отличия статистически не значимы.\n"
     ]
    }
   ],
   "source": [
    "chi2_test(df.loc[[\"Ампициллин\", \"Цефалексин\"]], bonf=2, silent=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "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>Рецидив \"+\"</th>\n",
       "      <th>Рецидив \"-\"</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Триметоприм/сульфаметоксазол</th>\n",
       "      <td>24</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ампициллин + Цефалексин</th>\n",
       "      <td>34</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Рецидив \"+\"  Рецидив \"-\"\n",
       "Триметоприм/сульфаметоксазол           24           21\n",
       "Ампициллин + Цефалексин                34            9"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 5.387, v=1, P=0.020\n",
      "Пограничный случай, есть основания задуматься над наличием различий.\n"
     ]
    }
   ],
   "source": [
    "new_df = df.pipe(combine_rows, [\"Ампициллин\", \"Цефалексин\"])\n",
    "\n",
    "display(new_df)\n",
    "\n",
    "barplot_ct(new_df)\n",
    "\n",
    "chi2_test(new_df, bonf=2, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Итак, Триметоприм/сульфаметоксазол превосходит как ампициллин, так и цефалексин, которые \n",
    "друг от друга не отличаются."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.5.**\n",
    "А. О’Нил и соавт. (A. O’Neil et al. A waterborn epidemic of acute infectious non-bacterial gastroenteritis in Alberta, Canada. Can.J. Public Health, 76:199—203, 1985) недавно сообщили о вспышке гастроэнтерита в маленьком канадском городке. Исследователи предположили, что источником инфекции была водопроводная вода. Они исследовали зависимость между количеством выпитой воды и числом заболевших. Какие выводы можно сделать из приводимых данных?\n",
    "\n",
    "Количество выпитой воды, стаканов в день|Число заболевших|Число не заболевших\n",
    "-|-|-\n",
    "Менее 1|39|121\n",
    "От 1 до 4|264|258\n",
    "5 и более|265|146"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "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>Заболел</th>\n",
       "      <th>Не заболел</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 1</th>\n",
       "      <td>39</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>От 1 до 4</th>\n",
       "      <td>264</td>\n",
       "      <td>258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5 и более</th>\n",
       "      <td>265</td>\n",
       "      <td>146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Заболел  Не заболел\n",
       "Менее 1         39         121\n",
       "От 1 до 4      264         258\n",
       "5 и более      265         146"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 74.973, v=2, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[39, 121], [264, 258], [265, 146]], index=[\"Менее 1\", \"От 1 до 4\", \"5 и более\"], columns=[\"Заболел\", \"Не заболел\"])\n",
    "\n",
    "display(df)\n",
    "\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "barplot_ct(df, idxname=\"Количество выпитой воды, стаканов в день\")\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Сравним группы попарно, чтоб убедиться в том, что все группы различны (это видно из последнего графика). Поправка Бонферрони на множественные сравнения = 3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "#### 1: \"Менее 1\" - \"От 1 до 4\""
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>Заболел</th>\n",
       "      <th>Не заболел</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 1</th>\n",
       "      <td>39</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>От 1 до 4</th>\n",
       "      <td>264</td>\n",
       "      <td>258</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Заболел  Не заболел\n",
       "Менее 1         39         121\n",
       "От 1 до 4      264         258"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 32.995, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "#### 2: \"Менее 1\" - \"5 и более\""
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>Заболел</th>\n",
       "      <th>Не заболел</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Менее 1</th>\n",
       "      <td>39</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5 и более</th>\n",
       "      <td>265</td>\n",
       "      <td>146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Заболел  Не заболел\n",
       "Менее 1         39         121\n",
       "5 и более      265         146"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 72.793, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "#### 3: \"От 1 до 4\" - \"5 и более\""
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>Заболел</th>\n",
