lectures/05_preprocessing-pipelines.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "![](img/330-banner.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "# Lecture 5: Preprocessing and `sklearn` pipelines\n",
    "\n",
    "UBC 2021-22\n",
    "\n",
    "Instructor: Giulia Toti"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "import time\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from IPython.display import HTML\n",
    "\n",
    "sys.path.append(\"code/.\")\n",
    "\n",
    "import mglearn\n",
    "from IPython.display import display\n",
    "from plotting_functions import *\n",
    "\n",
    "# Classifiers and regressors\n",
    "from sklearn.dummy import DummyClassifier, DummyRegressor\n",
    "\n",
    "# Preprocessing and pipeline\n",
    "from sklearn.impute import SimpleImputer\n",
    "\n",
    "# train test split and cross validation\n",
    "from sklearn.model_selection import cross_val_score, cross_validate, train_test_split\n",
    "from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import (\n",
    "    MinMaxScaler,\n",
    "    OneHotEncoder,\n",
    "    OrdinalEncoder,\n",
    "    StandardScaler,\n",
    ")\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from utils import *\n",
    "\n",
    "pd.set_option(\"display.max_colwidth\", 200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Announcements\n",
    "- Homework 3 is out. (Due date: Jan 31st) Please start early. \n",
    "- Homework 1 grades and solutions are out. Please do not share them with anyone or do not post them anywhere. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Learning outcomes\n",
    "\n",
    "From this lecture, you will be able to \n",
    "\n",
    "- explain motivation for preprocessing in supervised machine learning;\n",
    "- identify when to implement feature transformations such as imputation, scaling, and one-hot encoding in a machine learning model development pipeline; \n",
    "- use `sklearn` transformers for applying feature transformations on your dataset;\n",
    "- discuss golden rule in the context of feature transformations;\n",
    "- use `sklearn.pipeline.Pipeline` and `sklearn.pipeline.make_pipeline` to build a preliminary machine learning pipeline. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Motivation and big picture [[video](https://youtu.be/xx9HlmzORRk)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "- So far we have seen\n",
    "    - Three ML models (decision trees, $k$-NNs, SVMs with RBF kernel)\n",
    "    - ML fundamentals (train-validation-test split, cross-validation, the fundamental tradeoff, the golden rule)\n",
    "- Are we ready to do machine learning on real-world datasets?\n",
    "    - Very often real-world datasets need preprocessing before we use them to build ML models. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Example: $k$-nearest neighbours on the Spotify dataset\n",
    "\n",
    "- In lab1 you used `DecisionTreeClassifier` to predict whether the user would like a particular song or not. \n",
    "- Can we use $k$-NN classifier for this task? \n",
    "- Intuition: To predict whether the user likes a particular song or not (query point) \n",
    "   - find the songs that are closest to the query point\n",
    "   - let them vote on the target\n",
    "   - take the majority vote as the target for the query point\n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "spotify_df = pd.read_csv(\"data/spotify.csv\", index_col=0)\n",
    "train_df, test_df = train_test_split(spotify_df, test_size=0.20, random_state=123)\n",
    "X_train, y_train = (\n",
    "    train_df.drop(columns=[\"song_title\", \"artist\", \"target\"]),\n",
    "    train_df[\"target\"],\n",
    ")\n",
    "X_test, y_test = (\n",
    "    test_df.drop(columns=[\"song_title\", \"artist\", \"target\"]),\n",
    "    test_df[\"target\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean validation score 0.508\n"
     ]
    },
    {
     "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>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_score</th>\n",
       "      <th>train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000611</td>\n",
       "      <td>0.000243</td>\n",
       "      <td>0.507740</td>\n",
       "      <td>0.507752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000351</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.507740</td>\n",
       "      <td>0.507752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000343</td>\n",
       "      <td>0.000153</td>\n",
       "      <td>0.507740</td>\n",
       "      <td>0.507752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000359</td>\n",
       "      <td>0.000160</td>\n",
       "      <td>0.506211</td>\n",
       "      <td>0.508133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000379</td>\n",
       "      <td>0.000174</td>\n",
       "      <td>0.509317</td>\n",
       "      <td>0.507359</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fit_time  score_time  test_score  train_score\n",
       "0  0.000611    0.000243    0.507740     0.507752\n",
       "1  0.000351    0.000154    0.507740     0.507752\n",
       "2  0.000343    0.000153    0.507740     0.507752\n",
       "3  0.000359    0.000160    0.506211     0.508133\n",
       "4  0.000379    0.000174    0.509317     0.507359"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dummy = DummyClassifier(strategy=\"most_frequent\")\n",
    "scores = cross_validate(dummy, X_train, y_train, return_train_score=True)\n",
    "print(\"Mean validation score %0.3f\" % (np.mean(scores[\"test_score\"])))\n",
    "pd.DataFrame(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean validation score 0.546\n"
     ]
    },
    {
     "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>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_score</th>\n",
       "      <th>train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.002325</td>\n",
       "      <td>0.006984</td>\n",
       "      <td>0.563467</td>\n",
       "      <td>0.717829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.001870</td>\n",
       "      <td>0.006374</td>\n",
       "      <td>0.535604</td>\n",
       "      <td>0.721705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.001740</td>\n",
       "      <td>0.006395</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.708527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.001607</td>\n",
       "      <td>0.006033</td>\n",
       "      <td>0.537267</td>\n",
       "      <td>0.721921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.001491</td>\n",
       "      <td>0.006096</td>\n",
       "      <td>0.562112</td>\n",
       "      <td>0.711077</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fit_time  score_time  test_score  train_score\n",
       "0  0.002325    0.006984    0.563467     0.717829\n",
       "1  0.001870    0.006374    0.535604     0.721705\n",
       "2  0.001740    0.006395    0.529412     0.708527\n",
       "3  0.001607    0.006033    0.537267     0.721921\n",
       "4  0.001491    0.006096    0.562112     0.711077"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsClassifier()\n",
    "scores = cross_validate(knn, X_train, y_train, return_train_score=True)\n",
    "print(\"Mean validation score %0.3f\" % (np.mean(scores[\"test_score\"])))\n",
    "pd.DataFrame(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>acousticness</th>\n",
       "      <th>danceability</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>energy</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>key</th>\n",
       "      <th>liveness</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>tempo</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>valence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>0.229000</td>\n",
       "      <td>0.494</td>\n",
       "      <td>147893</td>\n",
       "      <td>0.666</td>\n",
       "      <td>0.000057</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0469</td>\n",
       "      <td>-9.743</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0351</td>\n",
       "      <td>140.832</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>654</th>\n",
       "      <td>0.000289</td>\n",
       "      <td>0.771</td>\n",
       "      <td>227143</td>\n",
       "      <td>0.949</td>\n",
       "      <td>0.602000</td>\n",
       "      <td>8</td>\n",
       "      <td>0.5950</td>\n",
       "      <td>-4.712</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1750</td>\n",
       "      <td>111.959</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.372</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     acousticness  danceability  duration_ms  energy  instrumentalness  key  \\\n",
       "842      0.229000         0.494       147893   0.666          0.000057    9   \n",
       "654      0.000289         0.771       227143   0.949          0.602000    8   \n",
       "\n",
       "     liveness  loudness  mode  speechiness    tempo  time_signature  valence  \n",
       "842    0.0469    -9.743     0       0.0351  140.832             4.0    0.704  \n",
       "654    0.5950    -4.712     1       0.1750  111.959             4.0    0.372  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "two_songs = X_train.sample(2, random_state=42)\n",
    "two_songs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[    0.        , 79250.00543825],\n",
       "       [79250.00543825,     0.        ]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "euclidean_distances(two_songs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's consider only two features: `duration_ms` and `tempo`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>duration_ms</th>\n",
       "      <th>tempo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>147893</td>\n",
