{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "<script>\n",
    "  function code_toggle() {\n",
    "    if (code_shown){\n",
    "      $('div.input').hide('500');\n",
    "      $('#toggleButton').val('Show Code')\n",
    "    } else {\n",
    "      $('div.input').show('500');\n",
    "      $('#toggleButton').val('Hide Code')\n",
    "    }\n",
    "    code_shown = !code_shown\n",
    "  }\n",
    "\n",
    "  $( document ).ready(function(){\n",
    "    code_shown=false;\n",
    "    $('div.input').hide()\n",
    "  });\n",
    "</script>\n",
    "<div align=\"right\">\n",
    "    <form action=\"javascript:code_toggle()\"><input type=\"submit\" id=\"toggleButton\" value=\"Show Code\"></form>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Singapore's Best Mobile Phone Plans\n",
    "This week, I'm taking a break from the HDB resale flat pricing model and turning my attention to a hotter topic: mobile phone plans.  \n",
    "  \n",
    "## Context\n",
    "This week, some of my colleagues noted an arbitrage opportunity in Singtel's offer for the iPhone XS / XS Max on its Combo 12 plan. The idea was to:  \n",
    "  \n",
    "1. Re-contract (or sign a new contract) on Singtel Combo 12 to get a free iPhone XS / XS Max (Singtel now charges \\$418++)  \n",
    "2. Sell the phone for approximately \\$1,800  \n",
    "3. Pay the monthly bill of \\$167.90 for Combo 12  (with a Corporate Individual Scheme [CIS] discount of 30%)  \n",
    "4. Repeat every 6 months\n",
    "  \n",
    "The 6-months cycle in table form:\n",
    "  \n",
    "|  Month  |  Income |              Expenditure             |   Profit   | Remarks                                                 |\n",
    "|:-------:|:-------:|:------------------------------------:|:----------:|---------------------------------------------------------|\n",
    "|    1    | \\$1,800 | \\\\$167.90 + \\\\$500 + \\\\$418 = \\\\$1087.90 | \\\\$1,132.10 | Re-contract on Combo 12, get a free iPhone, and sell it |\n",
    "|    2    |    -    |               \\$167.90               |  -\\$167.90 | Pay Combo 12 bill                                       |\n",
    "|    3    |    -    |               \\$167.90               |  -\\$167.90 | Pay Combo 12 bill                                       |\n",
    "|    4    |    -    |               \\$167.90               |  -\\$167.90 | Pay Combo 12 bill                                       |\n",
    "|    5    |    -    |               \\$167.90               |  -\\$167.90 | Pay Combo 12 bill                                       |\n",
    "|    6    |    -    |               \\$167.90               |  -\\$167.90 | Pay Combo 12 bill                                       |\n",
    "|  Total  | \\$1,800 |              \\$1,927.40              |  -\\$127.40 |                                                         |\n",
    "| Average |   \\$300 |               \\$321.23               |  -\\$21.23  |                                                         |  \n",
    "  \n",
    "By repeating this *forever*, one can pay an average monthly bill of \\$21.23 for Combo 12 perks. Of course, this means that you would be stuck with Singtel for as long as you want to enjoy the benefits, and that you'll eventually need to pay a full 24 months' worth of Combo 12 bills (or terminate your contract) to downgrade your plan. Nevertheless, you stand to gain from lower bills for more services assuming the following conditions hold:  \n",
    "  \n",
    "1. Singtel continues to offer the iPhone (the only phone with high resale value) at a low, stable price when signing up for the Combo 12 plan\n",
    "2. The Combo 12 plan continues to exist, and stays at the same price\n",
    "3. The resale value of the iPhone remains sufficiently high\n",
    "4. Apple comes up with a new model every year\n",
    "5. The CIS discount remains at 30% or better\n",
    "  \n",
    "Based on these assumptions, the deal was too risky for me, and I chose not to pursue it. After realising that the Singtel deal turned sour when Singtel increased the price for the iPhone XS on Combo 12 from \\\\$0 to \\\\$418, I was relieved that I hadn't jumped in. A colleague instead suggested that perhaps, I could do some analysis on the best mobile phone plan currently on offer. Hence, in this post, I do exactly that: I explore the determinants of mobile phone plan prices and outline an objective way to choose an ideal phone plan."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import required modules\n",
    "from IPython.display import Image  \n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pydea\n",
    "import pydotplus\n",
    "from scipy.spatial import ConvexHull\n",
    "import seaborn.apionly as sns\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn import tree\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from statsmodels.graphics.gofplots import qqplot\n",
    "from statsmodels.regression.linear_model import OLS, add_constant\n",
    "import warnings\n",
    "\n",
    "# Settings\n",
    "%matplotlib inline\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The Data\n",
    "I collected mobile phone plan data from the websites of Singtel, Starhub, M1, and Circles.Life. Specifically, I looked at postpaid mobile plans that were not restricted to any age group. I saved the monthly price, free talktime (in minutes), free SMS/MMS, data (in GB), provider, length of contract (in months), and whether or not the plan was a SIM-only plan. There were a total of 48 plans.  \n",
    "  \n",
    "Note that the maximum amount of talktime and SMS/MMS were 10,000 minutes and 10,000 respectively as per the Telcos' policies for unlimited talktime and SMS/MMS. The maximum amount of data was assumed to be 1,000 GB, although there was no cap on data in M1's policy. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>sms</th>\n",
       "      <th>data</th>\n",
       "      <th>provider</th>\n",
       "      <th>contract</th>\n",
       "      <th>sim_only</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>simonly10</td>\n",
       "      <td>36.05</td>\n",
       "      <td>150</td>\n",
       "      <td>500</td>\n",
       "      <td>12.0</td>\n",
       "      <td>singtel</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>simonly5</td>\n",
       "      <td>20.00</td>\n",
       "      <td>150</td>\n",
       "      <td>500</td>\n",
       "      <td>6.0</td>\n",
       "      <td>singtel</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>xs_sim12</td>\n",
       "      <td>24.00</td>\n",
       "      <td>200</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>s_sim12</td>\n",
       "      <td>34.00</td>\n",
       "      <td>400</td>\n",
       "      <td>0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>mysim20_12mth</td>\n",
       "      <td>20.00</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>5.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>mysim40_12mth</td>\n",
       "      <td>40.00</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>15.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>data_baller</td>\n",
       "      <td>48.00</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>circles_life</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>sweet_spot</td>\n",
       "      <td>28.00</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>circles_life</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             name  price  talktime  sms  data      provider  contract sim_only\n",
       "0       simonly10  36.05       150  500  12.0       singtel        12     True\n",
       "1        simonly5  20.00       150  500   6.0       singtel        12     True\n",
       "9        xs_sim12  24.00       200    0   9.0       starhub        12     True\n",
       "10        s_sim12  34.00       400    0  11.0       starhub        12     True\n",
       "27  mysim20_12mth  20.00       100  100   5.0            m1        12     True\n",
       "28  mysim40_12mth  40.00       100  100  15.0            m1        12     True\n",
       "45    data_baller  48.00       100    0  26.0  circles_life         0     True\n",
       "46     sweet_spot  28.00       100    0   6.0  circles_life         0     True"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read data\n",
    "mobile = pd.read_csv('mobile_plan_data.csv')\n",
    "\n",
    "# Remove unused columns\n",
    "mobile = mobile.drop(['perk', 'perk_value'], axis = 1)\n",
    "\n",
    "# Adjust \"unlimited\" figures\n",
    "mobile['talktime'][mobile.talktime == 43200] = 10000\n",
    "mobile['sms'][mobile.sms == 260000] = 10000\n",
    "mobile['data'][mobile.data == 999999] = 1000\n",
    "\n",
    "# Preview\n",
    "mobile.groupby('provider').head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploring the Data\n",
    "First, we explore the distributions of the various features. We note that most of the features were not normally distributed. Hence, we applied the log transformation to all numeric features, less the contract length, which we convert into a categorical feature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5c4882b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5c7446d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5c7a6668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5c979198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5cf8ce80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Linear Regression: Quantifying Linear Relationships\n",
    "Using Starhub and 24-month contracts as the reference categories for the service provider and contract features respectively, we obtain a regression with an adjusted R-squared value of **76%**. Here are the findings:  \n",
    "  \n",
    "1. A **1.00% increase** in free talktime corresponds to a **0.18% increase** in price.\n",
    "2. A **1.00% increase** in number of free SMS/MMS corresponds to a **0.07% *decrease*** in price (or simply negligible).\n",
    "3. A **1.00% increase** in data corresponds to a **0.16% increase** in price.\n",
    "4. A **SIM-only plan** is **0.69% cheaper** than a full mobile plan.\n",
    "5. M1 and Singtel plans are respectively **0.51% and 0.54% more expensive** than Starhub plans. Starhub plans and Circle.Life plans are competitively priced.\n",
    "6. There is **no significant difference in price** for plans with 24-month contracts, 12-month contracts, and plans without contracts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  price   R-squared:                       0.812\n",
      "Model:                            OLS   Adj. R-squared:                  0.767\n",
      "Method:                 Least Squares   F-statistic:                     18.23\n",
      "Date:                Sat, 29 Sep 2018   Prob (F-statistic):           3.09e-11\n",
      "Time:                        22:57:43   Log-Likelihood:                -9.8884\n",
      "No. Observations:                  48   AIC:                             39.78\n",
      "Df Residuals:                      38   BIC:                             58.49\n",
      "Df Model:                           9                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=========================================================================================\n",
      "                            coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------\n",
