06.06-Mean-Encoding.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Target guided encodings\n",
    "\n",
    "In the previous lectures in this section, we learned how to convert a label into a number, by using one hot encoding, replacing by a digit or replacing by frequency or counts of observations. These methods are simple, make (almost) no assumptions and work generally well in different scenarios.\n",
    "\n",
    "There are however methods that allow us to capture information while pre-processing the labels of categorical variables. These methods include:\n",
    "\n",
    "- Ordering the labels according to the target\n",
    "- Replacing labels by the target mean (mean encoding / target encoding)\n",
    "- Replacing the labels by the probability ratio of the target being 1 or 0\n",
    "- Weight of evidence.\n",
    "\n",
    "All of the above methods have something in common:\n",
    "\n",
    "- the encoding is **guided by the target**, and\n",
    "- they create a **monotonic relationship** between the variable and the target.\n",
    "\n",
    "\n",
    "### Monotonicity\n",
    "\n",
    "A monotonic relationship is a relationship that does one of the following:\n",
    "\n",
    "- (1) as the value of one variable increases, so does the value of the other variable; or\n",
    "- (2) as the value of one variable increases, the value of the other variable decreases.\n",
    "\n",
    "In this case, as the value of the independent variable (predictor) increases, so does the target, or conversely, as the value of the variable increases, the target value decreases.\n",
    "\n",
    "\n",
    "\n",
    "### Advantages of target guided encodings\n",
    "\n",
    "- Capture information within the category, therefore creating more predictive features\n",
    "- Create a monotonic relationship between the variable and the target, therefore suitable for linear models\n",
    "- Do not expand the feature space\n",
    "\n",
    "\n",
    "### Limitations\n",
    "\n",
    "- Prone to cause over-fitting\n",
    "- Difficult to cross-validate with current libraries\n",
    "\n",
    "\n",
    "### Note\n",
    "\n",
    "The methods discussed in this and the coming 3 lectures can be also used on numerical variables, after discretisation. This creates a monotonic relationship between the numerical variable and the target, and therefore improves the performance of linear models. I will discuss this in more detail in the section \"Discretisation\".\n",
    "\n",
    "===============================================================================\n",
    "\n",
    "## Mean Encoding or Target Encoding\n",
    "\n",
    "Mean encoding implies replacing the category by the average target value for that category. For example, if we have the variable city, with categories London, Manchester and Bristol, and we want to predict the default rate, if the default rate for London is 30% we replace London by 0.3, if the default rate for Manchester is 20% we replace Manchester by 0.2 and so on.\n",
    "\n",
    "\n",
    "## In this demo:\n",
    "\n",
    "We will see how to perform one hot encoding with:\n",
    "- pandas\n",
    "- Feature-Engine\n",
    "\n",
    "And the advantages and limitations of each implementation using the Titanic dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# to split the datasets\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# for encoding with feature-engine\n",
    "from feature_engine.categorical_encoders import MeanCategoricalEncoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>survived</th>\n",
       "      <th>sex</th>\n",
       "      <th>cabin</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>B5</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>C22</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>female</td>\n",
       "      <td>C22</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>male</td>\n",
       "      <td>C22</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>female</td>\n",
       "      <td>C22</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   survived     sex cabin embarked\n",
       "0         1  female    B5        S\n",
       "1         1    male   C22        S\n",
       "2         0  female   C22        S\n",
       "3         0    male   C22        S\n",
       "4         0  female   C22        S"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load dataset\n",
    "\n",
    "data = pd.read_csv(\n",
    "    '../titanic.csv',\n",
