kaggle-titanic.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "02ae35fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d34e8bd5",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"train.csv\")\n",
    "test = pd.read_csv(\"test.csv\")\n",
    "test_ids = test[\"PassengerId\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b2b5beeb",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "data.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b790aa32",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.hist(\"Age\")    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8db502ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.displot(data['Age'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "99a7e07b",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.countplot(x = 'Survived', data = data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5e4526a",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.countplot(x = 'Survived', data = data, hue = 'Pclass')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "60f7ccae",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.countplot(x = 'Pclass', data = data, hue = 'SibSp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7388e4b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.boxplot(x = 'Pclass', y = 'Age', data = data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af303956",
   "metadata": {},
   "source": [
    "# Data Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8c7807f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2da30b46",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fb64b0e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(data[data  ['Pclass'] == 1]['Age'].mean())\n",
    "print(data[data  ['Pclass'] == 2]['Age'].mean())\n",
    "print(data[data  ['Pclass'] == 3]['Age'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5edad7c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_na_values(cols):\n",
    "    age = cols[0]\n",
    "    pclass = cols[1]\n",
    "    \n",
    "    if pd.isnull(age):\n",
    "        if pclass == 1:\n",
    "            return round(data[data['Pclass'] == 1]['Age'].mean())\n",
    "        elif pclass == 2:\n",
    "            return round(data[data['Pclass'] == 2]['Age'].mean())\n",
    "        elif pclass == 3:\n",
    "            return round(data[data['Pclass'] == 3]['Age'].mean())\n",
    "    else:\n",
    "        return age\n",
    "\n",
    "data['Age'] = data[['Age', 'Pclass']].apply(fill_na_values, axis=1)\n",
    "test['Age'] = test[['Age', 'Pclass']].apply(fill_na_values, axis=1)\n",
    "\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6451c692",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9741d7bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.heatmap(data.isnull())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "24fa4eb0",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "data.drop(['Ticket', 'Name', 'PassengerId', 'Cabin'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "53fa434c",
   "metadata": {},
   "outputs": [],
   "source": [
    "test.drop(['Ticket', 'Name', 'PassengerId', 'Cabin'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4f9a545a",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce96c5e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "50ae2e05",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "\n",
    "cols = ['Sex']\n",
    "\n",
    "for col in cols:\n",
    "    data[col] = le.fit_transform(data[col])\n",
    "    test[col] = le.transform(test[col])\n",
    "    print(le.classes_)\n",
    "    \n",
    "# sex = pd.get_dummies(data['Sex'], drop_first=True)\n",