       "      <td>140.832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>654</th>\n",
       "      <td>227143</td>\n",
       "      <td>111.959</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     duration_ms    tempo\n",
       "842       147893  140.832\n",
       "654       227143  111.959"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "two_songs_subset = two_songs[[\"duration_ms\", \"tempo\"]]\n",
    "two_songs_subset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[    0.        , 79250.00525962],\n",
       "       [79250.00525962,     0.        ]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "euclidean_distances(two_songs_subset)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do you see any problem? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- The distance is completely dominated by the the features with larger values\n",
    "- The features with smaller values are being ignored. \n",
    "- Does it matter? \n",
    "    - Yes! Scale is based on how data was collected. \n",
    "    - Features on a smaller scale can be highly informative and there is no good reason to ignore them.\n",
    "    - We want our model to be robust and not sensitive to the scale. \n",
    "- Was this a problem for decision trees?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Scaling using `scikit-learn`'s [`StandardScaler`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html)\n",
    "\n",
    "- We'll use `scikit-learn`'s [`StandardScaler`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html), which is a `transformer`.   \n",
    "- Only focus on the syntax for now. We'll talk about scaling in a bit. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>acousticness</th>\n",
       "      <th>danceability</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>energy</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>key</th>\n",
       "      <th>liveness</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>tempo</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>valence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.697633</td>\n",
       "      <td>-0.194548</td>\n",
       "      <td>-0.398940</td>\n",
       "      <td>-0.318116</td>\n",
       "      <td>-0.492359</td>\n",
       "      <td>1.275623</td>\n",
       "      <td>-0.737898</td>\n",
       "      <td>0.395794</td>\n",
       "      <td>-1.280599</td>\n",
       "      <td>-0.617752</td>\n",
       "      <td>-0.293827</td>\n",
       "      <td>0.138514</td>\n",
       "      <td>-0.908149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.276291</td>\n",
       "      <td>0.295726</td>\n",
       "      <td>-0.374443</td>\n",
       "      <td>-0.795552</td>\n",
       "      <td>0.598355</td>\n",
       "      <td>-1.487342</td>\n",
       "      <td>-0.438792</td>\n",
       "      <td>-0.052394</td>\n",
       "      <td>0.780884</td>\n",
       "      <td>2.728394</td>\n",
       "      <td>-0.802595</td>\n",
       "      <td>-3.781179</td>\n",
       "      <td>-1.861238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.599540</td>\n",
       "      <td>1.110806</td>\n",
       "      <td>-0.376205</td>\n",
       "      <td>-0.946819</td>\n",
       "      <td>-0.492917</td>\n",
       "      <td>0.446734</td>\n",
       "      <td>-0.399607</td>\n",
       "      <td>-0.879457</td>\n",
       "      <td>0.780884</td>\n",
       "      <td>2.534909</td>\n",
       "      <td>0.191274</td>\n",
       "      <td>0.138514</td>\n",
       "      <td>0.575870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.307150</td>\n",
       "      <td>1.809445</td>\n",
       "      <td>-0.654016</td>\n",
       "      <td>-1.722063</td>\n",
       "      <td>-0.492168</td>\n",
       "      <td>0.170437</td>\n",
       "      <td>-0.763368</td>\n",
       "      <td>-1.460798</td>\n",
       "      <td>-1.280599</td>\n",
       "      <td>-0.608647</td>\n",
       "      <td>-0.839616</td>\n",
       "      <td>0.138514</td>\n",
       "      <td>1.825358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.634642</td>\n",
       "      <td>0.491835</td>\n",
       "      <td>-0.131344</td>\n",
       "      <td>1.057468</td>\n",
       "      <td>2.723273</td>\n",
       "      <td>0.170437</td>\n",
       "      <td>-0.458384</td>\n",
       "      <td>-0.175645</td>\n",
       "      <td>-1.280599</td>\n",
       "      <td>-0.653035</td>\n",
       "      <td>-0.074294</td>\n",
       "      <td>0.138514</td>\n",
       "      <td>-0.754491</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   acousticness  danceability  duration_ms    energy  instrumentalness  \\\n",
       "0     -0.697633     -0.194548    -0.398940 -0.318116         -0.492359   \n",
       "1     -0.276291      0.295726    -0.374443 -0.795552          0.598355   \n",
       "2     -0.599540      1.110806    -0.376205 -0.946819         -0.492917   \n",
       "3     -0.307150      1.809445    -0.654016 -1.722063         -0.492168   \n",
       "4     -0.634642      0.491835    -0.131344  1.057468          2.723273   \n",
       "\n",
       "        key  liveness  loudness      mode  speechiness     tempo  \\\n",
       "0  1.275623 -0.737898  0.395794 -1.280599    -0.617752 -0.293827   \n",
       "1 -1.487342 -0.438792 -0.052394  0.780884     2.728394 -0.802595   \n",
       "2  0.446734 -0.399607 -0.879457  0.780884     2.534909  0.191274   \n",
       "3  0.170437 -0.763368 -1.460798 -1.280599    -0.608647 -0.839616   \n",
       "4  0.170437 -0.458384 -0.175645 -1.280599    -0.653035 -0.074294   \n",
       "\n",
       "   time_signature   valence  \n",
       "0        0.138514 -0.908149  \n",
       "1       -3.781179 -1.861238  \n",
       "2        0.138514  0.575870  \n",
       "3        0.138514  1.825358  \n",
       "4        0.138514 -0.754491  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()  # create feature trasformer object\n",
    "scaler.fit(X_train)  # fitting the transformer on the train split\n",
    "X_train_scaled = scaler.transform(X_train)  # transforming the train split\n",
    "X_test_scaled = scaler.transform(X_test)  # transforming the test split\n",
    "pd.DataFrame(X_train_scaled, columns=X_train.columns).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### `fit` and `transform` paradigm for transformers\n",
    "- `sklearn` uses `fit` and `transform` paradigms for feature transformations. \n",
    "- We `fit` the transformer on the train split and then transform the train split as well as the test split. \n",
    "- We apply the same transformations on the test split. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### `sklearn` API summary: estimators\n",
    "\n",
    "Suppose `model` is a classification or regression model. \n",
    "\n",
    "```\n",
    "model.fit(X_train, y_train)\n",
    "X_train_predictions = model.predict(X_train)\n",
    "X_test_predictions = model.predict(X_test)\n",
    "```    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### `sklearn` API summary: transformers\n",
    "\n",
    "Suppose `transformer` is a transformer used to change the input representation, for example, to tackle missing values or to scales numeric features.\n",
    "\n",
    "```\n",
    "transformer.fit(X_train, [y_train])\n",
    "X_train_transformed = transformer.transform(X_train)\n",
    "X_test_transformed = transformer.transform(X_test)\n",
    "```  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- You can pass `y_train` in `fit` but it's usually ignored. It allows you to pass it just to be consistent with usual usage of `sklearn`'s `fit` method.   \n",
    "- You can also carry out fitting and transforming in one call using `fit_transform`. But be mindful to use it only on the train split and **not** on the test split. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- Do you expect `DummyClassifier` results to change after scaling the data? \n",
    "- Let's check whether scaling makes any difference for $k$-NNs. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score: 0.726\n",
      "Test score: 0.552\n"
     ]
    }
   ],
   "source": [
    "knn_unscaled = KNeighborsClassifier()\n",
    "knn_unscaled.fit(X_train, y_train)\n",
    "print(\"Train score: %0.3f\" % (knn_unscaled.score(X_train, y_train)))\n",
    "print(\"Test score: %0.3f\" % (knn_unscaled.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score: 0.798\n",
      "Test score: 0.686\n"
     ]
    }
   ],
   "source": [
    "knn_scaled = KNeighborsClassifier()\n",
    "knn_scaled.fit(X_train_scaled, y_train)\n",
    "print(\"Train score: %0.3f\" % (knn_scaled.score(X_train_scaled, y_train)))\n",
    "print(\"Test score: %0.3f\" % (knn_scaled.score(X_test_scaled, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- The scores with scaled data are better compared to the unscaled data in case of $k$-NNs.\n",
    "- I am not carrying out cross-validation here for a reason that we'll look into soon. \n",
    "- Note that I am a bit sloppy here and using the test set several times for teaching purposes. But when you build an ML pipeline, please do assessment on the test set only once.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Common preprocessing techniques\n",
    "\n",
    "Some commonly performed feature transformation include:  \n",
    "- Imputation: Tackling missing values\n",
    "- Scaling: Scaling of numeric features\n",
    "- One-hot encoding: Tackling categorical variables      \n",
    "    \n",
    "\n",