      "const                     2.7480      0.242     11.350      0.000       2.258       3.238\n",
      "talktime                  0.1880      0.033      5.721      0.000       0.121       0.254\n",
      "sms                      -0.0629      0.029     -2.153      0.038      -0.122      -0.004\n",
      "data                      0.1699      0.044      3.824      0.000       0.080       0.260\n",
      "sim_only                 -0.5321      0.194     -2.741      0.009      -0.925      -0.139\n",
      "provider_circles_life     0.4138      0.248      1.669      0.103      -0.088       0.916\n",
      "provider_m1               0.4632      0.239      1.941      0.060      -0.020       0.946\n",
      "provider_singtel          0.4928      0.232      2.121      0.041       0.022       0.963\n",
      "contract_0               -0.0903      0.246     -0.367      0.715      -0.588       0.407\n",
      "contract_12              -0.2682      0.216     -1.241      0.222      -0.706       0.169\n",
      "==============================================================================\n",
      "Omnibus:                        0.378   Durbin-Watson:                   2.141\n",
      "Prob(Omnibus):                  0.828   Jarque-Bera (JB):                0.287\n",
      "Skew:                           0.181   Prob(JB):                        0.866\n",
      "Kurtosis:                       2.885   Cond. No.                         68.0\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "# Configure features to use\n",
    "feats = ['talktime', 'sms', 'data', 'provider', 'contract', 'sim_only']\n",
    "\n",
    "# Extract data\n",
    "y_data = np.log(mobile['price'])\n",
    "X_data = mobile[feats]\n",
    "\n",
    "# Convert contract to categorical\n",
    "X_data.contract = X_data.contract.astype(str)\n",
    "\n",
    "# Get dummy features\n",
    "X_data = pd.get_dummies(X_data)\n",
    "\n",
    "# Take logs\n",
    "X_data.talktime[X_data.talktime == 0] = 1\n",
    "X_data.talktime = np.log(X_data.talktime)\n",
    "X_data.sms[X_data.sms == 0] = 1\n",
    "X_data.sms = np.log(X_data.sms)\n",
    "X_data.data = np.log(X_data.data)\n",
    "\n",
    "# Make starhub and 24-month contracts the reference categories\n",
    "X_data = X_data.drop(['provider_starhub', 'contract_24'], axis = 1)\n",
    "\n",
    "# Fit linear regression model\n",
    "lm_mobile = OLS(y_data, add_constant(X_data.astype(float)))\n",
    "lm_results = lm_mobile.fit()\n",
    "print(lm_results.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A Q-Q plot of the residuals suggests that our data transformations have worked: the assumption of normality of the residuals for our model is satisfied."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5cf1a3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Residual plot\n",
    "qqplot(lm_results.resid, line = 's')\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decision Trees: Quantifying Non-Linear Relationships\n",
    "Another way to view the problem is through a non-linear lens. That is, we use a decision tree to (1) establish rules on how to segregate the data and (2) identify the important features. We create a decision tree with a maximum depth of 3 (to avoid overfitting), and obtain the decision tree plotted below, which has an R-squared of **83%**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8313318671838421\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 7,
     "metadata": {
      "image/png": {
       "width": 1000
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get data\n",
    "y_data = mobile['price']\n",
    "X_data = mobile[feats]\n",
    "X_data = pd.get_dummies(X_data)\n",
    "\n",
    "# Decision tree\n",
    "tree_mobile = DecisionTreeRegressor(max_depth = 3)\n",
    "tree_mobile.fit(X_data, y_data)\n",
    "\n",
    "# Plot\n",
    "dot_data = tree.export_graphviz(\n",
    "    tree_mobile, feature_names = X_data.columns,\n",
    "    out_file=None, filled=True,\n",
    "    rounded=True, precision = 0\n",
    ")\n",
    "\n",
    "# Draw graph\n",
    "graph = pydotplus.graph_from_dot_data(dot_data)  \n",
    "\n",
    "# NOT RUN - compute adjusted R-squared\n",
    "yhat = tree_mobile.predict(X_data)\n",
    "SS_Residual = sum((y_data-yhat)**2)\n",
    "SS_Total = sum((y_data-np.mean(y_data))**2)\n",
    "r_squared = 1 - (float(SS_Residual))/SS_Total\n",
    "adjusted_r_squared = 1 - (1-r_squared)*(len(y_data)-1)/(len(y_data)-X_data.shape[1]-1)\n",
    "print(adjusted_r_squared)\n",
    "\n",
    "# Show graph\n",
    "Image(graph.create_png(), width = 1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examining the feature importances, we see that the only three things that matter are (1) talktime, (2) data, and (3) whether or not the plan is SIM-only, in that order."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Importance</th>\n",
       "      <th>Feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.482157</td>\n",
       "      <td>talktime</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>sms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.333866</td>\n",
       "      <td>data</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>contract</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.183976</td>\n",
       "      <td>sim_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>provider_circles_life</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>provider_m1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>provider_singtel</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>provider_starhub</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Importance                Feature\n",
       "0    0.482157               talktime\n",
       "1    0.000000                    sms\n",
       "2    0.333866                   data\n",
       "3    0.000000               contract\n",
       "4    0.183976               sim_only\n",
       "5    0.000000  provider_circles_life\n",
       "6    0.000000            provider_m1\n",
       "7    0.000000       provider_singtel\n",
       "8    0.000000       provider_starhub"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract coefficients\n",
    "tree_impt = pd.DataFrame(tree_mobile.feature_importances_)\n",
    "tree_impt['Feature'] = X_data.columns\n",
    "tree_impt = tree_impt.rename(columns = {0: 'Importance'})\n",
    "\n",
    "# View feature importances\n",
    "tree_impt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Envelopment Analysis: Identifying the Most Worthwhile Plans\n",
    "Thus far, we've looked at the most important determinants of price. The linear regression and decision tree models generally concur: the amount of talktime, amount of data, and whether the plan was SIM-only were the things that mattered most. Yet, all these insights are not actionable unless we come up with a hierarchy of the most worthwhile mobile plans. Hence, this is what we aim to do next through Data Envelopment Analysis (DEA).  \n",
    "  \n",
    "## DEA?\n",
    "DEA is a way to compare the efficiency of different business units. Although each business unit uses several inputs and outputs, DEA allows us to combine them into a single score. Then, we can use these scores to rank the various business units.\n",
    "  \n",
    "In this case, we can think of the outputs as talktime and data, and the sole input as cost. We can easily change the inputs and outputs to whatever is important to us. For example, we could add SMS/MMS as an output if we text a lot, and the contract time as an input because contracts tie us down to a single provider. For now, we calculate the efficiency scores for the mobile plans using talktime per dollar and data (in GB) per dollar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5d228e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract data\n",
    "df = mobile[['price', 'talktime', 'data']]\n",
    "\n",
    "# Calculate efficiency scores\n",
    "df['talktime_cost'] = df.talktime / df.price\n",
    "df['data_cost'] = df.data / df.price\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we want to know which plans give us the best value for our money. These plans have either: (1) the most data per dollar, (2) the most talktime per dollar, or some combination of both. To do that, we only need the outermost points of the graph above. Thus, we compute the convex hull and draw a line around the outermost points. This is the efficient frontier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5d239c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract efficiencies\n",
    "eff_data = df[['talktime_cost', 'data_cost']].copy()\n",
    "\n",
    "# Compute convex hull\n",
    "hull = ConvexHull(eff_data)\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Graphically, we can see which plans are the most efficient. The three \"most efficient\" plans are:  \n",
    "  \n",
    "* M1's mySIM 98 SIM-only 12-month plan that costs \\$98.00 (if you prefer data)\n",
    "* Starhub's \"M\" SIM-only 12-month plan that costs \\$44.00 (if you prefer talktime and you are okay with a contract)\n",
    "* Starhub's \"M\" SIM-only no-contract plan that costs \\$44.00 (if you prefer talktime and you are okay to pay for freedom with less data)\n",
    "  \n",
    "Any plan that lies along the green line (frontier) is considered as good as the three optimal plans. All plans that lie inside the frontier are inferior. Let's put it this way. If data is important to you, there is no better plan than M1's mySIM 98. If talktime is important to you, there is no better plan than Starhub's \"M\" SIM-only 12-month plan. If you want a balance of both, the tradeoff between talktime and data must follow the relationship indicated by the downward-sloping line.\n",
    "  \n",
    "Next, we do the math using the `pyDEA` package to arrive at the same conclusions and tease out more information: we get a ranking of the most efficient mobile plans, evaluated using talktime and data as outputs and price as the sole input."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## All Plans\n",
    "First, we perform the analysis for all plans. We divide each feature by the respective means (mean normalisation) to prevent the magnitude of each feature from distorting the efficiency scores. Then, we run the DEA algorithm (essentially a linear programme) to get the scores. The top 10 efficient plans are given below. The top 3 plans are the same as the ones derived using the graphical method. However, we also obtain the next best plans."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>provider</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>data</th>\n",
       "      <th>efficiency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>m_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>m_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>mysim98_12mth</td>\n",
       "      <td>m1</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>mysim_e118</td>\n",
       "      <td>m1</td>\n",
       "      <td>118.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0.831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>l_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>l_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>combo3</td>\n",