    "    usecols=['cabin', 'sex', 'embarked', 'survived'])\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# let's fill na in embarked\n",
    "\n",
    "data['embarked'].fillna('Missing', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now we extract the first letter of the cabin\n",
    "# to create a simpler variable for the demo\n",
    "\n",
    "data['cabin'] = data['cabin'].astype(str).str[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "survived :  2  labels\n",
      "sex :  2  labels\n",
      "cabin :  9  labels\n",
      "embarked :  4  labels\n"
     ]
    }
   ],
   "source": [
    "# let's have a look at how many labels each variable has\n",
    "\n",
    "for col in data.columns:\n",
    "    print(col, ': ', len(data[col].unique()), ' labels')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['B', 'C', 'E', 'D', 'A', 'n', 'T', 'F', 'G'], dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's explore the unique categories\n",
    "\n",
    "data['cabin'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['female', 'male'], dtype=object)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['sex'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['S', 'C', 'Missing', 'Q'], dtype=object)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['embarked'].unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Encoding important\n",
    "\n",
    "We calculate the target mean per category using the train set, and then use those mappings in the test set.\n",
    "\n",
    "Note that for the pandas implementation we need to keep the target in the training set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((916, 4), (393, 4))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's separate into training and testing set\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    data[['cabin', 'sex', 'embarked', 'survived']],  # this time we keep the target!!\n",
    "    data['survived'],  # target\n",
    "    test_size=0.3,  # percentage of obs in test set\n",
    "    random_state=0)  # seed to ensure reproducibility\n",
    "\n",
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore original relationship between categorical variables and target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3Xd8VfX9+PHXOwlhbwJhrzAkQVZkuFGBRKtWa1VwW8S2olKr1tZfrdpvl6PqV2kduBWpu7RqwIEDlb1MmBEEAoSwCSP7/fvjnOR7CRkn5N6ce5P38/HII/fe8zmf877njvc9n3M+n4+oKsYYYwxAlN8BGGOMCR+WFIwxxpSxpGCMMaaMJQVjjDFlLCkYY4wpY0nBGGNMGUsKYU5EzhaRrFqs/7SI/D6YMVWwDRWRhEqWXSUic0+w3vtF5LXaRVf/iMhLIvI/VSw/JCJ96jKmSuL4XEQmh0EcZ4jIuiDUU6vPYqSwpFAHROQHETnqfliz3Q91ixBs53oRmR/4mKr+XFX/GOxteaWqr6vq+LrebkNOKKraQlU3+h1HVUSkjYi84H4eckVkvYj8JhTbUtWvVHVAKOqujywp1J0LVbUFMBQYBvzW53iM8dNjQAvgJKA1cBHw/YlUJCIxQYyrwbOkUMdUNRuYg5McABCRxiLyiIhsEZGdbpNP04rWF5F7ROR799fVahG5xH38JOBpYIx7RLLfffyYpgYRuUlEMkVkr4jMFpEuActURH4uIhtEZJ+ITBcRcZcliMgXInJARHaLyL/KhXZeJesdc/TibuM2Edno1vOwiFT1PmwiIv9yn+8yERkSUFcXEXlHRHaJyCYRuc19PAX4HXCFuy9WishYEfkuYN1PRGRRwP35IvLjqup1l0UFvAZ7RORNEWnnLuvlPr/r3Ndyt4jcW9kTE5GmIvKoiGx29+v80tddRN5yf0UfEJEvRSSx3OodRORjd798ISI9y+3jBPf2S+7r8YFbdqGI9K0ipkq3W11dIjJORNa66z4FSGXbAU4BZqrqPlUtUdW1qvp2uf1Y9mUvAU1R7nvqaxF5TET2An8Ukf0ikhRQPk6co/OOEtDs4752b5d7zk+IyP+6t28QkTXu89soIjdX8RzqJ1W1vxD/AT8A57m3uwHfAU8ELH8cmA20A1oC/wH+4i47G8gKKPtToAtOQr8COAx0dpddD8wvt+2XgP9xb58D7AaGA42BJ4EvA8oq8F+gDdAD2AWkuMveAO51t9sEON3jesfE5Jad5z7XHsB6YHIl++1+oBC4DGgE3Alscm9HAUuB+4BYoA+wEZgQsO5rAXU1AY4CHYAYIBvY7u7vpu6y9h7qnQYscF/HxsAzwBvusl7u83vOrXMIkA+cVMnzmw58DnQFooFTgcbushvd2BrjvD9WlHtNc4Ez3eVPVLCPEwLK7gVGus/7dWBWFe/V6rZbYV3ufj0Y8Fr9Ciiq4rWdAWQANwD9yi0r3Y8xAY99XloXznuqCLjVjaMp8ALwp4DytwBp5T9DQE/gCNDKvR8N7ABGu/cvAPriJLSz3LLDK/os1tc/3wNoCH84SeGQ+0FW4FOgjbtMcL7Y+waUHwNscm9X+UYEVgAXu7evp+qk8DzwUMCyFjhfur3c+8qxX/ZvAve4t18BngW6VRBDVesdE5NbNiXg/i+BTyt5bvcDCwLuR7kf4DOAUcCWcuV/C7wYsO5r5ZZ/BVwKjAbmunGmAGOBVW6Z6updA5wbsKyzuw9j+L8vs24ByxcBV1bw3KJwEtEQD++fNm69rQNe01kBy1sAxUD3gH0cmBRmBJQ9H1jr8X1b0XYrrAu4ttxrJUAWlSeFpjhHc0vd/ZcJpLrLSvdjVUmh/Gt0HrAx4P7XwLUVfYaA+QHLxgHfV7EP3gdu9/JZrC9/1nxUd36sqi1x3lgDcX5ZAcQBzYCl7iHwfiDNffw4InKtiKwIKJsUUFd1ugCbS++o6iFgD84v1VLZAbeP4HzhANyN80FfJCIZInJjuborW68iWwNub3bjqrasqpbgfNF0wfnF16V0P7j74ndApyrq+gJn/5/p3v4c59fgWe59PNTbE3gvYNkanC/kwO162RcdcI5ejmtHF5FoEfmr20R1EOdHRek6pQL3yyGcX/CV7UdPr43H7VZWV5dyMSnHvs7HUNWjqvpnVR2Bc4T2JvBWaVOcB+Xr/gxoKiKj3Ka0ocB7law7E5jo3p7k3gdARFJFZIE4zav7cRKf189XvWBJoY6p6hc4v7gecR/ajfOLMVFV27h/rdU5KX0M983+HDAVaK+qbYB0/q/ttrohb7fjfKmV1tcc5wO5zUPc2ap6k6p2AW4G/iGVXIbqQfeA2z3cuKotK865h25u+a04R1NtAv5aqur5pSFXUFf5pPAFxyeF6urdivOLNnB5E1Wtdh+WsxvIw2mqKG8ScDHOr9/WOL+c4dg2+sD90gKnOa6q/eiFl+1WZke5mIRjX+dKqepB4M9Ac6A3zpEzOD+WSsWXX61cHSU4iWUizvP4r6rmVrLJt4CzRaQbcAluUhCRxsA7OJ/NTu7