    "We can have one lecture on each of them! In this lesson our goal is to getting familiar with them so that we can use them to build ML pipelines. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "In the next part of this lecture, we'll build an ML pipeline using [California housing prices regression dataset](https://www.kaggle.com/harrywang/housing). In the process, we will talk about different feature transformations and how can we apply them so that we do not violate the golden rule. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Imputation and scaling [[video](https://youtu.be/G2IXbVzKlt8)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Dataset, splitting, and baseline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "We'll be working on [California housing prices regression dataset](https://www.kaggle.com/harrywang/housing) to demonstrate these feature transformation techniques. The task is to predict median house values in Californian districts, given a number of features from these districts. If you are running the notebook on your own, you'll have to download the data and put it in the data directory. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "      <th>ocean_proximity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6051</th>\n",
       "      <td>-117.75</td>\n",
       "      <td>34.04</td>\n",
       "      <td>22.0</td>\n",
       "      <td>2948.0</td>\n",
       "      <td>636.0</td>\n",
       "      <td>2600.0</td>\n",
       "      <td>602.0</td>\n",
       "      <td>3.1250</td>\n",
       "      <td>113600.0</td>\n",
       "      <td>INLAND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20113</th>\n",
       "      <td>-119.57</td>\n",
       "      <td>37.94</td>\n",
       "      <td>17.0</td>\n",
       "      <td>346.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>3.4861</td>\n",
       "      <td>137500.0</td>\n",
       "      <td>INLAND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14289</th>\n",
       "      <td>-117.13</td>\n",
       "      <td>32.74</td>\n",
       "      <td>46.0</td>\n",
       "      <td>3355.0</td>\n",
       "      <td>768.0</td>\n",
       "      <td>1457.0</td>\n",
       "      <td>708.0</td>\n",
       "      <td>2.6604</td>\n",
       "      <td>170100.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13665</th>\n",
       "      <td>-117.31</td>\n",
       "      <td>34.02</td>\n",
       "      <td>18.0</td>\n",
       "      <td>1634.0</td>\n",
       "      <td>274.0</td>\n",
       "      <td>899.0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>5.2139</td>\n",
       "      <td>129300.0</td>\n",
       "      <td>INLAND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14471</th>\n",
       "      <td>-117.23</td>\n",
       "      <td>32.88</td>\n",
       "      <td>18.0</td>\n",
       "      <td>5566.0</td>\n",
       "      <td>1465.0</td>\n",
       "      <td>6303.0</td>\n",
       "      <td>1458.0</td>\n",
       "      <td>1.8580</td>\n",
       "      <td>205000.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "6051     -117.75     34.04                22.0       2948.0           636.0   \n",
       "20113    -119.57     37.94                17.0        346.0           130.0   \n",
       "14289    -117.13     32.74                46.0       3355.0           768.0   \n",
       "13665    -117.31     34.02                18.0       1634.0           274.0   \n",
       "14471    -117.23     32.88                18.0       5566.0          1465.0   \n",
       "\n",
       "       population  households  median_income  median_house_value  \\\n",
       "6051       2600.0       602.0         3.1250            113600.0   \n",
       "20113        51.0        20.0         3.4861            137500.0   \n",
       "14289      1457.0       708.0         2.6604            170100.0   \n",
       "13665       899.0       285.0         5.2139            129300.0   \n",
       "14471      6303.0      1458.0         1.8580            205000.0   \n",
       "\n",
       "      ocean_proximity  \n",
       "6051           INLAND  \n",
       "20113          INLAND  \n",
       "14289      NEAR OCEAN  \n",
       "13665          INLAND  \n",
       "14471      NEAR OCEAN  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "housing_df = pd.read_csv(\"data/housing.csv\")\n",
    "train_df, test_df = train_test_split(housing_df, test_size=0.1, random_state=123)\n",
    "\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Some column values are mean/median but some are not. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Let's add some new features to the dataset which could help predicting the target: `median_house_value`. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "train_df = train_df.assign(\n",
    "    rooms_per_household=train_df[\"total_rooms\"] / train_df[\"households\"]\n",
    ")\n",
    "test_df = test_df.assign(\n",
    "    rooms_per_household=test_df[\"total_rooms\"] / test_df[\"households\"]\n",
    ")\n",
    "\n",
    "train_df = train_df.assign(\n",
    "    bedrooms_per_household=train_df[\"total_bedrooms\"] / train_df[\"households\"]\n",
    ")\n",
    "test_df = test_df.assign(\n",
    "    bedrooms_per_household=test_df[\"total_bedrooms\"] / test_df[\"households\"]\n",
    ")\n",
    "\n",
    "train_df = train_df.assign(\n",
    "    population_per_household=train_df[\"population\"] / train_df[\"households\"]\n",
    ")\n",
    "test_df = test_df.assign(\n",
    "    population_per_household=test_df[\"population\"] / test_df[\"households\"]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "      <th>ocean_proximity</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6051</th>\n",
       "      <td>-117.75</td>\n",
       "      <td>34.04</td>\n",
       "      <td>22.0</td>\n",
       "      <td>2948.0</td>\n",
       "      <td>636.0</td>\n",
       "      <td>2600.0</td>\n",
       "      <td>602.0</td>\n",
       "      <td>3.1250</td>\n",
       "      <td>113600.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>4.897010</td>\n",
       "      <td>1.056478</td>\n",
       "      <td>4.318937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20113</th>\n",
       "      <td>-119.57</td>\n",
       "      <td>37.94</td>\n",
       "      <td>17.0</td>\n",
       "      <td>346.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>3.4861</td>\n",
       "      <td>137500.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>17.300000</td>\n",
       "      <td>6.500000</td>\n",
       "      <td>2.550000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14289</th>\n",
       "      <td>-117.13</td>\n",
       "      <td>32.74</td>\n",
       "      <td>46.0</td>\n",
       "      <td>3355.0</td>\n",
       "      <td>768.0</td>\n",
       "      <td>1457.0</td>\n",
       "      <td>708.0</td>\n",
       "      <td>2.6604</td>\n",
       "      <td>170100.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "      <td>4.738701</td>\n",
       "      <td>1.084746</td>\n",
       "      <td>2.057910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13665</th>\n",
       "      <td>-117.31</td>\n",
       "      <td>34.02</td>\n",
       "      <td>18.0</td>\n",
       "      <td>1634.0</td>\n",
       "      <td>274.0</td>\n",
       "      <td>899.0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>5.2139</td>\n",
       "      <td>129300.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>5.733333</td>\n",
       "      <td>0.961404</td>\n",
       "      <td>3.154386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14471</th>\n",
       "      <td>-117.23</td>\n",
       "      <td>32.88</td>\n",
       "      <td>18.0</td>\n",
       "      <td>5566.0</td>\n",
       "      <td>1465.0</td>\n",
       "      <td>6303.0</td>\n",
       "      <td>1458.0</td>\n",
       "      <td>1.8580</td>\n",
       "      <td>205000.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "      <td>3.817558</td>\n",
       "      <td>1.004801</td>\n",
       "      <td>4.323045</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "6051     -117.75     34.04                22.0       2948.0           636.0   \n",
       "20113    -119.57     37.94                17.0        346.0           130.0   \n",
       "14289    -117.13     32.74                46.0       3355.0           768.0   \n",
       "13665    -117.31     34.02                18.0       1634.0           274.0   \n",
       "14471    -117.23     32.88                18.0       5566.0          1465.0   \n",
       "\n",
       "       population  households  median_income  median_house_value  \\\n",
       "6051       2600.0       602.0         3.1250            113600.0   \n",
       "20113        51.0        20.0         3.4861            137500.0   \n",
       "14289      1457.0       708.0         2.6604            170100.0   \n",
       "13665       899.0       285.0         5.2139            129300.0   \n",
       "14471      6303.0      1458.0         1.8580            205000.0   \n",
       "\n",
       "      ocean_proximity  rooms_per_household  bedrooms_per_household  \\\n",
       "6051           INLAND             4.897010                1.056478   \n",
       "20113          INLAND            17.300000                6.500000   \n",
       "14289      NEAR OCEAN             4.738701                1.084746   \n",
       "13665          INLAND             5.733333                0.961404   \n",
       "14471      NEAR OCEAN             3.817558                1.004801   \n",
       "\n",
       "       population_per_household  \n",
       "6051                   4.318937  \n",
       "20113                  2.550000  \n",
       "14289                  2.057910  \n",
       "13665                  3.154386  \n",
       "14471                  4.323045  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### When is it OK to do things before splitting? \n",
    "\n",
    "- Here it would have been OK to add new features before splitting because we are not using any global information in the data but only looking at one row at a time. \n",
    "- But just to be safe and to avoid accidentally breaking the golden rule, it's better to do it after splitting. \n",
    "\n",