       "      <th>Не заболел</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>От 1 до 4</th>\n",
       "      <td>264</td>\n",
       "      <td>258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5 и более</th>\n",
       "      <td>265</td>\n",
       "      <td>146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Заболел  Не заболел\n",
       "От 1 до 4      264         258\n",
       "5 и более      265         146"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 17.540, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from itertools import combinations\n",
    "\n",
    "for i, (idx1, idx2) in enumerate(combinations(df.index, 2), start=1):\n",
    "    display(Markdown(f\"#### {i}: \\\"{idx1}\\\" - \\\"{idx2}\\\"\"))\n",
    "    \n",
    "    display(df.loc[[idx1, idx2]])\n",
    "\n",
    "    chi2_test(df.loc[[idx1, idx2]], silent=False, bonf=3)\n",
    "    \n",
    "    display(Markdown(\"---\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Связь заболеваемости с количеством выпитой воды статистически значима. \n",
    "\n",
    "Сравнив группы попарно (используя поправку Бонферрони), можно убедиться, что заболеваемость растет с количеством выпитой воды."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.6.**\n",
    "Как правило, качество исследования выше, а соответствие собираемых данных поставленному вопросу точнее, если данные собираются специально для этого исследования после его планирования. Р. и С. Флетчеры (R. Fletcher, S. Fletcher. Clinical research in general medical journals: а 30-year perspective. N. Engl. J. Med., 301:180—183, 1979) исследовали 612 работ, случайным образом выбранных из журналов Journal of American Medical Association, Lancet и New England Journal of Medicine, чтобы определить, собирали ли их авторы свои данные до или после планирования исследования. Вот что удалось обнаружить:\n",
    "\n",
    "_|1946|1956|1966|1976\n",
    "-|-|-|-|-\n",
    "Число рассмотренных работ|151|149|157|155\n",
    "Процент работ, где данные собирали после планирования исследования|76|71|49|44\n",
    "Процент работ, где данные собирали до планирования исследования|24|29|51|56\n",
    "\n",
    "\n",
    "Оцените статистическую значимость различия долей. Если различия есть, то можно ли сказать, что положение меняется к лучшему?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "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>1946</th>\n",
       "      <th>1956</th>\n",
       "      <th>1966</th>\n",
       "      <th>1976</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Процент работ, где данные собирали после планирования исследования</th>\n",
       "      <td>115</td>\n",
       "      <td>106</td>\n",
       "      <td>77</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Процент работ, где данные собирали до планирования исследования</th>\n",
       "      <td>36</td>\n",
       "      <td>43</td>\n",
       "      <td>80</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    1946  1956  1966  1976\n",
       "Процент работ, где данные собирали после планир...   115   106    77    68\n",
       "Процент работ, где данные собирали до планирова...    36    43    80    87"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 48.698, v=3, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[151, 149, 157, 155], [76, 71, 49, 44], [24, 29, 51, 56]], \n",
    "                  index=[\"Число рассмотренных работ\", \n",
    "                         \"Процент работ, где данные собирали после планирования исследования\", \n",
    "                         \"Процент работ, где данные собирали до планирования исследования\"], \n",
    "                  columns=[1946, 1956, 1966, 1976])\n",
    "\n",
    "df.loc[\"Процент работ, где данные собирали после планирования исследования\"] = \\\n",
    "       (df.loc[\"Число рассмотренных работ\"] * df.loc[\"Процент работ, где данные собирали после планирования исследования\"] / 100).round()\n",
    "df.loc[\"Процент работ, где данные собирали до планирования исследования\"] = \\\n",
    "       (df.loc[\"Число рассмотренных работ\"] * df.loc[\"Процент работ, где данные собирали до планирования исследования\"] / 100).round()\n",
    "df.drop(\"Число рассмотренных работ\", inplace=True)\n",
    "\n",
    "display(df)\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Из графиков можно предположить о схожести групп 1946 и 1956, а так же групп 1966 и 1976. Если это так, то можно их сруппировать и сравнить полученные группы между собой. \n",
    "Всего надо сделать 3 сравнени, т.о. поправка Бонферрони = 3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 0.732, v=1, P=0.392\n",
      "Отличия статистически не значимы.\n"
     ]
    }
   ],
