Titanic.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import statistics\n",
    "import os\n",
    "from sklearn.model_selection import KFold,cross_val_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import SVC, LinearSVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.linear_model import Perceptron\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "import xgboost as xgb\n",
    "\n",
    "#Pulling data\n",
    "Data_path = 'C:/Users/Imran/Documents/Data Science/Projects/Titanic/'\n",
    "Train_df = pd.read_csv(Data_path + 'train.csv')\n",
    "Test_df = pd.read_csv(Data_path + 'test.csv')\n",
    "\n",
    "#Combining both datasets, this allows for both datasets to analyzed and manipulated together\n",
    "Train_df['Dataset'] = 1\n",
    "Test_df['Dataset'] = 0\n",
    "Combined_df = Train_df.append(Test_df,sort=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>**Lets start by reviewing the feature details:**\n",
    "\n",
    "<br>**Passenger ID**:     A unique passenger ID number from 1-891\n",
    "<br>**Survived**:  \t    1 = survived, 0 = didn't survive\n",
    "<br>**Social Class**:     A measure social of status (1 = Upper Class, 2 = Middle, 3 = Lower)\n",
    "<br>**Name**:     \t    Name of the passenger\n",
    "<br>**Gender**:    \t    Male or Female\n",
    "<br>**Age**:      \t    Age in years\n",
    "<br>**Sib/Spouse**: \t    The number of siblings or spouses the passenger had on board\n",
    "<br>**Parents/children**: The number of parents or children the passenger had on board\n",
    "<br>**Ticket**:   \t    Ticket number/code\n",
    "<br>**Fare**:     \t    Price paid for the ticket\n",
    "<br>**Cabin**:    \t    Code of the Cabin the passenger was staying in\n",
    "<br>**Embarked**: \t    Port on which they boarded the titanic (C = Cherbourg, Q = Queenstown, S = Southampton)\n",
    "\n",
    "<br>**First Parameters to drop are:\n",
    "<br>**Passenger Id** – Just a unique number (1,2,3)\n",
    "<br>**Name** –         We can’t use this to classify or group the passenger.\n",
    "<br>**Ticket** –       Random combination of letters and numbers (A/5 21171, PC 17599)\n",
    "<br>**Cabin** –        Letter and number combination(C85, C123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "Combined_df = Combined_df.drop(['PassengerId','Name','Ticket','Cabin'],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>**Checking for Missing data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Survived    418\n",
       "Pclass        0\n",
       "Sex           0\n",
       "Age         263\n",
       "SibSp         0\n",
       "Parch         0\n",
       "Fare          1\n",
       "Embarked      2\n",
       "Dataset       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Combined_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>Age, Fare and Embarked have missing values, these will be dealt with by:\n",
    "<br>Age & Fare: Filling blanks with the average age of the sample\n",
    "<br>Embarked: Fill with the most common value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "Combined_df['Age'] = Combined_df['Age'].fillna(np.mean(Combined_df['Age']))\n",
    "Combined_df['Fare'] = Combined_df['Fare'].fillna(np.mean(Combined_df['Fare']))\n",
    "Combined_df['Embarked'] = Combined_df['Embarked'].fillna(statistics.mode(Combined_df['Embarked']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>Rechecking for Missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Survived    418\n",
       "Pclass        0\n",
       "Sex           0\n",
       "Age           0\n",
       "SibSp         0\n",
       "Parch         0\n",
       "Fare          0\n",
       "Embarked      0\n",
       "Dataset       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Combined_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "for feature in Combined_df:\n",
    "    if feature != 'Dataset':\n",
    "        if feature == 'Survived':\n",
    "            plt.hist(Combined_df[feature][Combined_df['Survived'].notnull()])\n",
    "        else:\n",
    "            plt.hist(Combined_df[feature])\n",
    "        plt.title('Distribution of {}'.format(feature))\n",
    "        plt.ylabel('Frequency')\n",
    "        plt.xlabel(feature)\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Initial thoughts on feature distribution:**\n",
    "<br>**Survived:**  \t    38% of passengers in the training dataset survived\n",
    "<br>**Social Class:**   The majority of passengers where of lower social status, while there was an equal spread between middle and upper class\n",
    "<br>**Gender:**    \t    64% of passengers were male\n",
    "<br>**Age:**      \t    Distribution of age generally follows a normal distribution with a right skew\n",
    "<br>**Sib/Spouse:** \t    The majority of passengers were travelling without siblings, the largest group size was 8\n",