       "      <td>singtel</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>m</td>\n",
       "      <td>starhub</td>\n",
       "      <td>88.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>combo6</td>\n",
       "      <td>singtel</td>\n",
       "      <td>95.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mysim98</td>\n",
       "      <td>m1</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0.456</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            name provider  price  talktime    data  efficiency\n",
       "0        m_sim12  starhub   44.0     10000    13.0       1.000\n",
       "1   m_nocontract  starhub   44.0     10000     5.0       1.000\n",
       "2  mysim98_12mth       m1   98.0     10000  1000.0       1.000\n",
       "3     mysim_e118       m1  118.0     10000  1000.0       0.831\n",
       "4        l_sim12  starhub   54.0     10000    19.0       0.821\n",
       "5   l_nocontract  starhub   54.0     10000     8.0       0.815\n",
       "6         combo3  singtel   68.9     10000     3.0       0.639\n",
       "7              m  starhub   88.0     10000     5.0       0.500\n",
       "8         combo6  singtel   95.9     10000     6.0       0.459\n",
       "9        mysim98       m1   98.0     10000    25.0       0.456"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract relevant data\n",
    "df = mobile[['price', 'talktime', 'data']]\n",
    "df = df / df.mean()\n",
    "\n",
    "# Configure inputs and outputs\n",
    "df_inputs = df[['price']]\n",
    "df_outputs = df[['talktime', 'data']]\n",
    "\n",
    "# Configure problem\n",
    "mobile_prob = pydea.DEAProblem(df_inputs, df_outputs, returns = 'CRS')\n",
    "\n",
    "# Solve problem\n",
    "results = mobile_prob.solve()\n",
    "\n",
    "# Attach results\n",
    "dea_data = mobile[['name', 'provider', 'price', 'talktime', 'data']].copy()\n",
    "dea_data['efficiency'] = round(results['Efficiency'], 3)\n",
    "\n",
    "# Sort by efficiency\n",
    "dea_data.sorted = dea_data.sort_values(by = ['efficiency', 'price', 'data'], ascending = [False, True, False])\n",
    "dea_data.sorted = dea_data.sorted.reset_index(drop = True)\n",
    "\n",
    "# Save\n",
    "top_plans = dea_data.sorted.copy()\n",
    "\n",
    "# View data\n",
    "dea_data.sorted.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## All Plans, Less M1's Unlimited Data Plans\n",
    "M1's data plans looked good because I set a high cap on data of 1,000 GB. However, lowering the cap will not change the results: M1's plan offer good value. And because they offer good value, they are also pretty expensive. For most of us who would not pay that much for unlimited data, I perform the analysis on all plans less M1's two unlimited data plans."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>provider</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>data</th>\n",
       "      <th>efficiency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>m_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>m_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>mysim50_12mth</td>\n",
       "      <td>m1</td>\n",
       "      <td>50.0</td>\n",
       "      <td>100</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>l_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>data_baller</td>\n",
       "      <td>circles_life</td>\n",
       "      <td>48.0</td>\n",
       "      <td>100</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0.903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>l_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>mysim98</td>\n",
       "      <td>m1</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0.652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>xl_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>119.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>xs_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>24.0</td>\n",
       "      <td>200</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>combo3</td>\n",
       "      <td>singtel</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.639</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            name      provider  price  talktime  data  efficiency\n",
       "0        m_sim12       starhub   44.0     10000  13.0       1.000\n",
       "1   m_nocontract       starhub   44.0     10000   5.0       1.000\n",
       "2  mysim50_12mth            m1   50.0       100  30.0       1.000\n",
       "3        l_sim12       starhub   54.0     10000  19.0       0.999\n",
       "4    data_baller  circles_life   48.0       100  26.0       0.903\n",
       "5   l_nocontract       starhub   54.0     10000   8.0       0.815\n",
       "6        mysim98            m1   98.0     10000  25.0       0.652\n",
       "7       xl_sim12       starhub  119.0     10000  33.0       0.649\n",
       "8       xs_sim12       starhub   24.0       200   9.0       0.641\n",
       "9         combo3       singtel   68.9     10000   3.0       0.639"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract relevant data\n",
    "df = mobile[['price', 'talktime', 'data']]\n",
    "df = df.drop([30, 38], axis = 0)\n",
    "\n",
    "# Graph data\n",
    "df_graph = df.copy()\n",
    "df_graph['talktime_cost'] = df_graph.talktime / df_graph.price\n",
    "df_graph['data_cost'] = df_graph.data / df_graph.price\n",
    "\n",
    "# Mean normalisation\n",
    "df = df / df.mean()\n",
    "\n",
    "# Configure inputs and outputs\n",
    "df_inputs = df[['price']]\n",
    "df_outputs = df[['talktime', 'data']]\n",
    "\n",
    "# Configure problem\n",
    "mobile_prob = pydea.DEAProblem(df_inputs, df_outputs, returns = 'CRS')\n",
    "\n",
    "# Solve problem\n",
    "results = mobile_prob.solve()\n",
    "\n",
    "# Attach results\n",
    "dea_data = mobile[['name', 'provider', 'price', 'talktime', 'data']].copy()\n",
    "dea_data['efficiency'] = round(results['Efficiency'], 3)\n",
    "\n",
    "# Sort by efficiency\n",
    "dea_data.sorted = dea_data.sort_values(by = ['efficiency', 'price', 'data'], ascending = [False, True, False])\n",
    "dea_data.sorted = dea_data.sorted.reset_index(drop = True)\n",
    "\n",
    "# View data\n",
    "dea_data.sorted.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plotting the graph:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5d180c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract efficiencies\n",
    "eff_data = df_graph[['talktime_cost', 'data_cost']].copy()\n",
    "\n",
    "# Compute convex hull\n",
    "hull = ConvexHull(eff_data)\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first thing we notice is the concentration of points on the far left of the graph. That's because the Telcos are competing primarily using data bundles. Along that dimension, M1 takes the top spot with its mySim 50 plan, which offers a whopping 30 GB. The second thing we notice is that Starhub is out there with 4 plans (M and L, 12-month and no-contract) owning the market segment focused on talktime - it has no close competitor.  \n",
    "  \n",
    "I've added labels for several other points that do not lie on the efficient frontier. Two of them are worth noting: Starhub's XL SIM-only plan and M1's mySim 98. I highlighted them because they offer a good balance between data and talktime."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bundle Plans\n",
    "I don't actually know what the alternative to SIM-only plans are called. Hence, I'll just call them bundle plans since the underlying marketing tactic is essentially bundling (of phones and mobile services)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>provider</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>data</th>\n",
       "      <th>efficiency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>combo3</td>\n",
       "      <td>singtel</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mysim_e118</td>\n",
       "      <td>m1</td>\n",
       "      <td>118.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>m</td>\n",
       "      <td>starhub</td>\n",
       "      <td>88.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>combo6</td>\n",
       "      <td>singtel</td>\n",
       "      <td>95.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>l</td>\n",
       "      <td>starhub</td>\n",
       "      <td>108.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.640</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         name provider  price  talktime    data  efficiency\n",
       "0      combo3  singtel   68.9     10000     3.0       1.000\n",
       "1  mysim_e118       m1  118.0     10000  1000.0       1.000\n",
       "2           m  starhub   88.0     10000     5.0       0.784\n",
       "3      combo6  singtel   95.9     10000     6.0       0.720\n",
       "4           l  starhub  108.0     10000     8.0       0.640"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract relevant data\n",
    "df = mobile[['price', 'talktime', 'data']][mobile.sim_only == False]\n",
    "df = df / df.mean()\n",
    "\n",
    "# Configure inputs and outputs\n",
    "df_inputs = df[['price']]\n",
    "df_outputs = df[['talktime', 'data']]\n",
    "\n",
    "# Data for graphics\n",
    "df_graph = mobile[['price', 'talktime', 'data']][mobile.sim_only == False]\n",
    "df_graph['talktime_cost'] = df_graph.talktime / df_graph.price\n",
    "df_graph['data_cost'] = df_graph.data / df_graph.price\n",
    "\n",
    "# Configure problem\n",
    "mobile_prob = pydea.DEAProblem(df_inputs, df_outputs, returns = 'CRS')\n",
    "\n",
    "# Solve problem\n",
    "results = mobile_prob.solve()\n",
    "\n",
    "# Attach results\n",
    "dea_data = mobile[['name', 'provider', 'price', 'talktime', 'data']].copy()\n",
    "dea_data['efficiency'] = round(results['Efficiency'], 3)\n",
    "\n",
    "# Sort by efficiency\n",
    "dea_data.sorted = dea_data.sort_values(by = ['efficiency', 'price', 'data'], ascending = [False, True, False])\n",
    "dea_data.sorted = dea_data.sorted.reset_index(drop = True)\n",
    "\n",
    "# View data\n",
    "dea_data.sorted.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once again, M1 is spoiling the market with a beast of a plan: mySim(e) 118 with unlimited data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5eb8bc18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract efficiencies\n",
    "eff_data = df_graph[['talktime_cost', 'data_cost']].copy()\n",
    "\n",
    "# Compute convex hull\n",
    "hull = ConvexHull(eff_data)\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bundle Plans, Less M1's Unlimited Data Plan\n",
    "Removing M1's unlimited data plan, we get a better picture of how the rest of the bundle plans fare: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>provider</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>data</th>\n",
       "      <th>efficiency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>combo3</td>\n",
       "      <td>singtel</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mysim_e90</td>\n",
       "      <td>m1</td>\n",
       "      <td>90.0</td>\n",
       "      <td>100</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>m</td>\n",
       "      <td>starhub</td>\n",
       "      <td>88.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>combo6</td>\n",
       "      <td>singtel</td>\n",