    "- Question: Should we remove `total_rooms`, `total_bedrooms`, and `population` columns? \n",
    "    - Probably. But I am keeping them in this lecture. You could experiment with removing them and examine whether results change. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### EDA "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "      <th>ocean_proximity</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6051</th>\n",
       "      <td>-117.75</td>\n",
       "      <td>34.04</td>\n",
       "      <td>22.0</td>\n",
       "      <td>2948.0</td>\n",
       "      <td>636.0</td>\n",
       "      <td>2600.0</td>\n",
       "      <td>602.0</td>\n",
       "      <td>3.1250</td>\n",
       "      <td>113600.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>4.897010</td>\n",
       "      <td>1.056478</td>\n",
       "      <td>4.318937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20113</th>\n",
       "      <td>-119.57</td>\n",
       "      <td>37.94</td>\n",
       "      <td>17.0</td>\n",
       "      <td>346.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>3.4861</td>\n",
       "      <td>137500.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>17.300000</td>\n",
       "      <td>6.500000</td>\n",
       "      <td>2.550000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14289</th>\n",
       "      <td>-117.13</td>\n",
       "      <td>32.74</td>\n",
       "      <td>46.0</td>\n",
       "      <td>3355.0</td>\n",
       "      <td>768.0</td>\n",
       "      <td>1457.0</td>\n",
       "      <td>708.0</td>\n",
       "      <td>2.6604</td>\n",
       "      <td>170100.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "      <td>4.738701</td>\n",
       "      <td>1.084746</td>\n",
       "      <td>2.057910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13665</th>\n",
       "      <td>-117.31</td>\n",
       "      <td>34.02</td>\n",
       "      <td>18.0</td>\n",
       "      <td>1634.0</td>\n",
       "      <td>274.0</td>\n",
       "      <td>899.0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>5.2139</td>\n",
       "      <td>129300.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>5.733333</td>\n",
       "      <td>0.961404</td>\n",
       "      <td>3.154386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14471</th>\n",
       "      <td>-117.23</td>\n",
       "      <td>32.88</td>\n",
       "      <td>18.0</td>\n",
       "      <td>5566.0</td>\n",
       "      <td>1465.0</td>\n",
       "      <td>6303.0</td>\n",
       "      <td>1458.0</td>\n",
       "      <td>1.8580</td>\n",
       "      <td>205000.0</td>\n",
       "      <td>NEAR OCEAN</td>\n",
       "      <td>3.817558</td>\n",
       "      <td>1.004801</td>\n",
       "      <td>4.323045</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "6051     -117.75     34.04                22.0       2948.0           636.0   \n",
       "20113    -119.57     37.94                17.0        346.0           130.0   \n",
       "14289    -117.13     32.74                46.0       3355.0           768.0   \n",
       "13665    -117.31     34.02                18.0       1634.0           274.0   \n",
       "14471    -117.23     32.88                18.0       5566.0          1465.0   \n",
       "\n",
       "       population  households  median_income  median_house_value  \\\n",
       "6051       2600.0       602.0         3.1250            113600.0   \n",
       "20113        51.0        20.0         3.4861            137500.0   \n",
       "14289      1457.0       708.0         2.6604            170100.0   \n",
       "13665       899.0       285.0         5.2139            129300.0   \n",
       "14471      6303.0      1458.0         1.8580            205000.0   \n",
       "\n",
       "      ocean_proximity  rooms_per_household  bedrooms_per_household  \\\n",
       "6051           INLAND             4.897010                1.056478   \n",
       "20113          INLAND            17.300000                6.500000   \n",
       "14289      NEAR OCEAN             4.738701                1.084746   \n",
       "13665          INLAND             5.733333                0.961404   \n",
       "14471      NEAR OCEAN             3.817558                1.004801   \n",
       "\n",
       "       population_per_household  \n",
       "6051                   4.318937  \n",
       "20113                  2.550000  \n",
       "14289                  2.057910  \n",
       "13665                  3.154386  \n",
       "14471                  4.323045  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The feature scales are quite different. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 18576 entries, 6051 to 19966\n",
      "Data columns (total 13 columns):\n",
      " #   Column                    Non-Null Count  Dtype  \n",
      "---  ------                    --------------  -----  \n",
      " 0   longitude                 18576 non-null  float64\n",
      " 1   latitude                  18576 non-null  float64\n",
      " 2   housing_median_age        18576 non-null  float64\n",
      " 3   total_rooms               18576 non-null  float64\n",
      " 4   total_bedrooms            18391 non-null  float64\n",
      " 5   population                18576 non-null  float64\n",
      " 6   households                18576 non-null  float64\n",
      " 7   median_income             18576 non-null  float64\n",
      " 8   median_house_value        18576 non-null  float64\n",
      " 9   ocean_proximity           18576 non-null  object \n",
      " 10  rooms_per_household       18576 non-null  float64\n",
      " 11  bedrooms_per_household    18391 non-null  float64\n",
      " 12  population_per_household  18576 non-null  float64\n",
      "dtypes: float64(12), object(1)\n",
      "memory usage: 2.0+ MB\n"
     ]
    }
   ],
   "source": [
    "train_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have one categorical feature and all other features are numeric features. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18391.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "      <td>18391.000000</td>\n",
       "      <td>18576.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-119.565888</td>\n",
       "      <td>35.627966</td>\n",
       "      <td>28.622255</td>\n",
       "      <td>2635.749677</td>\n",
       "      <td>538.229786</td>\n",
       "      <td>1428.578165</td>\n",
       "      <td>500.061100</td>\n",
       "      <td>3.862552</td>\n",
       "      <td>206292.067991</td>\n",
       "      <td>5.426067</td>\n",
       "      <td>1.097516</td>\n",
       "      <td>3.052349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.999622</td>\n",
       "      <td>2.134658</td>\n",
       "      <td>12.588307</td>\n",
       "      <td>2181.789934</td>\n",
       "      <td>421.805266</td>\n",
       "      <td>1141.664801</td>\n",
       "      <td>383.044313</td>\n",
       "      <td>1.892491</td>\n",
       "      <td>115083.856175</td>\n",
       "      <td>2.512319</td>\n",
       "      <td>0.486266</td>\n",
       "      <td>10.020873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-124.350000</td>\n",
       "      <td>32.540000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.499900</td>\n",
       "      <td>14999.000000</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.692308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-121.790000</td>\n",
       "      <td>33.930000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>1449.000000</td>\n",
       "      <td>296.000000</td>\n",
       "      <td>788.000000</td>\n",
       "      <td>280.000000</td>\n",
       "      <td>2.560225</td>\n",
       "      <td>119400.000000</td>\n",
       "      <td>4.439360</td>\n",
       "      <td>1.005888</td>\n",
       "      <td>2.430323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-118.490000</td>\n",
       "      <td>34.250000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>2127.000000</td>\n",
       "      <td>435.000000</td>\n",
       "      <td>1167.000000</td>\n",
       "      <td>410.000000</td>\n",
       "      <td>3.527500</td>\n",
       "      <td>179300.000000</td>\n",
       "      <td>5.226415</td>\n",
       "      <td>1.048860</td>\n",
       "      <td>2.818868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>-118.010000</td>\n",
       "      <td>37.710000</td>\n",
       "      <td>37.000000</td>\n",
       "      <td>3145.000000</td>\n",
       "      <td>647.000000</td>\n",
       "      <td>1727.000000</td>\n",
       "      <td>606.000000</td>\n",
       "      <td>4.736900</td>\n",
       "      <td>263600.000000</td>\n",
       "      <td>6.051620</td>\n",
       "      <td>1.099723</td>\n",
       "      <td>3.283921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>-114.310000</td>\n",
       "      <td>41.950000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>39320.000000</td>\n",
       "      <td>6445.000000</td>\n",
       "      <td>35682.000000</td>\n",
       "      <td>6082.000000</td>\n",
       "      <td>15.000100</td>\n",
       "      <td>500001.000000</td>\n",
       "      <td>141.909091</td>\n",
       "      <td>34.066667</td>\n",
       "      <td>1243.333333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          longitude      latitude  housing_median_age   total_rooms  \\\n",
       "count  18576.000000  18576.000000        18576.000000  18576.000000   \n",
       "mean    -119.565888     35.627966           28.622255   2635.749677   \n",
       "std        1.999622      2.134658           12.588307   2181.789934   \n",
       "min     -124.350000     32.540000            1.000000      2.000000   \n",
       "25%     -121.790000     33.930000           18.000000   1449.000000   \n",
       "50%     -118.490000     34.250000           29.000000   2127.000000   \n",