n60O8Pf96w5KCPx4HxonIUPfN/BzwmIh0BBCRriIyoYL1muN8GHa55W7AOVIotRPoJiKxlWx3JnCDiAx1PwB/Bhaq6g/VBSwiP3U/RAD73DiKq1uvEneJSFsR6Q7cDpQ/aR1ohIhc6p50nIbTRr8Ap1nmoIj8xj1hGy0iSSJyirveTqCXHHsS+xtgAE6b+CJVzcBJkqOAL90y1dX7NPCn0hO77gnNi2u6A9zX/QXg7+Kc2I4WkTHu69LSfZ57cL4Y/1xBFeeLyOnua/1HnNex0l/mHnnZbmU+ABIDXqvbOP6LvIyI/F5EThGRWBFpgvM+2A+sU9VdOD9Urnb3y41UnDzLm4lznu0qAn79l+fW/znwIs4PgDXuoliccym7gCIRSQXq/HJqv1lS8IH7pnwFKO1U9hucNtUF7mH7JzhfXuXXWw08CnyL86U3GKfttNRnOCfvskVkdwXrf+pu8x2cX3Z9gSs9hn0KsFBEDuGcFL9dVTd5XLe8f+O0Ja/A+TJ5vpqyV+AkomuAS1W1UFWLgQtxmgk24fzynoHzCxecX4MAe0RkGYCqHgaWARmqWuAu/xbYrKo5bpnq6n3Cff5zRSQXJ0GNOrHdwJ04Fx0sxmn++RvOZ/IVnGa1bcBqdxvlzQT+4K43AueLsLa8bLdCqrob5yKIv+IklX4c+948bhWcL+XdOEc444AL3KYwgJuAu9y6EnESenUxLMQ5yugCfFRN8Zk4R0RlycM9srgN54hjH84Rx+zqtlvfiHsCxZg6ISKKc7VJpt+xGGOOZ0cKxhhjyoQ0KYhIioisE6ez1D0VLO8hIvNEZLmIrBKR8yuqxxhjTN0IWfORiETjdEwah3MZ4WJgotsuXlrmWWC5qv5TRAYBH6pqr5AEZIwxplqhPFIYCWSq6kb3pN4snMvdAinQyr3dmtpfUmeMMaYWQjmQVFeO7WCSxfFXadyPcxXHrTiXW55XUUUiMgWYAtC8efMRAwcODHqwxhhTny1dunS3qlbYKTZQKJNCRR0+yrdVTQReUtVHRWQM8KqIJLnXcP/fSqrP4gyxQHJysi5ZsiQkARtjTH0lIpurLxXa5qMsju3RWNoTNdDPcK4JRlW/xen236C6lBtjTDgJZVJYDPQTkd5ur8srOb4jyBbgXCgb+rkJbm9dY4wxdS9kSUFVi3DG6JmDM2jYm6qaISIPishFbrFfAzeJyEqcoZmvV+tNZ4wxvgnpjEWq+iHOgFKBj90XcHs1cFooYzDGGOOd9Wg2xhhTxpKCMcaYMpYUjDHGlLGkEALzN+zmu6wDfodhjDE1ZkkhyNK3HeD6Fxcx8bkFZOZUNvGTMcaEJ0sKQZRXWMyv/rWCds1jadIoipteWcqBo4V+h2WMMZ5ZUgiih+esY0POIR7+6RD+efUIsvYd4bY3llNcYl0vjDGRwZJCkHyduZvn52/i2jE9Oat/HKf0ascDFyXxxfpdPDRnrd/hGWOMJyHtvNZQHDhayJ1vraRPh+b8NvWksscnjerB6h0HeOaLjQzq3IqLh3b1MUpjjKmeHSkEwf2zM8jJzeexK4bSNDb6mGX3/SiRkb3bcffbq+yKJGNM2LOkUEsfrNrBe8u3ces5CQzp3ua45bExUfzjquF0aNGYKa8uYVduvg9RGmOMN5YUamHnwTzuff87hnRrzS1jEyot16FFY565ZgT7jhTwi9eWUlBUUmlZY4zxkyWFE6Sq3PX2KvIKi/n7FUNpFF31rkzq2pqHLxvCks37+MPsdGwwWGNMOLKkcIJeW7CZL9fv4t7zT6JvXAtP61w4pAu/PLsvbyzaymsLt4Q4QmOMqTm7+ugEbNx1iD99uIYz+8dx9eieNVr31+MHsDY7lwdmZ9CvYwtG92kfoijrjyU/7GXv4QK/w6hWbEwUyb3a0aKxfaxM5LJ3bw0VFZfwqzdX0jgmmocvOxmRiqairlx0lPD4lUP58fSv+eXry5g99TS6tW0Womgjm6ryyNx1TJ/3vd+heBYbE8WZ/TowITGecYM60aZZrN8hGVMjlhRqaPq871m5dT9PTRpGp1ZNTqiOVk0a8dy1yfx4+tdMeWUpb/9iDM1i7aUIpKr88b9reOHrTUwc2Z2rRtXsiMwPB44W8umaHNLSd/DJmhxiooQxfdszITGe8Ymd6NjyxN4vxtQlCeUJTxFJAZ4AooEZqvrXcssfA8a6d5sBHVX1+Os6AyQnJ+uSJUtCEW61Vm7dz6X//IYLT+7M41cOq3V989blcONLizl/cGeemjisxkcd9VVJifL//p3OzIVbuP7UXvzhwkERtW9UlVVZB0jLyCYtPZtNuw8jAsk925KS1JmUpHi6tmnqd5imgRGRpaqaXG25UCUFEYkG1gPjgCxgMTDRnYKzovK3AsNU9caq6vUrKRwtKOaC//2Ko4XFpE07k9ZNGwWl3qe/+J6/frSWuyYMqPKy1oaiqLiEu99ZxbvLtvGLs/ty94QBEZUQylNV1u3MJS3dSRBrs52Rc0/u1pqUpHhSkzrTu0Nzn6M0DYHXpBDKNouRQKaqbnQDmgVcDFSYFICJwB9CGE+t/PWjNWzcfZjXJ48KWkIAuPnMPqzefpBH5q5jYHxLzj2pU9DqjjSFxSVM+9cKPli1gzvG9efWcxIiOiEAiAgD41sxML4V087rz6bdh90EsYOH0tbxUNo6BnRq6SSIwfEM6NQy4p+ziWyhPFK4DEhR1cnu/WuAUao6tYKyPYEFQDdVLa5g+RRgCkCPHj1GbN68OSQxV+aL9bu47oVF3Hhab+67cFDQ6z9aUMxPn/mGH3Yf4f1bTiWhY8ugbyPc5RcVM3Xmcj5evZPfnT+QKWf29TukkNu2/yhz3COIxZv3ogq92jcjJakzqUnxnNyttSUIEzTh0Hz0U2BCuaQwUlVvraDsb3ASwnHLyqvr5qP9RwoY/9iXtG7aiP/cejpNGkVXv9IJ2Lb/KBc/NZ+WTRrx/i2nBfVoJNwdLSjm5teW8uX6XTx4cSLXjunld0h1blduPnNXOwni2+/3UFSidGndhAluE9OInm2JjrIEYU5cODQfZQHdA+53A7ZXUvZK4JYQxnJCVJV7309n7+ECXrj+lJAlBICubZryz6tHMOm5Bdz2xnJeuP6UBvElcDi/iJ+9vJiFm/byt58M5opTevgdki/iWjbmqlE9uWpUT/YfKeAT9yqm1xdu4cWvf6BDi8aMT+xEalI8o/u0r7YHvTEnKpRHCjE4J5rPBbbhnGiepKoZ5coNAOYAvdVDMHV5pPDvFdu4fdaKOj0JPHPhFn733nfcfFafY4bhro8O5hVy/QuLWJl1gL9fPsSGFq/Aofwi5q3NIS09m3nrcjhSUEzrpo047yQnQZzer0NIf6yY+sP3IwVVLRKRqThf+NHAC6qaISIPAktUdbZbdCIwy0tCqEvb9x/l9++nM7xHG24+s0+dbbehzMGw73AB176wiLXZB3lq4jBSB3f2O6Sw1KJxDBcO6cKFQ7qQV1jMl+t3kZaezdzV2byzLIvmsdGMHdiR1KTOnD0gjubWm9rUUkj7KYRCXRwplJQo17ywkOVb9vPhbWfQq44vGSwoKuHq5xeycut+3v75qQzu1rpOtx9quw/lc/WMhWzcfZinrx7OOQMb7hVXJ6qgqIRvN+4hLX0HczN2sudwAY1jojizfxypSfGcO7ATrZs1nPNSpnq+n2gOlbpICi/M38SD/13NXy4dzMSR/rRx7z6Uz8VPfU2JKrOnnk5cy8a+xBFs2QfyuGrGArbtP8qMa0/h9H4d/A4p4hWXKIt/2FvWFyL7YB4xUcKpCR1IcXtTd2hRP94/5sRZUjhBG3bm8qMn53N6QgdmXJfs6yWB6dsOcNnT35DUpTUzbxpNbExkn1zM2neESc8tZM+hfF68YSQje7fzO6R6p6REWZm1n7T0bD5Kz2bL3iNECZzSqx0pSfGkJMXTubX1pm6ILCmcgIKiEi7959ds359H2rQzwmKsmv+s3M6tbyxn4sju/PmSwRF73foPuw9z1YyF5OYV8vKNIxnWo63fIdV7qsqaHbmkpe8gLSOb9TsPATCqdzuemjS83hx9Gm98P9EciZ78bAPp2w7y9NUjwiIhgDMHw5odB/nH598zqEtrrqnhUN3hIDMnl0nPLaSwuISZN40mqWv9OkcSrkSEQV1aMahLK+4YP4Dvdx0iLT2bpz7L5OoZC3ljymjaNbdRXM2xIrs9IoiWbt7H9HmZXDaiGylJ8X6Hc4xfjx/AOQM78sDsDBZs3ON3ODWyevtBrnhmASUK/7p5jCUEH/WNa8EtYxN4/rpkfthzmGueX8iBI4V+h2XCjCUFnA5Ud7y5gs6tm/KHEAxjUVulczD0aN+MX76+jKx9R/wOyZOVW/cz8bkFxMZE8ebNo+nfqeEN3xGOTk3owDPXjGDDzkNc9+IicvMsMZj/Y0kB+NOHa9iy9wh/v3wILZuE52V8pXMwFBaXMOWVpRwpKPI7pCot+WEvV89YSKumMbx58xj6eJyy1NSNswd05KlJw0jfdoAbX1oc9u8nU3cafFL4bO1OZi7cwpQz+jAqzKfG7BvXgv+dOIw12Qe56+1