       "      <td>95.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>l</td>\n",
       "      <td>starhub</td>\n",
       "      <td>108.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>comboxo</td>\n",
       "      <td>singtel</td>\n",
       "      <td>78.0</td>\n",
       "      <td>300</td>\n",
       "      <td>18.0</td>\n",
       "      <td>0.711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>mysim_e70</td>\n",
       "      <td>m1</td>\n",
       "      <td>70.0</td>\n",
       "      <td>100</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.647</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        name provider  price  talktime  data  efficiency\n",
       "0     combo3  singtel   68.9     10000   3.0       1.000\n",
       "1  mysim_e90       m1   90.0       100  30.0       1.000\n",
       "2          m  starhub   88.0     10000   5.0       0.851\n",
       "3     combo6  singtel   95.9     10000   6.0       0.812\n",
       "4          l  starhub  108.0     10000   8.0       0.776\n",
       "5    comboxo  singtel   78.0       300  18.0       0.711\n",
       "6  mysim_e70       m1   70.0       100  15.0       0.647"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract relevant data\n",
    "df = mobile[['price', 'talktime', 'data']][mobile.sim_only == False]\n",
    "df = df.drop(38, axis = 0)\n",
    "df = df / df.mean()\n",
    "\n",
    "# Configure inputs and outputs\n",
    "df_inputs = df[['price']]\n",
    "df_outputs = df[['talktime', 'data']]\n",
    "\n",
    "# Data for graphics\n",
    "df_graph = mobile[['price', 'talktime', 'data']][mobile.sim_only == False]\n",
    "df_graph = df_graph.drop(38, axis = 0)\n",
    "df_graph['talktime_cost'] = df_graph.talktime / df_graph.price\n",
    "df_graph['data_cost'] = df_graph.data / df_graph.price\n",
    "\n",
    "# Configure problem\n",
    "mobile_prob = pydea.DEAProblem(df_inputs, df_outputs, returns = 'CRS')\n",
    "\n",
    "# Solve problem\n",
    "results = mobile_prob.solve()\n",
    "\n",
    "# Attach results\n",
    "dea_data = mobile[['name', 'provider', 'price', 'talktime', 'data']].copy()\n",
    "dea_data['efficiency'] = round(results['Efficiency'], 3)\n",
    "\n",
    "# Sort by efficiency\n",
    "dea_data.sorted = dea_data.sort_values(by = ['efficiency', 'price', 'data'], ascending = [False, True, False])\n",
    "dea_data.sorted = dea_data.sorted.reset_index(drop = True)\n",
    "\n",
    "# View data\n",
    "dea_data.sorted.head(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAIECAYAAACABz+XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl8VOXZ//HPZIesJAHCFvZNQTZF4KgoirV3W6uPVXGrS639qRS1blRri221iNa1dS2trU8VfdRqbW+XKoJ4FKWKiMoue1hDEggEss3vj3MCQxxCgCRnknzfrxcvzpw5M+eaeyaZK9e9nFA4HEZEREREWq+4oAMQERERkWApIRQRERFp5ZQQioiIiLRySghFREREWjklhCIiIiKtnBJCERERkVYuIegARFoqx3GeBi6ttTsMbAOWA68Cj7quW3KE5+kBrIzY1dN13VVH8px1nCsHuBY4C+iF9ztkLfA28Jjrul81xnmjxLEK6O7fvNx13aeb4ryHIsr7cjCrXdftcYTn2PveO45zMvBuzR2u64YO97n8+2cBY/2bd7quO+VQYm1qjuOMAn4CnAzkAduBFcBzwNOu6+5oghhO5hDeA5EgqUIo0rRCQA5wPHA3sMxxnJOCDal+HMc5A1gM3AkMAzKBVGAAMBH4zHGc64OLUAQcx4l3HOdh4APgMqAHkAJ0AEYDDwPzHccZElSMIrFICaFI0ygEPgI+BtZF7G8PvOk4ztBAoqonx3FOA/4J5Pq7wnjJ4VJ/GyAReMBxnAubPsKYtAfvPY/8Vxpx//pa981v6gAP0Vfsi3XdQY4N0lPAT/H++AKvzRcAWyKO6Q3823Gc9k0cm0jMUpexSNN423XdCTU3HMcZDzyLl2ClAH8GhgcUW50cx0kF/oaX8AGsAr7nuu4X/v19gRl48a8AvgwgzJjjuu4GYFTkPsdx5uJVhwH+FOvdrpFc170m6BgOxnGcs4HLI3b9DbjGdd2djuPEAROAPwFtgJnAzqaPUiQ2KSEUCYDruv9xHOcq4GV/1zDHcU5yXfc9AL9yMRH4PtAHiAfWAG8Av3Ndd6N/3BTgV7WefqXjOLD/eLKjgUnAaUBXoByvwvc88IjrunvqCPcKoJO/XQ6c7Lru6ojXssxxnDOBe4BrI8dEOo4TAs4HfozXzdwWr7r0b+Ae13ULIk/kOM7xwC+AE/ES5ZXAP4B7Xdctqj0my/cXx3H+wgHGEjqO8w4wzr95n+u6N0fc14N94+a+dF13UETMNwDHAJV4Se7TeElcZR1t1SCO8P062HOn4HWnDvN3TQPKOMjnKNoYwlrvx0t+zFOAb+NVv1f6x87wu2jv8J8jFZgLXOW67vJa8Y0EfgachPcHUxHwIfBH13X/c5CX94uI7TeAy1zXDQO4rlsNPOs4TiYQdl338VrnzQZuBM4GeuK1+RfAX4E/R77vjuMkAj/CG0/bz2+/+cDjrus+7x/zNLXGEDuOUxOLxhJKzFGXsUhwXsH7kq9xOuytHi4FfgkMwfvyTMH74pkELHAcp099T+I4zq3A58BVeBNBkoA04FjgXuBdx3GS63iK/4nYnhGZDNZwXXe967oX10oGE4EX8QbxjwPaAcl43XWTgC8cxzkh4vgzABf4Lt74xGS88Yk/Bz7xk7fD8XzE9um17vtWxPYL/v+/9WMeidfuaXhVvceAfzmOk3SYcdRLA7xfB/MY+5LB/wC3HcFzReoM/Bcv+e/Kvvfv747j3IA3XOIcvCSvDXAK+/4gAsAfgzoXLyHvhFeV7oD3h9FbjuM8eKCTO46Tz/5V9mk1yWAk13Ufi5IMDgYW4rXFQLz3PQMYAzzhnzst4iF/wWvHQXjvTybe5JUZjuP8+UAxisQyJYQiAfG/rNyIXQP8/wuBrIjtBew/ZqsD8Dt/ex3wWa2n/gxvnFdNFakC72e9Gm/c2md4VZcao4Er6wj1mIjtD+o4rrZfs38yWQIs8eMAL0H8hz9zGbzB/vH+9na8ZLmmKpMK7PD3f4RXvanxtb8vcoxYpJcjnucYx3E6Rdy3X0LoOE4XvAS0xnq8LvKaxCLBdd3IczeGI32/DsivSl/m31wJTHBdt4r6fY4OZjReErcF2BCxPw64Hy9xWsb+3bSDHcc5xo/tFP+4murZVryqW2HE8dc5jnOg1x45SSSMV1U8KL9i+ipeQltjJbAp4vYpwCP+8WOAi2odG1nprvmsrAAW1TpdzRhMkZijhFAkWMUR2xkArut+ClwC9HddN9d13aGu63bDm91bY5x/7J/wurgine267ih/DBvAo3jLb3R0Xber67rDgI54S8Xs93wHkB6xXXDAoyI4jpMOXBex6yUgz3XdAXhdgbv9/bnA1f525BdyX9d1B+IljdcDP3Fdt9B13U9d1x3F/gnHb/zX++9osbiuuxV4J2JXTSU2ATjV37fQdd3FeAlNTULyFdDNdd2eQD5esvLj+rz+I3Sk71dUjuMci5d0A+wCznJddxvU+3NUH//CWw6oL95yRJGucl23H977H6mv//8v2Nf2v8f7vAzHe+1/jDj+ZqKL/JwWua67+wDH1XYxXhdxjfNc1+2F93l8OGL/Dx3H6c7+n9PXXdft5bpuF6C/H+eNAK7r/gbYb9yl3577jSsViRVKCEWClR2xHbke4ULgbsdxtjqOs8dxnNXsG79V+3F18r8YN+FV43Y4jrMLr/KSG3FYXc8XWXnLOeBR+xuN1y1Y46c1X9Cu67p44/Fq1FTpXo/Y95zjOFcA7VzXfch13Vfqed4DidZtPBo/CWdfd/EX7EtkBgDTHcc5C9jhuu6NruseyrqCh6UB3q8DeRGvGxfgOtd1Pz+ySKO6xnXdMtd1d7J/he5d13Wfgr1/8GyOuC/N7wKPTBRvBCr9MXeVeGP1avRzHKddlHNHfk4z/CEL9RGZXL/puu7/+XFWA7ew7+cyDm9Mp4tXqQYY5zjOw/5wh3Wu6050XTfyjzyRZkMJoUhA/FmPJ0bsWurvPw/4FG+8VQ7el2cC3hilwznPH/DGK56ANzZqE14lLHKpm7oGuc+N2P7WAY/aX+RyHtujVJmWRWx39P+/BJiMt2j3OGA6sMZxnE8cx7m4nuc9kH+wr5t5vD9xJPK1PA97k7FReJWeIrwZq/8AtjqO82/HcUYfYRwH1QDv14F0j9g+7bADrIPrupFVwbKI7TW1Do28r2ZtzkOZ5BhtDOU8oMrfTqD+VdTIz+qSyDv8yTuRsXf0P8vH4c1grsBb4uZ1vM/IDMdxBtbzvCIxRQmhSHDOw5tBXOMtx3Hi8bqp4vAqI2Nc1+3md0l951BP4M9WramuLAPyXdft6bpuDt4EhfqYEbF9ruM4A2of4DhOJ8dxXnUcp5e/q3a1pvZ6b5FfmlvAS8Zc173Hdd2+wFHATXjjA4cDzziOc3s94/0Gv2rzln+zvf+cNQnhZ67rLos4tsB13Yl4ieoJeLOnywADvOc4TmSltkE10Pt1IGXs62o/33Gc7x/h8zWkEvaNvQOvDdpE/Gsbebtmln0k/z1+M2LXL/zEfz+O4/zEcZxfRVQQIz+rfWodm4K3sDWRx7quu9R13Uvxqrbj8f6ASMCbDDPPcZx+B3vBIrFGCaFIAPwupkcjds13XXc23oSRmopZOf6SKP5agP9T6znMAZ4+07+/DftPCNmG/4XmOE5X9q849qtj5vJLeBNbwJv08c+aiQD+c/UHXgPOBBY6jnM135yM8FDN7Fx/ZvElEffN9Pf/2HGc5x3HGeG67iLXdX/P/jNgf1CP11uXyG7ji9g3I/WFyIMcx3nScZx7gU6uZzLwd//uBLzL9jWWI32/uvjHR/vd/gPg/0XcfuwAXa816tuuR8zvYp4Xset6oL1fsa3GG7+5CG9sYd9vPsNed7KvSngC8FTN7GDHceL8SvP9eEvjfOon4LMiHv/tmkTZb8N72Dc2MYw3wzvkOM4rjuP8AkhzXfdt/w+ImnGqqXjL7nyD4ziZfhxHMktcpFFoHUKRpnGa4y1KHMIblN414r7deGuagTejcifel0pbvEvbrcBbqiUNL8mq+TI5D7B4s28jves4zg7gSfZfs+94YJ3jOFvxlrAB70suhNcleQJed+1+XNet9r9IXbwxd33xlr5Z4j+2L/u6MNsCqa7rlvhdnzf6+y8AznAcZ7N/7prjS4BHHcfJxfvybQec5zjOBrxqVmQlMXJ2auRrvtdPQldTd5f2q3htnYI32L8madqbEDqOcyr7Jo7c6Lf9LmBw7Tgcx3kUL6nc4Lpu7QkZhytySZ/6vF//rPV467fxDex/VRRc17V+3K/iLePSCXiAfbOO6/oc3XX4L6nefsu+19MX77O/FO9npSZxvQK470BP4Lrux47j3Ik3wx28n6sJjuMsw0uWIyvV+XhV0//Fm9DSDa9tX3EcZzleQhx5/P+5rrvccZwf4bXf94EpfowQ/bNau01ruqRvZN8fGSIxQRVCkaZRc/3ikeyfDG4GTndddz6Av6RJ5GziNLzlNNLwvkAiq1z4jylk/7FP7fC+7E52XfcDvEWga3QEjsb7Y3Ai3xzbFZXrXZVkXK3j+7N/clcO3Oq6bs0X9m14s04j4+ofcfwO4Af+mKwSvIStRie8ZKumOlXN/l2mkcv1JPrPe0JdEwlc193BvokrNUn1p67rrog47Eu8tfTw4+yDV7WriXkb3qXRwOvWPp59a/odsUN9v/xZwl9E7MrwYx5Ux2l+yr6E5VK/Wl3n5+hQX8fhcF33NfZfWDoZLxGvSQb34C2TU+fEHn92723sv2TRUPZP7pYBp7mu+7XrurvwkrvIZWb61Dr+Q7x1IQHex1tSBryK+UD2TwZXAv/nby9i/yWDOvr/Tq7rNYgEQQmhSNMKs+/KCz8H+riuO6fWMffhVamWs+8KFZPwulnvw5t8UsH+S6+ci/dFVQZsxOu+es6/73y8ZGqzf78LnOG67pN+DFvxqmDb6grcdd1P8JKTyXizXmvWBfwcmAoMcl13WsTx5XhftD8EZuMlfTXd4I8Bg13Xfds/tsJ13cvxks7n8dbfK8dbH+9V4ETXdSMTxsn+cSV4S/f813+NB5tZWjuh3q+72B+bNgpvMsl/8Cq2u/Hei8eB4W6Uhbkb2KG+X+fjdXvu9OOdjTdLPSp/4kfkVUme9JcJgro/R43Odd278CZavYyXoJXjrQP5V2BYrc9AXc/zO7w/KKbjVV3L8X5e3sFL7Ia6rjsv4vj5eMnnNLykeDdehfUjvJ+9k11/0XXXdZfgJdzX47VVMd77sRivyn18xLG78Jbz+cx/zpqr9ERdIkkkSKFw+BsLuYuIiIhIK6IKoYiIiEgrp4RQREREpJVTQigiIiLSyikhFBEREWnllBCKiIiItHJKCCPccccdYbxlQRr1X1FRUZOcpzn9U5uoXdQuahe1i9olFv41s3ZpMEoII5SUlDTJeaqrq5vkPM2J2iQ6tUt0apfo1C7RqV2iU7tE11rbRQmhiIiISCsX09cyNsZk4K3YPwTvIu83WGu/ca1VERERETl8MZ0QAi8BD1hrf2KM6QfMMMacZq0tBjDG9Me77msXvMs2PWitfcO/bzrQO+K5nrHWTm/a8EVERERiX8wmhMaY4cBma60FsNYu9ZO8HwIP+4d9C/i5tXalMSYTeNsYMw/vGp99rbUnBRG7iIiISHMSswkhXnVvUa19XwLn1dyw1j4ccV8CkAyk4F3UvMIY8yqQBcwE7rbWVtQ+iTHmKryLnZObm0thYWFDvoaoKisrm+Q8zYnaJDq1S3Rql+jULtGpXaJTu0TXnNolJyenwZ4rlhPC9cD4Wvt6AGtrH2iMeQkYC0y21q43xnwLWAZcB1QAv8brWr6t9mOttU8CTwJMmjQp3JCNeyCFhYUN+ia2BGqT6NQu0aldolO7RKd2iU7tEl1rbZdYnmX8ETDEGDMYwBjTDrgGeLb2gdbac4BjgNONMenW2j9ba/+ftXaPtbYa+CXwnSaMXURERKTZiNmE0FpbBUwA7jHGfAhYvArgWmPMg1GOLwC+AsYbY7Jq3R0PtM6FhUREREQOIpa7jLHWrgRMlP3XAxhjngJ+Y61dY4zJBk4HZgDXGmNS8SqDANOA55smahEREZHmJWYrhPX0MHC/MeZd4O/AXdbaxcDvgCLgHeANYAVeUtgqXHTRRUGHICIiIs1ITFcID8ZauxD4QZT91XgLWt/b5EHFgL///e9Ner7y8nKmTp1KQUEBXbt25dZbbyUxMbFJYxAREZHD16wTwpZk1qxZvP7666SmprJx40a6detGXl4ea9euZcGCBQwdOpTvfve7DBs2DIA777yTiy++mLfeeovNmzcTHx9PdnY211xzDWeddRavvPIKl112GePHj2flypV8/PHHXHfddXz00UcUFRWRnJzMb3/726ix7NmzhylTppCcnMyuXbs4/vjjOeeccygpKeG+++5j+/btpKWlccstt5CZmcltt93G2WefjeM4vPfee9x+++1Mm9ZqCrIiIiLNnhLCGLJ69WpmzJgBwJAhQ5gxYwaXX345M2bMoLy8nMcff5wnnniCiooK1q9fT+/evXnrrbd48cUXiY+Pj/qcOTk5XHTRRcydO5fJkyczc+ZM4uLiOPPMM6muriYu7pujBqZNm8bZZ5/NyJEjAZg8eTLDhw/nl7/8JbfccguDBw9m2bJl/OxnP2P69OmsWrUKx3EAOOmkk3jwwW/M+REREZEYpoQwhowaNWrvds+ePRk4cCAAeXl5LFiwgIyMDNatW4frupx77rkA3HzzzfziF7+gqqqK0047jdNPPz3qc+bl5TF06NC9CWB2djY7duwgMzPzG3F88sknJCYmMnPmTMBbpHPNmjWUlpYyePBgAPr27cuuXbuorKwkHA7v9/jat0VERCS2KSFsJsLhMBMnTuTRRx9l27ZtPPbYYwCMGDGCMWPGAGCM4dRTTz2k54xm5MiRjB07ltGjRwMwf/58hg0bxvTp01myZAn9+/dn7dq1fPzxx1RVVZGRkcH69evp0qULa9euJSur9qo/IiIiEsuUEDYj3bt3Z8uWLQwfPpxQKMSePXu44447SE9Pp7KykjFjxhyw6/hQ3HTTTdx55528/fbblJWV0bdvX4YNG8YjjzzCgw8+SGVlJbm5uZx66qkkJydz9913M3nyZBITE6msrOSee+5pgFcrIiIiTSWk7r19Jk2aFH744YcPfuAROpLL4kyYMIHp06eTmpraILE8//zzzJkzZ+/tm266iR49ekQ99rXXXuPYY4+lU6dOfPTRR8yePZtbbrmlQeJorZcKOhi1S3Rql+jULtGpXaJTu0TXzNol1FBPpAphMzJlyhSOPfbYBksGAc4//3zOP//8eh3rOA6/+93vqKiooH379g2WDIqIiEiwlBA2I1OmTAn0/NnZ2dx7b6tc2lFERKRFa+5XKhERERGRI6SEUERERKSVU0IYoHA4zKatxVq3T0RERAKlMYQBWVOwmT/+9TU+XrCUkUP6ce2l3yO/c4egwxIREZFWSBXCAPxt0+f87rO32XZ0Et2GdWV9djk3T/1z0GGJiIhIK6UKYUCqqqpJSIiHjm1JHZLLnj7tWLOziPzUdkGHJiIiIq2MKoQB+GHHYyh5Yy3Ln/2CHZ9sZtWLi0lISuCBpe/xWsFXVFZXBR2iiIiItCJKCANy3+1XMvzo3mzaUsyA9PZc19NhZHY+72xaxrxt64IOT0RERFoRdRkHpFvn9kydfAWbC0vokJNJKBSiD50YmZNPz9RsANbvKqFjShoJcUd+fWIRERGRA1FCGKBQKETH3Kz99vVO866fuLuqgkdXfEBGQjIXdh9Ot7ZZ0Z5CRERE5IipyzhGpcQncmH+MHZVVfDAkvewBYuorK4OOiwRERFpgZQQxrCjM/O4dcApjMjuylublvL7JbMprdgTdFgiIiLSwqjLOMa1TUjiou7DGZLVmQXFBaQmJAUdkoiIiLQwqhA2E4My87io+3BCoRBF5bt4ZNn7rN9VEnRYIiIi0gIoIWyGisrL2LK7lN8vmc0bGxZTFdbYQhERETl8SgiboV5pOdw6cBzD2nXhjY1LeGDJexSUqVooIiIih0cJYTOVmpDEJT1GcEXP4yiu2M1bG5cGHZKIiIg0U5pU0swdk9WZXmk5EPZub9lTSmV1NZ3aZAQbmIiIiDQbSghbgLSE5L3br67/kkXbN/GtvP6c2rEv8SEVgUVERKRuyhZamAndhjI4sxN2w2IeXDqHDWXbgw5JREREYpwSwhYmLTGZy3oex2U9jmVb+S7uWzKbL0s2Bh2WiIiIxDB1GbdQQ9t1oXdaLq9vWEzP1GwAqsNh4kKhgCMTERGRWKMKYQuWnpjMeflDaJuQRFW4moeWzuHtTcu0bqGIiIjsRwlhK1FeXUVmYgr/KviKh5e+z6bdO4IOSURERGKEEsJWok18Ipf3PI5Luo9gy55S7l08i5mbllEdDgcdmoiIiARMCWErEgqFGJHdlckDxzEgowPu1lVUVFcFHZaIiIgETJNKWqGMxBR+1HMkpZXlJMcnUF5dxSfb1nJ8TndNOhEREWmFVCFspUKhEOmJ3oLWnxat4/m1C3hk2fts2V0acGQiIiLS1JQQCsdn53NR9+Fs3L2DaYtnMWvzCo0tFBERaUXUZSyEQiGOy+5Gv/Rcnl+zgFfWf8HG3TuYkD806NBERESkCSghlL0yE9vw417HM2/bWvLapANQUV1FfChOYwtFRERaMCWEsp9QKMTInPy9t19et5DNe0q5IH8YucmpAUYmIiIijUVjCKVOPVKzWb+rhGmL32XOlq81tlBERKQFUkIodTo+J59bB46jV2oOL61byKPLP6Bwz66gwxIREZEGpIRQDqpdUht+0nsUE/KHsnH3diqqK4MOSURERBqQxhBKvYRCIUbldGd4u64kxcUD8PamZQzL6kJOctuAoxMREZEjoQqhHJKaZLCofBf/2biUaYtn4m5dSVhjC0VERJotJYRyWNolteXWgafQPTWb/1v7OY8t/4Bt5RpbKCIi0hwpIZTDlp3Ulqt7j+bcbkNYvauIB5e8R0V1VdBhiYiIyCHSGEI5IqFQCCe3BwPSO7Bh93YS4+IJh8OUVu4hPTEl6PBERESkHlQhlAaRk9yWQZl5AMzbtpbffvUOH25drbGFIiIizUBMVwiNMRnAvcAQYAtwg7V2eT0f2wl4AOgBrASut9ZuaqRQJULvtBy6tc3i+bWfsaC4gAn5Q8lKahN0WCIiInIAsV4hfAl41Vo7CrgRmGGMyaq50xjT3xgzwxgzxxjzhjHmDH9/EvAKcL//2AeBV4wxMZ0AtxQ5yalc02cM53QdzNc7C5m6aCafbFsXdFgiIiJyAKFY7dIzxgwHbrTWXhSx72og0Vr7sH97EvCatXalMSYTeBs4AxgLjLDW3h7x2HsA11r7z1rnuQq4CiA3N3fEAw880MivDCorK0lIaB256baKMl7btoyhqR0ZktbxgMe1pjY5FGqX6NQu0aldolO7RKd2ia45tUtOTk6ooZ4rll9xb2BRrX1fAufV3KhJDH0JQDKQUsdj+9U+ibX2SeBJgEmTJoVzcnKOOPCDKSwspCnOEwtygOs7diGENwFlbuFqQoQYmd2NUGjf57g1tcmhULtEp3aJTu0SndolOrVLdK21XWK5y3g9kF9rXw9gbe0DjTEvAUuAh6216w/lsdL44kIhQqEQ4XCYBcUFPLdmPk99/RElFWVBhyYiIiLEdkL4ETDEGDMYwBjTDrgGeLb2gdbac4BjgNONMenAv4AfGGO6+Y/NB84E/t1EsUsUoVCIH/caxdldBrFsx1amLnqXedvWaiayiIhIwGI2IbTWVgETgHuMMR8CFphsrV1rjHkwyvEFwFfAeGvtduAK4Bn/sX8FrrDWljbdK5Bo4kIhxnbozS0DTiYvJZ2/r/6UlTu3BR2WiIhIqxbLYwix1q4ETJT91wMYY54CfmOtXWOMyQZOB2b4x3wGnNx00cqhaJ+Sxk/7nsCi7ZvolZZD4Z5CNpZtp2NK+n5jC0VERKTxxXRCWA8PA/cbY3KA3cBd1trFAcck9RQXCnG0v5h1YUUZTyz5gIEZHTiv2xAydJUTERGRJtOsE0Jr7ULgB0HHIUeuXUIK3+k0ELthEVMXzeScrscwvF0XVQtFRESaQMyOIZTWJS4UYlzHPtw84GTaJ6fxzOpP+MvKeVSFq4MOTUREpMVr1hVCaXk6pqRzXb8TmbV5OcUVu4kP6W8WERGRxqaEUGKOVy3su/f2mp1FzNy8nHO6HkN6YnKAkYmIiLRMKr9IzNuwewcLSzYydfFMPitaH3Q4IiIiLY4SQol5x+fkc1P/sWQnteXpVf/l6ZXzKK3YE3RYIiIiLYYSQmkWOrXJ4Pp+J/KdTgNZWLKB97Z+HXRIIiIiLYbGEEqzER+KY3xePwZl5pGTnApAQVkJGYkppCVobKGIiMjhUkIozU6nNhkAVIfDPL3yv+yqquC8bsdwTFbngCMTERFpntRlLM1WXCjEZT2PJSsxhT+vnMffVv2XnZXlQYclIiLS7CghlGatc5tMbuh/EmfkDeCzogKmLprJxt07gg5LRESkWVGXsTR78aE4zujUn8GZeczasoL2/vjCcDisS9+JiIjUgyqE0mJ0aZvJRd2HEx+KY1dlOfcvmc0XJRuCDktERCTmKSGUFmlnVTlV4TB/+vpj/nfVJ+zS2EIREZEDUkIoLVL75DR+1n8sp+f149Oi9UxdNJMvSzYGHZaIiEhMUkIoLVZCXBym00Bu6H8SqQlJvLNpGeFwOOiwREREYo4mlUiL161tFjf2H8uuqgpCoRDbK3azblcJR2V2DDo0ERGRmKAKobQKCXHxZCSmAPDOpmU8+fVcnl09n7KqioAjExERCZ4qhNLqfK/zUSTGxfPOpmUs3bGZ8/OHMTCjQ9BhiYiIBEYVQml1EuLi+W7no7i+30kkxyfwxIoPmbPl66DDEhERCYwSQmm1uqe246b+JzO+Yz8GZXYCoLK6OuCoREREmp66jKVVS4yL5zudBwLelU2e+nou2Ult+X6Xo0mJTww4OhERkaahCqGIr5owXdpkMrdwNfcsepcl2zcHHZIaMFYXAAAgAElEQVSIiEiTUEIo4osPxXFml6O5rt+JJMbF89iKD3lhzQJ2ayayiIi0cEoIRWrpkZrNTQNO5pQOvVlYsoEKjSsUEZEWTgmhSBRJcfF8v8sgbjvqVNITk6kOh5m5aTl7qiqDDk1ERKTBKSEUqUMbf2LJitJCXiv4kmmL32XZjq0BRyUiItKwlBCK1EPf9Fx+2vcEQqEQf1zu8tLaz1UtFBGRFkMJoUg99UrL4ZYBJ3NS+168v3UlT349N+iQREREGoTWIRQ5BElxCfxP18EMyepEVTgMeItZV4WrSY7Xj5OIiDRP+gYTOQy903L3bv9n01I+2baOC7oPo3daToBRiYiIHB51GYscob5puYQJ84dl7/OPdQspr9bYQhERaV6UEIocoT7pudwy4BSc3J7M3vI19y6exaqd24IOS0REpN6UEIo0gOT4BH7Q7Riu7TOG6nB47/hCERGR5kAJoUgD6pventuOOnXvWMJ3Ny1npaqFIiIS45QQijSw+JD3Y7WnqpL3t67k4aVzeHX9l5RXVwUcmYiISHRKCEUaSXJ8AjcPOJnROT14d/Ny7tPYQhERiVFKCEUaUUp8IuflD+Hq3qMpr67ikWUuJRW7gw5LRERkP1qHUKQJ9M/owOSBp7C8dCuZiSkAFO7ZRU5y24AjExERUYVQpMmkxCcyKLMTAEu2b+aur97mtYKvqNTYQhERCZgSQpEA5Ke2Y2RON97ZtIz7lsxmzc6ioEMSEZFWTAmhSADaxCcyIX8YV/UaRVlVBQ8uncMbGxYHHZaIiLRSSghFAnRUZkduHTCOY7O7khgXH3Q4IiLSSmlSiUjA2iYkcmH34YT9q5ssKC5g/a4STs/rH3BkIiLSWighFIkRoVAIgJWl25i1ZQVflGzEZPUih5yAIxMRkZZOXcYiMeasroO4stfxlFbu4c8bF/D6hsVUVlcHHZaIiLRgSghFYtCgzDxuHTiOo9vm8ubGJSwoLgg6JBERacHUZSwSo1ITkjgrtz8nJ/ejZ6rXbbxuVzGd2mTsvV6yiIhIQ1BCKBLjeqflArCzspw/LHfJSUrlwvxhdGmbGXBkIiLSUsR8QmiMyQDuBYYAW4AbrLXLI+7vDdwB5AOlwO3W2oVBxCrSmFITkrgwfxgvrP2c+5fO5vS8/pzWsa+qhSIicsRiPiEEXgIesNb+xBjTD5hhjDnNWlvs338JMM1a+5UxZgjwNDDCGBMHfATsjHiuG621nzRl8CIN6ZiszvRKy+GltQt5fcNiFhZv4Jo+Dm0TEoMOTUREmrGYTgiNMcOBzdZaC2CtXWqMmQ78EHjY3zcl4iHrgN3+9mDgI2vtxKaLWKTxpSUkc2nPYxla3JlF2zfRJj6mf4xFRKQZiPVvkt7Aolr7vgTOi9zhVwPPBS4HrvZ3nwZkG2PeAJKAp6y1z9U+gTHmKuAqgNzcXAoLCxv0BURTWVnZJOdpTtQm0dXVLl1JpmtqPtu2bWNbRRn/2raMM9r1pkNSahNH2fT0eYlO7RKd2iU6tUt0zaldcnIabp3aWE8I1wPja+3rAaytuWGMaQ/8FXgZ+J61tsK/qz8wy1r7pD8O8QVjzDZr7ZuRT2atfRJ4EmDSpEnhhmzcAyksLGzQN7ElUJtEV992Kdyxha1bdvOnTQs4I68/4zr2adFjC/V5iU7tEp3aJTq1S3SttV1i/RvjI2CIMWYwgDGmHXAN8GzEMY8CN1lr/xSRDGKtvcpP9rDWbgd+Sa3KokhL0S+9PZMHjGNQZh7/3rCIh5bOYWPZ9qDDEhGRZiKmK4TW2ipjzATgj34yCDDZWrvWGPOgtfZ64HSgvTEm8qHfB+KstUUR+5KACkRaqPTEZC7veRzzi9bz4trPmbl5ORd2Hx50WCIi0gzEdEIIYK1dCZgo+6/3/4+6GJsx5t/GmKesta8YY7KBX+MtTyPSog1r14U+abnE+9dG3rR7BwAdU9KDDEtERGJYzCeER+Ai4HfGmBvwlp75nbXWDTgmkSaRnpi8d/vldQtZUVqI6TSAkzv0Ic5PFEVERGq02ITQX6fw6oMeKNLCXdR9OP+3dgH/LPiKz0s2cEH+MFULRURkP7E+qUREjlBGYgpX9BzJJd1HsHl3KfctnsWS7ZuDDktERGJIi60Qisg+oVCIEdld6ZueyxsbltAjNRuA6nBYXcgiIqIKoUhrkpGYwnn5Q0iOT6CiuooHlsxm1uYVVIfDQYcmIiIBUkLYil122WUUFxcf/EBpkcqrq8hITOGV9V/wh2Xvs2VPadAhiYhIQJQQirRSqQlJXNnreC7MH8aG3TuYtmgWs1UtFBFplTSGsBkJh8NMmTKFoqIitm/fTt++fTnrrLOYNm0aWVlZ7Nixg9tvv53evXtz2WWXMXLkSLZu3crHH3/MrbfeynvvvUdBQQFHHXUU1157LQCvvfYan332GRs2bGD8+PFcfvnlFBcX8/Of/5ykpCS2b9/OBRdcwPjx47nqqqv41a9+RdeuXbn77rsZNWoUw4cP/8axp59+esAtJfUVCoUYmZNPv/T2vLB2Ae7WVYzJ7UFcKD7o0EREpAkpIWxGnnnmGXr06MGdd94JwLx587jhhhv4xz/+QWpqKkVFRfzoRz/i5ZdfZt26ddx0000MGjSIGTNm8NRTT/G3v/0NgO9+97t7E8LOnTtzySWXAHDppZcyduxY7r33Xq699loGDRpEOBxmwoQJDB8+nPvuu49rrrmGs88+m+TkZMaNG8fVV18d9djc3NxgGkkOS1ZSG37c63h2VpaTGBfPnqpKPi1az/E5+Zp0IiLSCqjLuBmZN2/eftW3Xr160a1bN1JTUwFo164d8fHxVFZWkpaWxqBBgwDIy8tj+PB9lzBLSNj3d8CIESP2bp922mnMnz+fDRs27H1sKBTihBNO4PPPPyczM5PzzjuPO+64gxtvvBHggMdK8xMKhUjzF7T+eNsanl/7GX9c7rJ1z86AIxMRkcamhLAZGT16NP/617/23l62bBkFBQWUlZUBUFRURHl5+X4J38EsXbp07/abb77JsGHD6NSpE4sWLQK8buo5c+YwaNAgtm/fzksvvcSvf/1rHnroIYADHivN2wm5PZmQP5T1u0qYtvhd3t+yUmMLRURaMHUZNyMXXnghv/nNb5g4cSJ79uyhW7duTJs2jauvvprMzEy2b9/O/ffff0jPuWjRIp577jm2bNmCMYZevXpx9913c9ttt5GUlERJSQlXXnkl7du35yc/+Ql333333jGE7777btRjO3To0EgtIE0lFAoxKqc7/dM7MGPNfF5c9zmb9uzgnK7HBB2aiIg0glBYf/XvNWnSpPDDDz/c6OcpLCwkJyen0c/TnKhNoouFdgmHw8wtXEP31Cw6t8mkvLqKhFBcoGMLY6FdYpHaJTq1S3Rql+iaWbs02C9iVQhFpE6hUIjRud333n5hzWdsr9jNhO7DyE5qG2BkIiLSUDSGsBkIh8Ns2lqMqrkStHA4TK+0HFbvKuKeRTP5YOsqfS5FRFoAVQhj3JqCzfzxr6/x8YKljBzSj2sv/R75nTVGT4IRCoUYk9tj79jCF9YuYEFxARPyh9EuqU3Q4YmIyGFShTDG3XzXdD79cgUd22fx6ZcruPmu6UGHJEJOcluu7jOGH3Q9hvVl26kMVwUdkoiIHAFVCGNc6a4y2mWlERcXR7usNEp3lgUdkggAcaEQJ7TvycicfJLi4gmHw/xn01KOy85XtVBEpJlRhTDGpbVtQ1FxKdXV1RQVl5LWVl+0EluS4rzL3G0t38nbm5Zxz6KZzC1crbGFIiLNiBLCGHff7Vcy/OjebNpSzIhBfbjv9iuDDkkkqvbJadwy4BS6tM1kxprPeGLFXIrLVdEWEWkO1GUc47p1bs/UyVewubCEDjmZhHRdWYlhucmpXNvH4f2tK3lt/Vc8tGwOvzjqNOJD+ttTRCSWKSFsBkKhEB1zs4IOQ6Re4kIhTmrfi4EZHdiyeyfxoTjC4TClleWk+9dKFhGR2KI/20WkUbRPTuOozI4AfFC4mrsXvcPHhWs0tlBEJAYpIRSRRtcvLZe8lHSeXTOfP339ESUVGlsoIhJLlBCKSKNrn5LGT/uewFldBrF0xxamLnqXz4oKgg5LRER8GkMoIk0iLhTi5A69OSqjI8+tmY/mR4mIxA4lhCLSpDr41cI4PyN8f8tK2sQnMrxdF82iFxEJiBJCEWlyNclgdTjMguIClpVuZUFxAed2O4b0xJSAoxMRaX00hlBEAhMXCnF1nzGc2fkovtq+iamL3uXTonWaiSwi0sSUEIpIoOJCIcZ17MtNA04mJ7ktf1v1CevKSoIOS0SkVVGXsYjEhLyUdK7rdyJLtm+hW1tvIfaNZdvJa5MRcGQiIi2fKoQiEjPiQ3F7F7MuKNvOtMWzeHrlPEor9gQcmYhIy6aEUERiUseUNL7daQALSzYwdfFMrVsoItKIlBDGgHA4zKatxRpILxIhPhTH+Lx+3NT/ZNoltuHpVfP468r/Uq2fExGRBqcxhAFbU7CZP/71NT5esJSRQ/px7aXfI79zh6DDEokZndpkcH3/k3hn0zJ2V1XuXbJGREQajiqEAbv5rul8+uUKOrbP4tMvV3DzXdODDkkk5sSH4jg9rz9ndjkagDW7S3hm1SfsrCwPODIRkZZBCWHASneV0S4rjbi4ONplpVG6qyzokERi3paKXXxWvJ6pi2aysHhD0OGIiDR7SggDlta2DUXFpVRXV1NUXEpa2zZBhyQS80akd+LG/mPJSExh+sqPVS0UETlCSggDdt/tVzL86N5s2lLMiEF9uO/2K4MOSaRZ6Nwmk5/1P4kz8vozv2g9H2xdFXRIIiLNliaVBKxb5/ZMnXwFmwtL6JCTSUgD5kXqLT4UxxmdBnBMVmc6JKcBsHZXMTlJbWmbkBRwdCIizcdhJYTGmOFAMjDPWlvZsCG1PqFQiI65WU12vosuuoi///3vTXY+kcbW2b+aSVW4mr+snEdldRXn5w/l6My8gCMTEWkeDpgQGmM6AT+z1t4csS8OeAkYA2wCQsaYb1tr1zV6pNJgGiMZrKysZOLEiXtv79ixgx49enDXXXcxd+5c/vKXvwBw5ZVXctxxxzX4+UXAqxhe3vM4nl09n6e+/ojjsrtxdpfBtE1IDDo0EZGYVleF8JfAslr7zgMGAn2ttduNMdcBvwUua5zwpMasWbN4/fXXSU1NZePGjXTr1o28vDzWrl3LggULGDp0KN/97ncZNmwYAHfeeScXX3wxb731Fps3byY+Pp7s7GyuueYazjrrLF555RUuu+wyxo8fz8qVK/n444+57rrr+OijjygqKiI5OZnf/va3UWPZs2cPU6ZMITk5mV27dnH88cdzzjnn8Pjjj+89ZuLEiUyYMIGCggIefPBBnnnmGcCrTj700EN06tSp8RtNWqVubbO4sf9Y3tq4hLc3LWPpji1M7OvQ3u9SFhGRb6orIfwOcHStfRcC06y12/3bjwNLGyMw+abVq1czY8YMAIYMGcKMGTO4/PLLmTFjBuXl5Tz++OM88cQTVFRUsH79enr37s1bb73Fiy++SHx8fNTnzMnJ4aKLLmLu3LlMnjyZmTNnEhcXx5lnnkl1dTVxcd+cdzRt2jTOPvtsRo4cCcDkyZMZPnw4PXv2BGDJkiWUl5czePBgnnjiCX70ox+RmOhVaK644gpee+01rrrqqsZoIhEAEuLiMJ0HMigrj/e3rCQnKRXwrgqkcboiIt/0jYTQGDMWOBnIAM42xoSAsH/sWOBdY8yl/r4QkGWM+SUwy1r7XlMF3hqNGjVq73bPnj0ZOHAgAHl5eSxYsICMjAzWrVuH67qce+65ANx888384he/oKqqitNOO43TTz896nPm5eUxdOjQvQlgdnY2O3bsIDMz8xtxfPLJJyQmJjJz5kzA6y5es2bN3oRwypQp3HvvvQDs3r2blJSUvY9t06YNZWVaa1GaRn7bdlzYvR0ApRV7eHzFh3yn81EMzNDVgEREIn0jIbTWzgZmG2N+aK39W81+Y8yJwHJr7QORxxtjpllrf934oUpdwuEwEydO5NFHH2Xbtm089thjAIwYMYIxY8YAYIzh1FNPPaTnjGbkyJGMHTuW0aNHAzB//vy9XdVz5syhR48edO3aFYBx48bx9NNPc+KJJwLw8ssv8+Mf//jwXqTIESitKqcyXM0TKz5kVE4+3+8yiDbxGlsoIgJ1dxnPMcacZ619wb99LfBy5AHGmJHAV40VnBya7t27s2XLFoYPH04oFGLPnj3ccccdpKenU1lZyZgxYw7YdXwobrrpJu68807efvttysrK6Nu3796EcOrUqTz77LN7jx08eDD9+/fnhz/8IQBjx45l0KBBRxyDyKHKS0nnxv5jeWPDYmZuXs7i7ZuZkD+MAaoWiogQOlAVyBjTBXgD2IDXfZwFjKwZP+h3Jb8FTLfWzmiacBvXpEmTwg8//HCjn6ewsJCcnJxGee4JEyYwffp0UlNTG+T5nn/+eebMmbP39k033USPHj2iHvvhhx/yxRdfHFYFsDHbpDlTu0R3pO2yauc2nl09n6ykNlzde3SLGVeoz0t0apfo1C7RNbN2abBfXgesEFpr1/sVwFPxxgvOttaWRhzSCW8dwhaRDLYEU6ZM4dhjj22wZBDg/PPP5/zzz6/XsaNHj97bjSwSy3qkZnPzgJPZXVVJKBSiuLyMTbt30F/VQhFppepcmNpaWwb86wD3FQC3NUZQcnimTJkSdAgizUZiXDyJcd4Qirc2LeWDrasYk9ODM7scRYrGFopIKxPTl64zxmQA9wJDgC3ADdba5RH39wbuAPKBUuB2a+1C/77jgbuANOAdYIq1tqJpX4GINAdndRlEclwCszYvZ9GOTVyQP4x+6e2DDktEpMl8c5G52PIS8Kq1dhRwIzDDGBN5jbdL8NZFHIeXGD4NYIzpBzwIXAyMBlYDjzZh3CLSjCTFxfP9Lkczqe+JJITieHT5B3y4dXXQYYmINJmYTQj96yVvttZaAGvtUmA68MOaY6y1U6y1NbOc1wG7/e2JwC+ttRuttWFr7ZPA0bWSSRGR/fRMy+bmAadwWse+HJ3ZEYDK6uqAoxIRaXx1Xcu4a8DXKO4NLKq170u8y+ft5V9f+VzgcuDqOh67GOgD/LfW468CrgLIzc2lsLCwIWKvU2VlZZOcpzlRm0SndomusdtldHJHKrbvZGu4lL9tXkjHxFTGZfUgKe7Il21qTPq8RKd2iU7tEl1zapeGnA1d1xjCecaYX1trH2uwsx2a9cD4Wvt6AGtrbhhj2gN/xVsf8XsRYwTX440rXHegx9bwq4dPgrfsTFNMNW9mU9qbhNokOrVLdE3VLpXVVfTcnct7W77m6/ISLsgfRp/03EY/7+HS5yU6tUt0apfoWmu71NVlPAY4yxgzxx+T19Q+AoYYYwYDGGPaAdcAz0Yc8yhwk7X2T7UmjPwF+JUxJsV/7Hhgh7V2U9OELiItQUJcPGd3HczEvicQIsQflru8tO5z9lRVBh2aiEiDqmsdwpXAt4wxlwGzjDF/BKZaa6uaIjBrbZUxZgLwRz8ZBJhsrV1rjHnQWns9cDrQ3hgT+dDvW2s/NMb8Be9qK1XASuCKpohbRFqe3mk53DzgZP69YRELigv4dt6AoEMSEWlQB7xSSSRjTAfgYWAg8CNr7X8P8pBmqSVcqaS5UptEp3aJLsh22V1VQUp8IlXhamZtXsGJ7XuSFBcbK3jp8xKd2iU6tUt0zaxdGuxKJfWaZWyt3WytnYC3tMuLxpjfG2PaNFQQIiLNRc2i1Uu2b+G1gq+YtngWX5c2jwHoIiIHckjLzlhr/wkMBlKBhcaY0xolKhGRGHdUZkeu7TOG6nCYR5a9zz/WfUF5tcYWikjzdMjrEFprd1hr/x/eMi+P+GP1RERanb7p7bl1wCmMye3B7C0r+PPX84IOSUTksBz2wtTW2jnAMKCg4cIREWlekuMTOLfbEK7pM4bT87wFGSqqqyivbpL5dyIiDeKIRkJba3cDtzdQLCIizVbktY9f37CYL0o2cmH3YfRIzQ4wKhGR+onZS9eJiDRX/dPbU1FdxUNL5/DP9V9SoWqhiMQ4JYQiIg2sf0YHbh14CqNyujNz83LuWzKLtbuKgw5LROSAlBCKiDSClPhEzs8fyk96j6aiuprqeqz5KiISlCNOCI0xHRsiEBGRlmhgRgduP+pUuqd6F1z6z8alrNlZFHBUIiL7q/ekEmPM2cD/AJ3ZtzJ2AtADyG/wyEREWoj4kPe3d1lVBe7Wlby+YTGnduzDt/L6kxAXH3B0IiL1rBAaY64BpgBvApuBN/AuZdcW+FZjBSci0pK0iU/klgHjODa7G//ZtIzfL5mtsYUiEhPq22V8HfA9a+3/An2AR621rwCTgJsbKzgRkZambUIiF3YfxlW9RrGrqoJHlr3PzsryoMMSkVauvglhlrV2jb/dyVpbCmCt/QA4qVEiExFpwY7K7MitA07h0h7HkpqQBEDhnl0BRyUirVV9E8ICY0yuv11ojOkKYIxJBjIbJTIRkRaubUISR2fmAfBFyUbu+uptXt+wmMrq6oAjE5HWpr4J4SPAlf72C3jXMD4OeByY2RiBiYi0Jj1TsxnergtvblzC/Utns25XSdAhiUgrUq+E0Fr7Z+AZ/+Y9wCrgT0AFcFWjRCYi0oqkJiRxcY8RXNlrJKUVe7h/yWze2bQs6LBEpJWo97Iz1tr1/v+VwA0Axpi2wBDgw0aJTkSklRmU2YmeA3N4ed1CkrQkjYg0kfouOzP7AHdVAX9uuHBERCQ1IYlLeozghNyeAHyybR1vblxCVVhjC0WkcRywQmiMSYq42dcYk8i+BalrdAE6NUZgIiKtXSjk/cr9emch7tZVLCzewIXdh9O5TUbAkYlIS1NXhXA3UOb/61jrdpl/+yvgsUaOUUSkVTu32xAu73kcxRW7+f2SWbylaqGINLADVgittXuTRWPMBmutKoEiIgEZktWZ3mk5vLR2IXbDYvJS0jkmq3PQYYlIC1HfSSVHNWoUIiJyUGkJyVza81hOKO1Br9QcANbuKqZzm4y910sWETkc9UoIrbVFxphUYBTQmX1jCROA71hrz2mk+EREpJbead51AnZU7OaRZe+Tl5LOhfnDSAw4LhFpvuqVEBpjhgCvAUuBbKAU2AacgL8EjYiINK30xBQuyB/Gi2s/594lsxmbmc93stupWigih6y+vzXuB35mrT0NWA9cYq09C/g1kNNYwYmISN2GtevC5IGncHRGR2YWr+KhpXPYU1UZdFgi0szUNyEcaK190d8eZK1d7W8/BVzW4FGJiEi9pSemcHnP4zg7pz/5bbNIjq/3NQdERID6J4SVxpia3zBhY0wKgLW2DMhrlMhERKTeQqEQg1Lb84NuQwDYWLadPyxz2bR7R8CRiUhzUN+E8FXg2/72h8ClAMaY7wCbGiEuERE5AkUVZRSUlXDf4lm8u2k51eFw0CGJSAyrb0L4a6CNvz0FuM0Y8yXeZesmNkJcIiJyBAZmdOTWgePon9GBVwu+5JFl77N5d2nQYYlIjKpXQmit3WKtfcHfXgb0By4Geltr5zRifCIicpgyE1P4Uc+RXNx9OBt372D2lhVBhyQiMeqgI4/9sYPfwVuDMBtvuZm5wL+ttZrKJiISw0KhEMdmd6Nvei7Jcd6v/IKy7SSG4mifkhZwdCISK+qsEBpjHGAFcCPetYsX+v//DFhmjBnV6BGKiMgRy0xsQ0q8t3T1i2s/Z9riWczevEJjC0UEqKNCaIzpA7wIXGGtfb3W3XcaY74NvGSMOdFa+3VjBikiIg3nhz1G8MLaBfxj/RcsKN7ABd2H0j5Z1UKR1qyuCuFtwJ1RkkEA/P13Arc3RmAiItI4spLa8ONex3NB/jAKdpdw7+JZrCgtDDosEQlQXWMIx3HwGcT/i5c4iohIMxIKhTg+J5/+6e15a+NSurXNAqA6HCYuFDrIo0WkpamrQphkrd1V14P9+5MaNiQREWkqWUltOC9/CElx8eypquS+xbOYs+VrjS0UaWV0BXQREQGgvLqK9MRkXlq3kEeXf0Dhnp1BhyQiTUQJoYiIAJCemMz/6z2a87sNZe2uYu5Z/C7vb1mpaqFIK1DXGMKOxpiCgzw+BOQ2YDwiIhKgUCjE6NzuDMhoz3NrPuPDwtWMyumucYUiLVxdCWGvJotCRERiSruktlzdezQ7q8pJiItjV2UFC0sKGJmdT0jJoUiLc8CE0Fq7uikDERGR2BIKhUhLSAZgbuEq/lnwFZ8WrWdC/lDaJbUNODoRaUgaQygiIgd1Soc+nNvtGFbu3MbURe/y4dZVhDW2UKTFUEIoIiIHFQqFcHJ7cuuAU8hvm8XzaxfwWsFXQYclIg2krjGEIiIi+8lJTuXqPmP4YOsq+qW3B6C8upLEULzGFoo0YwesEBpj3jTGpPnbpulCEhGRWBYXCnFC+550SEkjHA7zv6s+5cmv51JcXhZ0aCJymOrqMj7GWlvqb/+pKYIREZHmp296LitKC5m6eCYfFa7W2EKRZqiuLuPFxpjfAc8BCcaYvnjrDn6DtXZpYwQnIiKxLRQKcWL7XgzM6Mizq+fz3JrPWFC8gfPzh5CZ2Cbo8ESknupKCC8C7gKeBdoBbx3guDBas1BEpFXLTU5lYl+HOVv+f3t3Hh9Vdf9//DWTBbKQhQQIwSD7riC0qFUq+lWrpy5tv+oXbVVQvvy+iqLUBZFqtdaKSpW6tYVqtbYa26p1+7S0LohbgZbFqqyySiBASEgCSEhyf3/MQGM6hIDJ3JnM+/l45GHm3HtnPvNx7vDJOeeeu4Y3Slfp7iYicaapdQhLgHEAzrnVZtYnalGJiEjcCQYCnNK5Nyfm9yA1mITneczZspIT87urt1AkxjVr2RkVgyIi0lypwSQAtnxexRulq5i+7PjFkBIAACAASURBVC0W7tiouYUiMazZy8445wYAPwFOIzSXcB4w1cw+aqXYGr52FnA/MBTYBkw2s9UNtgeBM4DeZvZYa8cjIiKH1jUti5sGjObZDYv53fpFLK0o4cKioWSntPc7NBFppFk9hM65o4E3gPeAYYQKs3nA35xz0Zg/+DzwkpmdANwAFDvnchps/xNQCFzdIOagc26hc25ug58RUYhVRETCOrfP5Nq+J3N+t8GsqNzKI6ve1fxCkRjU3B7Ch4BbzOzpBm33O+c2A48ArbZOoXNuOLDVzAxCVzQ75x4HLgvHhZmdF973hgaHHgPMN7NrWis2ERE5tGAgwKmd+zAoqwsVNXsIBgLUex67a2vITGnnd3giQvMLwpFmdn6E9mJgZgvGE0lvYFmjto+Biw5x3OlAR+fcX4BUYLaZPdt4J+fcBGACQH5+PmVlZV8+4kOora2NyuvEE+UkMuUlMuUlsljPSzKQTxJlZWX8vXIT71Zu5Kzc3gxOz2/Vu5zEel78orxEFk95ycvLa7Hnam5BWOecSzWzmkbt7YD6Fosmsk2E5gc21APYeIjj+gNzzWxWeA7i751zO8xsTsOdzGwWMAtg0qRJXksm92DKyspa9H9iW6CcRKa8RKa8RBZPeflKRiorayp4sWwFn9ZWcmHRUDq0Um9hPOUlmpSXyBI1L82aQ0hojt6PI7RPCW9rTfOBoc65YwCcc7mE5go+09RBZjYhXOxhZpXA7Ry6V1FERKKgS/sOTOp3MucWDuLjylKmL3uTj3Zu9jsskYTV3B7CKcA859zLwB+BOuDbQD/gpFaKDQAzq3POjQEeDReDEJrPuNE5N9PMrnfO3Q6MBLo7514FZgBLzay8wVOlAvtaM1YREWm+pECQ/+rSl8FZXXhmw2ICkW+GJSJR0KyC0Mx2OeeOB64CLgg3vwVcbGatXmSZ2VoiXLhiZteH//ujxtucc68552ab2Z+ccx2BHwG3tXasIiJyeArSsri+39cJhucRzt36KTkp7RmW283nyEQSR7PXITSzWuDh8E88+C5wj3NuMrALuMfM3vM5JhERiWB/MVjveSytKGHtrh0MqyjhgqJjyUzWlcgira3ZBWG8MbMKQj2aIiISJ4KBANf0PYk3S1fzly3LWV1dxoVFxzI0p9Dv0ETatOZeVCIiIhIVSYEgZxT044b+o8lJac+v1y5ky55Kv8MSadPabA+hiIjEt8K0LCb3/zorq7ZTkJYFwJY9lQd+F5GWc0Q9hM65FOdc4i3SIyIiUZUUCDIwqzMAG3aXc+/yt3h63T/ZVdt4WVwR+TKa7CF0zp0LnE3ooozZ4dvGzQD+D0hyzs0DLjGz+FjSW0RE4la3tGy+UdCfv25ZyaqqbVzUfShDsrv6HZZIm3DQHkLn3J3Ag8AaYAfwmnPuISAdyAdygZWE7ycsIiLSmpICQc7qOoDv9z+FzOR2/GrNAp5ZvwjP8/wOTSTuNdVDeB0wyMxKAMILPv8DyDazz8Ntk4F1rR2kiIjIfkelZ/P9/qfwt9KVAK16H2SRRNHUHMI9+4tBADP7F1C+vxgMt9Ue4jlERERaXHIwyNldB3B21wEArKjcyu/WL2J3rW5IJXIkmirmIvXB17dWICIiIkdq8+dV/HPHZ9y7/E0+2VnqdzgicUe9eyIiEvdGd+7N9f1HkZaUwqw1f+eZ9YvVWyhyGJqaQ5jknOsLX7jbeMS2VolMRETkMHRPz+XG/qcwZ8sK3ihdTbe0LE7p3NvvsETiQlMF4R5gDl8s/nYDf2203+6WDkpERORIJAeT+GbhII7L7UaX9h0AKNlbRXpdFmlJKT5HJxK7DloQmlmPKMYhIiLSYgrTsgGora/j99uXkbRjBWO6H8eA8CLXIvJFmkMoIiJtVnIwiQvzB5IaTOYXn37AcxuW8Hmd5haKNHaoO5VkAaeb2QsN2v4FZDTcz8x6tU54IiIiX063dh24acBo/rx5OW9tXc3yyq1M6ncyuanpfocmEjOaulNJPrAQGN1oU164ADwNWA8MbbXoREREWkBKMInzug3mun6jGJDVmeyUNADd5UQkrKkh4x8Ar5rZpEbtAQAzWwfMBm5rndBERERaVo+MjvxP92EEAwEqavYwY8XbrKza5ndYIr5rqiA8B/hxhPaJDX5/HvhOi0YkIiISBbvraqipr+Wx1e/zh41L2VtX63dIIr5pqiBMN7Pyxo0N5xOa2V5AkzBERCTuFKZlc9OAUxndqTfvb1/HvcvfYpV6CyVBNVUQVobnER6Uc647sKNlQxIREYmO1GAS3zpqCNf2PZlgIMDb29b4HZKIL5oqCJ8C7jzE8dMJzSOUFuJ5HqXbKzTRWUQkinpl5nHzgNGM6T4MgO17d7G6arvPUYlET1PLzvwUeN059yQw1cw279/gnDsamEnoApOHWjXCBLKhZCuPPvUKC5auZOTQfky8/Fy6F2oRVRGRaEgNJpMaDP2zOGfLChbu2MioTj05p+sg2iU1uUqbSNw7aA+hmdUAZwLlwCfOubedc39wzr0HLCC0JM1/m5m6slrITXc/zqKPP6VLpxwWffwpN939uN8hiYgkpAuOOpavd+rFO9vWcv/yuXxaXeZ3SCKtqsk/eczsc2Cyc+524AQgH9gCfBDeJi2oevcecnMyCQaD5OZkUr1rj98hiYgkpHZJyXznqGM4Nrsrz25YzCOr3uW7Rw/nKx2L/A5NpFU0qw/czKqAv7VyLAkvMz2NHRVV5OV2oLyimo7ZHfwOSUQkofXpkM/NA07lr1tWMDCrCwC19fUkB3XnV2lb9ImOITOmjWf44N6UbqtgxJA+zJg23u+QREQSXrukZM7tNpiM5FTqvHoeWvUOL372ETX1WrdQ2g7Nko0hRYWdmH7LFWwt20nnvGwCgYDfIYmISAN1nkdReg5vb/uUTyq3cEn34fTM7Oh3WCJfmnoIY0wgEKBLfo6KQRGRGJQaTOLCoqFc3edr1IZ7C1/a9BE19XV+hybypaggFBEROUz9OnRiyoBTOTGvB0sqSqj36v0OSeRL0ZCxiIjIEWiflMJF3Yfyed0+2ielUFtfx7xtaxjVqRcpwSS/wxM5LOohFBER+RLaJ6UA8EllKS+XfML9y+eybpfu6irxRQWhiIhICzg2p5D/1/tEaupr+dnKd3hl08fs09xCiRMqCEVERFrIwKzOTBl4GsfndeeNrat5et0//Q5JpFk0h1BERKQFpSWlMKb7cQzNKSQ9KRWAmvpaggRI1txCiVEqCEVERFrB/jubALyy6RNWVW/nkqOPo3t6ro9RiUSmIWMREZFWNjCrC3vq9jFzxTu8VrKMWs0tlBijglBERKSVDcruwpQBpzKi41H8rXQlP13xNiV7dvodlsgBKghFRESiID05le8ePZz/7XU8e9VDKDFGcwhFRESiaHB2AQOyOpMUCPXJzNmygsFZXTgqPcfnyCSRqYdQREQkyvYXg9W1e3lv21oeWDGPP29eTm29boEn/lBBKCIi4pPM5HZMGXgax+V2Y86WFTy4ch6bdmtuoUSfCkIREREfZSSncmmPEVzRcyQ7933Ow6vf5fO6fX6HJQlGcwhFRERiwLE5Xemdmcf6XeW0T0rB8zx21Owhr12636FJAlAPoYiISIzISE5lUHZoQevFFSX8ZNnr/HXLCuo8zS2U1qWCUEREJAb165DPsdmF2OblzFwxj817Kv0OSdowFYQiIiIxKDO5HZf3/Apje3yVHfv2MGPF27y99VO/w5I2SnMIRUREYtiw3EL6ZObxx88+pF2S/tmW1qFPloiISIzLTGnH2J5fxfM8AP5etp7q2hpO7dz7wJqGIl+GPkUiIiJxIhAIALCmuoxXSz7hZyvfYcvnVT5HJW1BzPcQOueygPuBocA2YLKZrW6wPQicAfQ2s8catB8P3A1kAm8Ad5iZFnYSEZG4d3H34xiQ1YXnN37IjOVzcV0HMLpzH4LhglHkcMVDD+HzwEtmdgJwA1DsnGt4w8c/AYXA1fsbnHP9gJnA94ATgfXAY4iIiLQBgUCA4bnduGXgqQzK6sLLJZ+womqr32FJHIvpgtA5NxzYamYGYGYrgceBy/bvY2bnmdmvGx16DXC7mW0xM8/MZgGDGxWSIiIica1DSnvG9fwq1/Y9iYFZofULN+6uoD4811CkuWJ9yLg3sKxR28fARUdw3HKgD/CPho3OuQnABID8/HzKysqOONjmqq2tjcrrxBPlJDLlJTLlJTLlJbJEyEsOAcr2lrGzdi+Pbf4nBSkZnJfXj7yUtIMekwh5ORLxlJe8vLwWe65YLwg3EZof2FAPYGMzjusOfHao48K9h7MAJk2a5LVkcg+mrKysRf8ntgXKSWTKS2TKS2TKS2SJlJeOnseYlGG88Nm/mL1lMd8sHMTXO/WKOLcwkfJyOBI1LzE9ZAzMB4Y6544BcM7lEpor+Mwhjvs18EPnXPvwcWcAVWZW2prBioiI+CkQCPCVjkVMGXga/Tp05k+bPuKRVe+yr77O79AkxsV0QWhmdcAY4F7n3AeAAbeY2Ubn3EwA59ztzrlXge7OuVedc6PN7ANCReE7zrm/A1eEf0RERNq87JT2jO81kku6H0dRei4pwSS/Q5IYF+tDxpjZWsBFaL8+/N8fHeS4YqC4daMTERGJTYFAgJF53Q88/mz3Tl7a9BEXdR9Kp3aZPkYmsSimewhFRESkZVTs283GPRXcv3wu87atOXDXExFQQSgiIpIQhmR35ZYBp9E7M48XPvsXT2/9F9v37vI7LIkRKghFREQSRE5qGhN6ncCY7sPYUrOL97av9TskiRExP4dQREREWk4gEOCEvKPpXJvCUZ06A/DZ7grSklLIa5fhc3TiFxWEIiIiCSg7uR2pwWQ8z+O5DUso3VvNeYWD+Vp+D90TOQFpyFhERCSBBQIBrug1kp4ZHfnjZx/y89XvU7Z3t99hSZSpIBQREUlwuanp/F/vE7moaCgbdpdz3/I3Wb+r3O+wJIo0ZCwiIiIEAgG+lt+DAVmdeaN0Fd3SsgGo9zwNIScA9RCKiIjIAR1T07mwaCjJwSB76vZx3/K3+GD7Oq1b2MapIBQREZGIaurr6JDcjuc2LuUXn35AeY3mFrZVKghFREQkouyU9lzV52tccNSxrN21g+nL3uKD7evVW9gGqSAUEZFWN3fuXGbOnOl3GHIEgoEAJ3fqyc0DTuWo9Gz+XraeelQQtjW6qEREREQOKb9dBhP7nMTuuhqSAkGq9+3lk8pSvtqxiIAuOol76iEUERGRZgkGAmQmtwPg3e3reGbDYmavmU9FzR6fI5MvSwWhiIiIHLYzC/rx7W5DWFW1nXuXv8WCsg2aWxjHVBCKiIjIYQsGApzSuTc3DxxN1/YdeGbDYuZsWeF3WHKENIdQRESiQr1HbVOndplc0/dk3tm2hiHZXQHYW1dLajBJcwvjiHoIRUSk1fXs2ZPi4mKefvppv0ORVrC/tzCvXTqe5/HkuoX8as0Cdu773O/QpJnUQygiIq3u6KOPZv78+X6HIVHgAf0yO2Gbl3Hvsjf5zlHHMCL3KPUWxjj1EIqIiEiLCQYCnNqlDzcOGE2ndpn8dv0inli7gKp9e/0OTZqgglBERERaXJf2Hbiu3yjOKxzMht0VWsw6xmnIWERERFpFMBDgtC59OLlTT1KDSdR7HnO2rODk/J50SGnnd3jSgHoIRUQk6jzPo3R7ha48ThCpwSQANu3Zyeulq5i+/E0Wl2/yOSppSAWhiIhE1YaSrdwy/QnGXHMPt0x/gg0lW/0OyRdjx46loqLisI656667mDBhAhMmTOC5555rpchaT1F6Djf1P4W81HSeWvcPfr12IdWaWxgTVBCKiEhU3XT34yz6+FO6dMph0cefctPdj/sdUlx4/fXX6dKlC7NmzWLWrFmkp6f7HdIRKUjL4rp+ozin60A+2rmFx1a/r57iGKA5hHHG8zy2lu2kc162LuEXkbhUvXsPuTmZBINBcnMyqd4VH/fB9TyPO+64g/LyciorK+nbty/f+ta3uO+++8jJyaGqqopp06bRu3dvxo4dy8iRI9m+fTsLFixgypQpzJs3j5KSEgYNGsTEiRMBeOWVV1iyZAmbN2/mjDPOYNy4cVRUVDB16lRSU1OprKzk4osv5swzz+Sll17irLPO4rbbbqOmpoYrr7zS54wcuaRAkNML+jE4u4Dq2hoCgQB1Xj176vYduFeyRJcKwjiyoWQrjz71CguWrmTk0H5MvPxcuhd29jssEZHDkpmexo6KKvJyO1BeUU3H7A5+h9QsTz/9ND169ODOO+8EYOHChUyePJkXX3yRjIwMysvLufLKK3nhhRf47LPPuPHGGxkyZAjFxcXMnj2b3/zmNwCcc845BwrCwsJCLr30UgAuv/xyTjnlFO6//34mTpzIkCFD8DyPMWPGMHz4cDZu3EhFRQV33XUXu3bt4rvf/S5PPfUU2dnZ/iSkBXRNyzrw+xulq5i3bQ0XFg1laE6hj1ElJg0ZxxENs4hIWzBj2niGD+5N6bYKRgzpw4xp4/0OqVkWLlzImWeeeeBxr169KCoqIiMjA4Dc3FySkpKora0lMzOTIUOGAFBQUMDw4cMPHJec/O++mBEjRhz4/fTTT2fx4sVs3rz5wLGBQICTTz6ZDz/8kOTkZC6++GIAMjIyOO2001i0aFHrveEoG5LdlZyUNH69diFPrf0Hu2pr/A4poaggjCP/Mcyyu+lhFl3FJyKxqKiwE9NvuYLiR6Zyz5RxFBV28jukZjnxxBN59dVXDzxetWoVJSUl7NkT+i4uLy+npqbmCwXfoaxcufLA73PmzOG4446ja9euLFu2DAh9j7/zzjsMGTKEk046iffee+/A/gsXLmTgwIFf9m3FjMK0LCb3/zqu6wA+3FnCPcveZFllqd9hJQwNGceRwxlm0fCyiMSyQCBAl/wcv8M4LJdccgl33XUX11xzDXv37qWoqIj77ruPq666iuzsbCorK3nggQcO6zmXLVvGs88+y7Zt23DO0atXL37yk59w6623kpqays6dOxk/fjydO3fm6quvZurUqTz//PNUVlZywQUXUFBQ0Erv1h9JgSBnFvRnSHYBz6xfTFJA/VbRElDv0b9NmjTJe+ihh1r9dcrKysjLyzvs4zaWbOORp15mwdKVHD+sPxMvO/egf1n/z8R72LEzVDyWlVfRMbsDzz069cuG3mqONCdtnfISmfISmfISmfISWaznpd7zCIYvnnyjdBWd22VyTE7XVn/dWM9LIy12dal6COPI/mGW5lxlHK9X8YmIiAAHisHa+noWl2/isz07+UruUXz7qGPISE71Obq2R32xcWb/MMuhlpzJTE+jvKKa+vp6yiuqyUxPi1KEIiIiLSc5GOT6fl/nGwX9WVS+iXuXvclHO7f4HVabo4IwzjT3QpF4vYpPRESkseRgkLO7DuD7/b9OZnI7Hl8zn+17d/kdVpuiIeM4cjgXihzO8LKIiEg8OCo9h+/3P4VPq7eT3y603M/mPZVfWM9Qjox6COPI4a5D2NzhZYk/WlJIpPXpPItNycEg/bNCnSFrqsu4d/lbPLN+Ebtr9/kcWXxTD2Gc8DyP8spqcrN1oUii05JCIq1P51l86J6ewxld+vFG6SpWVG1jTPdhDMzq4ndYcUk9hHFgQ8lWbpn+BCvXfsaHy9ZSvWsPW7dVkJHW3u/QxAe6Y41I69N5Fh+Sg0l8s3Ag1/cfRVpSCr/89O/8YeNSv8OKSyoI48D+L6ahA3tRV1fPgiUrKdm6g/y8bDaUbPU7PImyw71jjYgcPp1n8aV7ei439j+F07v0JSdFq2ocCRWEcWD/F1NGensCQUhNSWL44N6sWrtJf7UmIC0pJNL6dJ7Fn+RgEucUDuKMgn4AfLRzM8UblvB5neYWNocKwjjQ8Itp795a0tq3I5j0779aNek5sWhJIZHWp/Ms/m35vJr5ZeuZvuwtVlRqNO1QdOu6BmL11nUNb1lXvrOK7A6ZdO2Sy+bScnZWVZOb3SHuJz3H2a2CouZgefE8L6GXFNLnJTLlJbIjzUtbP88S4fOybtcOnlm/mK17q/laXg/O6zaI9kkpTR4TZ3lpsQ+megjjwP41BYsfmcrLv7qDE47rT+m2CnZW7SK7Q4YmPSegppYU0lIZIi0jFpbu0vn85fTI6MiNA0Zzauc+fFC2jkXlm/wOKWZp2Zk4sf+LCTiw4PS4G39KZmaalqGRA7RUhkjbofO5ZaQGkzi/22BG5B5FYXgB63W7dlDQvsMhewsTiXoI49D+4rBDRromPcsXaKkMkbZD53PLOio9m2AgwN66Wmavmc+9y99iZdU2v8OKGSoI45gmPUtjWipDpO3Q+dw62iUlc2XPkSQFgjy2+n3+uHEpe+tq/Q7LdxoyjmO6X7E0lpmexo6KKvJyO1BeUU3H7A5+hyQiR0jnc+vplZnHzQNG81rJMuZtW8Oyyq1c128UWSmJe8MH9RDGuViY9CyxQ73GIm2HzufWlRpM5ttHHcM1fU9mYFZnOiS3A4jZC3jGjh1LRUVFqz2/eghF2hD1Gou0HTqfo6N3Zh69M0PLzJTt3c3sLUu4qP0wemfm+xxZdKkgFGljGl6RLiLxTedzdO2uq6HGq+ORVe8xqlMvvtl1IO2SjqxU8jyPO+64g/LyciorK+nbty/f+ta3uO+++8jJyaGqqopp06bRu3dvxo4dy8iRI9m+fTsLFixgypQpzJs3j5KSEgYNGsTEiRMBeOWVV1iyZAmbN2/mjDPOYNy4cQQCgRzgHqAGyAKe9Tzvr4FAoA9wTTicrcB0z/PqDxZvTBeEzrks4H5gKLANmGxmqxvt803gZiAF+CPwoJnFZn+vSAto64vlioj4pSg9hwkFx/H+51uYt20Nn1SWckn34+iVefgLVT/99NP06NGDO++8E4CFCxcyefJkXnzxRTIyMigvL+fKK6/khRde4LPPPuPGG29kyJAhFBcXM3v2bH7zm98AcM455xwoCAsLC7n00ksBuPzyy7niiit6ATcBj3qe91Eg9I9CcSAQWASMAm73PK8yEAhcC1wA/P5g8cb6HMLngZfM7ATgBqDYOXfgTyXn3KmEqt9zCL3xjsCtDba/5pyb2+Dnm9ENX6RlbSjZyi3Tn2DMNfdwy/Qn2FCi2zGJiLSk1GAS/110LBP7fI16z+OdbWuP6HkWLlzImWeeeeBxr169KCoqIiMjA4Dc3FySkpKora0lMzOTIUOGAFBQUMDw4cMPHJec/O++uxEjRhz4/fTTTwc4Dujqed5HAF5oAuS7wLGe5/3a87zK8O4esLmpeGO2IHTODQe2mpkBmNlK4HHgsga7TQYmmVmVmdUBPwTGhI/PBerMbHSDn9ei+y5EWpbWJRMRiY6+HToxZcCpXFh0LACln1extnpHs48/8cQTefXVVw88XrVqFSUlJezZE1o+qLy8nJqami8UfIeycuXKA7/PmTMHYDGwORAIDAQI9xCOAj4KP84KBAJPAg74Z1PPHctDxr2BZY3aPgYuavD4aODAELKZ1Tnntjnn8oFTgHrn3GtABqHexkcaDyc75yYAEwDy8/MpKytr8TfSWG1tbVReJ54oJ5E1zktFZRUd0ttRV1tLh/R2VFRWJWTe9HmJTHmJTHmJTHmJLFJe9gB/2r6cZbu3c0KHQkZnH01KMKnJ5/nGN77BjBkzGD9+PHv37qVbt27ceuutXHHFFWRlZVFVVcVtt91GWVkZNTU1B15z586d7Nq168Dj/dv27t3LwoULeeKJJygrK+P000/nt7/97ZpAIHAr8JNAIFADZAO/8jxvK0C4h3BsIBD4b0IjqvcdLN5YLgg3AWc0ausBbGzwuBToCpQ0aOsElAGDCVXOPyb0Pn8B/C8wq+ETmtms/W2TJk3yonFD6y9z4+y2On8szm4mHjWN85KT1YEdO0PrklVUVdExu0NC5k2fl8iUl8jaYl7mzp3Ln//8ZzIyMtiyZQtFRUUUFBSwceNGli5dyq9+9Styc3ObfI62mJeWcLC8XJ4zkpdLPuH97ev4tGYnlxw9nJ4ZHZt8rnvuuec/2kaNGvUfbeHBUADOP/98zj///P/Y9uyzz0Z8Dc/zyoGrGrcHAoFLgD94nrcP2AdUNRVrLBeE84EHnXPHmNm/wkPAVwMXNtjnCeBu59yVZlbvnBsPzA33Av6owX51zrkpwLM0Kgjjie5rKTOmjeeRp15mwdKVHD+sPxMvO9fvkETEJ+vXr6e4uBiAoUOHUlxczLhx4yguLua1117je9/7ns8Rti3tk1K4qGgoQ7O7UrxhCQ+tfIfLe3yVYbmFfod2MOuAxwOBwC5C8wfvbmrnmJ1DGJ4TOAa41zn3AWDALWa20Tk3M7xPMbAUmO+cmw+MAKbAgTmEDaUSqpDjluaPyf51yYofmco9U8ZRVNjJ75BExCcnnHDCgd979uzJwIEDgdBFCRoKbj39szozZeCpnNq5D/2zQt/BtfUHXc3FN57nve953mWe513led6PPM+ra2r/WO4hxMzWEpoI2bj9+ga/zwRmRjj8XufcEuDnQBrwU+DJ1ok0Ov7jvpa7dF/LRKR1yUTkUGL1bhttRfukFM7rNhgIFYMPrpzHgA6dOKvrgEPOLYxVMdtD2AKuBYqAt4E/AX8K9yjGrcz0NMorqqmvr6e8oprM9DS/QxIREUlodV49Rek5vLF1NT9d8TYbdpXjeR6l2yviqjAPxFOwrW3SpEneQw891Oqvc6QTeTeWbPuP+WNtZchQk5sjU14iU14iU14iU14iU14iO9K8LKsspXjDEir3fU5wzW4Wv7CUrw7p29rz/Vvs6tKYHjKWL9J9LUVERGLTwKwuTBlwGtf/8XckdWpP5wbz/Z97dKrf4R1SWx4ybpP2zx9TMSgiIhJb0pNT2DRnDdTU0/EbR4fm+++Oj/n+KghFREREWkhmehq1A1XmYwAAED5JREFU++rA8+Jqvr8KQhEREZEWMmPaeDIz0qjZV8uIIX2YMW283yE1i+YQioiIiLSQosJO9KouYF9tLZPPOyVupniph1BERESkJQUgJSU5bopBUA+hiIiISIsanFXgdwiHTQWhiIiISAs6rUsfv0M4bBoyFhEREUlwKghjWDze+kZERCTRPbzqXR5e9a7fYRwWDRnHqA0lW3n0qVdYsHQlI4f2a+1b34iIiEgCUw9hjLrp7sdZ9PGndGlw6xsRERGR1qCCMEZV795Dbk4mwWAwrm59IyIiIvFHBWGMykxPo7yimvr6+ri69Y2IiIjEHxWEMWrGtPEMH9yb0m0VcXXrGxERkUQ3OKsg7tYi1EUlMaqosBPTb7mCrWU76ZyXHVernYuIiCSyeFyHUAVhDAsEAnTJz/E7DBEREWnjNGQsIiIi0oLicR1CFYQiIiIiCU4FoYiIiEiCU0EoIiIikuBUEIqIiIgkOF1lLCIiItKC4m0NQlBBKCIiItKi4nEdQg0Zi4iIiCQ4FYQiIiIiLUjrEIqIiIhI3FFBKCIiIpLgVBCKiIiIJDgVhCIiIiIJTsvOiIiIiLQgrUMoIiIikuC0DqGIiIiIxB0VhCIiIiItSOsQioiIiEjcUUEoIiIikuBUEMY4z/Mo3V6B53l+hyIiIiJtlK4yjmEbSrby6FOvsGDpSkYO7cfEy8+le2Fnv8MSERGRNkY9hDHsprsfZ9HHn9KlUw6LPv6Um+5+3O+QRERE5BAGZxXE3VqE6iGMYdW795Cbk0kwGCQ3J5PqXXv8DklEREQOQesQSovKTE+jvKKa+vp6yiuqyUxP8zskERERaYNUEMawGdPGM3xwb0q3VTBiSB9mTBvvd0giIiJyCPG4DqGGjGNYUWEnpt9yBVvLdtI5L5tAIOB3SCIiItIGqSCMcYFAgC75OX6HISIiIm2YhoxjnNYhFBERkdamHsIYpnUIRUREJBrUQxjDtA6hiIhI/NE6hNKitA6hiIhI/NE6hNKitA6hiIiIREPM9hA657KA+4GhwDZgspmtbrTPN4GbgRTgj8CDZuaFt/UDHgDygcXAzWZWFb138OXNmDaeR556mQVLV3L8sP5MvOxcv0MSERGRQ9i/BuG1fU/2OZLmi+UewueBl8zsBOAGoNg5d2D9FefcqcA1wDnAKKAjcGt4Wx7wLKEi8gTgNeD30Q3/y9u/DmHxI1O5Z8o4igo7+R2SiIiItEGBWFzOxDk3HLjBzL7boO0qIMXMHgo/fjm8z6rw4yRgiZkd45z7PlBlZrMbHP8ccLeZfdjotSYAEwDy8/NHPPjgg6387qC2tpbk5JjtnPWFchKZ8hKZ8hKZ8hKZ8hKZ8hJZPOUlLy+vxe5YEavvuDewrFHbx8BFDR4fDRwYQjazOufcNudcfvj4ZyMc3w/4QkFoZrOAWQCTJk3y8vLyWuQNNKWsrIxovE48UU4iU14iU14iU14iU14iU14iS9S8xOqQ8Sage6O2HsDGBo9Lga6N9ukElDXzeBEREREhdgvC+cBQ59wxAM65XOBq4JkG+zwB3O2cC4b3GQ/MDV9U8gxwXfg4nHPHAgOAhdF7CyIiIiLxISYLQjOrA8YA9zrnPgAMuMXMNjrnZob3KQaWAvOdc/OBEcCU8LZ1hC4wsfDx9wAXm1l91N+MiIiISIyL1TmEmNlawEVov77B7zOBmQc5/g3gjVYLUERERKSNiMkeQhERERGJHhWEIiIiIglOBaGIiIhIglNBKCIiIpLgVBCKiIiIJDgVhCIiIiIJTgWhiIiISIJTQSgiIiKS4FQQioiIiCQ4FYQiIiIiCU4FoYiIiEiCU0EoIiIikuBUEIqIiIgkOBWEIiIiIglOBaGIiIhIglNBKCIiIpLgAp7n+R1DzHDObQPWR+Gl8oHtUXideKKcRKa8RKa8RKa8RKa8RKa8RBZPedluZme1yDN5nqefKP+cffbZ//A7hlj7UU6UF+VFeVFelJdY+EnUvGjIWERERCTBqSAUERERSXAqCP0xy+8AYpByEpnyEpnyEpnyEpnyEpnyEllC5kUXlYiIiIgkOPUQioiIiCS4ZL8DEBFpinMuF3gM6EHoO8uAO8wsoYc3nHP/BdwGeITyMtvMfuNvVCISrzRkHEXOuSzgfmAosA2YbGar/Y0qupxzAeB/gf8G0oC/AT8GUoA7gdHALmCqmS30KUzfOeeygV1mVuucOx64G8gE3iBUDO3zNcAocs79DviZmS0IPz4f+DuwFbgB+A6wD7jHzP7iW6BR5JxLBj4ERplZmXMuFXgTuBJYS4KdS865rsAYYI6ZfRLOx0Fz4JwbS+h7yAN+aWZPRz3oKIiQlwuAcYS+SxYBt5jZ3vD3csKcS43z0mhbBlBnZp+HH/cDHiC0NuFi4GYzq4pyyFGhIePoeh54ycxOIHTyFTvncnyOKdr6A52Bc4DTgAHAtwlN4l0DfA24FHjIOdfbryD95JzLB5YAR4W/jGYC3wNOJLRw+mM+hhdVzrlCYE/oVzfPOfdXoNbMSoEfAtnAKOBc4Drn3Cj/oo2qemAn0Cn8uCOh4qaMBDuXnHP/Q+gPpmOAQeHmg+YgXAyeApwK/BfwX86570Y57FbXOC/h75XjgYsIFco1wMTw7glzLh3k87J/W3vgLeCE8OM84FlCnTcnAK8Bv49qwFGkgjBKnHPDga1mZgBmthJ4HLjM18CizMyWm9mPzWyfmdUCm4AOQB8zm21mnpltBn7Ev7+sEka45+fnwLpw0zXA7Wa2JZybWcDgBPpD4jhC/0i9a2ZfJ3S+3OqcOxW4APihmdWZWSXwfWCyf6FGj5nVA9cCHzjnPgRWE/rDwSPBziUze87MrgA2wIF/xJvKwdXAJDOrMbO9wCRCuWxTGufFzLab2U1mtis83WI9UBvePWHOpcZ5aeRhoLTB48uBX5jZqvCxrwKVzrljWz/S6FNBGD29gWWN2j4G+vkQi++cc92dcz8PP/wEWN5ol0TNzQzgUf59C8VIn5vlQJ9oBuWjHOCvZvY6gJltIfQP1c3AlnBhtN9yoFf0Q4w+51wnQj3FJ5jZsYQ+JxMIvf9EP5cOlYP2DYf8wgVQepRi851zLt85dw/wFeAX4c9Swp5L+znnbgLeBv7ZoDmh/t1WQRg9m4Dujdp6ABujH4q/nHP/D7gXmGlmN6PcAOCcmwCsNrO5DZoTPTcbgc8bte0CKghNPWioCNgcjaBiwChChfIKgPAQ+i8JzYtL5M8LHPqcqQvPMQTAOdeOf/eUtWnOufOAJ4DnzewKM6shNM0gkc8lnHNnA/lm9ttGmxLq+1cFYfTMB4Y6546BA1dOXg0842tUUeacG0ZoIvzFDf4xKwFqnXOnh/dpD0wj9MWVEJxzJwPHmdkjjTb9GvhhOCc4584AqsIFQCKYD4x2zg0GcM6lEboI6XfA+865K8LtQeAu4Fd+BRpli4Azw/PC9uflUmAOCX4uNeP75DngBw0O+QGhz1Ob5pwrIDRs/m0z+8f+9nDPYMKeS865/oQutLk1wuZnCM2nzA3veyyhee9t8iItXWUcRc65noSGA3PDTVMb9Qa1ec65qcBYvvjX54uEvpAfBY4GkoD7zazNTt5tLHwlbQ9CV/gBDCE0L+x7hIZ2bgDqCF1Beo2ZlfkQpi+cc4MIXeWXRmh5lVlm9lT4asAHgGGEPjNPmdnD/kUaXeE/Dn5A6AKTdoTe/y/DRWLCnEvOuT6E5k/2AyoJFcu3cpAchOfp3gmcAQSAvxC6cr8u+tG3ngh5yQZSCX2H7PeBmU1NpHMpQl4WA8cSungNQtNxthP6nn03vLzTj8PbdgBXm9l62iAVhCIxyDk3FxhrZut8DkVEJGE4554Enky0zhrQkLFIrNpCgsxrEhGJIdv5z3nLCUE9hCIiIiIJTj2EIiIiIglOBaGIiIhIglNBKCIiIpLgVBCKiIiIJDgVhCIiIiIJTgWhiIiISIJTQSgiIiKS4FQQioiIiCQ4FYQiIiIiCU4FoYiIiEiCU0EoIiIikuBUEIqIiIgkOBWEIiIJxjk3zDmX43ccIhI7VBCKiCSeYYAKQhE5IOB5nt8xiIh8ac65J4EuZnZ2E/vcAFxrZj3Cj38G/NPMftOCcYwFzjKzMS31nIfx2o8B+8zsuoNsvxS4HugJ1AFrgAfNrDh6UYpILFIPoYi0JWc5506PtME5lwf8oGGbmV3XksWgn5xzo4CzgGkH2X4FMA44F/i+mXUCrgX+zzlXGLVARSQmqSAUkbbkNWCGcy7Sd9vtwLtRjicqnHPtgNnAVWZWfZDdJgK3m1nJ/gYzW2Bmoxu2iUhiSvY7ABGRFvQqcCFwOfDr/Y3Oub7h9u8AxQ3anwT+bma/CA/1fgWYBUwHhgP7gEvNbG54/7nAdDP7S4Pn6BF+jgLn3FXhY1Odc1uAH5vZI+H9RgAPAgOBWuAPwDQzqwpv/wpwP9AfSCJU3E4xs23NeN+3AwvNbE4T+3hAdjOeS0QSkHoIRaStuQG40zmX3qDtXuAeoOIQxw4CHiA0pFoA/IRQgdgsZvZz4DrgJTMraFAMjgIeBq4JD9X2AFYCf3POpTjnAsArwANmVhjePg/42qFe0zl3LDCW0NzApswGHg8XrV2b+55EJDGoIBSRNsXMlgBvEioMcc6dDAwGftGMw4cCF5rZ8vDj2UBReP7hl/Eo8D9m9mE4xr3hYvEdYDyhnrss4C/h7XvM7Ekze6mpJ3XOJQGPAzeaWVlT+5rZL4FLgNOAm51zy5xzM8I9nCKS4DRkLCJt0TRgkXNuNqEevylmts85d6jjXjez8v0PzKzWObcdyAOaLLgOxjmXDwwBFkZ4/WTgWTOrcM7dBbzunPsLoYL2H2ZWd4invx4oNbNnmxOLmb0JvBkeHl8FXAZ85Jz7hpm91+w3JSJtjgpCEWlzzGyTc+6XwBvAdjP7UzMPjbQOVx3/Hk2pIzS/73Dsf84BZnbQIWszm+6c+ylwPPBN4GHn3A/M7K+R9nfO9QJuAr56mPHsf733gPecc+uBqcA5R/I8ItI2aMhYRNqq+4AMwkPHLWQ7UNSo7dymDggP5X4IXNx4m3PujIZXRJvZPjN718xuJ1Ts/aSJp/4lcJeZbWxO4M657gfZtAJo15znEJG2Sz2EItImhZdf6dHCTzuP0Lp9zwGVwKWEevQafpfWAAXh+X25ZrYduAZ42Tm3A3ieUAF2GzDYzP7mnDsFuAi418w2OOdSgfOAtZGCcM6NA9KAx5oTtHOuM6HewJ8CTzVoHwjcSujqZxFJYOohFBFpvicI9fatJtSz1ofQYs+bnHMnhff5G9AB2Ao855wLmNm7wDeAScA24GNC378Xho+ZB3wEvOic20yoECwArm4cgHOuC6Gew/81s2bdasrMtgInheP9J/Az51wp8AzwsJn99nCSICJtj25dJyKSYMIXlcw1s3U+hyIiMUI9hCIiiWcJh16TUUQSiHoIRURERBKceghFREREEpwKQhEREZEEp4JQREREJMGpIBQRERFJcCoIRURERBKcCkIRERGRBKeCUERERCTB/X9l0ZR9t67BKwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1bb5cf74710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract efficiencies\n",
    "eff_data = df_graph[['talktime_cost', 'data_cost']].copy()\n",
    "\n",
    "# Compute convex hull\n",
    "hull = ConvexHull(eff_data)\n",
    "\n",
    "# CODE FOR CUSTOM GRAPHICS NOT INCLUDED"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that M1 offers 2 of the most value-for-money data plans, while Singtel offers the best talktime plan. Although Starhub provides the most value in the SIM-only market for talktime, Singtel takes top spot for bundle plans."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion\n",
    "In this post, I used linear regression to quantify the relationship between mobile plan parameters and their price, a decision tree to identify the most important parameters, and data envelopment analysis (DEA) to compute the most efficient plans for data and talktime, and the optimal tradeoff between these two priorities.  \n",
    "  \n",
    "As you can see from the variety of plans, the offering is extremely polarised. Plans deliver value either in terms of talktime or data. Thus, you only need to know your priority and your budget to make a good decision.  \n",
    "  \n",
    "## Data\n",
    "If your priority is data, you only need to calculate the most efficient data plans based on data per dollar. Here are the top 5 from the SIM-only and bundle categories. M1 has an excellent offering in both categories."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>sms</th>\n",
       "      <th>data</th>\n",
       "      <th>provider</th>\n",
       "      <th>contract</th>\n",
       "      <th>sim_only</th>\n",
       "      <th>data_cost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>mysim98_12mth</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>10.204082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mysim50_12mth</td>\n",
       "      <td>50.0</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>30.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>0.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>data_baller</td>\n",
       "      <td>48.0</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>circles_life</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>0.541667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>xs_sim12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>200</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>0.375000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>mysim40_12mth</td>\n",
       "      <td>40.0</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>15.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>0.375000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>mysim_e118</td>\n",
       "      <td>118.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>8.474576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>mysim_e90</td>\n",
       "      <td>90.0</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>30.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>comboxo</td>\n",
       "      <td>78.0</td>\n",
       "      <td>300</td>\n",
       "      <td>300</td>\n",
       "      <td>18.0</td>\n",
       "      <td>singtel</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>mysim_e70</td>\n",
       "      <td>70.0</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>15.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>0.214286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mysim_e40</td>\n",
       "      <td>40.0</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>5.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            name  price  talktime    sms    data      provider  contract  \\\n",
       "0  mysim98_12mth   98.0     10000  10000  1000.0            m1        12   \n",
       "1  mysim50_12mth   50.0       100    100    30.0            m1        12   \n",
       "2    data_baller   48.0       100      0    26.0  circles_life         0   \n",
       "3       xs_sim12   24.0       200      0     9.0       starhub        12   \n",
       "4  mysim40_12mth   40.0       100    100    15.0            m1        12   \n",
       "5     mysim_e118  118.0     10000  10000  1000.0            m1        24   \n",
       "6      mysim_e90   90.0       100    100    30.0            m1        24   \n",
       "7        comboxo   78.0       300    300    18.0       singtel        24   \n",
       "8      mysim_e70   70.0       100    100    15.0            m1        24   \n",
       "9      mysim_e40   40.0       100    100     5.0            m1        24   \n",