   "source": [
    "chi2_test(df[[1946, 1956]], bonf=3, silent=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 0.644, v=1, P=0.422\n",
      "Отличия статистически не значимы.\n"
     ]
    }
   ],
   "source": [
    "chi2_test(df[[1966, 1976]], bonf=3, silent=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "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>1946 + 1956</th>\n",
       "      <th>1966 + 1976</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Процент работ, где данные собирали после планирования исследования</th>\n",
       "      <td>221</td>\n",
       "      <td>145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Процент работ, где данные собирали до планирования исследования</th>\n",
       "      <td>79</td>\n",
       "      <td>167</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    1946 + 1956  1966 + 1976\n",
       "Процент работ, где данные собирали после планир...          221          145\n",
       "Процент работ, где данные собирали до планирова...           79          167"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 45.920, v=1, P=0.000\n",
      "Отличия статистически значимы.\n"
     ]
    }
   ],
   "source": [
    "new_df = df.pipe(combine_cols, [1946, 1956]).pipe(combine_cols, [1966, 1976])\n",
    "\n",
    "display(new_df)\n",
    "\n",
    "barplot_ct(new_df)\n",
    "\n",
    "chi2_test(new_df, bonf=3, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Разбиение таблицы показывает, что не отличаются 1946 от 1956 г. и 1966 от 1976 г.\n",
    "\n",
    "Объединенная группа 1946 и 1956 гг. отличается в лучшую сторону от объединенной группы 1966 и 1976 гг. Таким образом, между 1956 и 1966 г. ситуация изменилась к худшему."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.7.**\n",
    "\n",
    "Одна из причин инсульта — окклюзия сонной артерии. Чтобы выяснить, какое лечение — медикаментозное или хирургическое — дает в этом случае лучшие результаты, У. Филдс и соавт. (W. Fields et al. Joint study of extracranial arterial occlusion, V: Progress report of prognosis following surgery or nonsurginal treatment for transient ishemic attacks and cervical carotid artery lesions. JAMA, 211:1993—2003, 1970) сравнили долгосрочный прогноз у леченных двумя методами.\n",
    "\n",
    "\n",
    "Лечение|Повторный инсульт или смерть|Долгосрочное выживание\n",
    "-|-|-\n",
    "Хирургическое|43|36\n",
    "Медикаментозное|53|19\n",
    "\n",
    "Можно ли говорить о превосходстве одного из видов лечения?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "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>Повторный инсульт или смерть</th>\n",
       "      <th>Долгосрочное выживание</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Хирургическое</th>\n",
       "      <td>43</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Медикаментозное</th>\n",
       "      <td>53</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Повторный инсульт или смерть  Долгосрочное выживание\n",
       "Хирургическое                              43                      36\n",
       "Медикаментозное                            53                      19"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 5.185, v=1, P=0.023\n",
      "Пограничный случай, есть основания задуматься над наличием различий.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[43, 36], [53, 19]],\n",
    "                  columns=[\"Повторный инсульт или смерть\", \"Долгосрочное выживание\"],\n",
    "                  index=[\"Хирургическое\", \"Медикаментозное\"])\n",
    "\n",
    "display(df)\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Различия (в пользу хирургического лечения) статистически значимы."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.8.**\n",
    "В диагностике ишемической болезни сердца используют нагрузочную пробу, с помощью физической нагрузки вызывают ишемию миокарда, которую выявляют на ЭКГ. Существует другой метод, ишемию вызывают внутривенным введением дипиридамола, а выявляют с помощью эхокардиографии. Ф. Латтанци и соавт. (F. Lattanzi et al. Inhibition of dipyndamole-induced ishemia by\n",
    "antianginal therapy in humans: correlation with exercise electrocardiography. Circulation, 83:1256—1262, 1991) сравнили результаты двух методов у больных, получавших и не получавших антиангинальную терапию. Результаты приведены в таблице.\n",
    "\n",
    "Без антиангинальной терапии:\n",
    "\n",
    "_|Дипиридамол + эхокардиография \"+\"|\"-\"\n",
    "-|-|-\n",
    "**Нагрузка + ЭКГ \"+\"**|38|2\n",
    "\"-\"|14|3\n",
    "\n",
    "На фоне антиангинальной терапии:\n",
    "_|Дипиридамол + эхокардиография \"+\"|\"-\"\n",
    "-|-|-\n",
    "**Нагрузка + ЭКГ \"+\"**|21|6\n",