VhOtFAt9k7uaa5xcR17Ixb948hu7tbGa5cDQ+MZ4nrhzG0s37mPzyEvIKjxuL0jRADTop7DmUz91vf8fA+JbcMb6/3+F4MnZAR36TMpAPVu3gH5+H3zSV89blcMNLi+nerimzbh5tlz+GuQtO7syjlw/h2417uPnVpeQXWWJo6BpsUlBVfvfedxw8WshjVwylcUzkjB9z85l9uGhIFx6Zu45P1+z0O5wyczKymfLKEhI6tmDWlDFhcwWXqdolw7rxl0sG88X6XUyduZzC4hK/QzI+arBJ4Z1l25iTsZNfj+/PSZ1b+R1OjYgIf/vJySR2acXts1aQmZPrd0j8Z+V2fvn6MpK6Oh3t7FLHyHLlyB48eHEiH6/eybRZKyiyxNBgNciksHXvEe6fncHIXu2YfEbdDXYXTE1jo3nmmmSaNIripleWcuCof1eQvLVkK7fPWs6Inm159WfBnZnO1J1rx/Ti3vNP4oPvdnDX26soLgnPc1YmtBpcUiguUX791koAHr18SETPWVA6B0PWviPc9sZyXz7Ery7YzF1vr+K0hA68fMNIWtgonRHtpjP7cOf4/ry3fBv3vvcdJZYYGpwGlxSen7+RRZv28ocLB9WLq2JO6dWOBy5K4ov1u3hozto63faMrzby+/fTOXdgR567NpmmsZFzXsZUbuo5/bj1nARmLd7KA//JCNur3ExoNKifdWuzD/LInPWMH9SJy0Z08zucoPFjDoanPtvAI3PXc/7geB6/YljED9ZnjnXHuP7kF5Xw7JcbiY2J4nfnnxSx426ZmmkwSSG/qJhps1bQqmkMf7k0cgeWq8x9P0pk/c5D3P32Kvp0aBGyORhUlUfnruepeZlcMqwrD192MjE2NWS9IyL8NnUg+YXFPPfVJpo0iubX4wf4HZapAw3m0/zYxxtYm53L335yMu3r4djysTFR/OOq4XRo0Zgpry5hV25+0LehqvzpgzU8NS+TiSO78+hPh1hCqMdEhD9cmMiVp3Tnyc8yefLTDX6HZOpASD/RIpIiIutEJFNE7qmkzOUislpEMkRkZijiWLRpL898+T0TR3bn3JPq7yxfHVo05plrRrDvSAG/eG0pBUXBu6ywpET5/b/TmTF/E9ef2os/XzKYqAg+SW+8iYoS/nTJYC4d1pVHP17Pc19u9DskE2IhSwoiEg1MB1KBQcBEERlUrkw/4LfAaaqaCEwLdhy5eYXc8eYKurdtxv+7IPwGuwu2pK6tefiyISzZvI8/zE4PyknC4hLl7ndW8dqCLdx8Vh/+cOGgetf8ZioXHSU8dNnJXHByZ/704Rpe+fYHv0MyIRTKcwojgUxV3QggIrOAi4HVAWVuAqar6j4AVc0JdhB//O9qtu8/yls/H9NgJjUP5hwMhcUl3PHmSv6zcjvTzuvH7ef2s4TQAMVER/H4FUMpKCrhvn9nEBsdxZU+TVVrQiuUzUddga0B97PcxwL1B/qLyNciskBEUiqqSESmiMgSEVmya9cuzwHMycjmzSVZ/OLsvozo2bCmfgzGHAz5RcVMnbmM/6zczj2pA5l2Xn9LCA1Yo+gonpo0jLP6x/Hb977j3WVZfodkQiCUSaGib4/ybRkxQD/gbGAiMENE2hy3kuqzqpqsqslxcXGeNr4rN5/fvvudMxTEuZEx2F0w1XYOhrzCYm5+dSlzMnZy/4WD+PlZfUMUqYkkjWOieeaaEYzp054731rJf1dt9zskE2ShTApZQPeA+92A8u+gLODfqlqoqpuAdThJolZUld++u4pD+UU8fsXQBnsN/YnOwXCkoIgbX1rMF+t38ZdLB3P9ab1DHKmJJE0aRTPjumRG9GzLtFkrmJuR7XdIJohC+W25GOgnIr1FJBa4Ephdrsz7wFgAEemA05xU68sb/rV4K5+syeE3KQPp18Bn+6rpHAy5eYVc+/wiFmzcw98vH8JEazc2FWgWG8ML159CUtfW3DJzGfPWBf10oPFJyJKCqhYBU4E5wBrgTVXNEJEHReQit9gcYI+IrAbmAXepaq0mId685zAP/nc1p/Ztzw2n9qpNVfWG1zkY9h8p4OoZC1mxdT9PTRrOJcPqT69vE3wtmzTi5RtH0r9TS37+6lK+ztztd0gmCCTSxjVJTk7WJUuWVLisuES5/JlvWb8zl7RpZ9K1jU3wUkpVuX3WCv6zajszrk0+rr/G7kP5XD1jIRt3HeafVw+v1/05THDtPVzAxGcXsGXvEV6+cSQjezesizoihYgsVdXk6srVq8b2p7/4nqWb9/HHi5MsIZRT1RwMOw/mceWzC/hhz2Gev/74hGFMVdo1j+W1yaPo3KYJN760mOVb9vkdkqmFepMU0rcd4PFP1nPB4M5cPLSL3+GEpYrmYNi2/yhXPPMtO/Yf5aUbRnJGP29XdxkTKK5lY2ZOHk37FrFc+8Ii0rcd8Dskc4LqRVLIKyzmV/9aQdtmsfzPj5PsWvoqBM7B8PNXl3L509+y53ABr04exeg+7f0Oz0Sw+NZNeH3yKFo1acQ1zy9kbfZBv0MyJ6BeJIVH5qxjQ84hHrrsZNraNJDVKp2D4duNezhSUMQbN41meI+2fodl6oFubZsx86ZRxMZEcfWMhWTmHPI7JFNDEZ8UvsnczYz5m7hmdE/OHtDR73AixqRRPXhy4jDe+cWpJHUNzTDbpmHq2b45M28aDQhXzVjA5j2H/Q7J1EBEJ4UDRwu5862V9OnQnN+eP9DvcCLOhUO60Ceuhd9hmHqob1wLXp88ioKiEiY9t7DGPeqNfyI6KTwwO4Odufn8/YqhNIttGIPdGRMpBsS35NWfjSI3r5BJzy0k+0Ce3yEZDyI2KXywagfvLt/G1LEJDO1+3HBJxpgwkNS1NS/fOJK9hwuY9NwCcnItMYS7iEwKOQfzuPf97xjSrTVTz0nwOxxjTBWG9WjLizecwo4DeVw9YyF7Dxf4HZKpQkQmhbveXkVeYTF/v2IojWw6SGPC3im92vH8dcls3nOEa55fyIEjhX6HZCoRcd+oew4X8MX6Xfzu/JPoaydJjYkYpyZ04JlrRrBh5yGufXERuXmWGMJRxCWF7AN5nNGvQ61mEzPG+OPsAR15atIwMrYd4IYXF3M439tw7qbuRFxSEODhy4ZYr2VjItT4xHieuHIYy7bsY/LLS8grLPY7JBMg4pJCj/bNiG/dxO8wjDG1cMHJnXn08iEs2LSHm19dSn6RJYZwEXFJoUVj649gTH1wybBu/PXSwXyxfhe3vL6cwuISv0MyRGBSMMbUH1ec0oMHL07kkzU7mTZrBUWWGHxnP7uNMb66dkwvCopK+J8P1hAbE8UjPx1CdJSdM/RLSI8URCRFRNaJSKaI3FPB8utFZJeIrHD/JocyHmNMeJp8Rh/umjCA95Zv4973vqOkJLJmhKxPKj1SEJHvgIpeGQFUVU+uqmIRiQamA+OALGCxiMxW1dXliv5LVafWLGxjTH1zy9gE8guL+d/PMomNieKBixLtKkMfVNV89KNa1j0SyFTVjQAiMgu4GCifFIwxBoBfjetPXlEJz365kWaxMdyTaqMf17VKk4Kqbq5l3V2BrQH3s4BRFZT7iYicCawHfqWqW8sXEJEpwBSAHj161DIsY0y4EhF+mzqQ3Yfyee6rjUw5sw/tbOKsOlXtOQURGS0ii0XkkIgUiEixiHiZZ6+i477yzVH/AXq5TVGfAC9XVJGqPquqyaqaHBdncwgbU5+JCDee1pviEuXj1dl+h9PgeDnR/BQwEdgANAUmA096WC8L6B5wvxuwPbCAqu5R1Xz37nPACA/1GmPqucQurejWtilp6ZYU6pqnq49UNROIVtViVX0RGOthtcVAPxHpLSKxwJXA7MACItI54O5FwBpvYRtj6jMRITUpnvmZuzloA+fVKS9J4Yj7pb5CRB4SkV8BzatbSVWLgKnAHJwv+zdVNUNEHhSRi9xit4lIhoisBG4Drj+hZ2GMqXdSkuIpLFabOI5NAAAa30lEQVTmrc3xO5QGRVSrvh5YRHoCO4FY4FdAa+Af7tFDnUtOTtYlS5b4sWljTB0qKVFG/+VThvdoy9PXWMtybYnIUlVNrq6clx7Nw4EPVfUg8ECtIzPGGA+iooQJifG8tXQrRwuKaRob7XdIDYKX5qOLgPUi8qqIXCAiNjSGMaZOpCbFk1dYwhfrrQmprlSbFFT1BiABeAuYBHwvIjNCHZgxxozs3Y62zRrxkV2FVGe8Xn1UCHwEzAKW4vRMNsaYkIqJjmLcoE58tibH5lyoI146r6WIyEtAJnAZMAPoXOVKxhgTJKlJncnNL+KbzD1+h9IgeDlSuB54H+ivqtep6ofu5abGGBNypya0p2XjGOvIVke8nFO4UlXfD+h5bIwxdaZxTDTnnNSRuauzbRKeOlBpUhCR+e7/XBE5GPCX63HsI2OMCYqUxHj2HSlk0aa9fodS71WaFFT1dPd/S1VtFfDXUlVb1V2IxpiG7qwBcTRpFEVahjUhhZqXE81PiMiYugjGGGMq0iw2hrP7dyQtPdtmZQsxLyealwG/d6fUfFhEqu0mbYwxwZaSFE9Obj7Lt+73O5R6zcuJ5pdV9XycmdTWA38TkQ0hj8wYYwKcc1JHGkULaek7/A6lXvPUec2VAAwEegFrQxKNMcZUolWTRpyW0IG0jGyqG8jTnDgv5xRKjwweBNKBEap6YcgjM8aYclKT4tm69ygZ2+0CyFCpMimIiACHgDGqmqKqL6qqNegZY3wxblA8UYJ1ZAuhKpOCOsdoP1bV3XUUjzHGVKpd81hG9W5vl6aGkJdzCgtE5JSQR2KMMR6kDo4nM+cQmTm5fodSL3lJCmOBb0XkexFZJSLficgqL5W7g+mtcy9nvaeKcpeJiNrlrsaY6owfFA9YE1KoeJkwJ/VEKhaRaGA6MA7IAhaLyGxVXV2uXEuc+ZkXnsh2jDENS3zrJgzv0YaP0rOZek4/v8Opd7wcKWglf9UZCWSq6kZVLcCZi6GieRj+CDwE5HmK2BjT4KUkxZOx/SBb9hzxO5R6x0tS+AD4r/v/U2AjzoQ71ekKbA24n+U+VkZEhgHdVfW/VVUkIlNEZImILNm1a5eHTRtj6rPUJGdKlzl2wjnovPRoHqyqJ7v/++EcAcz3ULdUVF3ZQpEo4DHg1x5ieFZVk1U1OS4uzsOmjTH1Wfd2zUjs0oqPrHdz0NWkRzMAqroM8HI1UhbQPeB+N2B7wP2WQBLwuYj8AIwGZtvJZmOMFymJ8Szbsp+dB63lOZi89Gi+I+DvThGZCXhpw1kM9BOR3iISC1wJzC5dqKoHVLWDqvZS1V7AAuAiVV1yYk/FGNOQpA52rkKyJqTg8nKk0DLgrzHOuYWKThgfw52ycyowB1gDvKmqGSLyoIhcdOIhG2MMJHRsSd+45nZpapBVe0mqqj5QeltE2gL71eNoVKr6IfBhucfuq6Ts2V7qNMaYUqlJnfnnF9+z93AB7ZrH+h1OvVDVdJz3ichA93ZjEfkM+B7YKSLn1VWAxhhTmZSkeIpLlI9X29FCsFTVfHQFsM69fZ1bNg44C/hziOMyxphqJXZpRbe2Ta0JKYiqSgoFAc1EE4A3VLVYVdfgrSe0McaElIiQmhTP/MzdHMwr9DuceqGqpJAvIkkiEocz/tHcgGXNQhuWMcZ4k5IUT2GxMm9tjt+h1AtVJYXbgbdxZll7TFU3AYjI+cDyOojNGGOqNax7Wzq1asxH31kTUjBU2gykqgtxpt8s//hxVxQZY4xfoqKECYnxvLlkK0cLimkaG+13SBGtxj2ajTEm3KQkxpNXWMIX660JqbYsKRhjIt7I3u1o26wRH9lVSLVmScEYE/FioqMYN6gTn63JIb+o2O9wIpqnpCAip4rIJBG5tvQv1IEZY0xNpCZ1Jje/iG8y9/gdSkSrtr+BiLwK9AVWAKUpWIFXQhiXMcbUyKkJ7WnZOIa09GzGDuzodzgRy0sntGRgkNfxjowxxg+NY6I556SOzF2dzZ+Kk4iJttbxE+Flr6UD8aEOxBhjais1KZ59RwpZtGmv36FELC9HCh2A1SKyCMgvfVBVbfhrY0xYObN/HE0aRZGWkc2pCR38DicieUkK94c6CGOMCYZmsTGc3b8jaenZ3H9hIlFRFc0KbKriZT6FL+oiEGOMCYaUpHjSMrJZvnU/I3q29TuciONlOs7RIrJYRA6JSIGIFIvIwboIzhhjauqckzrSKFpIS9/hdygRycuJ5qeAicAGoCkw2X2sWiKSIiLrRCRTRO6pYPnPReQ7EVkhIvNFZFBNgjfGmPJaNWnE6QkdSMvIxi6adBwpKPJc1tM1W6qaCUS78ym8CJxd3ToiEg1MB1KBQcDECr70Z6rqYFUdCjwE/N1z5MYYU4mUpHi27j1KxnZr1CgpUS5/5lvP5b0khSMiEgusEJGHRORXQHMP640EMlV1o6oWALOAiwMLqGrgK9Ycp1OcMcbUyrhB8UQJNiMb8FF6NunbvCdHL0nhGrfcVOAw0B34iYf1ugJbA+5nuY8dQ0RuEZHvcY4UbquoIhGZIiJLRGTJrl27PGzaGNOQtWsey6je7UnLaNhJoai4hEfnrqN/pxae16k2KajqZkCAzqr6gKre4TYnVaeia8GOOxJQ1emq2hf4DfD/KonhWVVNVtXkuLg4D5s2xjR0qYPjycw5RGZOrt+h+ObtpVls3H2YO8cP8LyOl6uPLsQZ9yjNvT9URGZ7qDsL56iiVDdgexXlZwE/9lCvMcZUa0KiMxBDQ21Cyiss5vFPNjCsRxvGDerkeT0vzUf345wf2A+gqiuAXh7WWwz0E5He7jmJK4FjkomI9Au4ewHOFU7GGFNrnVo1YXiPNg12joVXv91M9sE87p4wEBHvnfi8JIUiVT1Q04BUtQjnPMQcYA3wpqpmiMiDIlI6RMZUEckQkRXAHcB1Nd2OMcZUJjWpMxnbD7JlzxG/Q6lTB/MKmf55Jmf068CYvu1rtK6nAfFEZBIQLSL9RORJ4Bsvlavqh6raX1X7quqf3MfuU9XZ7u3bVTVRVYeq6lhVzahR9MYYU4WUJKcJaU4DO+E848uN7D9SyN0TBtZ4XS9J4VYgEWcwvDeAg8C0Gm/JGGPqWPd2zUjs0oqPGlDv5l25+cyYv4kLBndmcLfWNV7fy9VHR1T1XlU9xb0C6F5VzTuhaI0xpo6lJMazbMt+dh5sGF9b0+dlkl9Uwh3j+5/Q+pUOiFfdFUY2dLYxJhKkDo7n0Y/XMycjm2vH9PI7nJDauvcIry/czE9HdKNvnPe+CYGqGiV1DE7nszeAhVTc78AYY8JaQseWJHRsQVp6/U8Kj3+yARHh9vP6VV+4ElU1H8UDvwOSgCeAccBuVf3ChtM2xkSSlMR4Fm7ay97DBX6HEjLrd+by3vIsrhvTk86tm55wPZUmBXfwuzRVvQ4YDWQCn4vIrSe8NWOM8UFKUjzFJcrHq+vvVUiPzFlH89gYfnl2Qq3qqfJEs4g0FpFLgdeAW4D/Bd6t1RaNMaaOJXZpRbe2Tett7+blW/Yxd/VObjqzD22bx9aqrqpONL+M03T0EfCAqqbXakvGGOMTESE1KZ6XvvmBg3mFtGrSyO+QgkZVeShtHe2bx/Kz03vXur6qjhSuAfoDtwPfiMhB9y/XZl4zxkSalKTOFBYr89bm+B1KUM3P3M23G/cw9ZwEmjeudoblalV1TiFKVVu6f60C/lqqaqtab9kYY+rQsO5t6NSqMR99V3+akEqPErq2acqkUT2CUqenmdeMMSbSRUUJExLj+Xx9To2mpwxnH6Vn8922A/xqXH8ax0QHpU5LCsaYBiMlMZ68whK+XB/5k3UVFZfwyNx19OvYgkuGHTd/2QmzpGCMaTBG9m5H22aN6sVw2u8sy2LjrsPcOWEA0VHB61tsScEY02DEREcxflA8n63JIb+o2O9wTljpBDpDu7dhfA0m0PHCkoIxpkFJSYonN7+IbzL3+B3KCXttwWZ2HMjj7pQBNZpAxwtLCsaYBuXUhPa0bBwTsR3ZcvMKmT7PmUDn1L4dgl6/JQVjTIPSOCaac07qyNzV2RQVl/gdTo0999Um9h0p5K4JA0JSf0iTgoikiMg6EckUkXsqWH6HiKwWkVUi8qmI9AxlPMYYA5CaFM++I4Us2rTX71BqZPehfJ7/aiPnD47n5G5tQrKNkCUFEYkGpgOpwCBgoogMKldsOZCsqicDbwMPhSoeY4wpdWb/OJo0iiItwqbpnD4vk7yiEn49PjRHCRDaI4WRQKaqblTVAmAWcHFgAVWdp6qlM2ovALqFMB5jjAGgWWwMZ/fvSFp6NiUl6nc4nmTtO8LrC7Zw2fATn0DHi1Amha44k/SUynIfq8zPcAbfO46ITBGRJSKyZNeuyO90YozxX+rgeHJy81m+db/foXjy+CcbQKjVBDpehDIpVHSdVIUpWUSuBpKBhytarqrPuvNDJ8fFxQUxRGNMQzV2YEcaRQtp6Tv8DqVaG3bm8u6yLK4d3ZMubU58Ah0vQpkUsoDuAfe7AdvLFxKR84B7gYtUNT+E8RhjTJlWTRpxekIH0jKyUQ3vJqRH5q6jWWwMvxxbuwl0vAhlUlgM9BOR3iISC1wJzA4sICLDgGdwEkL9Gs/WGBP2UpLi2br3KBnbw3c2gBVb9zMnYyc3ndGHdrWcQMeLkCUFVS0CpgJzgDXAm6qaISIPishFbrGHgRbAWyKyQkRmV1KdMcYE3bhB8URHSVh3ZHsoba0zgc4ZtZ9Ax4vaz8hQBVX9EPiw3GP3Bdw+L5TbN8aYqrRrHsuo3u1Iy8jmzhB1BquN+Rt28833e7jvR4NoEYQJdLywHs3GmAYtJSmezJxDZObk+h3KMVSVh+aspWubplw1OjgT6HhhScEY06BNSIwHCLsmpLT0bFZlHWDaef2CNoGOF5YUjDENWqdWTRjeo01YzbFQVFzCw3PXkdCxBZcOr9s+vZYUjDENXmpSZzK2H2TLniPVF64D7y7b5kygMz64E+h4YUnBGNPgpSQ5TUhzwmAspLzCYh77ZD1DurdhQmJwJ9DxwpKCMabB696uGYldWvFRGPRuLp1A5zcTgj+BjheWFIwxBmc47WVb9rPzYJ5vMeTmFfKPz7/n9IQOnJoQ/Al0vLCkYIwxhEcT0oyvNrH3cEHIJtDxwpKCMcYACR1bktCxhW+Xpu45