From 712cccbbb9aeb4186509341acf326de0f6d4ed25 Mon Sep 17 00:00:00 2001 From: Mingming GUO Date: Thu, 19 Jun 2025 11:38:38 +0000 Subject: [PATCH] Delete _exercice_Mingming_GUO.ipynb --- module3/exo3/_exercice_Mingming_GUO.ipynb | 574 ---------------------- 1 file changed, 574 deletions(-) delete mode 100644 module3/exo3/_exercice_Mingming_GUO.ipynb diff --git a/module3/exo3/_exercice_Mingming_GUO.ipynb b/module3/exo3/_exercice_Mingming_GUO.ipynb deleted file mode 100644 index 9a20ebe..0000000 --- a/module3/exo3/_exercice_Mingming_GUO.ipynb +++ /dev/null @@ -1,574 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Subject 6: Around Simpson's Paradox\n", - "\n", - "**Project completed by:** Mingming GUO\n", - "\n", - "### Prerequisites : Averaging and ratio calculation, simple graphical presentation techniques, possibly logistic regression\n", - "\n", - "In 1972-1974, in Whickham, a town in the northeast of England located about 6.5 km southwest of Newcastle upon Tyne, a survey of one-sixth of the electorate was conducted to inform research on thyroid and heart diseases (Tunbridge et al., 1977). A follow-up study was carried out twenty years later (Vanderpump et al., 1995). Some results concerned smoking and examined whether individuals were still alive during the second study.\n", - "\n", - "For simplicity, we will restrict ourselves to women and among them, the 1314 categorized as \"currently smoking\" or \"never smoked\". There were relatively few women in the initial survey who had quit smoking (162) and very few for whom information was unavailable (18). Survival at 20 years was determined for all women in the first survey.\n", - "\n", - "The data are available in this CSV file. Each line indicates whether the person smokes or not, if she was alive or deceased at the time of the second study, and her age at the time of the first survey.\n", - "\n", - "\n", - "\n", - "\n", - "## Instructions\n", - "\n", - "1. Represent in a table the total number of women alive and deceased over the period according to their smoking habits. Calculate the mortality rate in each group (smokers / non-smokers) as the ratio between the number of women who died in the group and the total number of women in that group. You may propose a graphical representation of these data and calculate confidence intervals if you wish. Why is this result surprising?\n", - "\n", - "2. Repeat question 1 (numbers and mortality rates) by adding a new category related to age group. For example, consider the following age classes: 18-34 years, 34-54 years, 55-64 years, over 65 years. Why is this result surprising? Can you explain this paradox? You may also propose a graphical representation to support your explanation.\n", - "\n", - "3. To avoid bias caused by grouping into arbitrary and irregular age brackets, try to perform logistic regression. By introducing a variable Death taking values 1 or 0 indicating if the individual died during the 20-year period, we can study the model Death ~ Age to analyze the probability of death as a function of age for smokers and non-smokers.\n", - "\n", - "Do these regressions allow you to conclude on the harmfulness of smoking? You can propose a graphical representation of these regressions (including confidence regions).\n", - "\n", - "---" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## I) Import data" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
SmokerStatusAge
0YesAlive21.0
1YesAlive19.3
2NoDead57.5
3NoAlive47.1
4YesAlive81.4
\n", - "
" - ], - "text/plain": [ - " Smoker Status Age\n", - "0 Yes Alive 21.0\n", - "1 Yes Alive 19.3\n", - "2 No Dead 57.5\n", - "3 No Alive 47.1\n", - "4 Yes Alive 81.4" - ] - }, - "execution_count": 7, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "import matplotlib.pyplot as plt\n", - "import numpy as np\n", - "import pandas as pd\n", - "plt.style.use('ggplot')\n", - "\n", - "data_file = \"Subject6_smoking.csv\"\n", - "data = pd.read_csv(data_file)\n", - "data.head()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## II) Check and handle missing data" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
agealivesmoker
count1314.0000001314.0000001314.000000
mean47.3593610.7191780.442922
std19.1606670.4495720.496921
min18.0000000.0000000.000000
25%31.3000000.0000000.000000
50%44.8000001.0000000.000000
75%60.6000001.0000001.000000
max89.9000001.0000001.000000
\n", - "