    "# data.drop = ['Sex'], axis=1, inplace=True)\n",
    "# data = pd.concat([data, sex], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42a6cdfc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "\n",
    "cols = ['Embarked']\n",
    "\n",
    "for col in cols:\n",
    "    data[col] = le.fit_transform(data[col])\n",
    "    test[col] = le.transform(test[col])\n",
    "    print(le.classes_)\n",
    "\n",
    "# embarked = pd.get_dummies(data['Embarked'], drop_first= True)\n",
    "# data.drop = ['Embarked'], axis=1, inplace=True)\n",
    "# data = pd.concat([data, embarked], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "be5cae1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "37394418",
   "metadata": {},
   "outputs": [],
   "source": [
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f1effb30",
   "metadata": {},
   "outputs": [],
   "source": [
    "test['Fare'].fillna(test['Fare'].mean(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "266913ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "625af36e",
   "metadata": {},
   "source": [
    "# Data Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ea290c3",
   "metadata": {},
   "source": [
    "Split into Train and Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "51f85462",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "y = data[\"Survived\"]\n",
    "X = data.drop(\"Survived\", axis=1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.1, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc906735",
   "metadata": {},
   "source": [
    "Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "id": "bce10a7f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.90      0.85      0.88        54\n",
      "           1       0.79      0.86      0.83        36\n",
      "\n",
      "    accuracy                           0.86        90\n",
      "   macro avg       0.85      0.86      0.85        90\n",
      "weighted avg       0.86      0.86      0.86        90\n",
      "\n",
      "[[46  8]\n",
      " [ 5 31]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "lr = LogisticRegression(random_state=0, max_iter=1000)\n",
    "lr.fit(X_train, y_train)\n",
    "\n",
    "lr_predictions = lr.predict(X_test)\n",
    "\n",
    "print(classification_report(y_test, lr_predictions))\n",
    "print(confusion_matrix(y_test, lr_predictions))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ed9fc0e",
   "metadata": {},
   "source": [
    "MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a58ff3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b364e59",
   "metadata": {},
   "source": [
    "Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca9e294a",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "svm = SVC()\n",
    "svm.fit(X_train, y_train)\n",
    "\n",
    "predictions = svm.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e161261b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "print(classification_report(y_test, predictions))\n",
    "print(confusion_matrix(y_test, predictions))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66a8c70b",
   "metadata": {},
   "source": [
    "GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c004becc",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "param_grid = {'C':[0.5, 1, 10, 50, 100, 1000], 'gamma':[1, 0.1, 0.001, 0.0001, 0.00001, 0.000001]}\n",
    "\n",
    "grid = GridSearchCV(SVC(), param_grid, refit=True, verbose=2)\n",
    "\n",
    "grid.fit(X_train, y_train)\n",
    "grid_predictions = grid.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "4f625380",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.85      0.87        54\n",