       "\n",
       "  sim_only  data_cost  \n",
       "0     True  10.204082  \n",
       "1     True   0.600000  \n",
       "2     True   0.541667  \n",
       "3     True   0.375000  \n",
       "4     True   0.375000  \n",
       "5    False   8.474576  \n",
       "6    False   0.333333  \n",
       "7    False   0.230769  \n",
       "8    False   0.214286  \n",
       "9    False   0.125000  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract top data plans\n",
    "data_plans = mobile.copy()\n",
    "\n",
    "# Calculate efficiency\n",
    "data_plans['data_cost'] = data_plans.data / data_plans.price\n",
    "\n",
    "# Sort plans\n",
    "data_plans = data_plans.sort_values(by = ['sim_only', 'data_cost'], ascending = [False, False]).groupby('sim_only').head(5)\n",
    "data_plans = data_plans.reset_index(drop = True)\n",
    "\n",
    "# View\n",
    "data_plans"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Talktime\n",
    "If talktime is your priority, the decision is easy. There are many plans with unlimited talktime. It is therefore only a matter of price, and Starhub provides the best offers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>sms</th>\n",
       "      <th>data</th>\n",
       "      <th>provider</th>\n",
       "      <th>contract</th>\n",
       "      <th>sim_only</th>\n",
       "      <th>talktime_cost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>m_sim12</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>227.272727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>m_nocontract</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>227.272727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>l_sim12</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>185.185185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>l_nocontract</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>185.185185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>mysim98_12mth</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>12</td>\n",
       "      <td>True</td>\n",
       "      <td>102.040816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>combo3</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>singtel</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>145.137881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>m</td>\n",
       "      <td>88.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>113.636364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>combo6</td>\n",
       "      <td>95.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>singtel</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>104.275287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>l</td>\n",
       "      <td>108.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>starhub</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>92.592593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mysim_e118</td>\n",
       "      <td>118.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>m1</td>\n",
       "      <td>24</td>\n",
       "      <td>False</td>\n",
       "      <td>84.745763</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            name  price  talktime    sms    data provider  contract sim_only  \\\n",
       "0        m_sim12   44.0     10000      0    13.0  starhub        12     True   \n",
       "1   m_nocontract   44.0     10000      0     5.0  starhub         0     True   \n",
       "2        l_sim12   54.0     10000      0    19.0  starhub        12     True   \n",
       "3   l_nocontract   54.0     10000      0     8.0  starhub         0     True   \n",
       "4  mysim98_12mth   98.0     10000  10000  1000.0       m1        12     True   \n",
       "5         combo3   68.9     10000  10000     3.0  singtel        24    False   \n",
       "6              m   88.0     10000      0     5.0  starhub        24    False   \n",
       "7         combo6   95.9     10000  10000     6.0  singtel        24    False   \n",
       "8              l  108.0     10000      0     8.0  starhub        24    False   \n",
       "9     mysim_e118  118.0     10000  10000  1000.0       m1        24    False   \n",
       "\n",
       "   talktime_cost  \n",
       "0     227.272727  \n",
       "1     227.272727  \n",
       "2     185.185185  \n",
       "3     185.185185  \n",
       "4     102.040816  \n",
       "5     145.137881  \n",
       "6     113.636364  \n",
       "7     104.275287  \n",
       "8      92.592593  \n",
       "9      84.745763  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract top plans\n",
    "tt_plans = mobile.iloc[[5, 6, 8, 9, 11, 12, 13, 16, 17, 24, 25, 29, 30, 34, 36, 37, 38, 45], ].copy()\n",
    "tt_plans = mobile.copy()\n",
    "\n",
    "# Calculate efficiency\n",
    "tt_plans['talktime_cost'] = tt_plans.talktime / tt_plans.price\n",
    "\n",
    "# Sort plans\n",
    "tt_plans = tt_plans.sort_values(by = ['sim_only', 'talktime_cost'], ascending = [False, False]).groupby('sim_only').head(5)\n",
    "tt_plans = tt_plans.reset_index(drop = True)\n",
    "\n",
    "# View\n",
    "tt_plans"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data and Talktime\n",
    "If it is a combination of data and talktime you seek, use the table below to pick a plan. These are the most efficient plans on offer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>provider</th>\n",
       "      <th>price</th>\n",
       "      <th>talktime</th>\n",
       "      <th>data</th>\n",
       "      <th>efficiency</th>\n",
       "      <th>sim_only</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>m_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>m_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>44.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>mysim98_12mth</td>\n",
       "      <td>m1</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>mysim_e118</td>\n",
       "      <td>m1</td>\n",
       "      <td>118.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0.831</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>l_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.821</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>l_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>54.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.815</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>combo3</td>\n",
       "      <td>singtel</td>\n",
       "      <td>68.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.639</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>m</td>\n",
       "      <td>starhub</td>\n",
       "      <td>88.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.500</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>combo6</td>\n",
       "      <td>singtel</td>\n",
       "      <td>95.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.459</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mysim98</td>\n",
       "      <td>m1</td>\n",
       "      <td>98.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0.456</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>l</td>\n",
       "      <td>starhub</td>\n",
       "      <td>108.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.407</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>xl_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>119.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.379</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>xl_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>119.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.371</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>max_plus</td>\n",
       "      <td>m1</td>\n",
       "      <td>228.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.193</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>xl</td>\n",
       "      <td>starhub</td>\n",
       "      <td>238.0</td>\n",
       "      <td>10000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.185</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>combo12</td>\n",
       "      <td>singtel</td>\n",
       "      <td>239.9</td>\n",
       "      <td>10000</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.183</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>s_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>34.0</td>\n",
       "      <td>400</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.069</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>mysim50_12mth</td>\n",
       "      <td>m1</td>\n",
       "      <td>50.0</td>\n",
       "      <td>100</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.059</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>xs_sim12</td>\n",
       "      <td>starhub</td>\n",
       "      <td>24.0</td>\n",
       "      <td>200</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.057</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>s_nocontract</td>\n",
       "      <td>starhub</td>\n",
       "      <td>34.0</td>\n",
       "      <td>400</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.057</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             name provider  price  talktime    data  efficiency sim_only\n",
       "0         m_sim12  starhub   44.0     10000    13.0       1.000     True\n",
       "1    m_nocontract  starhub   44.0     10000     5.0       1.000     True\n",
       "2   mysim98_12mth       m1   98.0     10000  1000.0       1.000     True\n",
       "3      mysim_e118       m1  118.0     10000  1000.0       0.831    False\n",
       "4         l_sim12  starhub   54.0     10000    19.0       0.821     True\n",
       "5    l_nocontract  starhub   54.0     10000     8.0       0.815     True\n",
       "6          combo3  singtel   68.9     10000     3.0       0.639    False\n",
       "7               m  starhub   88.0     10000     5.0       0.500    False\n",
       "8          combo6  singtel   95.9     10000     6.0       0.459    False\n",
       "9         mysim98       m1   98.0     10000    25.0       0.456     True\n",
       "10              l  starhub  108.0     10000     8.0       0.407    False\n",
       "11       xl_sim12  starhub  119.0     10000    33.0       0.379     True\n",
       "12  xl_nocontract  starhub  119.0     10000    15.0       0.371     True\n",
       "13       max_plus       m1  228.0     10000    13.0       0.193    False\n",
       "14             xl  starhub  238.0     10000    15.0       0.185    False\n",
       "15        combo12  singtel  239.9     10000    12.0       0.183    False\n",
       "16        s_sim12  starhub   34.0       400    11.0       0.069     True\n",
       "17  mysim50_12mth       m1   50.0       100    30.0       0.059     True\n",
       "18       xs_sim12  starhub   24.0       200     9.0       0.057     True\n",
       "19   s_nocontract  starhub   34.0       400     4.0       0.057     True"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract sim_only\n",
    "names = mobile.sim_only.copy()\n",
    "names.index = mobile.name\n",
    "\n",
    "# Map sim_only\n",
    "top_plans['sim_only'] = top_plans.name.map(names)\n",
    "\n",
    "# Sort plans\n",
    "top_plans = top_plans.sort_values(by = 'efficiency', ascending = False)\n",
    "top_plans = top_plans.reset_index(drop = True)\n",
    "\n",
    "# View table\n",
    "top_plans.head(20)"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}