       "75%     -118.010000     37.710000           37.000000   3145.000000   \n",
       "max     -114.310000     41.950000           52.000000  39320.000000   \n",
       "\n",
       "       total_bedrooms    population    households  median_income  \\\n",
       "count    18391.000000  18576.000000  18576.000000   18576.000000   \n",
       "mean       538.229786   1428.578165    500.061100       3.862552   \n",
       "std        421.805266   1141.664801    383.044313       1.892491   \n",
       "min          1.000000      3.000000      1.000000       0.499900   \n",
       "25%        296.000000    788.000000    280.000000       2.560225   \n",
       "50%        435.000000   1167.000000    410.000000       3.527500   \n",
       "75%        647.000000   1727.000000    606.000000       4.736900   \n",
       "max       6445.000000  35682.000000   6082.000000      15.000100   \n",
       "\n",
       "       median_house_value  rooms_per_household  bedrooms_per_household  \\\n",
       "count        18576.000000         18576.000000            18391.000000   \n",
       "mean        206292.067991             5.426067                1.097516   \n",
       "std         115083.856175             2.512319                0.486266   \n",
       "min          14999.000000             0.846154                0.333333   \n",
       "25%         119400.000000             4.439360                1.005888   \n",
       "50%         179300.000000             5.226415                1.048860   \n",
       "75%         263600.000000             6.051620                1.099723   \n",
       "max         500001.000000           141.909091               34.066667   \n",
       "\n",
       "       population_per_household  \n",
       "count              18576.000000  \n",
       "mean                   3.052349  \n",
       "std                   10.020873  \n",
       "min                    0.692308  \n",
       "25%                    2.430323  \n",
       "50%                    2.818868  \n",
       "75%                    3.283921  \n",
       "max                 1243.333333  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "- Seems like total_bedrooms column has some missing values. \n",
    "- This must have affected our new feature `bedrooms_per_household` as well. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "207"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "housing_df[\"total_bedrooms\"].isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLQAAANoCAYAAADZALN5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAD5bklEQVR4nOzdfbxlZV3//9d74k5UZEgdDcUZpTL4TWCNpslXBiwBuctQvIEMSDE0tQQUpAwVgwTT0jRRvClUiLwbwgDROWg1mIoQjIapMyKKCg0gw60Dn98fa21ms2efM2fO2edmn/N6Ph77sWZf67PWvq61z6xrr2td67pSVUiSJEmSJEnDYsFMZ0CSJEmSJEnaEjZoSZIkSZIkaajYoCVJkiRJkqShYoOWJEmSJEmShooNWpIkSZIkSRoqNmhJkiRJkiRpqNigpaGXZHmSSnLqTOelnySL2/x9ZIY+/6j284+aic+XpLlgquuaJCNJair2Pc7PryQjM/X5kuaH2f67vR9/S2+5JGuTrO1J+0h7HBfPTK40F9mgJc2Qfid6SdLcNJGLuCSnttssn7KMSZIkDamtZjoD0jzwQ+DXgNtmOiOSpFnrpcD2M50JSdImPg1cAdw40xkZcicDZ9BcG0kDYYOWNMWq6ufA/8x0PiRJs1dVXT/TeZAkbaqqbsMb05NWVTdio6AGzEcONWcl+b0kX05ye5I7knw1yR/1iXvgufgk+yVZleTOJD9N8t4km9wxT/KwJO9M8qMkdyW5MskL+j1j3zuGVuc98ATgCe266t5urMdMRluX5KFJ3tGTp8M2c4z2SfK5JP+X5O4k30xyUhIbuyVpM5I8L8n5Sb7XnndvSXJZkt/piTsVWNm+/cvu835XzEjve+Av27cru7YZadePOj7jZta9IMk32nP+D9t64yFjlPExSf6uLeM9SX6S5NwkS8Z5mCSpryRPbc+Z65OsS/KxJI/qE/dH7e/4O5L8LMmXkhzaJ27UMZr6rUuyIMkrknytPX/f0Z7rPpFk9664MX/fJ9k1yWeS3NZed1yY5EmjlPlPkvxPew7+XpK/SPKk0c7Z4ziG3fnYrf1d/7MkNyf5QJKHtnHPTXJFmmucHyY5eZT9bZvkxCRXt7G3td/R3qPE75PkP7Lx2umcJDuNEtvvO3hEe+3x5SQ/TnJvkuuTvD/JY8bYx5Ikr0ny7bZu+m6S127p8eva728m+fskq9vjd0eaa6lXJcko22zRd9mu+3CSG9py3pDmWnOTv3mNnxetmpOSHA+cBdwEfBS4F3ge8MEke1TVa/psdihwAPBZ4D+B5wDHATsBL+ra9y8AnwP+H/C1dv+PBf4R+MI4sncr8GbgT9v37+pad9U4tt9EkgXAvwLLgSvbvPwS8HHg86Ns8yfA3wE305T5FmAv4HTgacDvTyQvkjSPnA7cA1wO/JimLvg94JIkL6iqT7VxI8Bi4A/b2JFx7Psj7XJvmnpmbft+bZ/YcUlyDHAOzfn+Q8BdNOf6Xx0l/pfbvD4G+Dfgk8DjgcOB/ZI8vaq+O9H8SJrXngq8HrgM+AfgmcBLgCcm+e2qKoAk7wb+BPg+8H5gG+AFwGeSHF9VfzOJPPw1cALw32y8Xng8sC9wEbB6HPtYTPM44tXAB4ClwEHA7kl2r6q7OoFJ/ormsbsfAO8DfgF4JfBbkyhDxxLgP9q8nA08C3gZsEOSC2jqlM+0658H/FWSH1XVR7vytx1wKc01zlfb/TyU5hrpC0kO76rXSPIcmuP0c+ATNNcUB9J8p9vQHM/N+TWa66IvAv9CUy/tARwLPCfJb1TVLX22O4vmuuVf2zy/AHhXknuq6h/G8bm9Xk7zvX2pLdPDaa4F3wP8Mhuv24At/y6TPAO4GNgOWAGsAZ4M/DFNffrUqlo3gXyrqnz5GuoXTSNOAae2759Ec2L9IfCYrriH0VQ2BTyrK/2oNu1e4Old6dsB3wLuB3buSn95G//PQLrS/18bW8BRXemL27SP9OR7LbB2lDKd2m6zfDzrgGPatM8AC7rSn92m9+Zp9/YYXQE8ois9NCfuAp4/09+tL1++fM2WV29d06Yt6RO3CLgB+M7mtu9ZP9L8LHtQ2lh1Qd+6ZbR1wCOAn9E8NrOkK/3hNBdtBYz07GcVTYPds3rSn9HWIf8609+LL1++huvVdS4s4LCu9AU0N4YLeEabtnf7/mrgYV2xj6H5nf9z4Ild6R9p4xf3+dxN1gHraBpufqEndque38dH9fkt3TnPFnB8z/YfbtNf3JX2ZOA+4H979v3otix9z+fjOJ7d+XhlTxm+QXNt8lPgKV3rdgbuBv67Z1+nt/s5qSf9UTTXLTcBD+n6vtYAG4Cn9nxu53tcO47v4BHATn3KdWQb++ej7ON/gUVd6bu2fw/XTfDvche6rqG6ynJx+709YaLfJU3j3vfbv7df6/mMF7Tx75np/5vD+vKRQ81FR9CcgP66qn7cSayq9Wx8fOMP+2z38aq6oiv+buA8mkae3+jZPzQn2OqK/zLNSW8mHNku31hV93cSq+oL9O+h9QqaY/Qn1YwL0Ikv4I00J9YX9dlOktSqqjV90n4CfAp4UmbX1OSH0jRevb8731V1O/C23uAkvwE8HfhAVX2pe11VraLp2XtAkkdMaa4lzVWXV9UnO2/a36//2L5d1i47v9f/ov0d34n9MXAmzW/Zzu/yibqnqu7rTqiqDd2/jzfje8A7e9I+0i6XdaW9iKYR6Mye394/Bf52i3Lc33dpegp19ruBpldtgAur6htd634I/DuwW9phRtqnPf4YWF1VZ3TvuKpuoukR9Uiam+XQ9I5aDHyqqr7a87l/Md5MV9Vt1b9n0sdobsD8Tp91AG9r69vOfr7TlulXkjx8vJ/ftf313ddQbdoGml5qC4B9ulZt6Xd5EE2D2V9V1bd6PuMC4Ot43TVhPnKouWjPdjnSZ93Knphu3+iT1pmFY8eutD2A/6uqb/eJX0Xz2OJ02wNYV1Xf7LPu34Hf7Un7LZpGq0OSHNRnm7to7j5IkkbRju9xMs15//E0PXu7PZZJPCI4YHu0yy/3WffvfdI6j008Ls0YYL0eS/OD/pdpHr+XpC0xnt/de7bLkT6xIz0xE3E+8MdJrqR53O3LwFeqajyPynX8d29DCKNfP0AzrEmvVVvweWPlo3rSOjf2r+4T/2Oax+QW0eT3V2ny+/1Rzvm/3C6fTPOY31h1yhU0PbfGJcmzaR7pexrwi22+Oh47ymab+/u5fbyf3+ZhW+A1wAtpjsXDekK687Gl32WnPv3/Rjm2DwF+Mckjq+rmcWdagA1ampt2aJc/6V1RVbcluacrplu/OzGdk3H3ifXhNF1M+/npeDM5YI9g9DxtchxoxgULY99BeehkMyVJc1U76O1/AY+jaRD6N5p65H6aR2r2Bradqfz10elJ1a+eGq2egKZn16Fj7Ne6QtJEjOd39w7A3VX1sz6xP+6KmajX0DwKdjQbe6r+LMkHgVPapzU2Z0uuH6B5bK/XIK4f+h2jDeNYt3W77Jzz92Bjg00/nXP+qHVKVd2fZFwNM0kOp3ki5naaJ13W0txYh6aRa7R6dLzHfbw+STP+1//QjEF8U7u/xTQ9BbvzsaXfZefY9ntCqNtDacYh0xawQUtzUeekvYieH+ntoxHb0v/EPl630zxL3s+jJ7Hfbp07Pf1OyKM1xo322Yv6pP2M5tnvh1bVPVuePUma9/6IplfWG6vq9O4VSd5H06A1lSZST0D/umK0egLg5VX1wS3MmyQNws9oHt/eoU+j1qKumI4tOi9W1c+BM4AzkuxC81jZccDraK6TXzuJvPfq9Bh6FJveRBjU9cNkdI7jx6rqyDEjG6PWKe3ji49kY4+psfwlTQPWb1TXJCPtzIKvH8f2k5bkqTSNWRcDB3b3uEvyQjZtiNrS77JzbH+3qi6bfI7VzTG0NBdd1S6f1Wfd3j0xE3E1TbfQX+mz7hlbsJ/7GP0Owq3tcuc+654ySp52SrJbn3V79Un7r/azn7qZPEqS+utMyX5hd2L7I7xfXdAZo2VL7hyPtc2t7XJL6gloJjDpNVo9Ac04WpI0E65ql+P9TX9ru3zQebFtYBmr11FnDKWP0jRqrQcO2bKsblbnHPzbfdZtyfXDVPkWTUPNU9sZ3TdnrDrl6Yy/48yTgG/VpjPmPoXmUbzp0KnPL+rz+Ogz+8Rv6XdpfTqFbNDSXPRxmouAE5M80JMqyUNppoWFjYNOTsQn2uVb2wuXzv73Avbfgv2sAx7ZPrPd6+vt8si2Eu58xu/RPMrS69x2+bae+Gez6fhZAO+lOUZ/n2STZ9OTLErya+MphCTNU9e3y94fu39K/wunzqC3/RqgRjPqNm1vhe8AeyV5Yie9rfdO6bOvFTQXK8cmWdIV/7B+8VX1FZof4UcnObh3fZKt23pPkqZK5/f6m5Ns30lM8mia3jsbaH73d3R+P7+0Zz+vAZ7YnZBk2yT7sKkdaZ7muKvPusk4n6YH2Yndk2m05+xB9gSbkHYA9H8AfgU4rV+jVpLf6voe/oPm8cDfb3s4dWK2At66BR99PfDL7Xfa2ccODGag/C3JA/TU50meDhzbJ35Lv8vPAD8ATkrytN6VSR6S5Lc22Urj4iOHmnOq6jtJ3gj8NXBNkguAe4HnAUuA91bV5ZP4iHNoKsrDgSVJvkAzUOALgc/RdFntbd3vZyXN7CefTfIfNFPNXlRV19AMMvhVmsaof0/ynzSDMf5u+xnP7dnXR4A/AH4P+GqSzwO/1ObpojZPD6iqa5K8GngP8O0kn6OplBa2n7MXzfhaD5qJQ5L0gH8C3gC8p70ouoHmnP50+px3geuAG4EXJbmD9lGM3tmkelxOM4HH25L8Kk2D1PVV1bmAexfNeXxVW9dtCxxMc6Hxy907qqpbk/wpTR329STn0Vyw/T6wGujXw/clNHXViiRfpukJsQF4As1d+XU4gYikKVJVI+0j3McB1yb5NLANzW/wRwOv7+nZ82maMbH+KMnjgGtoevrsSXM+7X4U/CHAF5N8B/gKTYPDTjRjBm4N/M2Ay/KtJGfRNMRdk+RfaHrfHk7TEPdLjO/6YSq9iaYeOwl4XnveX0czVuQymsauxwJ3VtV9Sf6YZoD4y5N8gmb8pwNprrtuHOdnvoem8erKJJ+kqccOAH7UvqbDV2gmN3lRO9nLV2kaQA+huRl0WHfwln6XVXVPkhfQjLV5RZJLgW/StMUspvm7XMWWdYxQyx5ampOq6u00J5/v0Az0eBxwC3BsVb1qkvveQHPC+Tua8VP+lOZu/EvZOIvieGbWOA34EE0leypwOvCb7WcUzRSvnwB+rc3/DjS9s77au6O2e+zBNNMG/xLN3YGlNBcj/zJKOd5H03B1MU1X7j+jOXE/hObOysf7bSdJgqr6Ac05eQTYD3g5cA9NQ88ms/61dcfzaX7sHk1zzj+9N65nm2vb/f4MOKGNP7Zr/d/TjPWyvk3fh2Y8mBNG2d+HaG50XA8cQzNN+KeAF4wS/12ai8EzaMZDeTnwMprGrwuBSdWnkjQOr6I5v91C83v4aODbwGFVdWZ3YFXdCfwOzU2FvYBXAHfSPAa2tme/d9A03KyhOZe/jua39FXAflM0duBJNL/R76Ip1yE0T010ejRt0cx8g9YOgv8c4E9oHt98IU3vtmfQ3OT+Q7oGLa+qS2jqv6torjmOoenZ+zs0jVrj8e72835G8z0fSNOj6Tk0N/unXFXdR3Pd9VGaGQ7/hKYTxNE0DW79bNF32fZ63rON+eV2mz9oP+cfGXuiLo0hm87uKWmikvwTcCSwe1V9c6bzI0mSJGn2SvJHwAeBV1XVe2c6P5o4v8vpZ4OWNAFJHltVN/ak7UXTQ2sN8Kvlfy5JkiRJPDDG0rq2R1An7bE0j4k/AVhSVdePtr1mD7/L2cMxtKSJ+UCSX6LpVvszmjFEOmNnvcbGLEmSJEld/hB4TZIvAj+mGbrkQOARwGk2gAwVv8tZwgYtaWL+mea5/OfTnLhuoxms/fSq+s+ZzJgkSZKkWeffaQYA349mAPp7aQau/4eqemAG9iRH0QwWvjkfqaq1A8/lHJBkT5rJsjbnqqr6zAQ+YlzfpaaejxxKkiRJkjQLJBnhwTMyjmafqhqZ2twMp7ZR8MPjCP1oVR01tbnRVLJBC3jkIx9ZixcvnrL933HHHTz0oQ+dsv3PNMs33CzfcButfF//+tdvrqpHzUCW5q2prktmi7n+f6rXfCrvfCorWN7xsC6ZfpurS+b63+1cLp9lG05zuWwwPeWbyrrERw6BxYsX87WvbTLD9sCMjIywfPnyKdv/TLN8w83yDbfRypfk+9Ofm/ltquuS2WKu/5/qNZ/KO5/KCpZ3PKxLpt/m6pK5/nc7l8tn2YbTXC4bTE/5prIuWTBVO5YkSZIkSZKmgg1akiRJkiRJGio2aEmSJEmSJGmo2KAlSZIkSZKkoWKDliRJkiRJkoaKDVqSJEmSJEkaKlvNdAYkTZ3FJ13UN33tGQdOc04kDaN+55Djl27gKM8tkiRJc9IwXUPaQ0uSJEmSJElDxQYtSZIkSZIkDRUbtCRJkiRJkjRUbNCSJEmSJEnSULFBS5IkSZIkSUPFBi1JkiRJkiQNFRu0JEmSJEmSNFRs0JIkSZIkSdJQsUFLkiRJkiRJQ8UGLUmSJEmSJA0VG7QkSZIkSZI0VGzQkiRJkiRJ0lAZWINWkucn+WSS7ye5K8l1SU5P8vCeuIVJPpjk5iR3JLksydI++9suyZlJbmz3tyrJs/rELUhycpK1Se5OcnWSwwZVLkmSJEmSJM0ug+yhdQJwH/BGYH/gfcBxwOeTLABIEmBFu/7VwGHA1sDKJI/r2d85wMuBNwEHATcClyTZsyfurcCpwHuAA4ArgAuSPHeAZZMkSZIkSdIssdUA93VwVd3U9f7yJOuAjwLLgS8ChwB7AftW1UqAJKuANcDrgde0aXsALwGOqaoPt2mXA6uBt7T7IcmjaRrSzqiqs9rPXZlkV+AM4HMDLJ8kSZIkSZJmgYH10OppzOr4arvcuV0eAvyo05jVbncbcCFwaNd2hwA/B87vitsAnAfsl2TbNnk/YBvg3J7PPRdYmmTJxEojSZIkSZKk2WqqB4Xfu11+q13uDlzbJ241sEuSh3XFramqO/vEbQPs2hV3D/CdPnEAu00w35IkSZIkSZqlBvnI4YMk2Znm8cDLquprbfJOwNo+4eva5UJgfRt3yxhxO3Utb62q2kxcv/wdCxwLsGjRIkZGRkYLnbT169dP6f5nmuWbvY5fuqFvend5hrl842H5JEmSJGnumZIGrban1WeBDcDR3auA3sanTnrv+0HGbaKqzgbOBli2bFktX758c5tM2MjICFO5/5lm+Wavo066qG/62iOWP/DvYS7feFg+SZIkSZp7Bv7IYZLtaGYyfCKwX1Xd0LV6Hf17TS1sl7eMM25d13JhO3viWHGSJEmS5okkj0vy7iSrktyZpJIs7hO3MMkHk9yc5I4klyVZ2iduuyRnJrkxyV3tfp/VJ25BkpOTrE1yd5Krkxw2RcWUpHltoA1aSbYGPgk8DXhuVV3TE7KaZtyrXrsB11fV+q64JUm27xN3LxvHzFoNbAs8qU8cwDe3uBCSJEmSht2uwOE0N8y/3C+gvSm+AtgfeDVwGLA1zazpj+sJPwd4OfAm4CDgRuCSJHv2xL0VOBV4D3AAcAVwQZLnTrpEkqQHGViDVpIFwMeAZwOHVtUVfcJWADsn2btrux2Ag9t13XFbAy/oitsKeCFwaVXd0yZfTNPAdUTP5xwJXFtVayZVKEmSJEnD6EtVtaiqngtcMErMIcBewB9U1Seq6uI2bQHw+k5Qkj2AlwB/VlUfqKov0DSWXU8zZnAn7tHACcAZVXVWVa2sqlcAK4EzBl9ESZrfBjmG1t/TNEC9DbgjydO71t3QPnq4AlgFnJvkRJo7JifTjHn19k5wVV2V5HzgXW2vrzXAccASuhqvquqnSd4JnJzkduBKmkavfYFDB1g2SZIkSUOiqu4fR9ghwI+qamXXdrcluZDmWuI1XXE/B87vituQ5DzgpCTbtjfc96OZkf3cns85F/hQkiXecJekwRnkI4cHtMtTaBqtul8vgwcqloOAzwPvBT4N3AfsU1U/6Nnf0cCHgdOAi4DHA/tX1ZU9cae0Ma8FLgGeCRxeVRcOsGySJEmS5pbdgWv7pK8GdmknuurEramqO/vEbUPzeGMn7h42Do/SHQcbh0WRJA3AwHpoVdXiccatA45pX2PF3QW8rn2NFXcfTYPWaePKqCRJkiQ1k1Ct7ZPemVhqIbC+jbtljLidupa3VlXvLOy9cQ+S5FjgWIBFixYxMjIyaobXr18/5vphN5fLZ9mG01wuG/Qv3/FLN/SNnY3HYZCPHEqSJEnSsAjQ2/jUSZ/KuAepqrOBswGWLVtWy5cvHzV2ZGSEsdYPu7lcPss2nOZy2aB/+Y466aK+sWuPWN43fSYNdJZDSZIkSRoS6+jfa2phu7xlnHHrupYL29kTx4qTJA2ADVqSJEmS5qPVNONe9doNuL6q1nfFLUmyfZ+4e9k4ZtZqYFvgSX3iAL456RxLkh5gg5YkSZKk+WgFsHOSvTsJSXYADm7XdcdtTTOjeyduK5rZ1S9tZzgEuJimgesIHuxI4FpnOJSkwXIMLUmSJElzTpLnt//8zXZ5QJKbgJuq6nKahqpVwLlJTqR5xPBkmjGv3t7ZT1VdleR84F1JtgbWAMcBS+hqvKqqnyZ5J3ByktuBK2kavfYFDp26kkrS/GSDliRJkqS56IKe9+9tl5cDy6vq/iQHAWe167ajaeDap6p+0LPt0cDbaGZW3xG4Gti/qq7siTuFZmbE1wKPAa4DDq+qCwdSIknSA2zQkiRJkjTnVNWYswu2MeuAY9rXWHF3Aa9rX2PF3UfT6HXa+HMqSZoIx9CSJEmSJEnSULFBS5IkSZIkSUPFBi1JkiRJkiQNFRu0JEmSJEmSNFRs0JIkSZIkSdJQsUFLkiRJkiRJQ8UGLUmSJEmSJA0VG7QkSbNGkucn+WSS7ye5K8l1SU5P8vCeuIVJPpjk5iR3JLksydI++9suyZlJbmz3tyrJs/rELUhycpK1Se5OcnWSw6ayrJIkSZImzgYtSdJscgJwH/BGYH/gfcBxwOeTLABIEmBFu/7VwGHA1sDKJI/r2d85wMuBNwEHATcClyTZsyfurcCpwHuAA4ArgAuSPHewxZMkSZI0CFvNdAYkSepycFXd1PX+8iTrgI8Cy4EvAocAewH7VtVKgCSrgDXA64HXtGl7AC8BjqmqD7dplwOrgbe0+yHJo2ka0s6oqrPaz12ZZFfgDOBzU1ZaSZIkSRNiDy1J0qzR05jV8dV2uXO7PAT4Uacxq93uNuBC4NCu7Q4Bfg6c3xW3ATgP2C/Jtm3yfsA2wLk9n3susDTJkomVRpIkSdJUsUFLkjTb7d0uv9Uudweu7RO3GtglycO64tZU1Z194rYBdu2Kuwf4Tp84gN0mmG9JkiRJU8RHDiVJs1aSnWkeD7ysqr7WJu8ErO0Tvq5dLgTWt3G3jBG3U9fy1qqqzcT1y9+xwLEAixYtYmRkZLTQoXT80g2bpC16SP90YM6VH2D9+vVzslz9zKeyguWVJGnY2aAlSZqV2p5WnwU2AEd3rwJ6G5866b3vBxm3iao6GzgbYNmyZbV8+fLNbTJUjjrpok3Sjl+6gXdc0//nw9ojlk9xjqbfyMgIc+17Hc18KitYXkmShp2PHEqSZp0k29HMZPhEYL+quqFr9Tr695pa2C5vGWfcuq7lwnb2xLHiJEmSJM0SA+2h1U6X/gZgGbAH8BBgSVWt7YpZTDMTVT8Lq+rWrtjtaKZSPxLYEbgKeENVfanncxe0n/sK4DHAdcBbquqTky+VNPst7tOLQhpWSbYGPgk8DfidqrqmJ2Q18Jw+m+4GXF9V67vinpdk+55xtHYD7mXjmFmrgW2BJ/HgcbQ6Y2d9c6JlkSRJkjQ1Bt1Da1fgcJq741/eTOzpwDN6Xrf3xJwDvBx4E3AQcCNwSZI9e+LeCpwKvAc4ALgCuCDJcydYDmlOW3zSRQ+8rvnhbTaIadZob1B8DHg2cGhVXdEnbAWwc5K9u7bbATi4XdcdtzXwgq64rYAXApdW1T1t8sU0DVxH9HzOkcC1VTXaTRhJkiRJM2TQY2h9qaoWASR5Gf3voHd8b5QLFdrt9wBeAhxTVR9u0y6nuZP+Fprp2EnyaOAE4IyqOqvdfGWSXYEzgM9NrkiSpGn09zQNUG8D7kjy9K51N7SPHq4AVgHnJjmR5ibKyTRjXr29E1xVVyU5H3hX2+trDXAcsISuxquq+mmSdwInJ7kduJKm0Wtf4NApK6kkSZKkCRtoD62qun+AuzsE+Dlwftf+NwDnAfsl2bZN3o9m+vVze7Y/F1iaZMkA8yRJmloHtMtTaBqtul8vgwfqmoOAzwPvBT4N3AfsU1U/6Nnf0cCHgdOAi4DHA/tX1ZU9cae0Ma8FLgGeCRxeVRcOsnCSJEmSBmMmZzk8Pck/AHcAlwOn9IyTsjuwpmfcE2h6aG1D83jj6jbuHh487kknDpoxUHxcRJKGQFUtHmfcOuCY9jVW3F3A69rXWHH30TRonTaujEqSJEmaUTPRoHUP8H7gUuAm4MnAG4H/TPK0qvpWG7cTG2eq6raua31neWtV9U653hv3IEmOBY4FWLRoESMjI1teknFav379lO5/plm+mXf80g0T3nbRQ5rtZ3sZJ2oYvr/JmOvlkyRpKiUZAfYeZfUlVbX/VExqJUmavGlv0KqqG4E/7kr6cpKLaXpUnUJz8odmLJTeRqpOeu/78cT15uNs4GyAZcuW1fLlyzeb94kaGRlhKvc/0yzfzDtqEoO6H790A++4ZivWHrF8cBmaRYbh+5uMuV4+SZKm2CuBHXrSngH8DQ+eaASaSa160/pNanUgcCLwPeBVNJNaPaOqrhpEhiVJjZl85PABVfWDJP8OPLUreR2wS5/whV3rO8uFSdLTS6s3TpIkSZIeUFXf7E1L8nKa2W/P61k1kEmtJEmDMSsatFq9Pa1WA89Lsn3POFq70VQw3+mK2xZ4Eg8eR2u3drlJJSVJkiRJvZI8hGa23Qvb8Rq3RN9JrZKcB5yUZNuqumdwuZUEsLjP0yprzzhwBnKi6TbQWQ4nKskuNDNKfaUreQWwNU2F0onbimYq9Uu7KoOLaRq4juDBjgSurSoHhJckSZI0Hr8PPBz4aJ91pyfZkOS2JCuSLO1ZP55JrSRJAzLwHlpJnt/+8zfb5QFJbgJuqqrLk7yDpiFtFc2g8L8KnAzcD/xVZz9VdVWS84F3JdmaZiDG44AldDVeVdVPk7wTODnJ7cCVNI1e+wKHDrp8kiRJkuaslwI/Bf6tK23Qk1o9yJZMVjXXJ4OZy+WzbFOn3wRZg8rPTJdtqvUr32gTjs3G4zAVjxxe0PP+ve3ycmA5zR2K44CjaO5+3Ax8EXhzVV3Xs+3RwNtoplHfEbga2L+qruyJOwVYD7wWeAxwHXB4VV046dJIkiRJmvOS/BLwO8DfVtUDV3RTMKnVg2zJZFVzfTKYuVw+yzZ5/R4tbGzarDGoCa/m8vcG/cs32oRjs3ESsYE3aFXV5k7YHwI+NM593QW8rn2NFXcfTaPXaePMpiRJkiR1O5LmSZJ+jxs+yCQntZIkDcCsGENLkiRJkmbYS4Grq+rqccb3m9RqSZLte+J6J7WSJA2ADVqSJEmS5rUky2gGdd9s76w2fjKTWkmSBmAqxtCSJEmSpGHyUmAD8PHeFYOe1EqSNBg2aEmSJEmat9rGpxcDF1fVT/qETMWkVpKkSbJBS5IkSdK8VVU/Bx41xvqBT2olSZo8G7QkSZIkSfPS4pMu2iRt7RkHzkBOJG0pB4WXJEmSJEnSULGHliRJ6nuHWpIkSZqt7KElSZIkSZKkoWKDliRJkiRJkoaKDVqSJEmSJEkaKjZoSZIkSZIkaajYoCVJkiRJkqShYoOWJEmSJEmShspWM50BSZIkSZJmi8UnXdQ3fe0ZB05zTuan0Y6/1MsGLY3KE7kkSZIkSZqNbNDSFutt6Dp+6QaOOukiG7okSZIkSdK0cAwtSZIkSZIkDRV7aEmSJM1RDh8gSZLmKntoSZIkSZIkaajYoCVJkiRJkqShYoOWJEmSJEmShspAG7SSPC7Ju5OsSnJnkkqyuE/cwiQfTHJzkjuSXJZkaZ+47ZKcmeTGJHe1+31Wn7gFSU5OsjbJ3UmuTnLYIMsmSZIkSZKk2WHQg8LvChwOfB34MvCc3oAkAVYAS4BXA7cAJwMrk+xZVTd0hZ8DHAicCHwPeBVwSZJnVNVVXXFvBU4ATmk/+0XABUkOqqrPDbSEkiRJkiQNAScH0Vw26AatL1XVIoAkL6NPgxZwCLAXsG9VrWxjVwFrgNcDr2nT9gBeAhxTVR9u0y4HVgNvafdDkkfTNGadUVVntZ+xMsmuwBmADVqSJEmSJElzyEAbtKrq/nGEHQL8qNOY1W53W5ILgUNpG7TauJ8D53fFbUhyHnBSkm2r6h5gP2Ab4NyezzkX+FCSJVW1ZsKFkiRJkiRJAzdaDzJpPGZiUPjdgWv7pK8GdknysK64NVV1Z5+4bWgeb+zE3QN8p08cwG6TzrEkSZKkOSfJ8nbc397XrT1xAx0DWJI0eYN+5HA8dgLW9klf1y4XAuvbuFvGiNupa3lrVdVm4h4kybHAsQCLFi1iZGRkHFmfmPXr10/p/qfK8Us3jCtu0UOa2GEs43gMw/c33u+qH7+/4TbXyydJ0jR5DfDVrvcP/LiaojGAJUmTNBMNWgF6G5866VMZ9yBVdTZwNsCyZctq+fLlY4VPysjICFO5/6ly1Di7fx6/dAPvuGYr1h6xfGozNEOG4fsb73fVj9/fcJvr5ZMkaZp8q6quGGXdQMcAliQNxkw8criO/r2mFrbLW8YZt65rubC9czJWnCRJkiRtqb5jAAOdMYC74zYZAxg4D9gvybbTk11Jmh9moofWavrPfrgbcH1Vre+Ke16S7XvG0doNuJeNY2atBrYFnsSDx9HqjJ31zUFlXNLYnBZYkiQNqY8leSRwK3AJcFJVXd+uG2sM4JcmeVh7DTOeMYBXo6Hlb11pdpmJBq0VwNFJ9q6qywGS7AAcDHy8J+7NwAuAj7ZxWwEvBC5tZzgEuJimgeuINr7jSOBaZziUJEmSNIrbgHcAlwM/A54CvBFYleQpVfVTBj8G8INsydi+c33szJko32TGou0YT55n6rsbrXyDzMt4ynbND2/rm378JlMrDMagyjcf/89Nx9/MoAy8QSvJ89t//ma7PCDJTcBNbQPWCmAVcG6SE9k4qGKAt3f2U1VXJTkfeFeSrWmeUT+OZjDGI7rifprkncDJSW4HrqRp9NqXB3cBljQG7zhJkqT5pqq+AXyjK+nyJF8C/otmbKw/ZxaN7TvXx86cifJNZizajvGMRTsdZev/e77/Jf8gx88dT9kGcZy3xKDKNx//z432Xc3GMZenoofWBT3v39suLweWV9X9SQ4CzmrXbUfTwLVPVf2gZ9ujgbcBpwE7AlcD+1fVlT1xp9DcFXkt8BjgOuDwqrpwICWSJEmSNC9U1ZVJvg08tU3akjGAdxkjzrF9pWnizfr5YeANWlU15h2INmYdcEz7GivuLuB17WusuPtoGr1OG39O1W20//CSJEnSPNTd22rQYwBLkgZgJsbQkiRJkqRZKcky4FeAf26TBj0GsGaAN/CluWfBTGdAkqRuSR6X5N1JViW5M0klWdwnbmGSDya5OckdSS5LssnQokm2S3JmkhuT3NXu91l94hYkOTnJ2iR3J7k6yWFTVExJ0iyQ5GNJTkvy+0n2TXI8zaRTPwTe3YZ1jwH8oiT7tWmbjAEMdMYAflmSZwPn0YwB/JfTVihJmifsoaWB6XfXw2eUJU3ArsDhwNeBL9PnMY8kobmYWAK8mo0TjKxMsmdV3dAVfg5wIHAi8D3gVcAlSZ7RXnx0vBU4gWZcxq8DLwIuSHJQVX1uoCWUJM0W1wIvpqlLtgd+DHwK+MuquhlgisYA1hzidZA0M2zQkiTNNl+qqkUASV5G/3FLDgH2AvatqpVt7CqaGXFfTzMzFUn2AF4CHFNVH27TLqcZ5+Qt7X5I8miaxqwzquqs9jNWJtkVOAOwQUuS5qCqOh04fRxxAx0DWJI0eTZoSZpy3rXSlqiq+8cRdgjwo05jVrvdbUkuBA6lbdBq435O8whIJ25DkvOAk5Js245psh+wDXBuz+ecC3woyZKqWjPhQkmSJA3IIMYD8/e55gLH0JIkDaPdaR4T6bUa2CXJw7ri1vTMNtWJ24bm8cZO3D1sOgPV6na526RzLEmSNIcsPukiFp90Edf88LYH/i1NJ3toSZKG0U7A2j7p69rlQmB9G3fLGHE7dS1vraraTNyDJDkWOBZg0aJFjIyMjCPrs9PxSzeMK27RQ0aPHebyj2b9+vVDXa4t+a6GvaxbyvJK0oON1iBlzy3NVjZoSZKGUYDexqdO+lTGPUhVnQ2cDbBs2bJavnz5WOGz2lHjvKt6/NINvOOa/j8f1h6xfIA5mh1GRkaYi99rv+9q2Mu6pSyvpNluGHs8DWOeNbxs0JIkDaN19O81tbBd3tIVt8sYceu6lguTpKeXVm+cJEnSZvU27By/dMOYN49mcy8oG6k0W9mgJUkaRqvpP/vhbsD1VbW+K+55SbbvGUdrN+BeNo6ZtRrYFngSDx5HqzN21jcHlXFpNuh3cfKR/R86AzmRJEmaGAeFlyQNoxXAzkn27iQk2QE4uF3XHbc18IKuuK2AFwKXtjMcAlxM08B1RM/nHAlc6wyHkiRJ0uxiDy1J0qyT5PntP3+zXR6Q5Cbgpqq6nKahahVwbpITaR4xPJlmzKu3d/ZTVVclOR94V5KtgTXAccASuhqvquqnSd4JnJzkduBKmkavfYFDp66kkiRJkibCBi1J0mx0Qc/797bLy4HlVXV/koOAs9p129E0cO1TVT/o2fZo4G3AacCOwNXA/lV1ZU/cKTQzI74WeAxwHXB4VV04kBJJA9LvccHZPPaKJEnSVLBBS5I061TVmLMLtjHrgGPa11hxdwGva19jxd1H0+h12vhzKkmSZhsHMZfmB8fQkiRJkiRJ0lCxh5YkSZIkSTPIXmXTw8f25xZ7aEmSJEmSJGmo2KAlSZIkSZKkoeIjh5IkSZIkSRM02iOjPs44tWzQkiRJkiRJ85KNUcPLBi1JkiRJkqQBs7FsatmgJUnSLOQsPJIkSXOTv/MGY0YatJIsB1b2WXVbVe3YFbcQOBP4PeAhwCrgz6rqmp79bQe8FTgS2BG4CnhDVX1p0HnX1LH1WpIkSZIkjcdM99B6DfDVrvcbOv9IEmAFsAR4NXALcDKwMsmeVXVD13bnAAcCJwLfA14FXJLkGVV11ZSWQBMyWuOVJEmSJEnS5iyY4c//VlVd0fX6Wte6Q4C9gD+oqk9U1cVt2gLg9Z2gJHsAL6HpufWBqvoCcDhwPfCWaSuJJEmSpKGS5PlJPpnk+0nuSnJdktOTPLwrZnGSGuW1Y8/+tktyZpIb2/2tSvKsaS+YJM0DM92gNZZDgB9V1QOPJlbVbcCFwKE9cT8Hzu+K2wCcB+yXZNvpya4kSZKkIXMCcB/wRmB/4H3AccDnk/ReK50OPKPndXtPzDnAy4E3AQcBN9I8ObLnFOVfkuatmX7k8GNJHgncClwCnFRV17frdgeu7bPNauClSR5WVevbuDVVdWefuG2AXdt/S5IkSVK3g6vqpq73lydZB3wUWA58sWvd96rqitF21PXkyDFV9eE27XKaa5G30NyIlyQNyEw1aN0GvAO4HPgZ8BSauyKrkjylqn4K7ASs7bPtuna5EFjfxt0yRtxOg8u2JEmSpLmipzGrozPG785buLu+T44kOQ84Kcm2VXXPxHKqfhyXV1Np8UkXcfzSDRzl39msNSMNWlX1DeAbXUmXJ/kS8F80A8X/ORCg+myePu/HE/fglcmxwLEAixYtYmRkZFx5n4j169dP6f4H4filGzYfNIpFDxl9+9HKvSWfN9PHbj5/fwDv/thn+6Yv3fkRk8rHdB3TYfj+JmOul0+SpBmwd7v8Vk/66Un+AbiD5sb8KT2zr/vkiKQJG62Bdu0ZB05zTobHTD9y+ICqujLJt4Gntknr6N+7amG7vKUrbpcx4tb1WUdVnQ2cDbBs2bJavnz5BHI9PiMjI0zl/gdhMq3Oxy/dwDuu6f+ntPaI5ZP+vNH2MV3m8/c3ln7fy2z8Xofh+5uMuV4+SZKmU5KdaR4PvKxrwqp7gPcDlwI3AU+mebrkP5M8rao6DV8TenJkS260z/UbWVtavsnc1J1um7uJPMws29Sa7s43W9pZZSbNmgatVndvq9XAc/rE7AZc346f1Yl7XpLte+6G7AbcC3xnqjIrSZI0G/jYjTR5SR4GfBbYABzdSa+qG4E/7gr9cpKLaa5DTgGO7OyCCTw5siU32uf6jawtLd8wPQo20ZvIw8CyTa2p7AjQ7//caP+vZrqjST+z5q8uyTLgV4B/bpNWAEcn2buqLm9jdgAOBj7etekK4M3AC2gGbyTJVsALgUt9Tl2SJEnSWJJsR3Nd8URg76q6Yaz4qvpBkn9n49MlMMEnR+arfg3xxy/dwPLpz4qkITUjDVpJPgasAa6kmeHwKcDJwA+Bd7dhK4BVwLlJTqTpvnsyzR2Ot3f2VVVXJTkfeFeSrdv9HgcsAY6YjvJIkqT+FyeO+yBptmuvIT4JPA34nZ5xscbclAf3yPLJkQGwLpE0XjPVQ+ta4MXAq4HtgR8DnwL+sqpuBqiq+5McBJwFvBfYjqaBa5+q+kHP/o4G3gacBuwIXA3sX1VXTn1RJEmSBA5oq+GTZAHwMeDZwIFVdcU4t9sFeCbw6a5knxyRNHDWraObqVkOTwdOH0fcOuCY9jVW3F3A69qXJEmSJI3H39M0QL0NuCPJ07vW3VBVNyR5B7CA5ub6TcCv0jw5cj/wV53gYX9yZDb3jHKcQEn9zJoxtCRJkiRpmh3QLk9pX93eDJxK8yjhccBRwMOBm4EvAm+uqut6tvHJEUmaJjZoSZIkSZqXqmrxOGI+BHxonPvzyRFJmiY2aEmaELt+S5IkaXP8zShpqtigJY2Tg/FJkrSRF6mSenlekKbPbB73brrYoCXNMBvKJGn+8JwvaabZ6CRprrBBS7OeP/4laXC8kNForvnhbRzl3V5JkjQkbNDSlPLCSZIkSZpb/I0vaTawQUuaRlb+kiRJkqSpMN+ebrJBS5okB+OTNBk2dEuSJElbzgatecYLJ0mSJEnj4bWDpNnMBi1pyPjDQpIkSZI039mgJUmSpFF5I0WSpOE2Wl1+/NINfWc4HhY2aGlemG+D40mSJEmSNJctmOkMSJIkSZIkSVvCHloaWj4CIUmSJEnS/GSDliRJQ65fA/9seaR6vj7yvaU3XWbzdyhJkjQb2aAlSdKQsGeqJEmS1HAMLUmSJEmSJA0Ve2hJPXp7QMzUVKb2xJCkLTdfH3GcS/wOJUnSeNigNUfZGDKzPP6SNPXm+rl2rpcP5kcZJUnS1LBBS9KM8A68pLnCRpnN8xhJU8v/Y5LmIxu0NK9Z+UvSzNiS868N3ZIkSeplg5YkSdI4eBNkZvU7/qM1dk5lL+AtyYckSZo6c6ZBK8njgXcCvwsEuAz406q6fkYzJmkgpuoCwkcf1W0u1SU2vkjSzJhLdYkkzWZzokEryfbAF4F7gD8ECjgNWJnk16vqjqn43PFeLBy/dAPLpyID0jxnY5QGaabqEm3edDfOdX/eTM10q+FifaQO6xJJmj5zokELeDnwROBXq+o7AEn+G/hf4BXA38xg3gbGu+2aD/w71wyaF3XJfGWDw9w0iDpjuh9PHNS+NWtZl0jSNJkrDVqHAFd0Kg2AqlqT5D+AQ5kFFceW/KDxgl6aeV6EzEuzvi7R4Fnnzi9b0vvOvw1NkHWJJE2TudKgtTvw2T7pq4EXTHNetog/lqTBm82PC9lQNqsNbV0iabhM9+8/65hpZV0iSdMkVTXTeZi0JPcCf1NVJ/WknwacVFWbNNwlORY4tn37q8B1U5jFRwI3T+H+Z5rlG26Wb7iNVr4nVNWjpjszw2wI6pLZYq7/n+o1n8o7n8oKlnc8rEu20DTUJXP973Yul8+yDae5XDaYnvJNWV0yV3poQTPgYq+MGlx1NnD21GWnKxPJ16pq2XR81kywfMPN8g23uV6+GTBr65LZYr79zc2n8s6nsoLl1ZSasrpkrn+Pc7l8lm04zeWywfCXb8FMZ2BAbgF26pO+sF0nSdLmWJdIkibLukSSpslcadBaTfO8eq/dgG9Oc14kScPJukSSNFnWJZI0TeZKg9YK4OlJnthJSLIYeGa7bqbN9cdRLN9ws3zDba6XbzrN9rpktphvf3PzqbzzqaxgeTU1proumevf41wun2UbTnO5bDDk5Zsrg8I/FLgauAv4c5rn1t8KPBz49apaP4PZkyQNAesSSdJkWZdI0vSZEz20quoOYF/g28A/AR8D1gD7WmlIksbDukSSNFnWJZI0feZEDy1JkiRJkiTNH3Oih9ZskOR1SS5McmOSSnJqn5jHJjk9ydeS3JbkpiRfSPKszez7iUnubPe765QVYuw8DKx8SXZI8qYk/5nk/5Lc2v7796arPL2m4vtL8ntJvpHk7iTfT/LnSX5hygvTPy+bLV8b94dJPtnmt5J8ZJS4X0jyZ0muTXJHu99PJ/n1qSzHaAZdvjb2IUlOTfK/Se5J8pMk/5pkm6kqxxh5GXj5uraZ8fOLZo8k+yX5YpIft3/3NyT55yS7jbHN+9u/n3OnM6+DsCXlTfL0JBe3ddYdSa5J8qKZyPdEjLesSXZP8qkkP2rLuTrJ8Um2mqm8D0r7/VWS03rSFyb5YJKb2zJflmTpTOVzUPqVN8mzk5yb5LtJ7mqX70vy6JnMq8aW5PFJ/qX9/fmz9v/oLjOdry2V5HFJ3p1kVddvj8V94obu/2SS53f9BrsryXVprhse3hM3jGUbb/0xdGXrZy7VFUmWt2Xpfd3aEzd0ZeuwQWtwXg48GvjMGDG/CbwQ+CzwfOAo4G5gJMlBY2z3XuC2geRy4gZZvl2AVwKXA0e223wb+HSSVw064+M00O8vyX7AJ4GvAgcAf0szjsJfDTjf4zWe8kHzfTwJ+DzwszHi3gqc1e7vYOC17XYrkzxuknmdiIGWL8nWwL8BRwPvAH6X5m/2BmAmGiUH/f11mw3nF80eOwFfB/4EeA5wMs1sXVckeUJvcJLfBo5g/H9vs824ypvkQOBLwI+BlwCHAh8AtpvuDE/CZsua5JeAEeCJwJ/SnN8/A5wJvG26MzxISV4M7NEnPTQDde8PvBo4DNiamavPBmK08gJ/DPwicBpNmU8HDqH5O3jY9OVQ45Vke+CLwJOBPwT+APhlmr/Rh85k3iZgV+Bw4Bbgy/0Chvj/5AnAfcAbafL+PuA44PNJFsBQl2089cewlu1B5nBd8RrgGV2v3+msGPqyVZWvAbyABe1yK5rBH0/tE7MjsFVP2lbAdcCXRtnvS4Cf0PywLGDXYS8f8FBg+z7bfwG4ftjL16Z/A7i8J+1NwL3AY2Zj+brj2n/fAHxklLgfAR/vSXtyu+9XzIHynURzgf74mfh7nOrydcXMivOLr9n9An61/fs4vid9a+Bamh+2a4FzZzqvU1FemoGcfwq8a6bzNg1lPbZ9/ys9cecBN850fidRzh1pGiNf3JbvtK51h7Zp+3SlPQJYB/zdTOd9Csr7qD7xz2rjjpnpvPvq+32+lqahZNeutCXABuB1M52/LSxL9++Ul7V/d4t7Yoby/+Qo/7de2pZl32Eu2yjl7a0/hr5sc7GuAJa3+f6dMWKGsmydlz20BqSq7h9HzK1VtaEnbQNwFbBzb3yShcDf0LT43zqQjE7QIMtXVXdU1Z19dvE14JcmmdUJGWT5kjwe2BPoffzmn2guAA+YZHa32HjKtyVxwDZs2iPj1nY57eeVKSjfK4ELquoHE8/V4ExB+WbV+UWz3v+1y5/3pJ9I02PxHdObnSnXW94XAI9i7pUTNi1r55Hqfuf3Yf7N+HZgdVV9os+6Q4AfVdXKTkJV3QZcSPMjfxiNWt6quqlP/Ffb5Sa/RTUrHAJcUVXf6SRU1RrgPxiyv9Fx/k4Zyv+T4/y/NZRlG0Vv/TEXyjbf6oqOoS7bMP84mRPSjMfzDOBbfVa/Hfifqvqn6c3V4GymfL2eBfzP1OZosEYp3+7t8tru2PbHx53AqGPRDJH3AkcmOTTNmGhPbNNuAM6f2axNTjsmxeOB7yX5QDtWxd1pxkvbc4azN0hDf37R1EkzTt42SX4ZeD/NHcvzutY/ieYx6ldW1b0zlM2B2Ux596K5S7k0zbhZG5L8IMlfZobGRZyMzZT1AuBm4D1JlrTn9+fRPOI0lA16Sfai6SXxylFCdqenvm6tBnYZtsfwxlHefvZul+P5rabpN9bf6Fz4TdlrLv2f7P2/NdRl20z9Mexlm+t1xceS3Jdm/OqP58Fj8A112YZ+gM854FTgcTRjkDyg6z/VU2YgT4N0Kn3K1yvJscDTacYAGiansmn5dmqXt/SJv6Vr/dCqqjcluQf4FBsbxr8NLK+qdTOXs4Ho9BJ8A82dtRcB2wJvphkv7der6vqZytwgzKHzi6bOV2jGDQT4Ds3jEj/tWv8PwKe67+YNubHK+0vA9sDHacYP/DrN2BN/QfN4wp9Na04nb9SyVtVPkjyDZqzI77Uxncec3z7tOZ2kdjzE9wNnVdV1o4TtRPPIbK9OXbYQWD/43A3eOMvbu83DgXfRXHB/Zsoyp8nYif6/KdfR/H3ONXPi/2SSnYG3AJdV1dfa5GEv21h15dCWbY7XFbfR3JC6nKb39VNoxnlbleQp7fc3rGUD7KHVV5LfGWU2gN7XyCQ/5yU0Y/W8taq+3JW+Dc1/qndW1TcnVZj+nzuj5esTtxz4O+Cfqupjk/nMdn8zXb60y+q32WQ+s/3caSnfZvJwHE3vjNOAfWgeybkduDTNoMKT2fdMl69zXrwTOLiqPldVnwYOBB4CTGrigpku31SfXzRn/AHNTYaX0PwA+nzamaiSHAk8leZx1bli1PLSnBO2A95SVe+oqpGq+nOaQeFfleQRM5HhSRjru30UzY2KO2gmP9mH5jz/50neMCO5nZw30Jy3xxrQPkxRfT0DxlPeB6SZufITNI9Dvah3WAXNKnPlb3Q8hv7/ZNuj5bM045wd3b2K4S7bWHXlMJdtztYVVfWNqjqhqi6sqsur6l00g78vohkoHoa0bB320OrvP4FfG0dcv3GgxiXJwcBHgHOq6i97Vv8pTUvp3yXZsU3bvl0+PMnDq+r2iX42M1++7rin0syq8EXgjyb6eT1munyd1ux+PbF27Fo/UVNevrEk2Ql4J3Bmd9mTfJGmdf9EJtdjYUbLx8YxAf6je6y3qvpBkv9h8r2aZrp8f8rUnl80B1RV5/GIryT5N5r/2yclOYFm7LW/Bu7u+htaAGzdvr+jqnrH25rVRisvzYxwnXPC53s2u7RdvzvN/+uhsJmyvh5YDDyhqjo9QkbSPFr51iTnVNXN05zlCWkfpziFZuDpbZNs27V62/Zv9XaaOrlffd3p+dKvZ8ysM97yVtV9bfwC4KM0vQ0PrKr/nuYsa/xG692/kCH5+9xCQ/1/Msl2NNc2TwT2rqobulYPddk2U38MZdnmW10BUFVXJvk2zc1JGPKy2aDVR3sRO2VjOSV5Ns04FZ8GXtEnZDfgMcAP+6y7EriaZtDxCZkF5evELQUuoRlU/bBBXQDNgvKtbpe7A6u6tltM03AwqV4xU12+cfgVmkfwvtqdWFXrknyX8TXWjGoWlO97wF2Mfqdi3AOv9zMLyjel5xfNPVV1a5Lv0Ey3/kiaAdL/qn11ezzNdOzPY4gfXeopL2w8p/eeEzp3Lid1TphJfcq6FPhOV2NWx3/RTGqyK80YW8PgiTQ963onaIGmd+EJNDcoVtNMQ99rN5qZl2ftYxY9xlveq9q0fwBeCDy/qr4wHRnUhK1m4/is3XZjkr8pZ6mh/T/ZPrr2SeBpNLPKXdMTMrRl6zVKXTmMZZtvdUVHd6+soS6bjxxOs2wcm+ILwJGjzPZxBk03/+7XX7frjqRpQZ6Vxlk+0gwm+HmaxoODququ6cvlxI2nfO34Slez6bhhR9LMBPJvU53PKfbjdvm07sS259au9G8oGRptw+pFwP9L8tBOensH51fpacgbQkN7ftHMSLIIeDLwXZr//71/P/sAPwEua//97zOT08HoKS9sbJzbvyd0P+Bu+g+kOhT6lPXHwK5pZkHt9lvtcpjO71fR/28VmguXfWjGgFkB7JykM3gzSXYADm7XDYurGF95SfIOmnP90VX1menOqLbYCuDpaSbgAR64SfpMhutvdLyG8v9k2+vxY8CzgUOr6oo+YUNZtn761B/DWrarmF91BUmW0XRQ+EqbNNRls4fWgLR/GIvZ2Ei4W5Lnt//+XFXdmeTJNBfKNwNnAr+ZbHw0tXPiq6r/oacHR9fzyV/pnrZ3ugyyfEkeTdOYtQ3wl+2+uj/uG1V1z9SVZlODLF/rjcC/Jnk/zfgUT6EZc+pvq+rHTLPxlK+N242NM+Y8BHhCV9zlVXVTVa1N8q/AiUnupxlk8BdpHlXZFnjflBeoxyDL1/77L2l6JFzU/vDfrk27FXjPFBalrwF/f7Pu/KLZI8mnaXrq/TfN+Bi/QvMI8QbgHVV1NzDSZ7u7gZ9U1SbrZrPNlRegqq5N8hHgLe0Fy5U0j2m9jGYMxVl957JjPGWl6bVzBM14iGfSPG65nOYO9aer6gfTnO0Jq6pb6f+3CvD9zt9qkhU0vanPTXIizaMVJ9PcvR6agfC3oLxvAF4HfAj43yRP7wq/qaq+27sPzbgPAH8CfDbJn9P0qngr8AOaMTGHStfvks7g4gckuYnm7+9ymgvoYfw/+fc0Y8q+Dbij5//WDe2jh0NZtnHWH0NZtrleVyT5GLCG5vu7leaa9GSaG1TvbsOGsmwPqCpfA3jRjKdUo7wWtzFHjRFTm9l/Z9tdh718ND+OR43r7G9Yy9e1z9+n6al1D3A98CbgF2br99fGnTpG3PKuuO1pZvj6Js3gwTfSNPY9bS6Ur419GrCSZiyr22h6acza/39bWr6e/Xf+tmekfL5mz4tmYNSv0/zouRO4juaCafFmtlsLnDvT+Z+q8tLcgDmN5gLyXppZXV870/mforI+Hfhce16/g+ZRhD8HHjLTZRjQcSjgtJ60nWgaeNa1x+YLwB4zndepKC/Nhdto9cRHZjq/vkb9HneheZTtZzTj+Xxmc+fl2foa4+9vpCtm6P5PtvXgaGU7dcjLNt76Y+jKNkaZ50RdQdMw9d801zI/b3/HnA08dtjL1nmlLYAkSZIkSZI0FBxDS5IkSZIkSUPFBi1JkiRJkiQNFRu0JEmSJEmSNFRs0JIkSZIkSdJQsUFLkiRJkiRJQ8UGLUmSJEmSJA0VG7QkSZIkSZI0VGzQkiRJkiRJ0lCxQUuSJEmSJElDxQYtaZySrE2ydqbzIUmafpOpA2aq/kiyOEkl+ch0f7YkaXZKsrytG06dws8YSVJTtX+pwwYtzTqDOsn6Q16Sho91gCRJs9t0NIpJ47HVTGdAkiRJkiTNGS8Ftp/pTGjus0FLkiRJkiQNRFVdP9N50PzgI4eaVdpuqyvbt3/ZdmWt7mewk/x6kk8muSnJPUm+neS0JA/tijkKWNO+/cPu/SRZ3Mb8SpIzk1yV5JYkdydZneTPk2w94HKtbV87JfmHJD9Kcl+S5e36rZOckOSaJHe1+bk4yf8bZX+Lk3wkyY1J7k3y/SR/l+SRfWKrfY79cUnOT7IuyW1JLkiyqI15epIvJrm9Pa5/k2Srnv0sSPKKJF9r83dHku8l+USS3Qd5vCTNT3O1DujK1y8m+XCSnya5M8l/JNlnlNjHtOf177Xl/EmSc5MsGSX+lUm+1Zbje0lOAX5hlFjrJEmahHQ9cpdkn/Z8fkd7zjpnlPPfHyX5ahv3syRfSnJon7hT230vT3Jc17l9TVtH9Z4PH4gfa1/jKNPz2vPy97rO/Zcl+Z3efbL5urrvGFpJHpXk3W09cW9bb3ykUzf3xHbqqoe3dcqN7XH4am+eNH/ZQ0uzzQiwGPhD4PL2/QPaH9OX0PxI/2fgBmA5cArwu0n2rqq7gauAvwVeC1wNfKZrN7e2y98Hjga+CFwGbAvsDbwVWAb83uCKBe3+vwhsA3yyLcPPkgT4F+AQ4JvAu4GFwAuBlUleXFUXdHaS5FeBfwd+sS3XdcBvAK8GDkzy9Kq6qeezFwJfpjleHwaeAjwfeHyS1wGfB/4NOBs4APgz4DbgzV37+GvgBOC/gY8C9wKPB/YFLgJWT+roSNLcrgO2oTnXbk1zHn4k8GLg0iQHVtWlXeX8ZZqyP4bm3PxJmvPt4cB+7Xn+u13xbwH+Avgh8A80x+c1wDPGyI91kiRN3jOAk4EVNPXWM4BjgKcneVpV3QGQ5N3AnwDfB95Pc+59AfCZJMdX1d/02fcJwP8DzgP+lea8/Fbgl2nqyUE7HbinLcePgcfS1IWXJHlBVX2qjRthjLp6NEkeBVwBPJHmPP9x4FdoHk88MMkzq+rbPZttDVwK7EBT7y+kqTs/l2RZVf33BMqpuaSqfPmaVS+ai5MCTu1JXwB8B7gf2Ltn3TntNm/qSlvcpn1klM/ZGdimJy3AB9rt9upZtxZYO8EyrW33eWGfz/zDdt0lwFZd6b8G3EFz8fXwrvSVbfwf9OznTW36h3rSq329vSd9RZt+C3BgV/pDgRuB/wO27kpfB3wV+IWe/WwFPGKm/258+fI1N15zvA64tOc8/1RgA01vsgVd6atoLiqe1bOfZwA/B/61K+2Xu/axU1f6Y4Gf9DsG1km+fPnyNblXV11VwEt71v1dm/6W9v3e7furgYd1xT2G5kbEz4EndqWf2sbfCTy5K3074L/adfv2iV/eJ5+brGP0enZJn+0X0dx8+M4o5T+1d5t2/QhQPWkfbrf5i570l7bpX+xJ79RVn+w5/3fqqffP9N+Br5l/+cihhslewJOAz1TV5T3r3kjz43/cdyuq6odVdW9PWgHva99ORVfWk3o/k415PrGqNnTl5Vs0F1aPoO0pkGQXmgrkG1X1Tz37eTvNxcuLk2zTs249TYXW7Z/b5ZVVdVHX595Bc3d7J+BxPdvcU1X3dSdU1Yaqum2TkkrSYM2FOuBNPef5rwKfoml82wsgyW8ATwc+UFVf6snfKuCzwAFJHtEmv5imd9WZVbWuK/ZGml5qY7FOkqTJ+R+g9/z3ZuAu4Mj2fee8+hdVtb4TVFU/Bs6kaYg/os++/7Gq/qcr/m42njuP7BM/KVW1pk/aT2jqqSf1eyxwvNp64EU0Pb/e3vMZ/0jTs3qfJI/vs/nxVfXzrvcfo7mRs2yi+dHcYYOWhsme7XKkd0V7sv0f4IlJHj6enbXjb7y8feb91iT3t896f70NeewA8tztrqrq9wjEnsAt1b/L7EhXTPdypCeuU8ldQXP35ld7Vv9vVd3Zk/bjdnl1n8/trPulrrTzgWcmuTLJG5P8vz4XKZI0VfZslyO9K4akDvg5zZ31Xv/eLvdol7/VLh/Xjn3yoFebrwU0PbO6t/vyGPvuxzpJkibvP9qbIQ+oqv+jqZOWtHXSnu2qkT7bd9L27LOu3zm8t84YmDRjN/5tmrEp7+oaF+vVbchk6sUn09QHq6rqnj7rOzeq9uxJv7Wq1nYntDdbfgLsOIn8aI5wDC0Nkx3a5U9GWf9jmpP7DsDt49jfu4FX0jzL/ql2+3tpTo6vpRlfZJB6xxDp2AH431HW/bgrpns51jHojuv4WZ/YDeNY1z0w8mtojtXRwNs62yb5IHBKe/EiSVNl2OuA/6uq+/ukd8rTKd9O7fLQ9jWaziD4nZ5aPx1j3/1YJ0nS5I12Lu0+t+8A3F1V/c5vo50n++67qn6W5O5R4icsyU40N10eR9No9m80YxfeT9MTd28mVy9OtL4YrcftBkaZ+ETziw1aGiadSmDRKOsX9cSNKs1MSsfR3Al+RlXd1bXut2guZgatRkn/GeMv08COwZZqu/qeAZzRPmayD80xfB3NuWQqjpkkdQx7HfCLSRb0adQa7Tz/8qr64Dj22/mx/2g2vVAY7ViBdZIkDcKjRknvPv/9jOaRvR36NGqNdZ7cZN9tj6/teuI79Uq/Bp7xNnz9Ec3EGm+sqtN7PvN9NA1akzFj9YXmNh851GzUGQ+j96R8Vbt8Vu8GSR5N05X1e1XVuTM/2n4AltAM/ntZ94VM65lbmuFJugpYmOT/67Nu766Y7mW/Y7AtzaMqd9PMMjVlqur6qvoozQXEeppZVyRpEOZqHbA18LQ+6Xu1y86jdp3HEp8+zv12tvt/Y+x7S1yFdZIkjddvt7PDPiDJL9LUSWvaOumqdtUm50o2Pa9263cO75zrux/PvrVd7twn/il90vp5Uru8sDuxLVu/GXPHqmP7uY6mPnj6KI+Hd45Nv8fOpVHZoKXZqDOobe9J+d+B7wHPS/LbPetOo7lb8Y9dabeMsh+A69vlM7oroSS/QjP17nTq5Pn0JA9UCm1ejqW5+/5ZaH600zxj/ptJXtiznxNonm0/r88gv5OSZNsk+/RZtSNN9+PeC0JJmqi5XAe8JckDveOTPBX4fZqZnP4doKq+QtOodXSSg3t3kGTrJN0XOefRXFic2D4y0ol7LBPrpWSdJEnj92vAH/Sk/SXwEODc9n3nvPrmJNt3gtqbMa+neXzu4332/dIkT+6K346Ng8Kf2xXXGfvxyCQLuuJ/j+ZxwfHo1Iu9N3X+lP7jdY1WV/fVjpt1Pk298Gfd65IcAfwGMNLWK9K4+cihZqPraKboflGSO2ims6WqzkhyDM0z3V9M8s/tur1p7hx8na5ZM6pqfZKvAnu3Y2p8l+YRi/dV1Y+SfBp4HvDVJCtpBps9hGaq8sOmp6hAU8k9HzgI+EaSfwMWAi+kqQyP6OmefBzNhc/Hk7wA+DZNJbAfzbTtb5iCPD6E5ph/B/gK8AOacV4Opel18DdT8JmS5qe5WgfcCDwSuLI9zz+SZobCAl7R8yjiS4CVwIokX6a5c78BeALN3fl1NHf/qapvJ/kr4C+A/05yAc0d8xcCXwUO3MJ8WidJ0vhdCnygvQHxvzT10XLgm8BfA1TVSPvY3nHAtW39sw1wOM3j4q+vqu/22fcXga8kOY/mUbxDgF+hmf3wi11