Random_Forest.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Dependencies\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.externals import joblib\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass  Sex   Age  SibSp  Parch     Fare  Embarked\n",
       "0         0       3    0  22.0      1      0   7.2500         0\n",
       "1         1       1    1  38.0      1      0  71.2833         1\n",
       "2         1       3    1  26.0      0      0   7.9250         0\n",
       "3         1       1    1  35.0      1      0  53.1000         0\n",
       "4         0       3    0  35.0      0      0   8.0500         0"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#reading the dataset\n",
    "train_data = pd.read_csv(\"Resources/cleaned_train.csv\")\n",
    "#Display dataset used in scaling\n",
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " # Attempt 1: Without Scaling or Binning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "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>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass  Sex   Age  SibSp  Parch     Fare  Embarked\n",
       "0       3    0  22.0      1      0   7.2500         0\n",
       "1       1    1  38.0      1      0  71.2833         1\n",
       "2       3    1  26.0      0      0   7.9250         0\n",
       "3       1    1  35.0      1      0  53.1000         0\n",
       "4       3    0  35.0      0      0   8.0500         0"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Test is the X values for the data\n",
    "test = train_data.drop(labels='Survived', axis=1)\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(714,)"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Target is the y values for the data\n",
    "target = train_data['Survived']\n",
    "target.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split the data into train and test data\n",
    "X_train, X_test, y_train, y_test = train_test_split(test, target, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(535, 7)"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(179, 7)"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Value:  0.8212290502793296\n",
      "Minimum Value:  0.7430167597765364\n",
      "[16]\n"
     ]
    }
   ],
   "source": [
    "#Test random forest classifier with different numbers of decision trees\n",
    "#Save the number of trees and accuracy in a dictionary\n",
    "d = {}\n",
    "for i in np.arange(1,100):\n",
    "    clf = RandomForestClassifier(n_estimators=i)\n",
    "    clf_1 = clf.fit(X_train, y_train)\n",
    "    d.update({i:clf_1.score(X_test,y_test)})\n",
    "#finding the max and min accuracy scores\n",
    "my_dict = d\n",
    "key_max = max(my_dict.keys(), key=(lambda k: my_dict[k]))\n",
    "key_min = min(my_dict.keys(), key=(lambda k: my_dict[k]))\n",
    "\n",
    "print('Maximum Value: ',my_dict[key_max])\n",
    "print('Minimum Value: ',my_dict[key_min])\n",
    "\n",
    "#Find the corresponding number of trees where the result occurs.\n",
    "a = [key for key,values in d.items() if values == d[key_max]]\n",
    "print (a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x28ff668e080>]"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#At this point we noticed that the accuracy values and number of trees that they occur at change drastically with each iteration. This meant that\n",
    "#iterating over number of trees was not an effective method for optimizing accuracy. Since we know that accuracy in general increases as number of trees\n",
    "#increases, we plot the accuracy based on number of trees and look for the first region where the data becomes stable. We do this so we use the minimum \n",
    "#number of decision trees where the variance drops off. This maximizes accuracy per computation time because we want the prediction to run fast because it\n",
    "#is hosted on Heroku. Stability occurs around 40 trees regardless of number of re-runs.\n",
    "plt.plot(my_dict.keys(), my_dict.values())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.776536312849162"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Score the model with 50 trees but use KFold to account for variation\n",
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "clf.fit(X_train, y_train)\n",
    "clf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.7083333333333334, 0.8194444444444444, 0.75, 0.7916666666666666, 0.8028169014084507, 0.7746478873239436, 0.7746478873239436, 0.7887323943661971, 0.8169014084507042, 0.8873239436619719]\n",
      "0.7914514866979656\n"
     ]
    }
   ],
   "source": [
