Keras_Titanic_MLP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy\n",
    "import pandas as pd\n",
    "from sklearn import preprocessing\n",
    "numpy.random.seed(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df = pd.read_excel(\"data/titanic3.xls\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols=['survived', 'name', 'pclass', 'sex', 'age', 'sibsp', 'parch', 'fare', 'embarked']\n",
    "all_df = all_df[cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "msk = numpy.random.rand( len(all_df) ) < 0.8\n",
    "train_df = all_df[msk]\n",
    "test_df = all_df[~msk]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total: 1309\n",
      "train: 1034\n",
      "test: 275\n"
     ]
    }
   ],
   "source": [
    "print('total:', len(all_df))\n",
    "print('train:', len(train_df))\n",
    "print('test:', len(test_df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define the preprocess function  \n",
    "def preprocess_data(raw_df):\n",
    "    df = raw_df.drop( ['name'], axis=1 )\n",
    "    age_mean = df['age'].mean()\n",
    "    df['age'] = df['age'].fillna(age_mean)\n",
    "    fare_mean = df['fare'].mean()\n",
    "    df['fare'] = df['fare'].fillna(fare_mean)\n",
    "    df['sex'] = df['sex'].map( {'female':0, 'male':1} ).astype(int)\n",
    "    x_one_hot_df = pd.get_dummies( data=df, columns=[\"embarked\" ])\n",
    "    \n",
    "    nd_array = x_one_hot_df.values\n",
    "    feature = nd_array[:, 1:]\n",
    "    label = nd_array[:, 0]\n",
    "    \n",
    "    minmax_scale = preprocessing.MinMaxScaler( feature_range=(0,1) )\n",
    "    scaled_features = minmax_scale.fit_transform( feature )\n",
    "    \n",
    "    return scaled_features,label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_features, train_label = preprocess_data( train_df )\n",
    "test_features, test_label = preprocess_data( test_df )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.        , 0.38844819, 0.        , 0.        ,\n",
       "        0.41250333, 0.        , 0.        , 1.        ],\n",
       "       [0.        , 1.        , 0.00679502, 0.125     , 0.22222222,\n",
       "        0.2958059 , 0.        , 0.        , 1.        ],\n",
       "       [0.        , 0.        , 0.02151711, 0.125     , 0.22222222,\n",
       "        0.2958059 , 0.        , 0.        , 1.        ],\n",
       "       [0.        , 1.        , 0.40203823, 0.125     , 0.22222222,\n",
       "        0.2958059 , 0.        , 0.        , 1.        ],\n",
       "       [0.        , 0.        , 0.33408803, 0.125     , 0.22222222,\n",
       "        0.2958059 , 0.        , 0.        , 1.        ]])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_features[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 1., 0., 0., 0.])"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_label[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_7 (Dense)              (None, 40)                400       \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 30)                1230      \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 1)                 31        \n",
      "=================================================================\n",
      "Total params: 1,661\n",
      "Trainable params: 1,661\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add( Dense(units=40, input_dim=9, kernel_initializer='uniform', activation='relu') )\n",
    "\n",
    "model.add( Dense(units=30, kernel_initializer='uniform', activation='relu') )\n",
    "\n",
    "model.add( Dense(units=1, kernel_initializer='uniform', activation='sigmoid') )\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile( loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 930 samples, validate on 104 samples\n",
      "Epoch 1/30\n",
      " - 0s - loss: 0.6901 - acc: 0.5774 - val_loss: 0.6713 - val_acc: 0.7885\n",
      "Epoch 2/30\n",
      " - 0s - loss: 0.6681 - acc: 0.5957 - val_loss: 0.5915 - val_acc: 0.7885\n",
      "Epoch 3/30\n",
      " - 0s - loss: 0.6101 - acc: 0.6677 - val_loss: 0.4942 - val_acc: 0.8173\n",
      "Epoch 4/30\n",
      " - 0s - loss: 0.5454 - acc: 0.7634 - val_loss: 0.4624 - val_acc: 0.7788\n",
      "Epoch 5/30\n",
      " - 0s - loss: 0.5062 - acc: 0.7624 - val_loss: 0.4552 - val_acc: 0.7885\n",
      "Epoch 6/30\n",
      " - 0s - loss: 0.4911 - acc: 0.7645 - val_loss: 0.4399 - val_acc: 0.7788\n",
      "Epoch 7/30\n",
      " - 0s - loss: 0.4823 - acc: 0.7591 - val_loss: 0.4424 - val_acc: 0.7885\n",
      "Epoch 8/30\n",
      " - 0s - loss: 0.4789 - acc: 0.7581 - val_loss: 0.4344 - val_acc: 0.7885\n",