    "<br>**Parents/children:** The majority of passengers were travelling without parents or children, the largest group size was 6\n",
    "<br>**Fare:**     \t    The majority of fares were under 50 (assuming local currency or all have been converted), with a few premium tickets, the highest was 512\n",
    "<br>**Embarked:** \t    The majority of passengers embarked at Southampton\n",
    "\n",
    "\n",
    "<br>Now looking at survival rates by each feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "plt.clf()\n",
    "for feature in Combined_df[Combined_df['Survived'].notnull()]:\n",
    "    if feature not in ['Survived','Dataset']:\n",
    "        Chart_df = Combined_df[Combined_df['Survived'].notnull()][[feature,'Survived']].groupby(feature).agg(np.mean)\n",
    "        Chart_df.reset_index(inplace = True)\n",
    "        Chart_df['Survived'] = Chart_df['Survived']*100\n",
    "        plt.bar(Chart_df[feature],Chart_df['Survived'])\n",
    "        plt.title('Survival rate by {}'.format(feature))\n",
    "        plt.ylabel('Survival rate')\n",
    "        plt.xlabel(feature)\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Model Building</h1>\n",
    "<br>We'll now start the model building process. Before beginning we'll need to clean up the data.\n",
    "<br>Transformation notes:\n",
    "<br>Social Class,Sib/Spouse and Parents/children have been left as they are (numerical values from 0-9)\n",
    "<br>Age has been split into 6 intervals and assigned values from 0-5\n",
    "<br>Gender has been mapped to 0 for Males and 1 for Females\n",
    "<br>Fare has been split into 8 intervals and assigned values from 0-7\n",
    "<br>Embarked will be split into one column per location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "Combined_df['Age'] = pd.qcut(Combined_df['Age'], q=6,labels=list(range(0,6))).astype('int')\n",
    "Combined_df['Sex'] = Combined_df['Sex'].map({'male':0,'female':1})\n",
    "Combined_df['Fare'] = pd.qcut(Combined_df['Fare'], q=8,labels=list(range(0,8))).astype('int')\n",
    "Combined_df = pd.get_dummies(Combined_df,columns=['Embarked'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>**Model Selection:**\n",
    "<br>We'll now start the process of building our model. Firstly we'll fit a model using all available features to get a baseline measure of accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Algorithm</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>79.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LinearSVC</td>\n",
       "      <td>79.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>78.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>77.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>75.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Perceptron</td>\n",
       "      <td>75.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GaussianNB</td>\n",
       "      <td>74.98</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Algorithm  Score\n",
       "5  RandomForestClassifier  79.02\n",
       "4               LinearSVC  79.01\n",
       "0      LogisticRegression  78.34\n",
       "6  DecisionTreeClassifier  77.90\n",
       "1    KNeighborsClassifier  75.76\n",
       "3              Perceptron  75.20\n",
       "2              GaussianNB  74.98"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Train_df = Combined_df[Combined_df['Dataset']==1].drop('Dataset',axis=1)\n",
    "Test_df = Combined_df[Combined_df['Dataset']==0].drop(['Dataset','Survived'],axis=1)\n",
    "\n",
    "X_train = Train_df.drop('Survived',axis=1)\n",
    "Y_train = Train_df['Survived']\n",
    "\n",
    "Algorithms = [LogisticRegression(),KNeighborsClassifier(n_neighbors = 3),GaussianNB(),Perceptron(),LinearSVC(max_iter = 4000),RandomForestClassifier(n_estimators=100),DecisionTreeClassifier()]\n",
    "#Scoring will be done using 5-fold cross validation\n",
    "def produce_scoring_grid(Algorithms,X_train,Y_train):\n",
    "    Algorithm_Labels = [str(x).split('(')[0] for x in Algorithms]\n",
    "    Results = []\n",
    "    kf = KFold(5,shuffle = True,random_state = 42).get_n_splits(X_train.values)\n",
    "    for Algo in Algorithms:\n",
    "        range_score = cross_val_score(Algo,X_train.values,Y_train,scoring = 'accuracy',cv=kf)\n",
    "        Results.append(round(range_score.mean()*100,2))\n",
    "    Results_df = pd.DataFrame(list(zip(Algorithm_Labels,Results)),columns = ['Algorithm','Score']).sort_values('Score',ascending = False)\n",
    "    return Results_df\n",
    "\n",
    "produce_scoring_grid(Algorithms,X_train,Y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Accuracy is currently around 76-80% for most algorithms with Random Forest performing the best.\n",
    "\n",