    "#Unit testing was performed by executing kfold validation. This splits the data up into segments and uses 1 fold as test data\n",
    "#and k-1 folds as training data. This returns k results with each fold being used as testing data exactly once.\n",
    "#I then averaged the results to get a more accurate algorithm accuracy.\n",
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "kfold_test_scores = []\n",
    "# 10 folds of the data with the same random state\n",
    "kf = KFold(n_splits=10, random_state=42)\n",
    "#Get indices for each split (the data has 714 rows: the first split contains the slice indices from 0 to 72 as test data, and 73-713 as train data)\n",
    "for train_index, test_index in kf.split(test):\n",
    "    X_train, X_test = test.iloc[train_index], test.iloc[test_index]\n",
    "    y_train, y_test = target[train_index], target[test_index]\n",
    "    #Fit the model with this split, score it, and append the score to the list\n",
    "    clf.fit(X_train, y_train)\n",
    "    test_score = clf.score(X_test, y_test)\n",
    "    kfold_test_scores.append(test_score)\n",
    "#Print all scores, then average them\n",
    "print(kfold_test_scores)\n",
    "avg = (sum(kfold_test_scores)/len(kfold_test_scores))\n",
    "print(avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Attempt 2: Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Attempt 2\n",
    "\n",
    "#Create a scaler that scales the data down to a standard normal (Guassian) distribution (mean=0, standard deviation=1)\n",
    "X_scaler = StandardScaler().fit(test)\n",
    "X_scaled = X_scaler.transform(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_2, X_test_2, y_train_2, y_test_2 = train_test_split(X_scaled, target, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Value:  0.8156424581005587\n",
      "Minimum Value:  0.7541899441340782\n",
      "[20]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x28ff6762748>]"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Test random forest classifier with different numbers of decision trees\n",
    "#Save the number of trees and accuracy in a dictionary\n",
    "d_2 = {}\n",
    "\n",
    "for i in np.arange(1,100):\n",
    "\n",
    "    clf = RandomForestClassifier(n_estimators=i)\n",
    "    clf_2 = clf.fit(X_train_2, y_train_2)\n",
    "    d_2.update({i:clf_2.score(X_test_2,y_test_2)})\n",
    "my_dict_2 = d_2\n",
    "\n",
    "key_max_2 = max(my_dict_2.keys(), key=(lambda k: my_dict_2[k]))\n",
    "key_min_2 = min(my_dict_2.keys(), key=(lambda k: my_dict_2[k]))\n",
    "\n",
    "print('Maximum Value: ',my_dict_2[key_max_2])\n",
    "print('Minimum Value: ',my_dict_2[key_min_2])\n",
    "\n",
    "max_tree = [key for key,values in d_2.items() if values == d_2[key_max_2]]\n",
    "print (max_tree)\n",
    "\n",
    "plt.plot(my_dict_2.keys()  , my_dict_2.values())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7932960893854749"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Stable around 50 trees.\n",
    "#Score the model with 50 trees but use KFold to account for variation\n",
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "clf.fit(X_train_2, y_train_2)\n",
    "clf.score(X_test_2, y_test_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.6805555555555556, 0.8333333333333334, 0.7361111111111112, 0.8194444444444444, 0.8591549295774648, 0.8028169014084507, 0.7464788732394366, 0.8028169014084507, 0.8169014084507042, 0.8873239436619719]\n",
      "0.7984937402190925\n"
     ]
    }
   ],
   "source": [
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "kfold_test_scores = []\n",
    "# 10 folds of the data with the same random state\n",
    "kf = KFold(n_splits=10, random_state=42)\n",
    "#Get indices for each split (the data has 714 rows: the first split contains the slice indices from 0 to 72 as test data, and 73-713 as train data)\n",
    "for train_index, test_index in kf.split(X_scaled):\n",
    "    X_train_2, X_test_2 = X_scaled[train_index], X_scaled[test_index]\n",
    "    y_train_2, y_test_2 = target[train_index], target[test_index]\n",
    "    #Fit the model with this split, score it, and append the score to the list\n",
    "    clf.fit(X_train_2, y_train_2)\n",
    "    test_score = clf.score(X_test_2, y_test_2)\n",
    "    kfold_test_scores.append(test_score)\n",
    "#Print all scores, then average them\n",
    "print(kfold_test_scores)\n",
    "avg = (sum(kfold_test_scores)/len(kfold_test_scores))\n",
    "print(avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "#After running all methods, the scaler and regular methods are about equally accurate. I choose to save the scaler method (Accuracy=0.80)\n",