lM+MrzaSmhTPkO6hmUDHC0sKxhjjSkmMZ+Gmvew9XFDn254+73uOFhaHdAIdLywpGGOMKyUpnuIS5ePVdXu0sG3/UV5bsJnLRnQjoWPoJtDxwpKCMca4Eru0onu7pnXehPT4x+vdCXT61+l2K2JJwRhjXCJCSmI88zN3czCvsE62mZmTyzvLsrhmdE+6hngCHS8sKRhjTICUpM4UFivz1tbNFC+PzFnvTKBzdt862V51LCkYY0yAYd3b0KlVYz76LvRNSCu37ictI5vJZ/SmfYvGId+eF5YUjDEmQFSUMCExns/X53CkoCik23pozlraNY9l8hl9QrqdmrCkYIwx5aQkxZNXWMKX63eFbBvzN+zm68w93DI2oc4m0PEipElBRFJEZJ2IZIrIPRUsP1NElolIkYhcFspYjDHGq5G92tG2WaOQDaetqjw8Zy1dWjfhqlF1N4GOFyFLCiISDUwHUoFBwEQRGVSu2BbgemBmqOIwxpiaiomOYvygeD5bk0N+UXHQ65+Tkc3KrANMG9efJo3qbgIdL0J5pDASyFTVjapaAMwCLg4soKo/qOoqoCSEcRhjTI2lJMWTm1/EN5l7glpvUXEJj8xdT9+45lw6rGtQ6w6GUCaFrsDWgPtZ7mM1JiJTRGSJiCzZtSt0bXzGGFPq1IT2tGwcE/SObO8u30ZmziHumjCAmOjwO60byogqGghcT6QiVX1WVZNVNTkuLq6WYRljTPUax0Rz7kkdmbs6m6Li4DRm5BUW88QnGxjSrXXZ3NDhJpRJIQvoHnC/G7A9hNszxpigSkmKZ9+RQhZt2huU+l5fuIVt+49yd8pAXybQ8SKUSWEx0E9EeotILHAlMDuE2zPGmKA6q39HmjSKIi0Icywcyi9i+rxMTktoz2k+TaDjRciSgqoWAVOBOcAa4E1VzRCRB0XkIgAROUVEsoCfAs+ISEao4jHGmJpqGhvN2f07kpaeTUnJCbV+l5nx1Ub2Hi7g7gkDgxRdaIS0x4Sqfgh8WO6x+wJuL8ZpVjLGmLCUOjietIxslm/dz4iebU+oDmcCnU2kJPo7gY4X4Xfq2xhjwsjYgR1pFC2kpe844Tr+8fn3HCko4s4J/g+NXR1LCsYYU4VWTRpxekIHPkrPRrXmTUjb9h/l1QWb+cnwbiR0bBmCCIPLkoIxxlQjNakzWfuOkrH9YI3XfeKT9aAwbVz4HyWAJQVjjKnWeYM6ER0lNe7IlplziLeXZnF1mEyg44UlBWOMqUa75rGM6t2uxpemPjp3HU0bRXPL2PCYQMcLSwrGGONBSlI8mTmHyMzJ9VR+VdZ+PkrPZvIZfcJmAh0vLCkYY4wHpcNSeG1CenjOOto2a8TkM3qHMqygs6RgjDEedGrVhBE923qaY+GbzN18tWE3t4xNoGWTRnUQXfBYUjDGGI9SEuPJ2H6QLXuOVFpGVfnbnHV0ad2Eq0f3rMPogsOSgjHGeJSS5DQhzanihPOcjJ2s3LqfaeeF3wQ6XlhSMMYYj7q3a0Zil1Z8VEnv5uIS5ZG565wJdIaH3wQ6XlhSMMaYGkhNimfZlv3sPJh33LJ3l2WRmXOIO8eH5wQ6XkRm1MYY45OUpM7A8U1I+UXFPP7JBk7u1rqsmSkSWVIwxpgaSOjYgoSOLY67NPX1Be4EOhPCdwIdLywpGGNMDaUmxbNw0172Hi4A/m8CnVP7tuf0fuE7gY4XlhSMMaaGJiTGU1yifLzaOVp4/qtN7DlcwN0p4T2BjheWFIwxpoYSu7Sie7umpKVns/dwAc99tZEJiZ0YGuYT6HgR0qQgIikisk5EMkXkngqWNxaRf7nLF4pIr1DGY4wxwSAipCTGMz9zNw+lrXUm0Bk/wO+wgiJkSUFEooHpQCowCJgoIoPKFfsZsE9VE4DHgL+FKh5jjAmmlKTOFBYrsxZv5dLh3ejXKfwn0PEilEcKI4FMVd2oqgXALODicmUuBl52b78NnCuRfNreGNNgDOvehk6tGhMbHcW08/r5HU7QxISw7q7A1oD7WcCoysqoapGIHADaA7sDC4nIFGAKQI8ePUIVrzHGeBYVJdz3o0SOFhbTrW0zv8MJmlAmhYp+8Zef4NRLGVT1WeBZgOTk5JpPkmqMMSFwwcmd/Q4h6ELZfJQFdA+43w3YXlkZEYkBWgN7QxiTMcaYKoQyKSwG+olIbxGJBa4EZpcrMxu4zr19GfCZqtqRgDHG+CRkzUfuOYKpwBwgGnhBVTNE5EFgiarOBp4HXhWRTJwjhCtDFY8xxpjqhfKcAqr6IfBhucfuC7idB/w0lDEYY4zxzno0G2OMKWNJwRhjTBlLCsYYY8pYUjDGGFNGIu0KUBHJBdb5HYcHHSjXMztMWZzBEwkxgsUZbJES5wBVrXaAppBefRQi61Q12e8gqiMiSyzO4ImEOCMhRrA4gy2S4vRSzpqPjDHGlLGkYIwxpkwkJoVn/Q7AI4szuCIhzkiIESzOYKtXcUbciWZjjDGhE4lHCsYYY0LEkoIxxpgyEZUUROQSEVERGeh3LBURkWIRWSEiK0VkmYic6ndMlRGReBGZJSLfi8hqEflQRPr7HVeggP2Z4e7TO0Qk7N6zAXGW/t3jd0wVqSDOXn7HVBER6SQiM0Vko4gsFZFvReQSv+MKJCLtA/ZjtohsC7gf63d8tRFR5xRE5E2gM/Cpqt7vczjHEZFDqtrCvT0B+J2qnuVzWMdx58H+BnhZVZ92HxsKtFTVr3wNLkC5/dkRmAl8rap/8DeyYwXGGc4iIc5K3ps9gYtU9Ulfg6uEiNwPHFLVR/yOJRjC7ldXZUSkBXAa8DMiY96FVsA+v4OoxFigsPRDB6CqK8IpIZSnqjk483RPdb84TP10DlBQ7r25OVwTQiQQkV4iskZEnnOPuueKSNPKykdMUgB+DKSp6npgr4gM9zugCjR1Dx/XAjOAP/odUCWSgKV+B1FTqroR5z3b0e9YymkqxzbLXOF3QJUIjPM9v4OpRCKwzO8g6qF+wHRVTQT2Az+prGAkDXMxEXjcvT3LvR9ub56jqjoUQETGAK+ISJJNMRpU4XiUUPa6h7lIibOMiEwHTsc5ejjF73gi2CZVXeHeXgr0qqxgRCQFEWmPc1iZJCKKM72nisjd4fqFq6rfikgHIA7I8TuecjJw5sSOKCLSBygm/PanCZ4MAn7Fquot7ufI07g9plL5AbeLgYhvProMeEVVe6pqL1XtDmzC+QURltwrpKKBPX7HUoHPgMYiclPpAyJyioiE3UnxUiISBzwNPBWuPwRMUHwGNBGRXwQ81syvYBqiiDhSwGkq+mu5x94BJgHhdHK0qYiUHqIJcJ2qFvsZUEVUVd1L/B53L5/MA34Apvka2PFK92cjoAh4Ffi7vyFVKPB1B+fcV1helhru3Pfmj4HHRORuYBdwGPiNv5E1HBF1SaoxxpjQipTmI2OMMXXAkoIxxpgylhSMMcaUsaRgjDGmjCUFY4wxZSwpGFNLInK/iNxZweNdRORtP2Iy5kRFSj8FYyKOqm4nAnuOm4bNjhSMqYSIXCsiq9y5HF4VkQtFZKGILBeRT0SkU0DxISLymYhsKO0p7o5Ome7evl5E3hWRNLfMQ748KWOqYUcKxlRARBKBe4HTVHW3iLQDFBjt9rqdDNwN/Npd5WRgNNAcWC4iH1RQ7VBgGM44NOtE5ElV3Rrq52JMTVhSMKZi5wBvq+puAFXdKyKDgX+JSGcgFmf8rVL/VtWjwFERmQeMBFaUq/NTVT0AICKrgZ6AJQUTVqz5yJiKCc6RQaAncQbkGwzcDDQJWFa+bEXjx5QfqdJ+lJmwY0nBmIp9ClzuDtuO23zUGtjmLr+uXPmLRaSJW/5sYHFdBWpMMNkvFWMqoKoZIvIn4AsRKQaWA/cDb4nINmAB0DtglUXAB0AP4I+qul1EetVp0MYEgY2Saowxpow1HxljjCljScEYY0wZSwrGGGPKWFIwxhhTxpKCMcaYMpYUjDHGlLGkYIwxpsz/Bzocrzy1Qn5zAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# let's explore the relationship of the categories with the target\n",
    "\n",
    "for var in ['cabin', 'sex', 'embarked']:\n",
    "    \n",
    "    fig = plt.figure()\n",
    "    fig = X_train.groupby([var])['survived'].mean().plot()\n",
    "    fig.set_title('Relationship between {} and Survival'.format(var))\n",
    "    fig.set_ylabel('Mean Survival')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can see that the relationship between the target and cabin and embarked goes up and down, depending on the category.\n",
    "\n",
    "\n",
    "## Mean encoding with pandas\n",
    "\n",
    "\n",
    "### Advantages\n",
    "\n",
    "- quick\n",
    "- returns pandas dataframe\n",
    "\n",
    "### Limitations of pandas:\n",
    "\n",
    "- it does not preserve information from train data to propagate to test data\n",
    "\n",
    "We need to store the mappings separately if planning to use the model in production."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "cabin\n",
       "A    0.529412\n",
       "B    0.761905\n",
       "C    0.563380\n",
       "D    0.718750\n",
       "E    0.718750\n",
       "F    0.666667\n",
       "G    0.500000\n",
       "T    0.000000\n",
       "n    0.304843\n",
       "Name: survived, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's calculate the mean target value per category for cabin\n",
    "\n",
    "X_train.groupby(['cabin'])['survived'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'A': 0.5294117647058824,\n",
       " 'B': 0.7619047619047619,\n",
       " 'C': 0.5633802816901409,\n",
       " 'D': 0.71875,\n",
       " 'E': 0.71875,\n",
       " 'F': 0.6666666666666666,\n",
       " 'G': 0.5,\n",
       " 'T': 0.0,\n",
       " 'n': 0.30484330484330485}"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# and now let's do the same but capturing the result in a dictionary\n",
    "\n",
    "ordered_labels = X_train.groupby(['cabin'])['survived'].mean().to_dict()\n",
    "\n",
    "ordered_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# now, we replace the labels with the target means\n",
    "\n",
    "X_train['cabin'] = X_train['cabin'].map(ordered_labels)\n",
    "X_test['cabin'] = X_test['cabin'].map(ordered_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "501     0.304843\n",
       "588     0.304843\n",
       "402     0.304843\n",
       "1193    0.304843\n",
       "686     0.304843\n",
       "971     0.304843\n",
       "117     0.718750\n",
       "540     0.304843\n",
       "294     0.563380\n",
       "261     0.718750\n",
       "Name: cabin, dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's explore the result\n",
    "\n",
    "X_train['cabin'].head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we can turn the previous commands into 2 functions\n",
    "\n",
    "\n",
    "def find_category_mappings(df, variable, target):\n",
    "    return df.groupby([variable])[target].mean().to_dict()\n",
    "\n",
    "\n",
    "def integer_encode(train, test, variable, ordinal_mapping):\n",
    "\n",
    "    X_train[variable] = X_train[variable].map(ordinal_mapping)\n",
    "    X_test[variable] = X_test[variable].map(ordinal_mapping)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# and now we run a loop over the remaining categorical variables\n",
    "\n",
    "for variable in ['sex', 'embarked']:\n",
    "    mappings = find_category_mappings(X_train, variable, 'survived')\n",
    "    integer_encode(X_train, X_test, variable, mappings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cabin</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "      <th>survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>501</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.338957</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.338957</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>402</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.553073</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1193</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.187608</td>\n",
       "      <td>0.373494</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>686</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.373494</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         cabin       sex  embarked  survived\n",
       "501   0.304843  0.728358  0.338957         1\n",
       "588   0.304843  0.728358  0.338957         1\n",
       "402   0.304843  0.728358  0.553073         1\n",
       "1193  0.304843  0.187608  0.373494         0\n",
       "686   0.304843  0.728358  0.373494         1"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's see the result\n",
    "\n",
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# let's inspect the newly created monotonic relationship\n",
    "# between the variables and the target\n",
    "\n",
    "for var in ['cabin', 'sex', 'embarked']:\n",
    "    \n",
    "    fig = plt.figure()\n",
    "    fig = X_train.groupby([var])['survived'].mean().plot()\n",
    "    fig.set_title('Monotonic relationship between {} and Survival'.format(var))\n",