    "\"-\"|16|14\n",
    "\n",
    "Оцените различия между результатами двух методов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MultiIndex([(    'Без антиангинальной терапии', ...),\n",
       "            (    'Без антиангинальной терапии', ...),\n",
       "            ('На фоне антиангинальной терапии', ...),\n",
       "            ('На фоне антиангинальной терапии', ...)],\n",
       "           )"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.MultiIndex.from_product([[\"Без антиангинальной терапии\", \"На фоне антиангинальной терапии\"], [\"Дипиридамол + эхокардиография\"], [\"+\", \"-\"]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(data=[[38, 2, 21, 6], [14, 3, 16, 14]], \n",
    "                  columns=pd.MultiIndex.from_product([[\"Без антиангинальной терапии\", \"На фоне антиангинальной терапии\"], [\"impact \\\"+\\\"\", \"impact \\\"-\\\"\"]]),\n",
    "                  index=[\"Нагрузка + ЭКГ \\\"+\\\"\", \"Нагрузка + ЭКГ \\\"-\\\"\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">Без антиангинальной терапии</th>\n",
       "      <th colspan=\"2\" halign=\"left\">На фоне антиангинальной терапии</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>impact \"+\"</th>\n",
       "      <th>impact \"-\"</th>\n",
       "      <th>impact \"+\"</th>\n",
       "      <th>impact \"-\"</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Нагрузка + ЭКГ \"+\"</th>\n",
       "      <td>38</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Нагрузка + ЭКГ \"-\"</th>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>16</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Без антиангинальной терапии             \\\n",
       "                                    impact \"+\" impact \"-\"   \n",
       "Нагрузка + ЭКГ \"+\"                          38          2   \n",
       "Нагрузка + ЭКГ \"-\"                          14          3   \n",
       "\n",
       "                   На фоне антиангинальной терапии             \n",
       "                                        impact \"+\" impact \"-\"  \n",
       "Нагрузка + ЭКГ \"+\"                              21          6  \n",
       "Нагрузка + ЭКГ \"-\"                              16         14  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "#### 1. Без антиангинальной терапии"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Не выполнено условие применимости для матрицы 2х2, в ожидаемой матрице существует элемент < 5\n",
      "Критерий CHI2 не применим для этих данных.\n",
      "Применение точного критерия Фишера.\n",
      "P=0.151\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "#### 2. На фоне антиангинальной терапии"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 2.732, v=1, P=0.098\n",
      "Отличия статистически не значимы.\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "---"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i, (factor, new_df) in enumerate(df.groupby(level=0, axis=1), start=1):\n",
    "    display(Markdown(f\"#### {i}. {factor}\"))\n",
    "    sns.heatmap(new_df.droplevel(0,1), annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "    plt.show()\n",
    "    barplot_ct(new_df.droplevel(0,1))\n",
    "\n",
    "    chi2_test(new_df.droplevel(0,1), silent=False)\n",
    "\n",
    "    display(Markdown(\"---\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Для единообразия, можно использовать точный критерий Фишера и для случая \"На фоне антиангинальной терапии\". Что даст весьма схожий результат."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P=0.094\n"
     ]
    }
   ],
   "source": [
    "_, p = fisher_exact(df[\"На фоне антиангинальной терапии\"], alternative='two-sided')\n",
    "print(f\"P={p:.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Вывод: В обоих случаях различия статистически не значимы."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.9.**\n",
    "Д. Сакетт и М. Гент (D. Sackett, M. Gent. Controversy in counting and attributing events in clinical trials. N. Engl. J. Med., 301:1410—1412, 1979) сделали важное замечание относительно методики сбора данных в исследовании результатов медикаментозного и хирургического лечения окклюзии сонной артерии (задача 5.7). Так как изучался «долгосрочный прогноз», в исследование включали только тех больных, которые не умерли и у которых не было повторного инсульта во время госпитализации. В результате из рассмотрения были исключены 15 оперированных (5 из них умерли, а у 10 инсульт произошел вскоре после операции) и только 1 больной, лечившийся медикаментозно. Если учесть и этих 16 больных, то данные примут такой вид:\n",
    "\n",
    "Лечение|Повторный инсульт или смерть|Долгосрочное выживание\n",
    "-|-|-\n",