    "#Save the scaler and the svc model to pkl files\n",
    "joblib.dump(X_scaler, 'Resources/models/rf_scaler.pkl');\n",
    "joblib.dump(clf, 'Resources/models/rf.pkl');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Attempt - 3 (Binning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass  Sex   Age  SibSp  Parch  Fare  Embarked\n",
       "0         0       3    0  22.0      1      0     0         0\n",
       "1         1       1    1  38.0      1      0     0         1\n",
       "2         1       3    1  26.0      0      0     0         0\n",
       "3         1       1    1  35.0      1      0     0         0\n",
       "4         0       3    0  35.0      0      0     0         0"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Prepare Dataset to be used in binning\n",
    "binned_data = pd.read_csv('Resources/AG_train-test_v2.csv')\n",
    "#test_2 corresponds to our X data for binning method\n",
    "test_2 = binned_data.drop('Survived', axis = 1)\n",
    "test_2.drop('Unnamed: 0', axis = 1, inplace=True)\n",
    "binned_data.drop('Unnamed: 0', axis = 1, inplace=True)\n",
    "#target is the same as before\n",
    "target = binned_data['Survived']\n",
    "binned_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "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>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass  Sex   Age  SibSp  Parch  Fare  Embarked\n",
       "0       3    0  22.0      1      0     0         0\n",
       "1       1    1  38.0      1      0     0         1\n",
       "2       3    1  26.0      0      0     0         0\n",
       "3       1    1  35.0      1      0     0         0\n",
       "4       3    0  35.0      0      0     0         0"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_3, X_test_3, y_train_3, y_test_3 = train_test_split(test_2, target, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Value:  0.7966101694915254\n",
      "Minimum Value:  0.7005649717514124\n",
      "[16]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x28ff661a2b0>]"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "d_3 = {}\n",
    "\n",
    "for i in np.arange(1,100):\n",
    "\n",
    "    clf = RandomForestClassifier(n_estimators=i)\n",
    "    clf_3 = clf.fit(X_train_3, y_train_3)\n",
    "    d_3.update({i:clf_3.score(X_test_3,y_test_3)})\n",
    "my_dict_3 = d_3\n",
    "\n",
    "key_max_3 = max(my_dict_3.keys(), key=(lambda k: my_dict_3[k]))\n",
    "key_min_3 = min(my_dict_3.keys(), key=(lambda k: my_dict_3[k]))\n",
    "\n",
    "print('Maximum Value: ',my_dict_3[key_max_3])\n",
    "print('Minimum Value: ',my_dict_3[key_min_3])\n",
    "\n",
    "max_tree_3 = [key for key,values in d_3.items() if values == d_3[key_max_3]]\n",
    "print (max_tree_3)\n",
    "\n",
    "plt.plot(my_dict_3.keys()  , my_dict_3.values())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.751412429378531"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#The region becomes stable around 30 trees. Since we want to use the same number of trees for each test we use 50.\n",
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "clf.fit(X_train_3, y_train_3)\n",
    "clf.score(X_test_3, y_test_3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.5070422535211268, 0.6197183098591549, 0.5070422535211268, 0.4788732394366197, 0.5211267605633803, 0.5492957746478874, 0.5492957746478874, 0.5142857142857142, 0.5571428571428572, 0.5714285714285714]\n",
      "0.5375251509054325\n"
     ]
    }
   ],
   "source": [
    "#Confirm using KFold\n",
    "clf = RandomForestClassifier(n_estimators=50)\n",
    "kfold_test_scores = []\n",
    "# 10 folds of the data with the same random state\n",
    "kf = KFold(n_splits=10, random_state=42)\n",
    "#Get indices for each split (the data has 714 rows: the first split contains the slice indices from 0 to 72 as test data, and 73-713 as train data)\n",
    "for train_index, test_index in kf.split(test_2):\n",
    "    X_train_3, X_test_3 = test_2.iloc[train_index], test_2.iloc[test_index]\n",
    "    y_train_3, y_test_3 = target[train_index], target[test_index]\n",
    "    #Fit the model with this split, score it, and append the score to the list\n",
    "    clf.fit(X_train_3, y_train_3)\n",
    "    test_score = clf.score(X_test_3, y_test_3)\n",
    "    kfold_test_scores.append(test_score)\n",
    "#Print all scores, then average them\n",
    "print(kfold_test_scores)\n",
    "avg = (sum(kfold_test_scores)/len(kfold_test_scores))\n",
    "print(avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}