" - ], - "text/plain": [ - " age alive smoker\n", - "count 1314.000000 1314.000000 1314.000000\n", - "mean 47.359361 0.719178 0.442922\n", - "std 19.160667 0.449572 0.496921\n", - "min 18.000000 0.000000 0.000000\n", - "25% 31.300000 0.000000 0.000000\n", - "50% 44.800000 1.000000 0.000000\n", - "75% 60.600000 1.000000 1.000000\n", - "max 89.900000 1.000000 1.000000" - ] - }, - "execution_count": 8, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df = pd.DataFrame({\n", - " \"smoker\": data[\"Smoker\"].apply(lambda val: 1 if val == \"Yes\" else 0),\n", - " \"alive\": data[\"Status\"].apply(lambda val: 1 if val == \"Alive\" else 0),\n", - " \"age\": data[\"Age\"]\n", - "})\n", - "\n", - "df.describe()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## III) Initial data visualization" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Total number of women in the study: 1314\n", - "Number of women who died during the study: 369\n", - "Number of women still alive after the study: 945\n", - "Number of non-smoking women: 732\n", - "Number of smoking women: 582\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/opt/conda/lib/python3.6/site-packages/numpy/core/fromnumeric.py:51: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n", - " return getattr(obj, method)(*args, **kwds)\n" - ] - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3oAAAGsCAYAAACLq9VjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3X1cVGX+//H3AHIzIopAEogUmqamJaGWlmgQmqWyW1lW3mTm9qVNq+27prnaZppbmWbZtxuNMtdSqzW3NjU0pc0yFOxGXcXbdFeTu7wB73Cu3x/9nG0EmRllBA6v5+PR4+GZ+ZzrfK7rOnOYT+fMOTZjjBEAAAAAwDL8ajoBAAAAAED1otADAAAAAIuh0AMAAAAAi6HQAwAAAACLodADAAAAAIuh0AMAAAAAi6HQA4A6aNy4cWrWrJlsNpveeuutmk7HadiwYUpNTfVZfG10ySWX6Omnn67pNGqFnj17asSIEc5lK8zvr1mtP+7Ut/4CVkOhB+CcDBs2TDabTTabTQEBAYqPj9cDDzygoqKimk7NY08//bQuueSSmk7Da2vXrtUzzzyj119/Xfv27dMdd9xxwXOYN2+ebDZbhddffPFFLVq0yON2vI2vTqmpqRo2bFiNbLu+qMn5BYD6LqCmEwBQd11//fVauHChysvLtX79eo0YMUJ79uzRJ598cs5tnjhxQoGBgdWYpfXk5+fLz89PAwYMqJHtnzhx4qzvNW7c2Ku2vI1H3cL8AkDN4YwegHMWGBio6OhoNW/eXAMGDNDDDz+spUuX6ujRo5Kkn376ScOGDVNUVJQaNWqk7t27Kzs727n+qlWrZLPZ9Mknn+i6665TcHCwXn/9dUnS+vXr1adPH4WFhSk0NFRdunTR2rVrnet+9tln6t69u0JCQhQbG6t7773X5Wzi6UuOXn/9dcXHxyssLEwDBgxQQUGBJOmtt97Sn/70J+3evdt5ZvLJJ5+UJM2fP19du3ZV48aNFRkZqZtvvllbt2516XteXp6uueYaBQcHq3Xr1nr//fcrXMJ35MgRjR49WrGxsbLb7erUqZM+/PBDt+P69ttvq127dgoKClLz5s01fvx4lZeXO/s1ePBgORwOZ95nY7PZ9OKLL+rWW29Vw4YNFRMToxdeeMEl5sUXX9RVV12l0NBQRUdH684779S+ffvcztHgwYOd27DZbM4zY5Vd6rVgwQJdffXVCg4OVkREhG666SaVlJRUGn96+YUXXnCO26233qrCwkJnTG5urm666SZddNFFCg0NVefOnbV06VKXbV5yySWaMGGCRo8eraZNm6pZs2Z67LHHdOrUKed2VqxYobffftvZh1WrVkmSpkyZooSEBAUFBSkqKkq9e/d27tNnc/ToUY0YMUJhYWGKjIzUmDFj5HA4JEkTJ05UmzZtKqxz7733qmfPnmdt87PPPlPPnj3VtGlTNW7cWMnJyfrmm29cYnbu3Km0tDQFBwerRYsWmjVrVoXLJ8vLy/Xkk0/q0ksvVXBwsNq3b6/XXnutyv5UxpNxP9Ov5/ezzz6Tv7+/9uzZ4xKzYMECBQcH6+eff5bk/rhxJk/b9XZeS0pKdMcdd6hhw4Zq1qyZxo8fL2NMhbiXXnpJl19+uYKDg3XZZZdp8uTJzs+r9Mv4P/XUU2rZsqWCgoIUGxurhx56yPm+J8eJJ554Qm3btpXdbldcXJweeOABHTx40Pn+oUOHdO+99yo6OlpBQUGKi4vTo48+6lWenvYXQB1iAOAcDB061KSkpLi8Nm3aNCPJHDp0yJSVlZm2bdua3/72tyYnJ8fk5+ebp59+2gQGBppNmzYZY4z5/PPPjSTTpk0b89FHH5kdO3aYPXv2mB9++MHY7XZz5513mpycHLN161Yzf/58s2bNGmOMMStWrDAhISFm5syZZuvWreabb74xPXv2NNdff71xOBzO/MLCwsydd95pvv/+e/Pll1+aFi1amCFDhhhjjCkrKzNjxowxzZs3N/v27TP79u0zhw8fNsYY8+abb5q///3vZtu2bSY3N9f069fPtGrVyhw/ftwYY0xpaamJjo42t9xyi/n222/NV199Za699loTEhJiJk2aZIwxxuFwmJ49e5rk5GTzxRdfmO3bt5vXXnvNNGjQwGRlZZ11XD/++GPj5+dnpkyZYrZs2WLee+8906RJEzN+/HhjjDE///yzmTFjhvH393fmfTaSTHh4uJk5c6bZsmWLc70PPvjAGTNjxgzz2WefmR07dpg1a9aYa6+91vTo0cP5fmVztHv3bvPyyy8bSc4cfv7550r3izfffNMEBASYp556ymzcuNF8++23ZsaMGaagoKDS+KFDh5pGjRqZfv36me+++858/vnnplWrVqZfv34uOb311ltm48aNZsuWLeaJJ54wDRo0MFu2bHHGxMfHmyZNmphnnnnGbN261bz33nvG39/fvPnmm85xvP76683AgQOdfTh+/Lj54IMPTKNGjcySJUvM7t27TV5enpk+fbopKys76zjHx8ebRo0amT/96U/mX//6l5k7d66x2+1m2rRpxhhj9uzZY/z9/c2qVauc6xw6dMg0bNjQzJs376ztfvjhh2bhwoVmy5Yt5ocffjD33XefCQ8PN4WFhcaYX/axK6+80nTp0sWsXbvW5OXlmZtuusmEhYWZ++67z2VMO3ToYJYtW2Z27Nhh3nvvPdO4cWMze/Zsl5j4+Piz5uLpuCcnJ1fY9un5PXXqlImNjTVTpkxxaffmm282AwcONMYYj44bZ/Kk3XOZ1/T0dNOyZUuzYsUK88MPP5i7777bNGrUyGV/nThxomnRooX58MMPzY4dO8wnn3xi4uLinJ9XY4wZMmSIiYqKMnPnzjXbtm0zX331lXnhhReMMZ4fJyZNmmSys7PNzp07TVZWlmnTpo3zWGaMMQ899JDp2LGj+frrr83u3bvNl19+aV5//XWv8vSkvwDqFgo9AOfkzC/oGzduNAkJCaZr167GGGMyMzNNbGysOXnypMt6vXr1MqNHjzbG/LeImDt3rkvMPffcYzp27GhOnTpV6baTk5PNmDFjXF7bvXu3kWTy8vKc+UVGRppjx445Y5555hkTHR3tXJ40aZLbL7fGGFNUVGQkmX/+85/GGGNef/1107BhQ2dxY4wxmzdvNpKchd7nn39ugoKCXGKMMebee+81AwYMOOu2rrvuOnP77be7vDZjxgwTHBzsLDQzMzONv7+/27wlmXvuucfltUGDBpnu3bufdZ3c3Fwjyezdu9fZj8rm6J133jGV/b/CM/eLuLg48+CDD551e5UVemeO7bJly4wks3Xr1rO207FjR/P00087l+Pj412KQ2OM6d27t7nzzjudyykpKWbo0KEuMS+88IK57LLLzIkTJ866rTPFx8eb6667zuW1sWPHmtjYWOdyv379zN133+1cfvXVV03Tpk3N0aNHPd7OqVOnTJMmTZzF4fLly40kk5+f74wpKioyISEhzmJrx44dxmazmc2bN7u09ec//9lceeWVzuXHH3/c3HDDDR7nctqZ415VoWeMMWPGjDFt27Z1Lv/0008mICDAfPzxx8YYz44blXHXrrfzmp+fbySZ5cuXO187fvy4iYmJcfantLTUhISEmE8//dRl3bfffts0btzYpZ1FixZVup1zPU58+OGHJjAw0HmM7N+/f4V9+TRv8qyqvwDqHi7dBHDOVq1apdDQUIWEhOiKK65QQkKC5s+fL0nKycnR/v371aRJE4WGhjr/++KLL5Sfn+/STpcuXVyW169fr5SUFPn5VX6IysnJ0YwZM1zabdeunSS5tN22bVsFBQU5l2NjY/XTTz+57deGDRv0m9/8RpdeeqkaNWqkFi1aSJJ2794tSdq0aZPatm3r8vujyy+/XE2aNHHJ8cSJE4qNjXXJc968eRX6/2sbN25Ujx49XF5LTk7WsWPHtH37dre5n+naa691We7evbs2bdrkXF61apV69+6tuLg4NWrUSNddd51LX087c448ceDAAe3Zs0dpaWlerdeuXTuXse3evbskafPmzZKkgoICZWRkOMc8NDRUGzdurJDzVVdd5bLsyfwPHDhQJ0+eVHx8vIYNG6Z33nlHhw8fdptzZeP873//W4cOHZIk/e53v9MHH3zgvGT1jTfe0ODBgxUcHHzWNnfu3KnBgwerVatWCgsLU1hYmA4ePOiyH0ZGRqpVq1bOdZo2bepymei6detkjFFSUpLLfjhlyhSX/fCZZ57RihUrquyjp+NelaFDh2rz5s3KycmRJL377ruKiIhQ7969JXl33PCmXW/n9fRnpFu3bs7XAgMD1blzZ+fyxo0bdfToUd16660uuf7ud7/TwYMHVVBQoNzcXEk662fA0+PEhx9+qB49eigmJkahoaG6++67deLECe3fv1+SlJGRoffff19XXHGFRo8erU8//dR56bAneXrSXwB1DzdjAXDOunbtqrffflsBAQG6+OKLXYoqh8Ohtm3b6m9/+1uF9ex2u8tyw4YNK8RU9dszh8OhMWPGOH8n9mvR0dHOf595Uxebzeb2NydlZWVKS0vTddddpzfffNPZXvv27V1uQlJVfqdzbNy4sfOL56+5u9nMmW2fztndNj3x6/7/+OOP6tu3rwYPHqwJEyYoMjJSe/fuVWpqaoUbrlQ2R56qjrx/bdiwYfrxxx/17LPP6tJLL1VISIjuvPPOCjlXNv+nv/yeTWxsrP71r3/p888/18qVKzVp0iSNGTNGa9euVVxcnMc5nrmf3XTTTWrWrJneeecd9ejRQ+vXr9fbb79dZRu33HKLIiMjNWvWLMXFxSkwMFDXXXed1/uhJK1Zs6bC587befF03KvStm1bJSUlae7cuercubPmzp2ru+66SwEBAc58PT1ueNOut/Pq7jhxOldJWrRokVq3bl3h/aZNm3rUhrvjxNq1a3X77bdr7Nixeu655xQeHq6vv/5aQ4cOdY5979699eOPP2rZsmVatWqV7rnnHnXo0EErVqzwKE9P+gug7uGMHoBzFhISolatWumSSy5xKfIkKSkpSTt27FBYWJhatWrl8l9MTEyV7V599dXKyso665fypKQkbdy4sUK7rVq1UmhoqMf5BwYGOm/OcdrmzZtVUFCgyZMnq1evXmrbtq1KSkpcvgi1a9dOmzdvdrkZwpYtW5w3fTid488//6xjx45VyPH0GcLKtG/fXqtXr3Z5LTs7WyEhIUpISPC4b6d9/fXXLstfffWV2rZtK+mXswlHjx7VjBkz1L17d7Vp08ajM57Sf7+Enjl+v3bRRRepefPmWrZsmVc5b9682XkmTPqlSJHkzDs7O1sZGRnq37+/OnTooIsvvlg7duzwahtS5fMvSUFBQerTp4+effZZff/99yorK9PixYurbKuycY6JiVFYWJgkyc/PTyNGjNAbb7yhN954Q926dVP79u3P2l5RUZE2bdqkxx9/XL1791a7du0UHBysAwcOOGPatWungoICbdu2zflaSUmJy42Drr76akm/FPVn7octW7assk9nqq5xHzJkiN577z19++23ys3N1dChQ53vnc9xo6p2Je/m9fTcnN73pF/uNvvrgqx9+/YKDg7Wjh07Kj0W+fv7KzExUZK0fPnySrfjyXHin//8pyIjI/X000+ra9euat26tfbu3VuhraZNm2rQoEF67bXX9Mknn2j16tXatGmTR3l60l8AdQ+FHgCfuPvuu3XppZfq5ptv1vLly7Vr1y7n89/cfWn+4x//qPz8fN19991at26dtm/frkWLFumrr76SJD311FP66KOP9Mgjj2jDhg3avn27li5dqvvuu8/t3RF/7dJLL9X+/fv11VdfqbCwUGVlZYqPj1dQUJBeeuklbd++XStWrNDo0aNdzn7cfffdCg0N1ZAhQ/Tdd99p7dq1uu+++xQSEuKMu+GGG5Samqrf/va3+tvf/qYdO3Zo/fr1eumll/TGG2+cNaexY8fqgw8+0NSpU7V161YtXLhQTz75pP7whz+c02MnPv74Y7388svKz8/XSy+9pAULFuiRRx6RJF122WWy2WyaNm2adu7cqcWLF+upp57yeOwkacmSJSooKNCRI0cqjZs4caJee+01TZo0SZs3b9bGjRv18ssvu9xF80w2m01DhgzRDz/8oOzsbD344IO6+eabddlll0mS2rRpo7/+9a/6/vvvtWHDBg0aNKjKgrOqPqxfv17bt29XYWGhTp48qTlz5uiNN97Qt99+q927d+uvf/2rDh8+7Lw0+Gw2bNigJ598Ulu3btX8+fP14osvOsf5tPvuu0//+te/NHv2bI0cObLK9sLDwxUVFaU33nhDW7du1VdffaVBgwYpJCTEGZOamqorr7xSQ4YMUU5Ojr799lsNHjxYAQEBzv2wVatWGj58uO6//36988472rZtm7799lu9+eab+stf/uJsa+zYsUpJSakyp+oa90GDBungwYMaNmyYOnbsqCuvvNL53vkcN6pq19t5bdWqlfr3768HH3xQn3/+uTZt2qQRI0a4XO4ZGhqqcePGady4cXr55Ze1ZcsWbdy4Ue+9957GjBnjbOfuu+9WRkaG5s2bp+3btysnJ0cvvviiJM+OE23atFFBQYHmzJmjHTt2aO7cuXrllVdc8n3iiSf04YcfasuWLcrPz9df//pXhYaGqkWLFh7n6a6/AOqgmvt5IIC6rLK7bp6psLDQPPDAAyYmJsY0aNDAxMTEmPT0dJObm2uM+e+NPvbs2VNh3bVr15qUlBRjt9tNaGio886Cp2VnZ5uUlBQTGhpq7Ha7ufzyy83o0aOdN3GoLL8zbyBy4sQJM2jQIBMeHm4kmYkTJxpjjFm0aJFp1aqVCQoKMldddZVZtWqV8ff3N5mZmc51c3NzTdeuXU1gYKBp1aqVWbRokYmKijLPP/+8M+b0nT0vueQS06BBA9OsWTPTu3dvs2LFiirH7a233jKXX365c8zGjRvncnMKb27GMn36dDNgwAATEhJioqOjzbPPPusS8/LLL5vmzZub4OBg0717d/Ppp58aSebzzz83xlQ9R6NHjzYXXXSRsdlszhtBVDbu8+bNMx07djSBgYGmadOmpm/fvqakpKTS+NPLzz33nImOjjbBwcEmPT3dHDhwwBnz3XffmWuvvdYEBweb+Ph4M2vWrAo3VomPj3feGOe0++67zyQnJzuXt2/fbq6//nrTsGFDZ58/+OADc+2115omTZqYkJAQ0759e5e7U1YmPj7ejBs3zgwbNsw0atTIhIeHm8cee8yUl5dXiE1PTzeNGzeu8m6Pp61atcp07NjRBAUFmdatW5v333/ftGzZ0rmfGvPLzVZSU1NNUFCQad68uXn55ZdN586dze9//3tnTHl5ufnLX/5i2rRpYxo0aGAiIiJMjx49zMKFC50xntx105Nxd3czll+PgySXz8tp7o4bVTlbu+cyr4WFheb22283drvdREZGmscff9wMGTKkQn9mz55trrzyShMUFGSaNGliunTpYl555RXn+ydOnDDjx4838fHxpkGDBiY2NtblxjKeHCfGjx9vLrroImO3281NN91k5s+fbySZnTt3GmOMeeqpp0z79u1Nw4YNTVhYmOnRo4f54osvvMrT0/4CqDtsxnBhNgCcr927d+uSSy7RkiVL1K9fv5pOR9IvZ8beeecd3XPPPTWdiseGDRumvXv3Kisrq6ZT8YkuXbqoa9eueumll3zS/uHDh9W8eXM9/fTTLs9qAwDUP9yMBQDOwbx58xQbG6tLL71Uu3fv1h//+EfFx8d7fYdJ1A8HDhzQRx99pNzcXL377rvV1u6SJUsUEBCgtm3b6sCBA/rzn/8sm82mgQMHVts2AAB1E4UeAJyDoqIiTZw4Uf/+97/VtGlTde/eXYsWLapwUxpAkpo1a6bw8HC9+OKLXt8EpSplZWV66qmntGvXLjVs2FBXX321/vnPf6pZs2bVtg0AQN3EpZsAAAAAYDHcdRMAAAAALIZCDwAAAAAsps79Ru8///lPTadwQURGRlb5nCnUDsxT3cA81Q3MU+3HHNUNzFPdwDzVDbVxnmJiYjyK44weAAAAAFgMhR4AAAAAWAyFHgAAAABYTJ37jR4AAACA2scYo2PHjsnhcMhms9V0OtXip59+0vHjxy/4do0x8vPzU3Bw8DmPJYUeAAAAgPN27NgxNWjQQAEB1ikxAgIC5O/vXyPbLi8v17FjxxQSEnJO63PpJgAAAIDz5nA4LFXk1bSAgAA5HI5zXp9CDwAAAMB5s8rlmrXJ+YwphR4AAAAAWAznVgEAAABUO8eS+dXanl//uzyK+/TTTzVixAitXr1arVq10p49ezR06FCtXLlS3377rd5//31NmjSpWnOrjTijBwAAAMAyFi9erC5duuijjz6q8N6VV15ZL4o8iUIPAAAAgEWUlpZq3bp1ev755yst9NasWaMhQ4bI4XCoa9euOnjwoPO97t27q6CgQEVFRbr//vvVt29f9e7dWzk5OReyC9WGQg8AAACAJSxdulQ9e/ZUy5Yt1aRJE33//feVxvn5+al3795aunSpJCk3N1fNmzdXVFSUJkyYoPvvv1//+Mc/NGfOHD322GMXsgvVhkIPAAAAgCUsXrxYAwYMkCQNGDBAixcvPmtsv379tGTJEknSRx99pP79+0uSvvjiCz3xxBO68cYbNWTIEB05ckRHjhzxffLVjJuxAAAAAKjziouLtWbNGm3ZskU2m02nTp2SzWbT0KFDK41PSkrSrl27VFRUpGXLlmn06NGSfnke4JIlSxQSEqKAgACVl5dfyG5UG87oAQAAAKjzPvnkE91666365ptvtHbtWq1bt04tWrTQvn37Ko232Wzq06ePnnzySV122WVq2rSpJCk5OVlvvfWWM+6HH364EOlXO87oAQAAAKh2nj4Oobp89NFHevDBB11e69u3r1566aWzrtO/f3/17dtX06dPd742adIkjRs3TqmpqTp16pS6dOmiv/zlLz7L21dsxhjjLmjDhg3KzMyUw+FQSkqK0tPTXd43xigzM1N5eXkKCgpSRkaGEhISnO87HA49/vjjatq0qR5//HFJ0pEjRzR9+nQVFBQoKipKjzzyiEJDQ90m/J///MfbPtZJkZGRKiwsrOk04AbzVDcwT3UD81T71bc5eve7Ao9jB3WM8mEm3qlv81RXWXGeysrKZLfbazqNalXTl25WNqYxMTEerev20k2Hw6E5c+Zo3Lhxmj59ur788kvt3bvXJSYvL0/79+/XzJkzNXLkSM2ePdvl/X/84x+KjY11eW3x4sXq0KGDZs6cqQ4dOlT5Q0kAAAAAgOfcFnrbtm1TdHS0mjVrpoCAAHXr1q3CsyTWrVunHj16yGazqXXr1iotLVVJSYkkqaioSLm5uUpJSXFZJycnR8nJyZJ+uQ62rj6fAgAAAABqG7e/0SsuLlZERIRzOSIiQvn5+RViIiMjXWKKi4sVHh6ut956S/fcc4+OHj3qss7BgwcVHh4uSQoPD9ehQ4cq3X5WVpaysrIkSVOnTnXZjpUFBATUm77WZcxT3cA81Q3MU+1X3+bIbi/1OLY2jUt9m6e6yorz9NNPPykgwHq3AKnJPgUFBZ3zfuI268p+wmez2TyKWb9+vRo3bqyEhARt3LjxnBJMTU1Vamqqc9lq1zKfjRWv27Yi5qluYJ7qBuap9qtvc1RWVuZxbG0al/o2T3WVFefp+PHj8vf3r+k0qlVN/0bv+PHjFfYTT3+j57bQi4iIUFFRkXO5qKjIeSbu1zG/TuB0zNdff61169YpLy9PJ06c0NGjRzVz5kyNGjVKjRs3VklJicLDw1VSUqKwsDCPEgYAAAAAVM3tb/Ratmypffv26cCBAyovL9eaNWuUlJTkEpOUlKTs7GwZY7R161bZ7XaFh4frrrvu0quvvqpZs2bp4Ycf1hVXXKFRo0Y511m9erUkafXq1ercubMPugcAAAAA9Y/bM3r+/v4aPny4Jk+eLIfDoV69eikuLk7Lly+XJKWlpalTp07Kzc3VqFGjFBgYqIyMDLcbTk9P1/Tp07Vy5UpFRkbq0UcfPf/eAAAAAKgVvHlEiSc8eYxJXFycLr/8cpWXl8vf31+333677r//fvn5uT2/VaOmTZumhg0b6oEHHqi2Nj36ZWFiYqISExNdXktLS3P+22azacSIEVW20b59e7Vv39653KhRI02YMMGbXAEAAADgrIKDg/XZZ59J+uW3sw8++KAOHz6sxx57rIYzu/Bqd2kLAAAAAOcgMjJSzz77rDIzM2WM0alTpzRp0iT17dtXqampeuedd5yxr7zyilJSUpSamqopU6ZIknbt2qU777xTffr00W9+8xtt27ZNkrR8+XLdcsstSktL0x133KGCgl/OXH711Ve68cYbdeONNyotLU1HjhyRJP3f//2fc5vPP/+8c5svvviirr/+et1xxx3avn17tfffevc/BQAAAABJ8fHxMsaosLBQy5YtU6NGjfSPf/xDx48fV3p6upKTk7Vt2zYtXbpUH3/8sUJCQpzPA//jH/+o559/Xi1atFBubq7Gjh2rRYsWqUuXLvr73/8um82m+fPn65VXXtHEiRP16quvasqUKercubNKS0sVFBSk1atXa+fOnfrkk09kjNGwYcP09ddfy263a8mSJVq+fLnKy8vVp08fdezYsVr7TqEHAAAAwLJOPwpu9erV2rx5sz755BNJ0uHDh7Vz50598cUXuuOOOxQSEiLpl2d8l5aWav369RoxYoRz/RMnTkiS9u3bp//5n//RgQMHdOLECbVo0UKS1LlzZ/35z3/Wb37zG910002KiYnR6tWrtXr1aufP3srKyrRz504dOXJEffr0cW7zxhtvrPZ+U+gBAAAAsKTdu3fLz8/P+dDxp59+Wj179nSJ+fzzzys8J9zhcCgsLEwrV66s8By9P/3pTxo5cqTS0tK0Zs0avfDCC5Kk3//+90pJSdHKlSvVr18/LViwQMYY/f73v9fgwYNd2njjjTcqbLO68Rs9AAAAAJZTVFSkxx9/XPfee69sNpuSk5M1d+5cnTx5UpK0fft2lZWVKTk5We+9956OHj0qSSopKVGjRo0UFxc3rb/3AAAgAElEQVSnJUuWSPrlrODGjRslSYcOHVJ0dLQkadGiRc7t7dq1S23bttWDDz6oK6+8Utu2bVPPnj21YMEClZaWSvrlbGBhYaGuueYaLV26VEePHtWRI0ecN5CpTpzRAwAAAFDtPHkcQnU7duyYbrzxRufjFW677TaNHDlSknTXXXdpz5496tOnj4wxatq0qd5880316tVLGzdu1E033aQGDRrohhtu0NixY/Xyyy9r3LhxeuGFF1ReXq4BAwaoffv2+sMf/qDf/e53io6OVmJiovbs2SNJmj17ttasWSM/Pz+1bt1avXr1UlBQkPLz89W/f39Jkt1u10svvaQOHTqoX79+SktLU/PmzdW1a9dqHwubOX3RaR3xn//8p6ZTuCAiIyNVWFhY02nADeapbmCe6gbmqfarb3PkzTPAauIL7dnUt3mqq6w4T2VlZbLb7TWdRrUKCAiocOnmhVTZmMbExHi0LpduAgAAAIDFUOgBAAAAgMVQ6AEAAAA4b3XsF2F1wvmMKYUeAAAAgPPm5+dXo79ns5ry8nL5+Z17ucZdNwEAAACct+DgYB07dkzHjx/3+TPiLpSgoCAdP378gm/XGCM/Pz8FBwefcxsUegAAAADOm81mU0hISE2nUa3q8t1RuXQTAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALCagphNA5Y68N1uOsjKPYv363+XjbAAAAADUJZzRAwAAAACLodADAAAAAIvx6NLNDRs2KDMzUw6HQykpKUpPT3d53xijzMxM5eXlKSgoSBkZGUpISNCJEyc0ceJElZeX69SpU7rmmms0cOBASdLChQu1YsUKhYWFSZIGDRqkxMTEau4eAAAAANQ/bgs9h8OhOXPmaPz48YqIiNDYsWOVlJSk5s2bO2Py8vK0f/9+zZw5U/n5+Zo9e7amTJmiBg0aaOLEiQoODlZ5ebkmTJigq666Sq1bt5Yk3Xzzzerfv7/vegcAAAAA9ZDbSze3bdum6OhoNWvWTAEBAerWrZtycnJcYtatW6cePXrIZrOpdevWKi0tVUlJiWw2m4KDgyVJp06d0qlTp2Sz2XzTEwAAAACAJA/O6BUXFysiIsK5HBERofz8/AoxkZGRLjHFxcUKDw+Xw+HQmDFjtH//fvXu3VuXXXaZM27ZsmXKzs5WQkKChgwZotDQ0Arbz8rKUlZWliRp6tSpLtuxsjI/P9ntdo9iQ+vJmNRGAQEB9WafrMuYp7qBear96tsc2e2lHsfWpnGpb/NUVzFPdUNdnie3hZ4xpsJrZ56VqyrGz89Pzz33nEpLS/X888/rxx9/VIsWLZSWlqbbbrtNkrRgwQLNnTtXGRkZFdpJTU1Vamqqc7mwsNBdypYQ7HCozMPHKxyrJ2NSG0VGRtabfbIuY57qBuap9quuOXr3u4JqyMbVoI5R1d6mp3+Hpdr1/YTPUt3APFUvT48r3h4rauM8xcTEeBTn9tLNiIgIFRUVOZeLiooUHh5eIebXA1BZTMOGDdWuXTtt2LBBktSkSRP5+fnJz89PKSkp2r59u0cJAwAAAACq5rbQa9mypfbt26cDBw6ovLxca9asUVJSkktMUlKSsrOzZYzR1q1bZbfbFR4erkOHDqm09JfLHk6cOKHvv/9esbGxkqSSkhLn+t98843i4uKqs18AAAAAUG+5vXTT399fw4cP1+TJk+VwONSrVy/FxcVp+fLlkqS0tDR16tRJubm5GjVqlAIDA52XYJaUlGjWrFlyOBwyxujaa6/V1VdfLUmaN2+edu3aJZvNpqioKI0cOdKH3QQAAACA+sOj5+glJiZWeMZdWlqa8982m00jRoyosF58fLyeffbZStt86KGHvMkTAAAAAOAht5duAgAAAADqFgo9AAAAALAYCj0AAAAAsBgKPQAAAACwGAo9AAAAALAYCj0AAAAAsBgKPQAAAACwGI+eowfgwnMsme9RnF//u3ycCQAAAOoazugBAAAAgMVQ6AEAAACAxVDoAQAAAIDFUOgBAAAAgMVQ6AEAAACAxVDoAQAAAIDFUOgBAAAAgMVQ6AEAAACAxVDoAQAAAIDFUOgBAAAAgMVQ6AEAAACAxQTUdAIAANQn735X4HHsoI5RPswEAKoHx7XaiTN6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMQE1nQDqF8eS+R7F+fW/y8eZAAAAANblUaG3YcMGZWZmyuFwKCUlRenp6S7vG2OUmZmpvLw8BQUFKSMjQwkJCTpx4oQmTpyo8vJynTp1Stdcc40GDhwoSTpy5IimT5+ugoICRUVF6ZFHHlFoaGj19xAAAAAA6hm3l246HA7NmTNH48aN0/Tp0/Xll19q7969LjF5eXnav3+/Zs6cqZEjR2r27NmSpAYNGmjixIl67rnn9Oyzz2rDhg3aunWrJGnx4sXq0KGDZs6cqQ4dOmjx4sU+6B4AAAAA1D9uC71t27YpOjpazZo1U0BAgLp166acnByXmHXr1qlHjx6y2Wxq3bq1SktLVVJSIpvNpuDgYEnSqVOndOrUKdlsNklSTk6OkpOTJUnJyckV2gQAAAAAnBu3l24WFxcrIiLCuRwREaH8/PwKMZGRkS4xxcXFCg8Pl8Ph0JgxY7R//3717t1bl112mSTp4MGDCg8PlySFh4fr0KFDlW4/KytLWVlZkqSpU6e6bMfKyvz8ZLfbPYoNrUNjcsRifQoICPDZPmm1sapJvpwnVJ/6Mk92e6nHsbVtPKprjrwZA0/5Yqzq6lzVl89SXWeleaoNnxVPc/B2+3V5ntwWesaYCq+dPivnSYyfn5+ee+45lZaW6vnnn9ePP/6oFi1aeJxgamqqUlNTncuFhYUer1uXBTscKisr8yj2WB0aE4fF+hQZGemzfdJqY1WTfDlPqD71ZZ48PbZLte9vXnXNkTdj4ClfjFVdnav68lmq66w0T7Xhs+JpDt5uvzbOU0xMjEdxbi/djIiIUFFRkXO5qKjIeSbu1zG/HoDKYho2bKh27dppw4YNkqTGjRurpKREklRSUqKwsDCPEgYAAAAAVM1todeyZUvt27dPBw4cUHl5udasWaOkpCSXmKSkJGVnZ8sYo61bt8putzsvxywt/eU06okTJ/T9998rNjbWuc7q1aslSatXr1bnzp2ru28AAAAAUC+5vXTT399fw4cP1+TJk+VwONSrVy/FxcVp+fLlkqS0tDR16tRJubm5GjVqlAIDA5WRkSHplzN1s2bNksPhkDFG1157ra6++mpJUnp6uqZPn66VK1cqMjJSjz76qA+7CQCAtb37XYHHsYM6RvkwE6D6eLNfe6ou7f+e9t+bPnGsqD88eo5eYmKiEhMTXV5LS0tz/ttms2nEiBEV1ouPj9ezzz5baZuNGjXShAkTvMkVAAAAAOABjwo91D+OJfM9jvXrf5cPMwEAAADgLbe/0QMAAAAA1C0UegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMTwwHTgLTx4af8Rul1LTL0A2AAAAgOc4owcAAAAAFkOhBwAAAAAWQ6EHAAAAABZDoQcAAAAAFkOhBwAAAAAWQ6EHAAAAABZDoQcAAAAAFkOhBwAAAAAWQ6EHAAAAABZDoQcAAAAAFkOhBwAAAAAWQ6EHAAAAABYTUNMJAACA2und7wpclu32UpWVlVUaO6hj1IVICdXgzHk9G+a07oyVp3mifuGMHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMd92sBo4l8z2O9et/lw8zsQ7GFAAAADh3nNEDAAAAAIuh0AMAAAAAi6HQAwAAAACLodADAAAAAIuh0AMAAAAAi6HQAwAAAACLodADAAAAAIvx6Dl6GzZsUGZmphwOh1JSUpSenu7yvjFGmZmZysvLU1BQkDIyMpSQkKDCwkLNmjVLP//8s2w2m1JTU9W3b19J0sKFC7VixQqFhYVJkgYNGqTExMRq7h4AAAAA1D9uCz2Hw6E5c+Zo/PjxioiI0NixY5WUlKTmzZs7Y/Ly8rR//37NnDlT+fn5mj17tqZMmSJ/f38NHjxYCQkJOnr0qB5//HF17NjRue7NN9+s/v37+653tZDHDwK3232biIXwcHUAdcHpY5U5frHbWFubK3ydDgDA4txeurlt2zZFR0erWbNmCggIULdu3ZSTk+MSs27dOvXo0UM2m02tW7dWaWmpSkpKFB4eroSEBElSSEiIYmNjVVxc7JueAAAAAAAkeXBGr7i4WBEREc7liIgI5efnV4iJjIx0iSkuLlZ4eLjztQMHDmjnzp1q1aqV87Vly5YpOztbCQkJGjJkiEJDQytsPysrS1lZWZKkqVOnumyntjjig7Nvfn5+snvYbqgPxsSbPnmzfV+MlTeqO1c/Pz+f7ZOejpUv5t9qAgICauWxA66sPk+nP9MNHA3cxgb+/1hvxsNuL/U41tN2z2yzqr9NvsrVU77Yd3wxpheCu8+Sp/2qDX2q6X3Fl2PlyTHPF/33xrkeK6qjTW/5aq7q8t8mt4WeMabCazabzauYY8eOadq0aRo2bJjzD0RaWppuu+02SdKCBQs0d+5cZWRkVGgnNTVVqampzuXCwkJ3KV9wjrKyam/TbrerzMN2j/lgTLzpkzfb98VYeaO6c7Xb7T7bJz0dK1/Mv9VERkbWymMHXFl9nk5/pk+ePOk2tvz/x3ozHp7+zfCm3TPbrOpvk69y9ZQv9h1fjOmF4O6z5Gm/akOfanpf8eVYeXLM80X/vXGux4rqaNNbvpqr2vi3KSYmxqM4t5duRkREqKioyLlcVFTkcqbudMyvB+DXMeXl5Zo2bZquv/56de3a1RnTpEkT+fn5yc/PTykpKdq+fbtHCQMAAAAAqua20GvZsqX27dunAwcOqLy8XGvWrFFSUpJLTFJSkrKzs2WM0datW2W32xUeHi5jjF599VXFxsbqlltucVmnpKTE+e9vvvlGcXFx1dQlAAAAAKjf3F666e/vr+HDh2vy5MlyOBzq1auX4uLitHz5ckm/XILZqVMn5ebmatSoUQoMDHRegrllyxZlZ2erRYsW+t///V9J/32Mwrx587Rr1y7ZbDZFRUVp5MiRPuwmAAAAANQfHj1HLzExscIz7tLS0pz/ttlsGjFiRIX1Lr/8ci1cuLDSNh966CFv8gQAAAAAeMjtpZsAAAAAgLrFozN6AAAAAOqXd78rqNE2B3WMqvbt1ycUejhvjiXzazqFes2b8ffrf1ed2b6n7fqiT7VBTc+rFTGmAID6hEs3AQAAAMBiKPQAAAAAwGIo9AAAAADAYij0AAAAAMBiKPQAAAAAwGIo9AAAAADAYij0AAAAAMBiKPQAAAAAwGJ4YDpwnngIs+e8GSsrqu/99xXGFQCAijijBwAAAAAWQ6EHAAAAABZDoQcAAAAAFkOhBwAAAAAWQ6EHAAAAABZDoQcAAAAAFkOhBwAAAAAWQ6EHAAAAABZDoQcAAAAAFhNQ0wngwnIsmV/TKQAAAADwMc7oAQAAAIDFUOgBAAAAgMVQ6AEAAACAxVDoAQAAAIDFcDMWAABqqXe/K6jpFOAhb+ZqUMcoH2YCd/hcob7gjB4AAAAAWAyFHgAAAABYDIUeAAAAAFgMv9ED6hHHkvkex/r1v8uHmdQMb/rvKSuOk7c8HVfGCgCAC4czegAAAABgMRR6AAAAAGAxFHoAAAAAYDEUegAAAABgMRR6AAAAAGAxHt11c8OGDcrMzJTD4VBKSorS09Nd3jfGKDMzU3l5eQoKClJGRoYSEhJUWFioWbNm6eeff5bNZlNqaqr69u0rSTpy5IimT5+ugoICRUVF6ZFHHlFoaGj19xAAAAAA6hm3Z/QcDofmzJmjcePGafr06fryyy+1d+9el5i8vDzt379fM2fO1MiRIzV79mxJkr+/vwYPHqzp06dr8uTJWrZsmXPdxYsXq0OHDpo5c6Y6dOigxYsX+6B7AAAAAFD/uC30tm3bpujoaDVr1kwBAQHq1q2bcnJyXGLWrVunHj16yGazqXXr1iotLVVJSYnCw8OVkJAgSQoJCVFsbKyKi4slSTk5OUpOTpYkJScnV2gTAAAAAHBu3F66WVxcrIiICOdyRESE8vPzK8RERka6xBQXFys8PNz52oEDB7Rz5061atVKknTw4EHn++Hh4Tp06FCl28/KylJWVpYkaerUqS7bqS2O2O3V3qafn5/sHrYb6sWY+CLXuqS6x8qbefLF9n3J01zrQp4BAQGKjIz0Sa61YU69ycEXPO2XuzxPz5Ovtu8NX4zp6TwbOBq4jQ308efK03G220tdlqs65nkzd2e2Wx188f3AF3lKvsn119x9ljztV234zuWrOahu5zJWnhzz6kr/fcUXxxVv5+pc/zbVBm4LPWNMhddsNptXMceOHdO0adM0bNgwr74US1JqaqpSU1Ody4WFhV6tfyE4ysqqvU273a4yD9s95sWY+CLXuqS6x8qbefLF9n3J01zrQp6RkZEqLCz0Sa61YU69ycEXPO2XuzxPz5Ovtu8NX4zp6TxPnjzpNrbcx58rT8f5zONbVcc8b+bOm+Omp3zx/cAXeUq+/y7j7rPkab9qw3cuX81BdTuXsfLkmFdX+u8rvjiueDtX5/q3yZdiYmI8inN76WZERISKioqcy0VFRS5n6k7H/HoAfh1TXl6uadOm6frrr1fXrl2dMY0bN1ZJSYkkqaSkRGFhYR4lDAAAAAComttCr2XLltq3b58OHDig8vJyrVmzRklJSS4xSUlJys7OljFGW7duld1uV3h4uIwxevXVVxUbG6tbbrmlwjqrV6+WJK1evVqdO3euxm4BAAAAQP3l9tJNf39/DR8+XJMnT5bD4VCvXr0UFxen5cuXS5LS0tLUqVMn5ebmatSoUQoMDFRGRoYkacuWLcrOzlaLFi30v//7v5KkQYMGKTExUenp6Zo+fbpWrlypyMhIPfrooz7sJgAAAADUHx49Ry8xMVGJiYkur6WlpTn/bbPZNGLEiArrXX755Vq4cGGlbTZq1EgTJkzwJlcAAAAAgAfcXroJAAAAAKhbKPQAAAAAwGIo9AAAAADAYij0AAAAAMBiPLoZC2o3x5L5NZ0CPMRcVT9PxvSI3V7jD3aHd/isWNPpeTXHL64yztbmiguRDgBYGoUegCq9+11Ble+f/sJ2Z9C+C5EOgGrg7nPtqzbdFXgXii/6X995M6aDOkb5MBMAp3HpJgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDM/RQ73CQ5hhRZ7u137976r2NlHzzJYfqr1NHlgOAHUfZ/QAAAAAwGIo9AAAAADAYij0AAAAAMBiKPQAAAAAwGIo9AAAAADAYij0AAAAAMBiKPQAAAAAwGIo9AAAAADAYij0AAAAAMBiAmo6AQC1k2PJfEmSOX5xDWcCqzi9T53NEbtdjrKyC5QNAADWxhk9AAAAALAYCj0AAAAAsBgKPQAAAACwGAo9AAAAALAYCj0AAAAAsBjuugkAQDV497uCKt/nDrY4zd2+ctqgjlE+zqRmeNr/uuRc+mS3l6qMOw1fcN7MVV3/DHJGDwAAAAAshkIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALMajxyts2LBBmZmZcjgcSklJUXp6usv7xhhlZmYqLy9PQUFBysjIUEJCgiTplVdeUW5urho3bqxp06Y511m4cKFWrFihsLAwSdKgQYOUmJhYXf0CAAAAgHrLbaHncDg0Z84cjR8/XhERERo7dqySkpLUvHlzZ0xeXp7279+vmTNnKj8/X7Nnz9aUKVMkST179lSfPn00a9asCm3ffPPN6t+/fzV2BwAAAADg9tLNbdu2KTo6Ws2aNVNAQIC6deumnJwcl5h169apR48estlsat26tUpLS1VSUiJJateunUJDQ32TPQAAAACgArdn9IqLixUREeFcjoiIUH5+foWYyMhIl5ji4mKFh4dX2fayZcuUnZ2thIQEDRkypNKCMCsrS1lZWZKkqVOnumyntjhit1d7m35+frL7oF1Ur/owTw0cDTyKq83j4Mt5CvXimOSLY4U3anuutenz5M1YnWa3l1b5/okGnn2WJCnQi3Hwpt3z3f75zpGnuXrT/9O8+X7gbq5qi3P9zhMQEFDlup7234pjWpvUpmNebVXT+2BkZKTbz1Nt5rbQM8ZUeM1ms3kdc6a0tDTddtttkqQFCxZo7ty5ysjIqBCXmpqq1NRU53JhYaG7lC84R1lZtbdpt9tV5oN2Ub3qwzydPHnSo7jaPA6+nKdjXhyTfHGs8EZtz7U2fZ68GavT3OVuPPwsSVK5F+PgTbvnu/3znSNPc/Wm/6d58/2gtuxn7pzrd57IyMgq1/W0/1Yc09qkNh3zaqua3gcLCwvdfp5qQkxMjEdxbi/djIiIUFFRkXO5qKiowpm6iIgIlwGoLOZMTZo0kZ+fn/z8/JSSkqLt27d7lDAAAAAAoGpuz+i1bNlS+/bt04EDB9S0aVOtWbNGo0aNcolJSkrS0qVL1b17d+Xn58tut7st9EpKSpwx33zzjeLi4s6jGwBQMxxL5td0Ch6rS7nWd2bLD7Vy+ycaNKhwVs7W5ooLkRIAwEtuCz1/f38NHz5ckydPlsPhUK9evRQXF6fly5dL+uUSzE6dOik3N1ejRo1SYGCgyyWYM2bM0KZNm3T48GE98MADGjhwoG644QbNmzdPu3btks1mU1RUlEaOHOm7XgIAAABAPeLRc/QSExMrPOMuLS3N+W+bzaYRI0ZUuu7DDz9c6esPPfSQpzkCAAAAALzg9jd6AAAAAIC6hUIPAAAAACyGQg8AAAAALIZCDwAAAAAshkIPAAAAACyGQg8AAAAALMajxysAAABUxhcPd/emTR7YDgCV44weAAAAAFgMZ/QA1GrvHb/Yo7g7g/bV6PZ9xVf9AgCgtnv3u4KaTqFO44weAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWAyFHgAAAABYDIUeAAAAAFgMhR4AAAAAWExATScAwBreO36xx7F3Bu3zYSZwx5u58gbzipr07ncFNZ1CtfOmT4M6RvkwEwB1EWf0AAAAAMBiKPQAAAAAwGIo9AAAAADAYjz6jd6GDRuUmZkph8OhlJQUpaenu7xvjFFmZqby8vIUFBSkjIwMJSQkSJJeeeUV5ebmqnHjxpo2bZpznSNHjmj69OkqKChQVFSUHnnkEYWGhlZj1wAAAACgfnJ7Rs/hcGjOnDkaN26cpk+fri+//FJ79+51icnLy9P+/fs1c+ZMjRw5UrNnz3a+17NnT40bN65Cu4sXL1aHDh00c+ZMdejQQYsXL66G7gAAAAAA3BZ627ZtU3R0tJo1a6aAgAB169ZNOTk5LjHr1q1Tjx49ZLPZ1Lp1a5WWlqqkpESS1K5du0rP1OXk5Cg5OVmSlJycXKFNAAAAAMC5cXvpZnFxsSIiIpzLERERys/PrxATGRnpElNcXKzw8PCztnvw4EHn++Hh4Tp06FClcVlZWcrKypIkTZ061WU7tcURu73a2/Tz85PdB+2ietWHeWrgaFDtbXozZp5uv6o2z2eefNF/b/hi//JVn84319r0eQo9h781dntple+faFCz+1J1sNlsalDL+hFYS/aZmvbr70cBAQFVfl9yt69W1qY7nraJ/6pNxzxULjIy0u3nqTZzW+gZYyq8ZrPZvI45V6mpqUpNTXUuFxYWVku71clRVlbtbdrtdpX5oF1Ur/owTydPnqz2Nr0ZM0+3X1Wb5zNPvui/N3yxf/mqT+eba236PB07h7817nI3NbwvVYcGDRrU+GfiTOW1ZJ+pab/+fhQZGVnl9yVPP2fefOeqLZ/duqQ2HfNQucLCQrefp5oQExPjUZzbSzcjIiJUVFTkXC4qKqpwpi4iIsJlACqLOVPjxo2dl3eWlJQoLCzMo4QBAAAAAFVzW+i1bNlS+/bt04EDB1ReXq41a9YoKSnJJSYpKUnZ2dkyxmjr1q2y2+1uC72kpCStXr1akrR69Wp17tz5PLoBAAAAADjN7aWb/v7+Gj58uCZPniyHw6FevXopLi5Oy5cvlySlpaWpU6dOys3N1ahRoxQYGKiMjAzn+jNmzNCmTZt0+PBhPfDAAxo4cKBuuOEGpaena/r06Vq5cqUiIyP16KOP+q6XAAAAAFCPePQcvcTERCUmJrq8lpaW5vy3zWbTiBEjKl334YcfrvT1Ro0aacKECZ7mCQBAreRYMl+SZI5fXMOZAADwX24v3QQAAMD/a+/+Y5uq/z2Ov85+Wxag6wZzMNTOYTAQIikBQVhkE2/EH8QQIoQQJARkRCLmGqdRNBHmEoHFycz+gIwY4kUS47wSI2QYtwT+2IAuIlE2CD9EwLGtjm2wja69f3lqYmAAAA6qSURBVJBvL2O6drLSc86ej7/W7vP5nM/nvNutr5zTcwDAWgh6AAAAAGAzBD0AAAAAsBmCHgAAAADYDEEPAAAAAGyGoAcAAAAANkPQAwAAAACbIegBAAAAgM0Q9AAAAADAZgh6AAAAAGAzBD0AAAAAsJmEWE8AwP23r+fBWE8BERpKrV5JvhLFmQyff1pTYiBRt27d6vecVdYExNr//Hwt9LPD0aUbN24M65gArIcjegAAAABgMwQ9AAAAALAZgh4AAAAA2AxBDwAAAABshqAHAAAAADZD0AMAAAAAm+H2CgAA3CXwv1/GegoAANwTjugBAAAAgM0Q9AAAAADAZgh6AAAAAGAzBD0AAAAAsBmCHgAAAADYDEEPAAAAAGyGoAcAAAAANkPQAwAAAACb4YbpAADAsoKnf4m4rfHY1CjOBADMhSN6AAAAAGAzBD0AAAAAsBmCHgAAAADYDEEPAAAAAGyGoAcAAAAANkPQAwAAAACbIegBAAAAgM1wHz2MSPt6HhyWcRIDibp165Yk6ZXkK8My5t0inWu0tm8Vg+2nO+sE+4rGe2W4/lYAAHC/RRT0GhoaVFlZqUAgoPz8fC1evLjf74PBoCorK+X1epWcnKzCwkK53e5B++7fv1+HDx/W6NGjJUnLli3TjBkzhnNtAAAAADAihQ16gUBAu3fv1nvvvSeXy6V33nlHHo9HEydODLXxer26evWqysrK1NTUpF27dqm4uDhs30WLFunFF1+M3uoAAAAAYAQK+x29M2fOKDMzU+PHj1dCQoLmzJmj+vr6fm2OHTum+fPnyzAMTZ48WV1dXfL5fBH1BQAAAAAMr7BH9Nra2uRyuUKPXS6XmpqaBrRJT0/v16atrS1s34MHD6q2tlZut1srV65UamrqgO1XV1erurpaklRSUtJvO2bR6XAM+5hxcXFyRGFc3JYYSByWcQzDUGLi7bGiVa9I5zqU7Q/X+v+t+z3XO+tkZ5Hu12jV/163/3d1isZrxUrvFbOx+nspaYT8X+UzhDVQJ/NLT09XQkKCKfNHJMIGvWAwOOA5wzAiajNY34ULF2rJkiWSpK+++kpffPGFCgsLB7QvKChQQUFB6HFLS0u4Kd93gRs3hn1Mh8OhG1EYF7cN14U5EhP//yIf0apXpHMdyvZjfWGS+z3XO+tkZ5Hu12jti3vd/t/VKRqvFSu9V8zG6u8l/wj5v8pnCGugTubX0tKi9PR00+WPrKysiNqFPXXT5XKptbU19Li1tVVOp3NAmzt3wH/aDNZ37NixiouLU1xcnPLz83X27NmIJgwAAAAAGFzYoJeTk6MrV66oublZfr9fR48elcfj6dfG4/GotrZWwWBQjY2Ncjgccjqdg/b1+Xyh/nV1dcrOzh7mpQEAAADAyBT21M34+HitXr1aW7duVSAQ0NNPP63s7GwdOnRI0u1TMJ944gmdOHFCGzduVFJSUugUzH/qK0l79+7V+fPnZRiGMjIytHbt2iguEwAAAABGjojuozdjxowB97hbuHBh6GfDMLRmzZqI+0rS66+/PpR5AgAAAAAiFPbUTQAAAACAtUR0RA8AAAC4F8HTv0TUznhsapRnAowMHNEDAAAAAJsh6AEAAACAzRD0AAAAAMBmCHoAAAAAYDMEPQAAAACwGYIeAAAAANgMQQ8AAAAAbIagBwAAAAA2ww3TAQAA/qVIbwIuWedG4HZcE6JT12i9ViIdl9ff4DiiBwAAAAA2Q9ADAAAAAJsh6AEAAACAzRD0AAAAAMBmuBiLye3reTDitq8kXxn2cYcy5lAMZV2RitZcIxWtWkVj+7AnO74GorEmO+4nAADuxhE9AAAAALAZgh4AAAAA2AxBDwAAAABshqAHAAAAADbDxVgAAADuEDz9S0zHNR6bGpXtj3RDqWusa8BrBcOBI3oAAAAAYDMEPQAAAACwGYIeAAAAANgMQQ8AAAAAbIagBwAAAAA2Q9ADAAAAAJsh6AEAAACAzRD0AAAAAMBmuGE6AAAYEaJ1I3SriMb67bpPo3HD8rvH7E1MVPDWrSHNK9yYw8WOdR3KmuxyI3qO6AEAAACAzRD0AAAAAMBmCHoAAAAAYDMEPQAAAACwGYIeAAAAANgMQQ8AAAAAbCai2ys0NDSosrJSgUBA+fn5Wrx4cb/fB4NBVVZWyuv1Kjk5WYWFhXK73YP27ezsVGlpqa5du6aMjAxt2rRJqampw7w8AAAAABh5wh7RCwQC2r17t959912VlpbqyJEjunTpUr82Xq9XV69eVVlZmdauXatdu3aF7VtVVaVp06aprKxM06ZNU1VVVRSWBwAAAAAjT9igd+bMGWVmZmr8+PFKSEjQnDlzVF9f36/NsWPHNH/+fBmGocmTJ6urq0s+n2/QvvX19crLy5Mk5eXlDRgTAAAAAPDvhD11s62tTS6XK/TY5XKpqalpQJv09PR+bdra2gbt297eLqfTKUlyOp26fv36326/urpa1dXVkqSSkhJlZWVFurb757X/jsqwaZKiM3L0xmX7w89Kc42Gkb5+O6KmQIz81/RYzyC2Yr3+WG/fSqKxr+5hTFPmjwiEPaIXDAYHPGcYRkRtIukbTkFBgUpKSlRSUjKkflZXVFQU6ykgAtTJGqiTNVAn86NG1kCdrIE6WYOV6xQ26LlcLrW2toYet7a2ho7E3dmmpaVlQJvB+o4ZM0Y+n0+S5PP5NHr06HtbCQAAAABAUgRBLycnR1euXFFzc7P8fr+OHj0qj8fTr43H41Ftba2CwaAaGxvlcDjkdDoH7evxeFRTUyNJqqmp0cyZM6OwPAAAAAAYeeI//PDDDwdrEBcXp8zMTH322Wf64YcfNG/ePM2ePVuHDh3S2bNnlZOTo8zMTDU