    "Хирургическое|58|36\n",
    "Медикаментозное|54|19\n",
    "\n",
    "Что теперь можно сказать о предпочтительности одного из видов лечения? Какое сравнение более верно — с учетом этих 16 больных или без их учета (как в задаче 5.7)? Почему?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "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>Повторный инсульт или смерть</th>\n",
       "      <th>Долгосрочное выживание</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Хирургическое</th>\n",
       "      <td>58</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Медикаментозное</th>\n",
       "      <td>54</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Повторный инсульт или смерть  Долгосрочное выживание\n",
       "Хирургическое                              58                      36\n",
       "Медикаментозное                            54                      19"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 2.273, v=1, P=0.132\n",
      "Отличия статистически не значимы.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[58, 36], [54, 19]],\n",
    "                  columns=[\"Повторный инсульт или смерть\", \"Долгосрочное выживание\"],\n",
    "                  index=[\"Хирургическое\", \"Медикаментозное\"])\n",
    "\n",
    "display(df)\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Теперь статистически значимых различий нет."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5.10.**\n",
    "Распространенность болезни X равна 10%. Болезнью Y страдает 1000 человек, болезнью Z — также 1000 человек. Болезнь X с равной вероятностью поражает страдающих болезнями Y и Z. Вероятность госпитализации при этих болезнях разная: для болезни X она составляет 40%, Y — 50%, Z — 20%. Посмотрим, сколько больных с разными сочетаниями болезней окажется в больнице.\n",
    "\n",
    "Из 1000 человек, страдающих болезнью Y, болезнь X имеют 10%, то есть 100 человек. Из них 50% (50 человек) будут госпитализированы в связи с болезнью Y, из оставшихся 50 человек в связи с болезнью X госпитализируют 40%, то есть 20 человек. Таким образом, в больнице окажется 70 больных с сочетанием болезней Y и X.\n",
    "\n",
    "Из 900 человек, страдающих болезнью Y, но не X, будут госпитализированы 50%, то есть 450 человек.\n",
    "\n",
    "Такой же расчет для болезни Z показывает, что в больницу попадет 52 человека с сочетанием болезней Z и X, а с болезнью Z, но не Х, — 180 человек.\n",
    "\n",
    "Исследователь, работающий в больнице в которую попали все госпитализированные, обнаружил следующую связь.\n",
    "\n",
    "_|С болезнью X|Без болезни X\n",
    "-|-|-\n",
    "Болезнь Y|70|450\n",
    "Болезнь Z|52|180\n",
    "\n",
    "Оцените статистическую значимость различий частоты болезни X среди страдающих болезнями Y и Z. Можно ли по этим данным судить о связи болезней Y и Z с болезнью X? (Приведенный пример заимствован из работы: D. Mainland. The risk of fallacious conclusions from autopsy data on the incidence of diseases with applications to heart disease. Am. Heart J., 45:644—654, 1953)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "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>С болезнью X</th>\n",
       "      <th>Без болезни X</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Болезнь Y</th>\n",
       "      <td>70</td>\n",
       "      <td>450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Болезнь Z</th>\n",
       "      <td>52</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           С болезнью X  Без болезни X\n",
       "Болезнь Y            70            450\n",
       "Болезнь Z            52            180"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "chi^2 = 8.812, v=1, P=0.003\n",
      "Отличия статистически значимы.\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[70, 450], [52, 180]],\n",
    "                  columns=[\"С болезнью X\", \"Без болезни X\"],\n",
    "                  index=[\"Болезнь Y\", \"Болезнь Z\"])\n",
    "\n",
    "display(df)\n",
    "sns.heatmap(df, annot=True, cmap=\"YlOrBr\", fmt='g')\n",
    "barplot_ct(df)\n",
    "\n",
    "chi2_test(df, silent=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Различия статистически значимы: в больнице среди страдающих болезнью Z доля больных Х выше, чем среди страдающих болезнью Y. \n",
    "\n",
    "Эти различия обусловлены исключительно разной вероятностью госпитализации при этих болезнях."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.7 ('kaggle')",
   "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": "3.9.7"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "c87114ee6b6cf384294f90c5f2b9fb4cb1520b5534c8c2c109b7c3f59f5d4f57"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}