      "Epoch 9/30\n",
      " - 0s - loss: 0.4734 - acc: 0.7645 - val_loss: 0.4274 - val_acc: 0.7885\n",
      "Epoch 10/30\n",
      " - 0s - loss: 0.4697 - acc: 0.7753 - val_loss: 0.4252 - val_acc: 0.8077\n",
      "Epoch 11/30\n",
      " - 0s - loss: 0.4657 - acc: 0.7602 - val_loss: 0.4219 - val_acc: 0.8173\n",
      "Epoch 12/30\n",
      " - 0s - loss: 0.4639 - acc: 0.7731 - val_loss: 0.4187 - val_acc: 0.8365\n",
      "Epoch 13/30\n",
      " - 0s - loss: 0.4624 - acc: 0.7806 - val_loss: 0.4184 - val_acc: 0.8365\n",
      "Epoch 14/30\n",
      " - 0s - loss: 0.4607 - acc: 0.7753 - val_loss: 0.4187 - val_acc: 0.8365\n",
      "Epoch 15/30\n",
      " - 0s - loss: 0.4627 - acc: 0.7849 - val_loss: 0.4200 - val_acc: 0.8269\n",
      "Epoch 16/30\n",
      " - 0s - loss: 0.4563 - acc: 0.7796 - val_loss: 0.4163 - val_acc: 0.8365\n",
      "Epoch 17/30\n",
      " - 0s - loss: 0.4550 - acc: 0.7882 - val_loss: 0.4163 - val_acc: 0.8365\n",
      "Epoch 18/30\n",
      " - 0s - loss: 0.4542 - acc: 0.7892 - val_loss: 0.4116 - val_acc: 0.7981\n",
      "Epoch 19/30\n",
      " - 0s - loss: 0.4537 - acc: 0.7957 - val_loss: 0.4213 - val_acc: 0.8077\n",
      "Epoch 20/30\n",
      " - 0s - loss: 0.4509 - acc: 0.7892 - val_loss: 0.4124 - val_acc: 0.8173\n",
      "Epoch 21/30\n",
      " - 0s - loss: 0.4508 - acc: 0.7914 - val_loss: 0.4128 - val_acc: 0.8077\n",
      "Epoch 22/30\n",
      " - 0s - loss: 0.4494 - acc: 0.7978 - val_loss: 0.4135 - val_acc: 0.8269\n",
      "Epoch 23/30\n",
      " - 0s - loss: 0.4523 - acc: 0.7860 - val_loss: 0.4148 - val_acc: 0.8269\n",
      "Epoch 24/30\n",
      " - 0s - loss: 0.4489 - acc: 0.7860 - val_loss: 0.4162 - val_acc: 0.8269\n",
      "Epoch 25/30\n",
      " - 0s - loss: 0.4491 - acc: 0.7914 - val_loss: 0.4194 - val_acc: 0.8269\n",
      "Epoch 26/30\n",
      " - 0s - loss: 0.4502 - acc: 0.7968 - val_loss: 0.4175 - val_acc: 0.8269\n",
      "Epoch 27/30\n",
      " - 0s - loss: 0.4460 - acc: 0.7957 - val_loss: 0.4138 - val_acc: 0.8077\n",
      "Epoch 28/30\n",
      " - 0s - loss: 0.4500 - acc: 0.7849 - val_loss: 0.4126 - val_acc: 0.8077\n",
      "Epoch 29/30\n",
      " - 0s - loss: 0.4458 - acc: 0.7935 - val_loss: 0.4169 - val_acc: 0.8269\n",
      "Epoch 30/30\n",
      " - 0s - loss: 0.4524 - acc: 0.7903 - val_loss: 0.4161 - val_acc: 0.8269\n"
     ]
    }
   ],
   "source": [
    "train_history = model.fit( x=train_features, y=train_label, validation_split=0.1, epochs=30, batch_size=30, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as pyplot\n",
    "def my_show_train_history(input_train_history, input_train, input_validation):\n",
    "    pyplot.plot(input_train_history.history[input_train] )\n",
    "    pyplot.plot(input_train_history.history[input_validation])\n",
    "    pyplot.title('Train History')\n",
    "    pyplot.ylabel(input_train)\n",
    "    # pyplot.ylabel('Accuracy')\n",
    "    pyplot.xlabel('Epoch')\n",
    "    pyplot.legend(['my_train','my_validation'], loc='upper left')\n",
    "    pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_show_train_history( train_history, 'acc', 'val_acc' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_show_train_history( train_history, 'loss', 'val_loss' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "275/275 [==============================] - 0s 54us/step\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate( x=test_features, y=test_label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8109090913425793"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# insert 'probability' into dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_features, all_label = preprocess_data( all_df )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_probability = model.predict( all_features )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.97696316],\n",
       "       [0.609896  ],\n",
       "       [0.9721666 ],\n",
       "       [0.40561366],\n",
       "       [0.97113806]], dtype=float32)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_probability[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd = all_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.insert( len(all_df.columns), 'probability', all_probability)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>name</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",
       "      <th>probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1299</th>\n",
       "      <td>0</td>\n",
       "      <td>Yasbeck, Mr. Antoni</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14.4542</td>\n",
       "      <td>C</td>\n",