2NmrPnj1qaGjQunXrlJaW9o99Jcntduvbb7/V119/rY6ODr366qtKSkq6H2u2jP/cogLmRp2sgTpZA3UyP2pkDdTJGqiTNVi1Tkbw775IBwAAAACwrLCnbgIAAAAArIWgBwAAAAA2E/Y+eoi+lpYWlZeX66+//pJhGCooKNBzzz2nzs5OlZaW6tq1a8rIyNCmTZuUmpoa6+mOSL29vfrggw/k9/vV19en2bNna+nSpdTIpAKBgIqKipSWlqaioiLqZEIbNmxQSkqK4uLiFB8fr5KSEupkQl1dXaqoqNDvv/8uwzC0fv16ZWVlUScTuXz5skpLS0OPm5ubtXTpUuXl5VEnEzlw4IB+/PFHGYah7OxsFRYWqre3lxqZzPfff6/Dhw8rGAwqPz9fixYtsvT/Jr6jZwI+n08+n09ut1s3b95UUVGR3nrrLf30009KTU3V4sWLVVVVpc7OTq1YsSLW0x2RgsGgenp6lJKSIr/fr82bN2vVqlWqq6ujRiZ04MABnT17NvR+2rt3L3UymQ0bNujjjz/ud2sd6mQ+O3fu1JQpU5Sfny+/36+enh5988031MmkAoGA1q1bp+LiYh08eJA6mURbW5vef/99lZaWKikpSTt27NCMGTN06dIlamQiFy9e1Keffqri4mIlJCSouLhYa9as0eHDhy1bJ07dNAGn0xm6ms8DDzygCRMmqK2tTfX19crLy5Mk5eXlqb6+PpbTHNEMw1BKSookqa+vT319fTIMgxqZUGtrq06cOKH8/PzQc9TJGqiTudy4cUO//vqrFixYIElKSEjQqFGjqJOJnTx5UpmZmcrIyKBOJhMIBNTb26u+vj719vbK6XRSI5P5448/lJubq+TkZMXHx2vKlCmqq6uzdJ04ddNkmpubde7cOT366KNqb28P3WDe6XTq+vXrMZ7dyBYIBPT222/r6tWrevbZZ5Wbm0uNTGjPnj1asWKFbt68GXqOOpnT1q1bJUnPPPOMCgoKqJPJNDc3a/To0fr888914cIFud1urVq1ijqZ2JEjRzR37lxJ/N0zk7S0NL3wwgtav369kpKSNH36dE2fPp0amUx2drb27dunjo4OJSUlyev1Kicnx9J1IuiZSHd3t7Zv365Vq1bJ4XDEejq4S1xcnD755BN1dXVp27ZtunjxYqynhLscP35cY8aMkdvt1qlTp2I9HQzio48+Ulpamtrb27VlyxZlZWXFekq4S19fn86dO6fVq1crNzdXlZWVqqqqivW08A/8fr+OHz+u5cuXx3oquEtnZ6fq6+tVXl4uh8OhHTt2qLa2NtbTwl0mTpyol156SVu2bFFKSooeeughxcVZ++RHgp5J+P1+bd++XfPmzdOsWbMkSWPGjJHP55PT6ZTP5+v3XRbEzqhRo/T444+roaGBGpnM6dOndezYMXm9XvX29urmzZsqKyujTiaUlpYm6fbfuZkzZ+rMmTPUyWRcLpdcLpdyc3MlSbNnz1ZVVRV1Mimv16tHHnlEY8eOlcRnCDM5efKkxo0bF6rBrFmz1NjYSI1MaMGCBaHT1b/88ku5XC5L18naMdUmgsGgKioqNGHCBD3//POh5z0ej2pqaiRJNTU1mjlzZqymOOJdv35dXV1dkm5fgfPkyZOaMGECNTKZ5cuXq6KiQuXl5XrjjTc0depUbdy4kTqZTHd3d+jU2u7ubv3888+aNGkSdTKZsWPHyuVy6fLly5Juf1idOHEidTKpO0/blPgMYSbp6elqampST0+PgsEgnyFMrL29XdLtK+LX1dVp7ty5lq4TV900gd9++02bN2/WpEmTZBiGJGnZsmXKzc1VaWmpWlpalJ6erjfffNMyl3O1mwsXLqi8vFyBQEDBYFBPPvmklixZoo6ODmpkUqdOndJ3332noqIi6mQyf/75p7Zt2ybp9umBTz31lF5++WXqZELnz59XRUWF/H6/xo0bp8LCQgWDQepkMj09PVq/fr127twZ+uoH7ydz2b9/v44ePar4+Hg9/PDDeu2119Td3U2NTGbz5s3q6OhQQkKCVq5cqWnTpln6vUTQAwAAAACb4dRNAAAAALAZgh4AAAAA2AxBDwAAAABshqAHAAAAADZD0AMAAAAAmyHoAQAAAIDNEPQAAAAAwGb+D3e1xYPDWkQzAAAAAElFTkSuQmCC\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "print(\"Total number of women in the study: \" + str(len(df[\"alive\"])))\n", - "print(\"Number of women who died during the study: \" + str(np.sum(df[\"alive\"] == 0)))\n", - "print(\"Number of women still alive after the study: \" + str(np.sum(df[\"alive\"])))\n", - "print(\"Number of non-smoking women: \" + str(np.sum(df[\"smoker\"] == 0)))\n", - "print(\"Number of smoking women: \" + str(np.sum(df[\"smoker\"])))\n", - "\n", - "df_smoker = df[df[\"smoker\"] == 1]\n", - "df_nonsmoker = df[df[\"smoker\"] == 0]\n", - "df_alive = df[df[\"alive\"] == 1]\n", - "df_dead = df[df[\"alive\"] == 0]\n", - "\n", - "def mortality_rate(df):\n", - " return np.sum(df[\"alive\"] == 0) / len(df[\"alive\"])\n", - "\n", - "def smoking_rate(df):\n", - " return np.sum(df[\"smoker\"] == 1) / len(df[\"smoker\"])\n", - "\n", - "fig, ax = plt.subplots(figsize=(15,7))\n", - "ax.hist(df_smoker['age'], bins=int(np.max(df_smoker['age']) - np.min(df_smoker['age'])),\n", - " density=True, alpha=0.5, label=\"Smokers\")\n", - "ax.hist(df_nonsmoker['age'], bins=int(np.max(df_nonsmoker['age']) - np.min(df_nonsmoker['age'])),\n", - " density=True, alpha=0.5, label=\"Non-smokers\")\n", - "plt.legend()\n", - "plt.title(\"Percentage of participants by age: smokers vs non-smokers\")\n", - "\n", - "fig, ax = plt.subplots(figsize=(15,7))\n", - "ax.hist(df_alive['age'], bins=int(np.max(df_alive['age']) - np.min(df_alive['age'])),\n", - " density=True, alpha=0.5, label=\"Alive\")\n", - "ax.hist(df_dead['age'], bins=int(np.max(df_dead['age']) - np.min(df_dead['age'])),\n", - " density=True, alpha=0.5, label=\"Deceased\")\n", - "plt.legend()\n", - "plt.title(\"Percentage of participants by age: alive vs deceased\")\n", - "\n", - "fig, ax = plt.subplots(figsize=(15,7))\n", - "ax.boxplot([df_alive[\"age\"], df_dead[\"age\"]])\n", - "plt.xticks([1, 2], [\"Alive\", \"Deceased\"])\n", - "plt.title(\"Boxplot of age distribution by survival status\")\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## IV) First analysis" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mortality rate among non-smoking women: 0.31420765027322406\n", - "Mortality rate among smoking women: 0.23883161512027493\n", - "Smoking rate among deceased women: 0.37669376693766937\n", - "Smoking rate among alive women: 0.4687830687830688\n" - ] - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "print(\"Mortality rate among non-smoking women: \" + str(mortality_rate(df_nonsmoker)))\n", - "print(\"Mortality rate among smoking women: \" + str(mortality_rate(df_smoker)))\n", - "print(\"Smoking rate among deceased women: \" + str(smoking_rate(df_dead)))\n", - "print(\"Smoking rate among alive women: \" + str(smoking_rate(df_alive)))\n", - "\n", - "tab = pd.DataFrame({\n", - " \"Smoker\": [mortality_rate(df_smoker)*100, (1 - mortality_rate(df_smoker))*100],\n", - " \"Non-Smoker\": [mortality_rate(df_nonsmoker)*100, (1 - mortality_rate(df_nonsmoker))*100]\n", - "}, index=[\"Deceased\", \"Alive\"])\n", - "\n", - "plt.bar([\"Smoker\", \"Non-Smoker\"], tab.loc[\"Deceased\"], label='Deceased')\n", - "plt.bar([\"Smoker\", \"Non-Smoker\"], tab.loc[\"Alive\"], bottom=tab.loc[\"Deceased\"], label='Alive')\n", - "plt.legend()\n", - "plt.title(\"Mortality by smoking status\")\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## V) Analysis by age group" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Proportion of women over 65 among smokers: 8%\n", - "Proportion of women over 65 among non-smokers: 26%\n" - ] - } - ], - "source": [ - "df[\"age_group\"] = df[\"age\"].apply(lambda a: 1 if a < 35 else 2 if a < 55 else 3 if a < 65 else 4)\n", - "\n", - "df_smoker = df[df[\"smoker\"] == 1]\n", - "df_nonsmoker = df[df[\"smoker\"] == 0]\n", - "\n", - "mortality_smoker = [mortality_rate(df_smoker[df_smoker[\"age_group\"] == i]) for i in range(1,5)]\n", - "mortality_nonsmoker = [mortality_rate(df_nonsmoker[df_nonsmoker[\"age_group\"] == i]) for i in range(1,5)]\n", - "\n", - "age_labels = [\"18-34 years\", \"34-54 years\", \"55-64 years\", \"over 65 years\"]\n", - "bar_width = 0.25\n", - "r1 = range(len(mortality_smoker))\n", - "r2 = [x + bar_width for x in r1]\n", - "\n", - "fig, ax = plt.subplots(figsize=(15,10))\n", - "plt.bar(r1, mortality_smoker, width=bar_width, label='Smokers')\n", - "plt.bar(r2, mortality_nonsmoker, width=bar_width, label='Non-smokers')\n", - "plt.xticks([r + bar_width/2 for r in range(len(mortality_smoker))], age_labels)\n", - "plt.legend()\n", - "plt.title(\"Mortality of smokers and non-smokers by age group\")\n", - "plt.show()\n", - "\n", - "print(f\"Proportion of women over 65 among smokers: {int(100 * np.sum(df_smoker['age_group'] == 4) / len(df_smoker['age_group']))}%\")\n", - "print(f\"Proportion of women over 65 among non-smokers: {int(100 * np.sum(df_nonsmoker['age_group'] == 4) / len(df_nonsmoker['age_group']))}%\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## VI) Analysis using logistic regression" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model accuracy: 82.98%\n", - "Confusion matrix:\n", - "[[ 51 37]\n", - " [ 19 222]]\n" - ] - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "from sklearn.linear_model import LogisticRegression\n", - "from sklearn.model_selection import train_test_split\n", - "from sklearn.metrics import confusion_matrix\n", - "\n", - "X = df[[\"smoker\", \"age\"]]\n", - "y = df[[\"alive\"]]\n", - "\n", - "x_train, x_test, y_train, y_test = train_test_split(X, y, random_state=0)\n", - "\n", - "logistic_model = LogisticRegression(random_state=0, solver='liblinear')\n", - "logistic_model.fit(x_train, y_train.values.ravel())\n", - "\n", - "accuracy = logistic_model.score(x_test, y_test)\n", - "print(f\"Model accuracy: {accuracy * 100:.2f}%\")\n", - "\n", - "conf_matrix = confusion_matrix(y_true=y_test, y_pred=logistic_model.predict(x_test))\n", - "print(\"Confusion matrix:\")\n", - "print(conf_matrix)\n", - "\n", - "model_output = logistic_model.predict_proba(x_test)\n", - "# Probability of death (alive=0)\n", - "probas_pred = model_output[:, logistic_model.classes_ == 0].flatten()\n", - "\n", - "fig, ax = plt.subplots(figsize=(15,10))\n", - "plt.scatter(x_test[x_test['smoker'] == 1][\"age\"], probas_pred[x_test['smoker'] == 1], label='Smoker')\n", - "plt.scatter(x_test[x_test['smoker'] == 0][\"age\"], probas_pred[x_test['smoker'] == 0], label='Non-smoker')\n", - "plt.legend()\n", - "plt.title(\"Predicted probability of death by age and smoking status\")\n", - "plt.xlabel(\"Age\")\n", - "plt.ylabel(\"Probability of death\")\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Answers to Questions\n", - "\n", - "**Why is the initial result surprising?** \n", - "At first glance, the mortality rate among non-smokers is higher than that among smokers, which contradicts expectations since smoking is generally harmful. This is a classic example of Simpson's paradox.\n", - "\n", - "**Why is this result surprising after considering age groups? Can you explain this paradox?** \n", - "After stratifying by age groups, it becomes evident that mortality is higher for smokers within each age group, but the overall mortality appears higher for non-smokers because the non-smoking group has a greater proportion of older women, who naturally have higher mortality rates. This confounding factor creates the paradox.\n", - "\n", - "**Does logistic regression allow concluding on the harmfulness of smoking?** \n", - "Yes, logistic regression accounting for age and smoking status shows smokers have a higher probability of death at any given age, supporting the conclusion that smoking is harmful." - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "base", - "language": "python", - "name": "base" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.12.2" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} -- 2.18.1