      "           1       0.79      0.83      0.81        36\n",
      "\n",
      "    accuracy                           0.84        90\n",
      "   macro avg       0.84      0.84      0.84        90\n",
      "weighted avg       0.85      0.84      0.85        90\n",
      "\n",
      "[[46  8]\n",
      " [ 6 30]]\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, grid_predictions))\n",
    "print(confusion_matrix(y_test, grid_predictions))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1057943b",
   "metadata": {},
   "source": [
    "K neighbours "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "24bb3a65",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.76      0.94      0.84        54\n",
      "           1       0.87      0.56      0.68        36\n",
      "\n",
      "    accuracy                           0.79        90\n",
      "   macro avg       0.82      0.75      0.76        90\n",
      "weighted avg       0.80      0.79      0.78        90\n",
      "\n",
      "[[51  3]\n",
      " [16 20]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "error_list = []\n",
    "\n",
    "for i in range(1,40):\n",
    "    knn = KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    knn_predictions = knn.predict(X_test)\n",
    "    error_list.append(np.mean(knn_predictions != y_test))\n",
    "\n",
    "i = np.argmin(error_list)\n",
    "knn = KNeighborsClassifier(n_neighbors=i)\n",
    "knn.fit(X_train, y_train)\n",
    "knn_predictions = knn.predict(X_test)\n",
    "\n",
    "print(classification_report(y_test, knn_predictions))\n",
    "print(confusion_matrix(y_test, knn_predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "0b3c15b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD4CAYAAADiry33AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAy1ElEQVR4nO3deXSc9ZXg/e/Vvq+WvEi2VVqCMWCzyBLgYLDJAgkJoTPphhA6ncCAAybp6ZnuZGbOoac73Wem37ffebvDTkOSCUmgE0K6CYGQBGNwAC/ygrGxjTbLkrxI1r5LVXXnj6oyslSSnpJKqpJ0P+fooHq2un4oXf10n98iqooxxpiFKybSARhjjJldluiNMWaBs0RvjDELnCV6Y4xZ4CzRG2PMAhcX6QCCWbJkiRYVFUU6DGOMmTf27dt3TlXzgu2LykRfVFREVVVVpMMwxph5Q0QaJtpnpRtjjFngLNEbY8wCZ4neGGMWOEv0xhizwFmiN8aYBc4SvTHGLHCW6I0xZoFbVIn+9aNnaWzvj3QYxhgzpxZNoh92e9n64308vbMu0qEYY8ycWjSJvqGtjxGPcqZ7MNKhGGPMnFo0ib6mpReAlp6hCEdijDFza9Ek+upAou+2RG+MWVwWTaIPtOhbe4awdXKNMYvJokn0gRb9sMdL18BIhKMxxpi5sygSvcer1LX2siwjCfC16o0xZrFYFIm+qaOfIbeXa0tyAXsga4xZXBZFog/U5685n+iti6UxZvFYFIk+UJ+/tnQJYD1vjDGLy6JI9DUtveSnJ7IiM4nk+Fgr3RhjFpVFkeirW3opzU9DRMjPSLREb4xZVBZ8oldValt6KctPAyA/PZEWmwbBGLOILPhEf6Z7kN4hN6XnE32Sda80xiwqjhK9iNwkIsdFpEZEvhNk/60ickhEDopIlYh83Om5sy3Q46Y0Px2AvHQr3RhjFpcpE72IxAKPAjcDa4E7RGTtmMNeB9ar6uXA14GnQzh3VlWfDSR6f4s+I5HeITf9w+65DMMYYyLGSYu+AqhR1TpVHQaeB24dfYCq9upHE8ikAur03NlW09pLVko8S9ISAF/pBmx0rDFmbv3PV47ymX/eidvjnfP3dpLoC4DGUa+b/NsuICK3icgx4Nf4WvWOz/Wff6+/7FPV2trqJHZHas72Uprn63EDvtIN2OhYY8zcOds9yA/ePsEHp7v51aFTc/7+ThK9BNk2bvpHVf2lqq4BvgB8N5Rz/ec/parlqlqel5fnICxnalp7KVuadv51fiDR26ApY8wcefLNOjyqrMpJ4ZHtNXi8czuDrpNE3wSsHPW6EJjwV5KqvgWUiMiSUM8Nt7beIdr7hinJC5LobRoEY8wcONc7xE/3NHDr5Sv4q5suora1j98cPjOnMThJ9HuBMhFxiUgCcDvw0ugDRKRU/LUREbkSSADanJw7mwI9bsqWpp/flp2SQFyMWOnGGDMnnt5Zz5DbywObS7n50uWU5KXy8PZqvHPYqp8y0auqG9gGvAYcBX6mqkdEZKuIbPUf9kXgsIgcxNfL5k/UJ+i5s/DvCKq65cIeNwAxMeLrYmmlG2PMLOvsH+bZd0/w2cuWU5KXRmyM8MDmUo6d6eH1Yy1zFkeck4NU9RXglTHbnhj1/T8A/+D03LlS09JLakIsKzKTLtien55opRtjzKz7/tsn6Bv2sG1L6fltn1+/gn/6fTUPb6/mExfnn+8oMpsW9MjYmpZeSvLTxt3IPBsda4yZZd2DI/zw7Xo+tXYpa5ZlnN8eFxvD/TeUcKipi7eqz81JLAs+0Y8u2wTkZyRaojfGzKpn322ge9DNg1vKxu37oysLWZGZxMOvV8/JGtYLNtF3D45wpnsweKJPT6Stb5iRCAxcMMYsfP3Dbp7eWccNF+VxWWHmuP0JcTFsvaGEqoYOdtW1z3o8CzbR1wZ63OSnj9sXGDR1rtda9caY8PvJrpN09I8Ebc0H/HH5SvLSE3l4e/Wsx7NgE32wHjcBgWkQrOeNMSbcBkc8PLWzjmtLcrlqdfaExyXFx3LfpmLeqW1jX8PstuoXbKKvbeklIS6GldnJ4/bl2zQIxphZ8q97G2ntGZq0NR/w5cpV5KQm8PD2mlmNacEm+uqWXoqXpBIXO/6fmJ9ho2ONMeE37PbyxJu1lK/O5urinCmPT0mI4+6Pu9hxvJX3m7pmLa4Fm+gDXSuDWZKWiIiVbowx4fWL/U2c7hrkwRvLHPeP/9NrVpORFDertfoFmegHRzw0dvSfXz5wrPjYGHJSEqx0Y4wJG7fHy2M7alhXmMmmsiWOz0tPiufPNrr47QdnOXame1ZiczQydr6pbe1FNfiD2IC89ERarXRjjHFo/8kOfvXexHMytnQP0dg+wEO3XBLyaNevbyzimZ11PLK9hke+fOVMQx1nQSb6mkm6VgbkZ9joWGOMMx6v8p9/9h5NHf0kxcdOeNx1ZUu4cU1+yNfPSkngrmuK+P3RswyOeCZ9j+lYsIk+RqBoScqEx+SnJ1J9tmcOozLGzFcvHzpF/bk+Hr/zSm6+bPmsvMe3bizjLz99EbEx4Z/7ZsEm+qLcVBLjJv6tmJ/umwbB61ViZuHGGmMWBq9XefSNGsry0/j0Jctm7X2SE8Lbih9tQT6MrZ6kx01AXnoibq/S0T88R1EZY+aj335whg/P9rJtS+m8bRQuuEQ/4vFy4lzfhD1uAs6PjrU6vTFmAqrKw9trcC1J5ZZ1KyIdzrQtuETf0NaH26uT9riB0YOmLNEbY4J743gLR051840bSmaldj5XFlyid9LjBkYvEm5dLI0x46kq33u9hoKsZG67oiDS4czIgkv01Wd9ib4kP3XS46x0Y4yZzNs1bRxs7OQbN5QQH2QqlflkfkcfRE1rLwVZyaQkTN6hKDkhlvTEOOtLb4wJ6uHt1SzNSORL5YWRDmXGFlyirz4bfFWpYPJspSljTBB76tvZXd/OfZtKJu2mPV84SvQicpOIHBeRGhH5TpD9d4rIIf/XOyKyftS+/yQiR0TksIg8JyJJY88PF49XqW3tnbLHTYAtEm6MCebh7dUsSUvgjopVkQ4lLKZM9CISCzwK3AysBe4QkbVjDqsHrlfVdcB3gaf85xYA3wTKVfVSIBa4PXzhX6i5Y4Aht9dxiz4/Pclq9MaYCxxs7GRn9Tnuua54VgcxzSUnLfoKoEZV61R1GHgeuHX0Aar6jqp2+F/uAkYXteKAZBGJA1KAiWcFmqGaVt+UBmVLHZZu0hNp6R6ak8V5jTHzwyPbq8lKiecrV6+OdChh4yTRFwCNo143+bdN5G7gVQBVbQb+ETgJnAa6VPW3wU4SkXtFpEpEqlpbW53EPk6gx01p3uRdKwPy0xMZGPHQO+Se1vsZYxaWI6e6+P3RFr6+0UVa4sKZIcZJog82SiBoE1hENuNL9N/2v87G1/p3ASuAVBH5SrBzVfUpVS1X1fK8vDwnsY9T09JLXnoimSnxjo63QVPGmNEefaOG9MQ4vnptUaRDCSsnv7KagJWjXhcSpPwiIuuAp4GbVbXNv/kTQL2qtvqPeRG4FvjxTIKeSHVLL6V5zso2cOEi4SUhnDeREY+XWJF5Ox+GMZNRVTxeDbo8pxMer9LeN/ncUlkp8bPWZ93rVdomef/Gjn5ePXyGB24oJTPZWWNxvnCS6PcCZSLiAprxPUz98ugDRGQV8CJwl6p+OGrXSeBqEUkBBoAbgapwBD6WqlLb0sttVzofwfbRIuEz73kzOOLhU///W3zi4qU89Lmxz6qNmf/+8oVD1J/r44Wt14S8sAbAN587wK/fPz3pMRuKsvn51munG+Kk/vKFQ/xif9Okx6QkxPL1j7tm5f0jacpEr6puEdkGvIav18z3VfWIiGz1738CeAjIBR7zfwDc/jLMbhF5AdgPuIED+HvkhJvHq/z15y/BNckc9GMFWvTh6Ev/86pGTrb38+NdDdy7qZhlmbPWi9SYOef1Kq8fPUtH/whvHG9hy5qlIZ3/walufv3+aW69fAXlRcEXzX6n5hy/OXKGrv4Rx+VXp2pbe3nxQBOfuWwZ15RMvMzfxcvSyUlNCOt7RwNHTxtU9RXglTHbnhj1/T3APROc+9fAX88gRkfiYmP4D1eFNoItIzmOhLiYGSf6YbeXx3fUUpafRt25Pp58q5a//twlM7qmMdGkprWXjv4RAL73eg2bL8oPqVX/6I4a0hLj+NvPXzphEi/LT+PVw2eoamjnxotD+0UylcfeqCUxLoa/vfVSlqQlhvXa88GCGxkbChHxD5qaWaL/5YEmTnUN8t8+ezFfuLyA5/actBG3ZkHZXd8OwH3XF3OwsZO3a9qmOOMjNS09vPL+af70mtWTttQvX5lFQmwMe/zvFS6N7f3828Fm7qhYtSiTPCzyRA8zHx3rW/m9lssKMrnhY3ncv7mEIbeXZ/5QH8YojYms3XVtLMtI4i8++TGWZiTyve3Vjs997I1akuJiuXuK2ndSfCyXr8xiV5gT/eNv1hIrwn2bSsJ63fnEEn16Ei3d0299/+rQKRra+tm2pRQRoSQvjVvWreDZd0/QMUUPA2PmA1VlT307lcU5JMbFct+mEt9cMHVTt+ob2vr49/dOcWflKnIdtKYrXDkcbu6iL0xjW053DfBCVRNfKi9c1M/NFn2iz5tB6cbrVR7ZXsNFS9P55Kia4gObS+gb9vCDd06EKUpjIqehrZ+WniEqXL6HqL4SSAKPvFEz5bmP76glNka4d1Oxo/eqcOXg8Sr7T3ZMfbADT75Zh1eVrdcv3tY8WKInPz2RroERBkc8IZ/76uEz1Lb2jVtLcs2yDD59yVJ+8HY93YMj4QzXmDkXqJlX+hN9ckIs91xXzM7qcxyYJCE3dw7wi/1N3L5hJfkZzlrTV67OJjZGwlKnb+kZ5Lk9J7ntigJW5jjvjbcQWaL3j44N9eGpby3JaorzUvnMZcvH7d+2uYyeQTfPvtsQljiNiZRd9W3kpiZcMKjwK1evJislnke2T9yqf/LNWgDuC6E1nZYYx6UrMthdN/NE/8zOekY8Xu7fXDrja813luinudLU74+2cOxMDw/cUBp0LcnLCjPZfFEeT++sC1u90ZhI2FPfToUr54LulGmJcXx9o4vXj7VwuLlr3Dkt3YM8v7eRL15ZSEFWckjvV1mcy8HGzmn9lR3Q3jfMs7sa+Nz6FbiWTL7a3GKw6BN9XnqgRe+8542q8sj2albmJPP5yydeGX7bljI6+kf46e6TM47TmEho7hygqWPgfH1+tK9eW0R6YhyPBqnVP/VWHR6vcv8NobemK4pyGPZ4ea+xczohA/CDt+vpH/awzVrzgCX6aZVu3qo+x3tNXdx/Q+mk83JctTqbjaW5PPlW3YxaJ8ZEyl5/rTxYos9Mjuer1xbx6uEzfHi25/z2tt4hfrL7JLeuX8Gq3NBr4xuKchBh2nX6roERfvj2CW6+dBllS53NZLvQLfpEn5uaSIw4L92oKg+/Xs3yzCS+eOXUI3G3bS7jXO8Q/7q3ccpjjYk2u+vbSE+KY82yjKD7v/5xFykJsTw2qlX/zB/qGXR7pl0bz0yJ56Kl6ecHaYXqR++coGfIzbYt1poPWPSJPjZGWJKW6Lgv/e76dqoaOth6fQkJcVPfvquLc9hQlM0Tb9Yy7PbONFxj5tTu+nYqinKCPocCyElN4CtXr+al905Rf66Prv4RfvRuA5+5bLnjld6CqXTlsK+hgxFPaD8zfUNunnm7nhvX5HPJisxpv/9Cs+gTPfjKN05Hxz68vZq89ET+ZMPKqQ/GN83Cti1lnO4anHLmPGOiSWvPEHWtfUHLNqPdc52L+NgYHt9Rww/fOUHvkHvGtfHK4lwGRjxBH/RO5se7GujsH7HW/BiW6IG8NGeDpvY1dPB2TRv3XldMUrzztSQ3lS1hfWEmj+2owR1iC8WYSNl7YuL6/Gj56UncUbGKF/c38/Qf6vjk2qVcvDx4qcepDf4ZLkOp0w+OePiXnXVcV7aEK1Zlz+j9FxpL9DhfJPyR7dVkp8Rz59WhrQwfaNU3tg/w7wdnbclcs8C9cayF3xyefD73ybx+9CwvH3L++dtd10ZKQiyXFkxdArnv+mJEoGfQzYNhaE3npSdSnJcaUp3+uT0nOdc7zINbymb8/gvNwlkUcQbyMxJp6x3C49UJa5GHm7t443grf/npi0hJCP22feLifFbnpvDK+6f5YojTKRsD8E+//5Dqll4qXblkhzhnes/gCH/xs/cYcnu4ujjX0SyOu+vbuWp1tqMVn5ZnJvPgljLOdg+yrjArpNgmUunK4eVDpyf9uQwYcnt48s06Klw5U/4FshhZix7fNAhe9XULm8jD26vJSIrjT6+Z3srwIsK1JbnsOdGOxxt0yV1jJtXcOUD/sIfvvx36zKjP7mqga2CEIbeXp3dOfX5n/zDHz/ZQMcEiIcF888Yy/v62y0KObSKVrlx6Bt0cO9M95bEv7GviTPcg37TWfFCW6IG8KUbHHj/Tw2tHzvJnG12kJ01/5ZtQPrjGjDY44uFc7zCxMcIP3z5B14DzOZT6h908vbOe6z+Wx2cvW86z756gs3/ymVWrTnSgOnV9fjYF3nuqOv2Ix7fwz+Urs9hYmjsXoc07luiZetDUI2/UkJoQy9c3Fs3ofZx+cI0Zq6ljAICvXVtEz5CbH4UwM+pPd5+kvW+YB7eUsm1LKX3DHr7/9uTn765vIyEuhvUrs6Yf9AytyEqmMDt5ynlv/u1AM00dAzzonyrcjGeJnskXCa9t7eXlQ6e465oislJmtpZk4INrid6EqrnTl+g/dckytqzJ55m36x3NoTQ44uGpt+q4pjiX8qIc1izL4FNrl/LDKWZW3VPfzuUrs0LqXTYbKlw57DnRjmrwcqfHqzy2o5a1yzPYsiZ/jqObPyzR89F8N8EGTQXWmrznuvCsDF/hymFP/cQfXGOCafa36Auzk9m2pZTO/hF+vGvqmVF/XtVIS8/QBT1hHtxSRvckM6v2Drk5fKr7/LTEkVTpyqG9b5ja1t6g+3/9/mnqz/VZa34KluiBxLhYslLix9XoA2tNfrliddjWmrzalUvbJB9cY4Jp7uwnLkZYmpHElauy+XjpEv5l5+RzKA27vTzxZh1Xrc7mmpKPateXFWZyw0V5PPOHevqHx/9VsL+hA49Xo6L3SqXLF3ewbpa+hX+qKctP49OXLJvr0OYVR4leRG4SkeMiUiMi3wmy/04ROeT/ekdE1o/alyUiL4jIMRE5KiLXhPMfEC7B1o59bIdvrUmnq+M4Efjhme48HmZxau4YYFlm0vluhg9uKeVc7zDP7Zl4ZtRfHmiiuXPg/DKXoz24pYz2vuGgM6vurm8jNka4MgoGHa3OTSE/PTFonf63H5zlw7O94xb+MeNNmehFJBZ4FLgZWAvcISJrxxxWD1yvquuA7wJPjdr3z8BvVHUNsB44Go7Aw23skoKnuwZ4YV8jf7whvGtNBj64Vqc3oWjuHLhgXvfK4lwqinJ48s06htzjW/VjF60f66rV2VxbEnxm1T317VxWkElqYuSH2YhI0HJnYOEf15JUblk38VThxsdJi74CqFHVOlUdBp4Hbh19gKq+o6qBNcV2AYUAIpIBbAKe8R83rKqdYYo9rMYuEv7km3WoEva1JgMf3N11Vqc3zjV3DFCQfeECHg/eWMqZ7kFe2Dd+DqWxi9YH8+CWMlp7hvhZ1Uczqw6OeHivsSsq6vMBla4cznQP0tg+cH7bjuOtHDnVzTduKJlyMJVxlugLgNFz7Db5t03kbuBV//fFQCvwAxE5ICJPi0jQ5V5E5F4RqRKRqtbWVgdhhVd+eiKtPUOo6vm1Jv/oygIKs8O/1mSwD64xExnxeDnTPUjhmJWaPl66hPUrs3h8R+0FszwGFq1fs+zCRevHuro4h/LV2Tyx46OZVQ82djLs8UZFfT6g4nydvg3wtea/t72agqxkbrtislRkApwk+mC/LoM2RUVkM75E/23/pjjgSuBxVb0C6APG1fgBVPUpVS1X1fK8vPF/as62vPREhj1eugZGeDqw1uQ0VsdxorL4wg+uMZM50zWIVxnXohcRvrmllKaOAf7tQPP57YFF6x/YPHntWkR48MYyTnUN8qJ/ZtXdde2IQHkII2JnW1l+Gtkp8eefa71T28aBk51844YSR9MzGGeJvgkYPSdvITBuZiQRWQc8Ddyqqm2jzm1S1d3+1y/gS/xRJ7BK/fEzPfx4VwOfX7+Collaa7I0z/fBtTq9cSIwWKoga/xfl1vW5LN2eQaP7ajF49UpF60fa1PZEtYVZvLYjlrcHi97TrRx8bIMMpOnPwI83GJihA1FOed/Xr73ejVLMxL5UrnNGeWUk0S/FygTEZeIJAC3Ay+NPkBEVgEvAnep6oeB7ap6BmgUkYv8m24EPghL5GEWGDT1D785Rv+whwdmca3JwAfXet4YJwKDpca26MHfKt9SSv25Pl4+dGrKReuDn1/GyfZ+frG/iX0NHVFVtgmocOVwsr2fl947xe76du7bVEJiXGQHc80nUz5WV1W3iGwDXgNige+r6hER2erf/wTwEJALPOZ/8ONW1XL/JR4EfuL/JVEHfC38/4yZCyT6/Sc752StyQpXDr/94CynuwZYnjn+B9iYgMBgqRVZwXt/ffqSZZTlp/HoGzUkx8dOuWj9WJ+4OJ81y9L5u5ePMjjijaoHsQGB/vT//cX3WZKWwB0VoU0Vvtg5KnCp6iuq+jFVLVHVv/dve8Kf5FHVe1Q1W1Uv93+Vjzr3oL/2vk5VvzCqd05UCZRugDlZnSbwwZ3L8s3AsCfkpdlM5DV39pOfnjhhCzYmRti2pZQPz/Y6WrR+rECrvsc/pcKGKEz0Fy9PJy0xjp4hN/dcV0xygrXmQ2FPMvzSEuPITI6fs7Um167IIC0xbs4S/bDby83//BZ/+6uorJyZSTR3ju9aOdZnL1uOa0kqyzOT+KMrQ++JctOlyyjNT+NjS9PCNgo8nOJiY6h05ZCVEs9Xrp7eVOGLWeRHRESRn/7HygsGpcym2BihvCh7zur0L+5v4kRbPx5tmZP3M+HT3DEw5SpPcbEx/OjrFXi8Oq3adWyM8MOvbWDEE71jO/7+tsvoHRohLQoGcs031qIf5ZIVmTOeoTIUFa4calp6OTfJgifhEBglGSPQ2D7AqU7rvz9feL3Kqc7BKVv0ACtzUmbUU6wwOwXXLPU0C4dlmUmU5s/us7OFyhJ9BAUeelWdmN1W/UvvneJkez/fvNG3+o5165w/zvUOMezxjhssZUwoLNFH0GUFWSTFx8xq+cbjVR55wzdKctvmUtKT4qxb5zzS2DFx10pjnLJEH0EJcTFcuSp7yhV0ZuLVw6epa+1j25ZS4mJj/ANPbETufHG+D32QwVLGOGWJPsIqXDkcPdMd0hqgTgXmPCnJS+XmS5eff7/a1r5Zfy5gwqPZWvQmDCzRR1iFKwdV2NcQ/lb974+e9Y2S3PzRKElbt3Z+ae7sJzM53nqamBmxRB9hV6zMJj5Wwl439815UsPq3BQ+v/6jUZKXFWSSHB9riX6eaO4YoNBa82aGLNFHWHJCLOsLs8Jep3/zw1beb+7i/htKiBs1SjI+NoarVs9d/30zM2MXHDFmOizRR4EKVw6Hm7voGxq/fud0BFrzvvm6x8/wV+HK4diZbrr6w/9cwISPqgZdcMSYUFmijwIVrhzcXuXAyc6wXO/dujb2NXSw9fpiEuLG/y8OPBeomoXnAiZ8ugZG6Bv2WIvezJgl+ihw1epsYoSwdXt8ZHsN+emJfKl8ZdD9l6/MIiF2dvvvm5kLzENvNXozU5boo0B6UjyXFmSyKwyJd19DO+/UtnHvpmKS4oPPeZIUH8vlK7Ms0Ue5yRYcMSYUluijREVRDgcbOxkc8czoOg9vryEnNYEvV04+X3e4nwuY8JtswRFjQmGJPkpUuHIYdns51NQ17Wscaupkx/FW7rnORUrC5P2uK1w5eLzK/pNRuTyAwde1Mjk+luyU6FnWz8xPNgojSmzwL8b8N786wqqc6f2pfvxsD5nJ8dzlYL7uK1dnExsj7K5r57qyuV+M3UytubOfguxk/Ku2GTNtluijRHZqArdvWMn+kx3UtvZO6xpxMcK3b1pDetLULcC0xDguLci0gVNRrLnTBkuZ8LBEH0X+1xfXzen7Vbpy+OHbJxgc8Uz44NZETnPHAOsLsyIdhlkArEa/iFUU5TDs8fJeY2ekQzFj9A+76egfsQexJiws0S9iG4pyELEJzqLR+VkrbbCUCQNHiV5EbhKR4yJSIyLfCbL/ThE55P96R0TWj9kfKyIHROTlcAVuZi4zJZ6LlqZbf/ooZIOlTDhNmehFJBZ4FLgZWAvcISJrxxxWD1yvquuA7wJPjdn/LeDozMM14XZ1cS77GjoY8XgjHYoZpckWHDFh5KRFXwHUqGqdqg4DzwO3jj5AVd9R1UCH7F3A+Zm0RKQQ+CzwdHhCNuFU4cphYMTD4ebp99834dfcMUB8rJCfnhjpUMwC4CTRFwCNo143+bdN5G7g1VGv/wn4K2DSJqOI3CsiVSJS1dra6iAsEw6B/vtWp48uzZ0DLM9MJibG+tCbmXOS6IN90jTogSKb8SX6b/tf3wK0qOq+qd5EVZ9S1XJVLc/LswE8cyUvPZHivFSr00eZ5o5+exBrwsZJom8CRk+DWAicGnuQiKzDV565VVUD0zBuBD4vIifwlXy2iMiPZxSxCbtKVy57T7Tj8Qb9/W0ioLnT5qE34eMk0e8FykTEJSIJwO3AS6MPEJFVwIvAXar6YWC7qv5XVS1U1SL/edtV9Sthi96ERaUrh55BN8fOdEc6FAMMu7209AxZjxsTNlMmelV1A9uA1/D1nPmZqh4Rka0istV/2ENALvCYiBwUkapZi9iEnS0YHl1Odw2gan3oTfg4mgJBVV8BXhmz7YlR398D3DPFNXYAO0KO0My6FVnJFGYns6e+na9tdEU6nEXv/GApa9GbMLGRsQbwter31LejanX6SDs/WMr60JswsURvALjalUtb3/C0Z8404dPUOYAILMtMinQoZoGwRG+Aj+r01s0y8po7BlianhR0YXdjpsM+SQaA1bkp5KcnsrvOEn2kBRYcMSZcLNEbAETE6vRRorlzwHrcmLCyRG/OqyzO5Uz3II3tA5EOZdHyeJXTnYPWojdhZYnenFd5vk7fNsWRZra09Azi9qq16E1YWaI355XmpZGdEm8DpyKo2eahN7PAEr05LyZG2FCUYz1vIsgWHDGzwRK9uUBlcS4n2/s53WV1+kho9i84ssJKNyaMLNGbC1TavDcR1dQxQE5qAikJjmYnMcYRS/TmAhcvzyAtMc4SfYRY10ozGyzRmwvExgjlRdmW6CPEFhwxs8ESvRmnwpVDdUsvbb1DkQ5lUVFVW3DEzApL9GacSlcuAHtPWKt+LrX3DTM44rUWvQk7S/RmnMsKMkmKj7FulnMs0OPGWvQm3CzRm3ES4mK4cpXV6eea9aE3s8USvQmqwpXDB6e76RoYiXQoi0azLThiZoklehNUpSsXVdjXYK36udLcOUBaYhwZydaH3oSXfaJMUFesyiI+Vthd386WNUsjHc6C8KN3T/C7D85OuP/YmR4KspIRkTmMyiwGjlr0InKTiBwXkRoR+U6Q/XeKyCH/1zsist6/faWIvCEiR0XkiIh8K9z/ADM7kuJjWV+YZXX6MDnbPcjfvXyU+nN99A65g34VZifzJxtWRjpUswBN2aIXkVjgUeCTQBOwV0ReUtUPRh1WD1yvqh0icjPwFFAJuIH/rKr7RSQd2CcivxtzrolSFa4cnnqrjv5htw3Jn6En36zDo8pP77maVblWgzdzy0mLvgKoUdU6VR0GngduHX2Aqr6jqh3+l7uAQv/206q63/99D3AUKAhX8GZ2VRbn4vYq+xs6Ix3KvHaud4if7mngC5cXWJI3EeEk0RcAjaNeNzF5sr4beHXsRhEpAq4Adgc7SUTuFZEqEalqbW11EJaZbVetziZGYI8tRDIjT++sZ8jt5f7NJZEOxSxSThJ9sCdDQRcVFZHN+BL9t8dsTwN+Afy5qnYHO1dVn1LVclUtz8vLcxCWmW1piXFcWpBpA6dmoLN/mGffPcEt61ZQkpcW6XDMIuUk0TcBo58QFQKnxh4kIuuAp4FbVbVt1PZ4fEn+J6r64szCNXOtoiiHA42dDI54Ih3KvPT9t0/QN+zhAWvNmwhykuj3AmUi4hKRBOB24KXRB4jIKuBF4C5V/XDUdgGeAY6q6v8OX9hmrlQW5zLs9nKoqSvSocw73YMj/PDtej59yVLWLMuIdDhmEZsy0auqG9gGvIbvYerPVPWIiGwVka3+wx4CcoHHROSgiFT5t28E7gK2+LcfFJHPhP+fYWbLhqJswOr00/Hsuw10D7rZtrks0qGYRc5RnzlVfQV4Zcy2J0Z9fw9wT5Dz/kDwGr+ZJ7JSElizLJ3d9e1si3Qw80j/sJund9ax+aI8LivMjHQ4ZpGzKRDMlCpcOexr6MDt8UY6lHnjJ7tO0tE/wrYt1po3kWeJ3kyp0pVL/7CHw6eCdpgyYwyOeHhqZx0bS3O5anV2pMMxxhK9mdoGl9XpQ/Gvextp7Rmy2ryJGpbozZTy05MoXpJq8944MOz28sSbtWwoyubq4pxIh2MMYIneOFThymFPfTteb9CxcsbvF/ubON01yLYtZTYLpYkaluiNI5XFOXQPujl2pifSoUQtt8fLYztqWF+YyaayJZEOx5jzbEpC40iFf8HwPfVtrF0xPwf/tHQPUtvaN2vX39fQTmP7AA/dcom15k1UsURvHCnISqYgK5l9Jzv5s42RjiZ0Ix4vX3ryXRra+mf1fdYuz+ATF+fP6nsYEypL9MaxNcvSqT47P0s3Lx08RUNbPw/dspaLl8/eXyQfW5pmrXkTdSzRG8dK89PYWXMOt8dLXOz8ebzj8SqP7qjh4uUZfG1jkSVis+jMn59WE3Gl+WkMu700dgxEOpSQvPL+aepa+9i2udSSvFmULNEbx0rzffOp17T0RjgS57xe5ZHtNZTmp3HzpcsiHY4xEWGJ3jgWSPTVLfOnTv+7o2c5fraHBzaXEBNjrXmzOFmiN46lJ8WzLCNp3rToVX2t+dW5KXxu3YpIh2NMxFiiNyEpW5o2bxL9jg9beb+5i/tvKJlXD4+NCTf79JuQlOT5Er1qdE+FoKo8/Ho1BVnJ3HZFYaTDMSaiLNGbkJTmp9E/7OFU12CkQ5nUu7Vt7D/Zydbri0mIs4+5WdzsJ8CEpGye9Lx5eHsN+emJfKl85dQHG7PAWaI3ITnf8yaKR8hWnWjn3bo27t1UTFJ8bKTDMSbiLNGbkOSmJZKTmkBta/S26B/eXkNuagJfrlwV6VCMiQqW6E3ISvPSqD4bnYn+UFMnb37Yyt3XuUhJsBk+jAGHiV5EbhKR4yJSIyLfCbL/ThE55P96R0TWOz3XzD+lS9OojtKeNw9vryEzOZ67rl4d6VCMiRpTJnoRiQUeBW4G1gJ3iMjaMYfVA9er6jrgu8BTIZxr5pnSvDS6BkY41zsc6VAucPR0N7/74Cxf21hEelJ8pMMxJmo4adFXADWqWqeqw8DzwK2jD1DVd1S1w/9yF1Do9Fwz/5Qtjc6eN4/vqCUtMY6vXeuKdCjGRBUnib4AaBz1usm/bSJ3A6+Geq6I3CsiVSJS1dra6iAsEykfTW4WPT1v3B4v24+18Ln1K8hMsda8MaM5SfTBZoIKWpwVkc34Ev23Qz1XVZ9S1XJVLc/Ly3MQlomUZRlJpCXGRVWL/ujpHnqH3FxTkhvpUIyJOk66JTQBo0edFAKnxh4kIuuAp4GbVbUtlHPN/CIilOT7HshGi931vo9cRVFOhCMxJvo4adHvBcpExCUiCcDtwEujDxCRVcCLwF2q+mEo55r5qSw/uiY321PfzurcFJZlJkU6FGOizpSJXlXdwDbgNeAo8DNVPSIiW0Vkq/+wh4Bc4DEROSgiVZOdOwv/DjPHSvPTaOkZomtgJNKh4PUqe060W2vemAk4GlGiqq8Ar4zZ9sSo7+8B7nF6rpn/Rs95c9Xq7IjGUt3SS2f/CJXFVp83JhgbGWumJZp63uzx1+crXdaiNyYYS/RmWgqzU0iMi4mKOv3u+naWZyZRmJ0c6VCMiUqW6M20xMYIxXmR73mjquyub6fClYOIrQlrTDCW6M20RUPPmxNt/bT2DFHpsvq8MROxRG+mrTQ/jaaOAfqH3RGLIVCfr7D6vDETskRvpi3Q86autS9iMeyubyc3NYGSvNSIxWBMtLNEb6bt/GpTEex5s7vO6vPGTMUSvZm21bmpxMVIxOr0TR39NHcOWLdKY6Zgid5MW0JcDKtzUyK22tTeE+0AVNiDWGMmZWutmRkpy0/nQ4elm/pzfVO2/tevzCQ/3dl8NXvq28lIiuOiZemOjjdmsbJEb2akND+N3x09y7DbS0LcxH8gDgx7+NIT73Kud2jS661fmcW/3X+to5r77rp2NhTlEBtj9XljJmOJ3sxI2dI0PF7lRFsfH1s6ccv6uT0nOdc7xD/ffjkleWlBj3nzw1b+39eOs7P6HJs+NvmaBC09g9Sd6+P2ipWTHmeMsURvZiiQtKvP9k6Y6IfcHp58q5ZKVw63Xj7x4mRlS9P48a4GHtleM2Wi31vvW7nS6vPGTM0expoZKclLQ2Ty9WN/XtXE2e4hHtxSNum1EuNiuW9TMXtOtLOrrm3SY/fUt5GSEMslKzKmFbcxi4klejMjyQmxFGYnT9iXfsTj5fEdtVyxKouNpVO3vm+vWMWStEQe2V4z6XG769u5anU28bH2ETZmKvZTYmasLD99whb9Lw8009w5wINbSh09YE2Kj+XeTS7+UHOO/Sc7gh7T2T/MsTM91n/eGIcs0ZsZK81Po+5cHx7vheu+e7zKY2/UcMmKDDZflO/4endWriY7JX7CVv3eE1afNyYUlujNjJXmpzHs9tLY3n/B9pcPneJEW7/j1nxAamIcd3/cxfZjLRxu7hq3f099GwlxMawrzJxx7MYsBpbozYx9NOfNR+Ubr1d5ZHsNH1uaxqfWLgv5mn96bRHpSXFBW/W769u5fGUWSfGx0w/amEXEEr2ZsdJR68cGvHbkDNUtvTywuZSYaQxoykiK52vXFvGbI2c4fuajB729Q24ON3dxtdXnjXHMUaIXkZtE5LiI1IjId4LsXyMi74rIkIj8lzH7/pOIHBGRwyLynIg4G99u5o2MpHiWZiSe73mjqjy8vQbXklRuWbdi2tf92kYXqQmxPPrGR636fQ0deNXq88aEYspELyKxwKPAzcBa4A4RWTvmsHbgm8A/jjm3wL+9XFUvBWKB28MQt4kyZfnp1Ppb9NuPtfDB6W7uv6FkRtMTZKcm8JVrVvPyoVPUtfquvae+jbgY4crVWeEI25hFwUmLvgKoUdU6VR0GngduHX2Aqrao6l5gJMj5cUCyiMQBKcCpGcZsolCpf1nBQGu+MDuZL1wx8ShYp+75eDEJcTE8tqMW8M1vc2lBJikJNqjbGKecJPoCoHHU6yb/timpajO+Vv5J4DTQpaq/DTVIE/1K89PoG/bw86omDjZ28o0bSsIymCkvPZE7KlbxywPNVJ/t4b2mTiqLrT5vTCic/CQG+9tbg2wbf6JINr7WvwtYAaSKyFcmOPZeEakSkarW1lYnlzdRJPBA9u9+/QHLMpL4D1cVhu3a920qIVaEB587wIhHbaCUMSFykuibgNFTBBbivPzyCaBeVVtVdQR4Ebg22IGq+pSqlqtqeV7e5BNamegTWD+2e9DNfdcXkxgXvq6PyzKT+FJ5IcfO9CACV622RG9MKJwk+r1AmYi4RCQB38PUlxxe/yRwtYikiG/EzI3A0emFaqJZbloi2SnxLElL4PYNq8J+/a3XlxAXI1y8LIPM5PiwX9+YhWzKJ1qq6haRbcBr+HrNfF9Vj4jIVv/+J0RkGVAFZABeEflzYK2q7haRF4D9gBs4ADw1O/8UE2nfvmkN+RmJJCeEfyDTypwU/u4Ll7IkLTHs1zZmoRNVR+X2OVVeXq5VVVWRDsMYY+YNEdmnquXB9tnIWGOMWeAs0RtjzAJnid4YYxY4S/TGGLPAWaI3xpgFzhK9McYscJbojTFmgbNEb4wxC1xUDpgSkVagYYLdS4BzcxhOqCy+mbH4Zsbim5n5HN9qVQ06UVhUJvrJiEjVRKO/ooHFNzMW38xYfDOzUOOz0o0xxixwluiNMWaBm4+JPtpnv7T4ZsbimxmLb2YWZHzzrkZvjDEmNPOxRW+MMSYEluiNMWaBmzeJXkRuEpHjIlIjIt+JdDxjicgJEXlfRA6KSFSsmiIi3xeRFhE5PGpbjoj8TkSq/f/NjrL4/oeINPvv40ER+UyEYlspIm+IyFEROSIi3/Jvj4r7N0l80XL/kkRkj4i854/vb/zbo+X+TRRfVNy/UXHGisgBEXnZ/3pa929e1OhFJBb4EPgkvsXK9wJ3qOoHEQ1sFBE5AZSratQMthCRTUAv8CNVvdS/7f8B2lX1f/l/YWar6rejKL7/AfSq6j9GIqZRsS0HlqvqfhFJB/YBXwD+jCi4f5PE98dEx/0TIFVVe0UkHvgD8C3gj4iO+zdRfDcRBfcvQET+AigHMlT1lun+/M6XFn0FUKOqdao6DDwP3BrhmKKeqr4FtI/ZfCvwf/zf/x98ySEiJogvKqjqaVXd7/++B9+i9gVEyf2bJL6ooD69/pfx/i8leu7fRPFFDREpBD4LPD1q87Tu33xJ9AVA46jXTUTRh9pPgd+KyD4RuTfSwUxiqaqeBl+yAPIjHE8w20TkkL+0E7HSUoCIFAFXALuJwvs3Jj6IkvvnLzscBFqA36lqVN2/CeKDKLl/wD8BfwV4R22b1v2bL4legmyLqt++wEZVvRK4GXjAX5YwoXscKAEuB04D/18kgxGRNOAXwJ+ranckYwkmSHxRc/9U1aOqlwOFQIWIXBqpWIKZIL6ouH8icgvQoqr7wnG9+ZLom4CVo14XAqciFEtQqnrK/98W4Jf4yk3R6Ky/vhuo87ZEOJ4LqOpZ/w+gF/gXIngf/bXbXwA/UdUX/Zuj5v4Fiy+a7l+AqnYCO/DVv6Pm/gWMji+K7t9G4PP+Z3/PA1tE5MdM8/7Nl0S/FygTEZeIJAC3Ay9FOKbzRCTV/0AMEUkFPgUcnvysiHkJ+Kr/+68C/x7BWMYJfIj9biNC99H/sO4Z4Kiq/u9Ru6Li/k0UXxTdvzwRyfJ/nwx8AjhG9Ny/oPFFy/1T1f+qqoWqWoQv321X1a8w3funqvPiC/gMvp43tcB/j3Q8Y2IrBt7zfx2JlviA5/D9+TmC76+iu4Fc4HWg2v/fnCiL71ngfeCQ/0O9PEKxfRxfefAQcND/9ZlouX+TxBct928dcMAfx2HgIf/2aLl/E8UXFfdvTKw3AC/P5P7Ni+6Vxhhjpm++lG6MMcZMkyV6Y4xZ4CzRG2PMAmeJ3hhjFjhL9MYYs8BZojfGmAXOEr0xxixw/xdqms+yMarLQAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(1,40), error_list)\n",
    "np.argmin(error_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ededf54",
   "metadata": {},
   "source": [
    "Decision Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "b9f08c37",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.80      0.83        54\n",
      "           1       0.73      0.83      0.78        36\n",
      "\n",
      "    accuracy                           0.81        90\n",
      "   macro avg       0.80      0.81      0.81        90\n",
      "weighted avg       0.82      0.81      0.81        90\n",
      "\n",
      "[[43 11]\n",
      " [ 6 30]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "dt = DecisionTreeClassifier().fit(X_train, y_train)\n",
    "\n",
    "dt_predictions = dt.predict(X_test)\n",
    "\n",
    "print(classification_report(y_test, dt_predictions))\n",
    "print(confusion_matrix(y_test, dt_predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "8678fccc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.83      0.86        54\n",
      "           1       0.77      0.83      0.80        36\n",
      "\n",
      "    accuracy                           0.83        90\n",
      "   macro avg       0.83      0.83      0.83        90\n",
      "weighted avg       0.84      0.83      0.83        90\n",
      "\n",
      "[[45  9]\n",
      " [ 6 30]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "rfc = RandomForestClassifier().fit(X_train, y_train)\n",
    "\n",
    "rfc_predictions = rfc.predict(X_test)\n",
    "\n",
    "print(classification_report(y_test, rfc_predictions))\n",
    "print(confusion_matrix(y_test, rfc_predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a4e9ace",
   "metadata": {},
   "source": [
    "# Kaggle Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "6ec8c6a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_preds = lr.predict(test)\n",
    "\n",
    "df = pd.DataFrame({\"PassengerId\": test_ids.values,\n",
    "                  \"Survived\" : submission_preds,\n",
    "                 })\n",
    "\n",
    "df.to_csv(\"submission_lr.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "id": "8ee7618a",
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_preds = knn.predict(test)\n",
    "\n",
    "df = pd.DataFrame({\"PassengerId\": test_ids.values,\n",
    "                  \"Survived\" : submission_preds,\n",
    "                 })\n",
    "df.to_csv(\"submission_knn.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "380295d8",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}