       "      <td>0.224009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1300</th>\n",
       "      <td>1</td>\n",
       "      <td>Yasbeck, Mrs. Antoni (Selini Alexander)</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14.4542</td>\n",
       "      <td>C</td>\n",
       "      <td>0.765482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1301</th>\n",
       "      <td>0</td>\n",
       "      <td>Youseff, Mr. Gerious</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>45.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2250</td>\n",
       "      <td>C</td>\n",
       "      <td>0.158982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1302</th>\n",
       "      <td>0</td>\n",
       "      <td>Yousif, Mr. Wazli</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2250</td>\n",
       "      <td>C</td>\n",
       "      <td>0.227004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1303</th>\n",
       "      <td>0</td>\n",
       "      <td>Yousseff, Mr. Gerious</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14.4583</td>\n",
       "      <td>C</td>\n",
       "      <td>0.226095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>0</td>\n",
       "      <td>Zabour, Miss. Hileni</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>14.5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14.4542</td>\n",
       "      <td>C</td>\n",
       "      <td>0.767593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1305</th>\n",
       "      <td>0</td>\n",
       "      <td>Zabour, Miss. Thamine</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14.4542</td>\n",
       "      <td>C</td>\n",
       "      <td>0.696737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1306</th>\n",
       "      <td>0</td>\n",
       "      <td>Zakarian, Mr. Mapriededer</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>26.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2250</td>\n",
       "      <td>C</td>\n",
       "      <td>0.244171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1307</th>\n",
       "      <td>0</td>\n",
       "      <td>Zakarian, Mr. Ortin</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2250</td>\n",
       "      <td>C</td>\n",
       "      <td>0.241578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1308</th>\n",
       "      <td>0</td>\n",
       "      <td>Zimmerman, Mr. Leo</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.8750</td>\n",
       "      <td>S</td>\n",
       "      <td>0.144389</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      survived                                     name  pclass     sex   age  \\\n",
       "1299         0                      Yasbeck, Mr. Antoni       3    male  27.0   \n",
       "1300         1  Yasbeck, Mrs. Antoni (Selini Alexander)       3  female  15.0   \n",
       "1301         0                     Youseff, Mr. Gerious       3    male  45.5   \n",
       "1302         0                        Yousif, Mr. Wazli       3    male   NaN   \n",
       "1303         0                    Yousseff, Mr. Gerious       3    male   NaN   \n",
       "1304         0                     Zabour, Miss. Hileni       3  female  14.5   \n",
       "1305         0                    Zabour, Miss. Thamine       3  female   NaN   \n",
       "1306         0                Zakarian, Mr. Mapriededer       3    male  26.5   \n",
       "1307         0                      Zakarian, Mr. Ortin       3    male  27.0   \n",
       "1308         0                       Zimmerman, Mr. Leo       3    male  29.0   \n",
       "\n",
       "      sibsp  parch     fare embarked  probability  \n",
       "1299      1      0  14.4542        C     0.224009  \n",
       "1300      1      0  14.4542        C     0.765482  \n",
       "1301      0      0   7.2250        C     0.158982  \n",
       "1302      0      0   7.2250        C     0.227004  \n",
       "1303      0      0  14.4583        C     0.226095  \n",
       "1304      1      0  14.4542        C     0.767593  \n",
       "1305      1      0  14.4542        C     0.696737  \n",
       "1306      0      0   7.2250        C     0.244171  \n",
       "1307      0      0   7.2250        C     0.241578  \n",
       "1308      0      0   7.8750        S     0.144389  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd[-10:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>name</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",