    "<br>**Model Refinement:**\n",
    "Now that we've established a base working model (with around 80% accuracy) we can now go about fine tuning our model for optimal performance, for this we'll evaluate 2 methods:\n",
    "<br>**Feature engineering:**\n",
    "Quite a few entries for this problem on Kaggle have engineered various features, such as attempting to pull a persons first and last name, given the features available and the minimal impact on survival rates, I don't personally believe that any additional features can be created.\n",
    "<br>**Feature selection:**\n",
    "The majority of Kaggle entries have included all features and simply allowed the ML algorithm to adjust around this, I believe this works fine for larger datasets, but for smaller datasets (the training set has 891 entries) any unnecessary features will simply introduce noise into the model, and hence reduce the accuracy\n",
    "<br>From our earlier charts the 3 most significant features (in terms of influence on survival rates) were Social Class, Gender and Embarked, from running different simulations,these 3 features produced the greatest increase in accuracy from our base model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>81.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>81.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>77.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LinearSVC</td>\n",
       "      <td>77.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>77.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GaussianNB</td>\n",
       "      <td>76.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Perceptron</td>\n",
       "      <td>75.42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Algorithm  Score\n",
       "5  RandomForestClassifier  81.15\n",
       "6  DecisionTreeClassifier  81.15\n",
       "1    KNeighborsClassifier  77.78\n",
       "4               LinearSVC  77.67\n",
       "0      LogisticRegression  77.33\n",
       "2              GaussianNB  76.66\n",
       "3              Perceptron  75.42"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "produce_scoring_grid(Algorithms,X_train[['Sex','Pclass','Embarked_C','Embarked_Q','Embarked_S']],Y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now to try and drive some understanding and results from our findings:\n",
    "<br>**Gender:**\n",
    "<br>Females were more likely to survive on the titanic, given that when evacuating the ship there were a limited number of lifeboats, there will be priority groups in terms of who to evacuate first. Given that in these types of scenarios women and children are usually evacuated first, this would help explain why women were more likely to survive.\n",
    "<br>This does raise the question of why Age wasn't as significant on survival rate. Well looking at the distribution of passenger ages the majority of passengers were aged between 20-40, generally there were a minimal amount of children, and as our age variable is transformed into a range, the lack of variation in the older age ranges results in minimal improvments to the model by including it. Retrospectively I've since reviewed Age by spliting into Adult/Children (Age <16 coded as Children), the survival rate for Children is 54% vs 36% for Adults. Therefore we would also include this feature in our model.\n",
    "<br>**Pclass:**\n",
    "<br>Those of higher social status were more likely to survive on titanic, again given the limited number of lifeboats, those of higher social status are more likely to be able to use thier influence to secure themselves a place on the lifeboats compared to those of lower social status.\n",
    "<br>**Embarked:**\n",
    "<br>Those who boarded in Cherbourg had a 20% higher survival rate than those who boarded in Queenstown or Southampton. Now in terms of providing a logical reason, this isn't as clear cut as the previous 2. The ordering of the route of the titanic was from Southampton, Cherbourg, Queenstown then on route to New York. So why did the passengers in the middle pick up point have a 20% higher survival rate? Personally I don't have an exact answer to this, my only possible guess could be due to cabin locations, if the cabin locations were assigned as and when each passenger boarded the ship then it's likely that these passengers were all assigned cabins in a specific region of the ship, and this region could have been closer to a specific exit point hence increasing their chances of survival. Cabin information is available but is missing for 3 quaters of the training dataset, if more cabin information was available then it would have been worth investigating this further alongside a map of the ship to understand if certain locations resulted in an increased survival rate."
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}