    "    fig.set_ylabel('Mean Survived')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mean Encoding creates linear relationships between the mean target and the categories.\n",
    "\n",
    "### Note\n",
    "\n",
    "Replacing categorical labels with this code and method will generate missing values for categories present in the test set that were not seen in the training set. Therefore it is extremely important to handle rare labels before-hand. I will explain how to do this, in a later notebook."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mean Encoding with Feature-Engine\n",
    "\n",
    "If using Feature-Engine, instead of pandas, we do not need to keep the target variable in the training dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((916, 3), (393, 3))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's separate into training and testing set\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    data[['cabin', 'sex', 'embarked']],  # predictors\n",
    "    data['survived'],  # target\n",
    "    test_size=0.3,  # percentage of obs in test set\n",
    "    random_state=0)  # seed to ensure reproducibility\n",
    "\n",
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_enc = MeanCategoricalEncoder(\n",
    "    variables=['cabin', 'sex', 'embarked'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MeanCategoricalEncoder(variables=['cabin', 'sex', 'embarked'])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# when fitting the transformer, we need to pass the target as well\n",
    "# just like with any Scikit-learn predictor class\n",
    "\n",
    "mean_enc.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cabin': {'A': 0.5294117647058824,\n",
       "  'B': 0.7619047619047619,\n",
       "  'C': 0.5633802816901409,\n",
       "  'D': 0.71875,\n",
       "  'E': 0.71875,\n",
       "  'F': 0.6666666666666666,\n",
       "  'G': 0.5,\n",
       "  'T': 0.0,\n",
       "  'n': 0.30484330484330485},\n",
       " 'sex': {'female': 0.7283582089552239, 'male': 0.18760757314974183},\n",
       " 'embarked': {'C': 0.553072625698324,\n",
       "  'Missing': 1.0,\n",
       "  'Q': 0.37349397590361444,\n",
       "  'S': 0.3389570552147239}}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# in the encoder dict we see the target mean assigned to each\n",
    "# category for each of the selected variables\n",
    "\n",
    "mean_enc.encoder_dict_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['cabin', 'sex', 'embarked']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this is the list of variables that the encoder will transform\n",
    "\n",
    "mean_enc.variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cabin</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>501</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.338957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.338957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>402</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.553073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1193</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.187608</td>\n",
       "      <td>0.373494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>686</th>\n",
       "      <td>0.304843</td>\n",
       "      <td>0.728358</td>\n",
       "      <td>0.373494</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         cabin       sex  embarked\n",
       "501   0.304843  0.728358  0.338957\n",
       "588   0.304843  0.728358  0.338957\n",
       "402   0.304843  0.728358  0.553073\n",
       "1193  0.304843  0.187608  0.373494\n",
       "686   0.304843  0.728358  0.373494"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = mean_enc.transform(X_train)\n",
    "X_test = mean_enc.transform(X_test)\n",
    "\n",
    "# let's explore the result\n",
    "X_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "If the argument variables is left to None, then the encoder will automatically identify all categorical variables. Is that not sweet?\n",
    "\n",
    "The encoder will not encode numerical variables. So if some of your numerical variables are in fact categories, you will need to re-cast them as object before using the encoder.\n",
    "\n",
    "Finally, if there is a label in the test set that was not present in the train set, the encoder will through and error, to alert you of this behaviour."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "feml",
   "language": "python",
   "name": "feml"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}