       "      <th>probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Miss. Helen Loraine</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.972167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Mrs. Hudson J C (Bessie Waldo Daniels)</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>25.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.971138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>0</td>\n",
       "      <td>Evans, Miss. Edith Corse</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>36.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.6792</td>\n",
       "      <td>C</td>\n",
       "      <td>0.977837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>0</td>\n",
       "      <td>Isham, Miss. Ann Elizabeth</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>50.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28.7125</td>\n",
       "      <td>C</td>\n",
       "      <td>0.975898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>286</th>\n",
       "      <td>0</td>\n",
       "      <td>Straus, Mrs. Isidor (Rosalie Ida Blun)</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>63.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>221.7792</td>\n",
       "      <td>S</td>\n",
       "      <td>0.962791</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     survived                                             name  pclass  \\\n",
       "2           0                     Allison, Miss. Helen Loraine       1   \n",
       "4           0  Allison, Mrs. Hudson J C (Bessie Waldo Daniels)       1   \n",
       "105         0                         Evans, Miss. Edith Corse       1   \n",
       "169         0                       Isham, Miss. Ann Elizabeth       1   \n",
       "286         0           Straus, Mrs. Isidor (Rosalie Ida Blun)       1   \n",
       "\n",
       "        sex   age  sibsp  parch      fare embarked  probability  \n",
       "2    female   2.0      1      2  151.5500        S     0.972167  \n",
       "4    female  25.0      1      2  151.5500        S     0.971138  \n",
       "105  female  36.0      0      0   31.6792        C     0.977837  \n",
       "169  female  50.0      0      0   28.7125        C     0.975898  \n",
       "286  female  63.0      1      0  221.7792        S     0.962791  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd[ (pd['survived']==0) & (pd['probability']>0.9) ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>name</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",
       "      <th>probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Allen, Miss. Elisabeth Walton</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>29.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>211.3375</td>\n",
       "      <td>S</td>\n",
       "      <td>0.976963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Allison, Master. Hudson Trevor</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>0.9167</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.609896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Miss. Helen Loraine</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.972167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Mr. Hudson Joshua Creighton</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.405614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Mrs. Hudson J C (Bessie Waldo Daniels)</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>151.5500</td>\n",
       "      <td>S</td>\n",
       "      <td>0.971138</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   survived                                             name  pclass     sex  \\\n",
       "0         1                    Allen, Miss. Elisabeth Walton       1  female   \n",
       "1         1                   Allison, Master. Hudson Trevor       1    male   \n",
       "2         0                     Allison, Miss. Helen Loraine       1  female   \n",
       "3         0             Allison, Mr. Hudson Joshua Creighton       1    male   \n",
       "4         0  Allison, Mrs. Hudson J C (Bessie Waldo Daniels)       1  female   \n",
       "\n",
       "       age  sibsp  parch      fare embarked  probability  \n",
       "0  29.0000      0      0  211.3375        S     0.976963  \n",
       "1   0.9167      1      2  151.5500        S     0.609896  \n",
       "2   2.0000      1      2  151.5500        S     0.972167  \n",
       "3  30.0000      1      2  151.5500        S     0.405614  \n",
       "4  25.0000      1      2  151.5500        S     0.971138  "
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd[:5]"
   ]
  },
  {
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}