x/0lzvv9d4N+T/CfNxCG/C3wOeO44yvFPNOfr97Q3DG6gmUXw6TQzzfbWJaPW1WN8xutp6uwz2s+4si3P84D/ozk+0papKl++Zt0L+G2aGZvW0/zYr651e9IM4Pt/NAP4fgf4K+BhffbzZJqLk9s6+wEWt+seDryLZlDZu4HVNIPMLmnjPtKzr7XA2gmWZ8xtaX6Av6HNw900XYcvAfYeJf6JNBcdnUGMrwfeAzy6T2zR3PHoTV/erju1z7pT23XLe/J3KU0Fdw/wI+Bi4Dkz/ffiy5evufWaq3UA8IvAh2kG+r2L5iJkn1G2+UXgdJqLortoLma+BZwDPLtP/CtpZtW6h6Yh6c9pHiHZ4rJYJ/ny5cvX2K/ucxawL/AfwB3AzcCHgEf1xAd4Oc3Nlzvb+u3LwO/32fcD57yec/tampsXW/XZ5tE0vbxuafOxkuYx9wedP3vz3rOPpwCfb/dxW3vef2q/fbTxY9XVI93ve/L5nraeuLetN/4RWNIndtS6anP1mK/580rVaGOCSpIkSZKkbkmW0zQavbmqTh3wvk+leWxxn6oaGeS+pbnGMbQkSZIkSZI0VGzQkiRJkiRJ0lBxUHhpAtpuxsvHETpiV2FJmlusAyRJkmaeDVrSxCynebZ9PEamLhuSpBmwHOsASZq32psVmaJ9n0ozELukzXBQeOCRj3xkLV68eIu3u+OOO3joQx86+AzNER6f0XlsxubxGdt4js/Xv/71m6vqUdOUJWFdMht5bKeWx3fqzJZja10y/axLpp/HbuI8dhM3n47dVNYl9tACFi9ezNe+9rUt3m5kZITly5cPPkNzhMdndB6bsXl8xjae45Pk+9OTG3VYl8w+Htup5fGdOrPl2FqXTD/rkunnsZs4j93EzadjN5V1iYPCS5IkSZIkaajYoCVJkiRpTknyzCSXJvlpkp8luTLJMT0xC5N8MMnNSe5IclmSpX32tV2SM5PcmOSuJKuSPKtP3IIkJydZm+TuJFcnOWwqyylJ85kNWpIkSZLmjCS/DlwGbA28HDgM+CpwTpLj2pgAK4D9gVe3MVsDK5M8rmeX57T7eRNwEHAjcEmSPXvi3kozmPd7gAOAK4ALkjx3sCWUJIFjaEmSJEmaW14E/AJwcFWtb9M+n2QP4KXA+4BDgL2AfatqJUCSVcAa4PXAa9q0PYCXAMdU1YfbtMuB1cBb2v2Q5NHACcAZVXVW+5krk+wKnAF8bkpLLEnzkD20JEmSJM0l2wA/B+7qSb+Vjdc/hwA/6jRmAVTVbcCFwKFd2xzS7uv8rrgNwHnAfkm2bZP3az/33J7PPBdYmmTJJMojSerDBi1JkiRJc8lH2uXfJfmlJDsmeTnwbOCd7brdgWv7bLsa2CXJw7ri1lTVnX3itgF27Yq7B/hOnziA3SZSEEnS6HzkcJosPumivulrzzhwmnMiSRpW1iWStHlVdW2S5cCngVe2yT8H/riqzmvf7wSs7bP5una5EFjfxt0yRtxOXctbq6o2E7eJJMcCxwIsWrSIkZGR0UJHtX79+r7bXfPD2zZJW7rzI7Z4/3PZaMdOm+exmziP3WDYoCVJkiRpzkjyy8AnaXpH/THNo4eHAv+Q5O6q+hgQoLfxiTa99/0g4zZRVWcDZwMsW7asli9fvrlNNjEyMkK/7Y7qcyNk7RFbvv+5bLRjp83z2E2cx24wbNCSJEmSNJf8FU2PrIOq6udt2heS/CLwt0k+QdNzql+vqYXtstMrax2wyxhx67qWC5Okp5dWb5wkaUAcQ0uSJEnSXLIUuLqrMavjv4BfBB5N03tr9z7b7gZc3zU74mpgSZLt+8Tdy8Yxs1YD2wJP6hMH8M0tLYQkaWw2aEmSJEmaS34M7Jlkm5703wLupukttQLYOcnenZVJdgAObtd1rAC2Bl7QFbcV8ELg0qq6p02+mKaB64iezzwSuLaq1ky2UJKkB/ORQ0mSJElzyXuAC4ALk7yXZgytQ4AXA++sqnuTrABWAecmOZHmEcOTaca8entnR1V1VZLzgXcl2RpYAxwHLKGr8aqqfprkncDJSW4HrqRp9NqXZvwuSdKA2UNLkjSrJbk4SSU5rSd9YZIPJrk5yR1JLkuytM/22yU5M8mNSe5KsirJs/rELUhycpK1Se5OcnWSw6aybJKkwauqfwGeS/MI4AdpBojfC3gVcGIbcz9wEPB54L00MyLeB+xTVT/o2eXRwIeB04CLgMcD+1fVlT1xp7QxrwUuAZ4JHF5VFw64iJIk7KElSZrFkrwY2KNPemgeA1kCvJqNd9ZXJtmzqm7oCj8HOJDmIuZ7NBc0lyR5RlVd1RX3VuAEmguSrwMvAi5IclBVfW7QZZMkTZ2q+jfg3zYTsw44pn2NFXcX8Lr2NVbcfTQNWqeNFSdJGgx7aEmSZqUkOwLvpP8FxCE0d9v/oKo+UVUXt2kLgNd37WMP4CXAn1XVB6rqC8DhwPXAW7riHk3TmHVGVZ1VVSur6hXASuCMqSifJEmSpImzQUuSNFu9HVhdVZ/os+4Q4EdVtbKTUFW3ARfy4LFKDqGZuv38rrgNwHnAfkm2bZP3A7YBzu35nHOBpUmWTLIskiRJkgbIBi1J0qyTZC/gpcArRwnZHbi2T/pqYJckD+uKW1NVd/aJ2wbYtSvuHjZOv94dBxunXZckSZI0CziGliRpVmlnkXo/cFZVXTdK2E7A2j7p69rlQmB9G3fLGHE7dS1vraraTFxvXo8FjgVYtGgRIyMjo2R3dOvXrx/3dscv3dA3fSKfOx9sybHVlvP4Th2PrSRJm2eDliRptnkD8BDgbWPEBOhtfOqkT2Xcg1TV2cDZAMuWLavly5ePFd7XyMgI493uqJMu6pu+9ogt/9z5YEuOrbacx3fqeGwlSdo8G7QkSbNGkl1oZhl8GbBt1xhXtO93BG6n6TnVr9fUwnbZ6ZW1DthljLh1XcuFSdLTS6s3TpIkSdIs4BhakqTZ5InAdjSDsd/S9YJmFsJbgKU0Y1vt3mf73YDrq2p9+341sCTJ9n3i7mXjmFmrgW2BJ/WJA/jmRAojSZIkaWrYoCVJmk2uAvbp84KmkWsfmkaoFcDOSfbubJhkB+Dgdl3HCmBr4AVdcVsBLwQurap72uSLaRq4jujJz5HAtVW1ZgBlkyRJkjQgPnIoSZo1qupWYKQ3PQnA96tqpH2/AlgFnJvkRJqeWyfTjHn19q79XZXkfOBd7WDza4DjgCV0NV5V1U+TvBM4OcntwJU0jV77AocOupySJEmSJscGLUnS0Kmq+5McBJwFvJfmMcVVwD5V9YOe8KNpBpg/DdgRuBrYv6qu7Ik7hWZmxNcCjwGuAw6vqgunqhySJEmSJsYGLUnSrFdVm8w2WFXrgGPa11jb3gW8rn2NFXcfTaPXaRPPqSRJkqTp4BhakiRJkiRJGio2aEmSJEmSJGmo+MihJElDbvFJF22StvaMA2cgJ5IkSdL0sIeWJEmSJEmShooNWpIkSZIkSRoqNmhJkiRJkiRpqNigJUmSJEmSpKFig5YkSZIkSZKGirMcToF+s01JkiRJkiRpMKa0h1aSi5NUktN60hcm+WCSm5PckeSyJEv7bL9dkjOT3JjkriSrkjyrT9yCJCcnWZvk7iRXJzlsKssmSZIkSZKkmTFlDVpJXgzs0Sc9wApgf+DVwGHA1sDKJI/rCT8HeDnwJuAg4EbgkiR79sS9FTgVeA9wAHAFcEGS5w6oOJIkSZIkSZolpqRBK8mOwDuB1/VZfQiwF/AHVfWJqrq4TVsAvL5rH3sALwH+rKo+UFVfAA4Hrgfe0hX3aOAE4IyqOquqVlbVK4CVwBlTUT5JkiRJkiTNnKnqofV2YHVVfaLPukOAH1XVyk5CVd0GXAgc2hP3c+D8rrgNwHnAfkm2bZP3A7YBzu35nHOBpUmWTLIskiRJkiRJmkUG3qCVZC/gpcArRwnZHbi2T/pqYJckD+uKW1NVd/aJ2wbYtSvuHuA7feIAdht/7iVJkiRJkjTbDXSWwyRbA+8Hzqqq60YJ2wlY2yd9XbtcCKxv424ZI26nruWtVVWbievN67HAsQCLFi1iZGRklOyObv369X23O37phnHvYyKfOyxGOz7y2GyOx2dsHh9JkiRJ891AG7SANwAPAd42RkyA3sanTvpUxj1IVZ0NnA2wbNmyWr58+VjhfY2MjNBvu6NOumjc+1h7xJZ/7rAY7fjIY7M5Hp+xeXwkSZIkzXcDe+QwyS7AKcBfANsm2bEdHJ6u979A03OqX6+phe2y0ytrc3HrupYL29kTx4qTJEmSNE8keW6SLyVZn+RnSb6WZN+u9QuTfDDJzUnuSHJZkqV99rNdkjOT3JjkriSrkjyrT9yCJCcnWZvk7iRXJzlsqsspSfPVIMfQeiKwHc1g7Ld0vaCZhfAWYCnN2Fa799l+N+D6qlrfvl8NLEmyfZ+4e9k4ZtZqYFvgSX3iAL45kcJIkiRJGk5JXgF8Fvg68DzgBcAFwPbt+gArgP2BVwOHAVsDK5M8rmd35wAvB94EHATcCFySZM+euLcCpwLvAQ4ArgAuSPLcwZZOkgSDfeTwKmCfPukraRq5zqFphFoBHJ1k76q6HCDJDsDBwMe7tlsBvJmm8vloG7cV8ELg0qq6p427mKaB64g2vuNI4NqqWjOIwkmSJEma/ZIsBt4FnFhV7+padUnXvw8B9gL27cy+nmQVsAZ4PfCaNm0P4CXAMVX14Tbtcpqb6m9p90OSR9PcxD+jqs5qP2Nlkl2BM4DPDbqckjTfDaxBq6puBUZ609snAb9fVSPt+xXAKuDcJCfS9Nw6mWbMq7d37e+qJOcD72oHm18DHAcsoWm86sT9NMk7gZOT3A5cSdPotS9w6KDKJ0mSJGkoHAPcD/zDGDGHAD/qNGYBVNVtSS6kuYZ4TVfcz4Hzu+I2JDkPOCnJtu2N9v1oZmI/t+dzzgU+lGSJN9olabAG+cjhuFTV/TRddT8PvBf4NHAfsE9V/aAn/Gjgw8BpwEXA44H9q+rKnrhT2pjX0tx5eSZweFVdOFXlkCRJkjQr7QX8D/CiJN9NsiHJd5K8qitmd+DaPtuuBnZJ8rCuuDVVdWefuG2AXbvi7mHjsCjdcbBxOBRJ0oAMepbDTVTVJrMNVtU6mjsnx2xm27uA17WvseLuo2nQOm3iOZUkSZI0B/xS+zoTeCPwXZphTN6TZKuq+luayafW9tm2M6HUQmB9G3fLGHE7dS1vrare2dd74zaR5FjgWIBFixYxMjIyWuio1q9f33e745du2CRtIvufy0Y7dto8j93EeewGY8obtCRJkiRpGi0AHg4cVVWfatO+2I6tdXKSv6MZ7qS38Yk2vff9IOM2UVVnA2cDLFu2rJYvX765TTYxMjJCv+2OOumiTdLWHrHl+5/LRjt22jyP3cR57AZj2h85lCRJkqQp9H/t8vM96ZcCi4DH0vSc6tdramG77PTK2lzcuq7lwnb2xLHiJEkDYoOWJGlWSbJfki8m+XGSe5LckOSfk+zWE7cwyQeT3JzkjiSXJVnaZ3/bJTkzyY1J7kqyKsmz+sQtSHJykrVJ7k5ydZLDprKskqQpsXqU9E5j0/1tzO59YnYDrq+q9V37WpJk+z5x97JxzKzVwLbAk/rEAXxzfFmXJI2XDVqSpNlmJ+DrwJ8Az6GZCXd34IokTwBo74CvAPYHXg0cBmxNM0X643r2dw7wcuBNNJOS3AhckmTPnri3AqcC7wEOAK4ALkjy3MEWT5I0xT7dLvfrSd8PuKGqfkxTh+ycZO/OyiQ7AAe36zpW0NQvL+iK24pmVvVL2xkOAS6maeA6ggc7ErjWGQ4lafAcQ0uSNKtU1SeAT3SnJfkvmhmrng+8g2Ya9b2AfTtTridZBawBXk873XqSPYCXAMdU1YfbtMtp7qS/pd0PSR4NnACcUVVntR+7MsmuwBnA56aqvJKkgfscsBJ4f5JHAt+jqT+eQzOLOjQNVauAc5OcSPOI4ck0vbje3tlRVV2V5HzgXUm2pqlnjgOW0NV4VVU/TfJOmjG6bgeupGn02hc4dArLKknzlj20JEnDoDMeys/b5SHAjzqNWQBVdRtwIQ++cDik3eb8rrgNwHnAfkm2bZP3o5l+/dyezz0XWJpkyYDKIUmaYu1Mg79Hc65/M/CvwNOBI6rqI23M/TS9dj8PvJemV9d9wD5V9YOeXR4NfJhmRvWLgMcD+1fVlT1xp7QxrwUuAZ4JHF5VFw62hJIksIeWJGmWSvILwC8AT6DpJfVjmosTaB5BvLbPZquBlyZ5WDv+ye7Amqq6s0/cNsCubBxH5R42joXSHQfNGCg+LiJJQ6Kqfga8qn2NFrMOOKZ9jbWvu4DXta+x4u6jadA6bUvzK0nacjZoSZJmq68Av9n++zs0jxf+tH2/E7C2zzadWaQWAuvbuFvGiNupa3lre1d/rLgHSXIscCzAokWLGBkZGaUoo1u/fv24tzt+6YZx73cieZlrtuTYast5fKeOx1aSpM2zQUuSNFv9AbAD8ESa8a0+n2SvqlpLM8ZJb+MTbJzBqvv9IOMepKrOBs4GWLZsWS1fvnys8L5GRkYY73ZHnXTRuPe79ogtz8tcsyXHVlvO4zt1PLaSJG2eY2hJkmalqvpWVX2lHST+2cDDgJPa1evo32tqYbu8ZZxx67qWC9vZE8eKkyRJkjQL2KAlSZr1qupWmscOd22TOuNe9doNuL4dP6sTtyTJ9n3i7mXjmFmrgW2BJ/WJA/jmhDMvSZIkaeBs0JIkzXpJFgFPBr7bJq0Adk6yd1fMDsDB7Tq64rYGXtAVtxXNVOqXVtU9bfLFNA1cR/BgRwLXVpUDwkuSJEmziGNoSZJmlSSfBq4E/hv4GfArwJ8BG4B3tGErgFXAuUlOpHnE8GSaMa/e3tlXVV2V5HzgXUm2ppmp8DhgCV2NV1X10yTvBE5Ocnv7+S8E9gUOnbrSSpIkSZoIG7QkSbPNFcDhwPHANsAPgBHg9HZAeKrq/iQHAWcB7wW2o2ng2qeqftCzv6OBt9FMo74jcDWwf1Vd2RN3Cs3MiK8FHgNcBxxeVRcOtniSJEmSJssGrRm2uM+MVWvPOHAGciJJs0NV/TXw1+OIWwcc077GirsLeF37GivuPppGr9PGnVlJkiRJM8IxtCRJkiRJkjRUbNCSJEmSJEnSULFBS5IkSZIkSUPFBi1JkiRJkiQNFRu0JEmSJEmSNFRs0JIkSZIkSdJQsUFLkiRJkiRJQ8UGLUmSJEmSJA0VG7QkSZIkSZI0VGzQkiRJkiRJ0lCxQUuSJEmSJElDZauZzoAkSdrU4pMumuksSJIkSbOWPbQkSZIkSZI0VGzQkiRJkiRJ0lDxkUNJkuag0R5ZXHvGgdOcE0mSJGnw7KElSZIkSZKkoWKDliRJkiRJkoaKDVqSJEmSJEkaKjZoSZIkSZrTklycpJKc1pO+MMkHk9yc5I4klyVZ2mf77ZKcmeTGJHclWZXkWX3iFiQ5OcnaJHcnuTrJYVNZNkmar2zQkiRJkjRnJXkxsEef9AArgP2BVwOHAVsDK5M8rif8HODlwJuAg4AbgUuS7NkT91bgVOA9wAHAFcAFSZ47oOJIklo2aEmSJEmak5LsCLwTeF2f1YcAewF/UFWfqKqL27QFwOu79rEH8BLgz6rqA1X1BeBw4HrgLV1xjwZOAM6oqrOqamVVvQJYCZwxFeWTpPnMBi1JkiRJc9XbgdVV9Yk+6w4BflRVKzsJVXUbcCFwaE/cz4Hzu+I2AOcB+yXZtk3eD9gGOLfnc84FliZZMsmySJK62KAlSZIkac5JshfwUuCVo4TsDlzbJ301sEuSh3XFramqO/vEbQPs2hV3D/CdPnEAu40/95KkzdlqpjMgSZIkSYOUZGvg/cBZVXXdKGE7AWv7pK9rlwuB9W3cLWPE7dS1vLWqajNxvXk9FjgWYNGiRYyMjIyS3dGtX7++73bHL92wSdpE9j+XjXbstHkeu4nz2A2GDVqSpFkjyfOBFwPLgEfTjE/yKeCvqur2rriFwJnA7wEPAVbRjG1yTc/+tqMZoPdIYEfgKuANVfWlnrgFwBuAVwCPAa4D3lJVnxx0GSVJ0+INNPXD28aICdDb+NRJn8q4B6mqs4GzAZYtW1bLly8fK7yvkZER+m131EkXbZK29ogt3/9cNtqx0+Z57CbOYzcYNmhJkmaTE2gasd4I3AA8hWa2qH2S/HZV3d81K9USmlmpbgFOppmVas+quqFrf+cABwInAt8DXkUzK9Uzquqqrri3tp99CvB14EU0s1IdVFWfm6rCSpIGL8kuNOfzlwHbdo1xRft+R+B2mp5T/XpNLWyXnV5Z64Bdxohb17VcmCQ9vbR642bU4j6NXABrzzhwmnMiSZNjg5YkaTY5uKpu6np/eZJ1wEeB5cAX2Tgr1b6dgXyTrALW0MxK9Zo2rTMr1TFV9eE27XKasUze0u5nk1mp2s9dmWRXmlmpbNCSpOHyRGA7Nh2cHZrz/Qk0N0xWA8/pE7MbcH1VrW/frwael2T7nnG0dgPuZeOYWauBbYEn8eBxtDpjZ31zy4siSRqNg8JLkmaNnsasjq+2y53bpbNSSZLGchWwT58XNOf2fWganFYAOyfZu7Nhkh2Ag9t1HSuArYEXdMVtBbwQuLSq7mmTL6Zp4DqiJz9HAtdW1ZoBlE2S1LKHliRptutcaHyrXY41K9VLkzysvas+nlmpVjO+Wam8CJGkIVFVtwIjvenNE+t8v6pG2vcraMZgPDfJiWx8hD3A27v2d1WS84F3tYPNrwGOo3n0/YiuuJ8meSdwcpLbgStpGr325cE3XAbumh/e1ne8LEmay2zQkiTNWkl2pnk88LKq+lqbPCtmpWrzN60zUw3CfJpRxxmEppbHd+p4bKdHOy7jQcBZwHtpHlNcBexTVT/oCT+aZoD502gmGbka2L+qruyJO4WmDnotGycZObyqLpyqckjSfGWDliRpVkryMOCzwAaaC4kHVjELZqWC6Z+ZahDm0+xWziA0tTy+U8djOzWqapPzelWtA45pX2NtexfwuvY1Vtx9NI1ep008p5Kk8XAMLUnSrJNkO5oxS54I7Nczc+GWzEo1Vtwms1JtJk6SJEnSLGGDliRpVmnHJ/kk8DTguVV1TU9IZ9yrXv1mpVqSZPs+caPNStUbB85KJUmSJM06NmhJkmaNJAuAjwHPBg6tqiv6hDkrlSRJkjTPOYaWJGk2+XuaBqi3AXckeXrXuhvaRw/nzKxUkiRJkiZmYD20kjw/ySeTfD/JXUmuS3J6kof3xC1M8sEkNye5I8llSZb22d92Sc5McmO7v1VJntUnbkGSk5OsTXJ3kquTHDaockmSptUB7fIUmkar7tfLoJmVCjgI+DzNrFSfBu5j9FmpPkwzOO9FwOMZfVaq02hmpboEeCbOSiVJkiTNWoPsoXUCcD3wRuAG4CnAqcA+SX67nRY3NHfWlwCvZuNd9ZVJ9uwZ9Pcc4EDgROB7wKuAS5I8o6qu6op7a/vZpwBfB14EXJDkoKr63ADLJ0maYlW1eJxxzkolSZIkzWODbNA6uKpu6np/eZJ1wEeB5cAXgUOAvYB9q2olQJJVNI+BvB54TZu2B/AS4Jiq+nCbdjnNwL1vafdDkkfTNGadUVVntZ+7MsmuwBmADVqSJEmSJElzzMAeOexpzOr4arvcuV0eAvyo05jVbncbcCEPHqfkEODnwPldcRuA84D9kmzbJu8HbAOc2/O55wJLkyyZWGkkSZIkSZI0W031LIedGai+1S53B67tE7ca2CXJw7ri1lTVnX3itgF27Yq7h41Tr3fHwcYp1yVJkiRJkjRHTNksh0l2pnk88LKq+lqbvBOwtk/4una5EFjfxt0yRtxOXctbq6o2E9cvf8cCxwIsWrSIkZGR0UJHtX79+r7bHb90wxbvq9tE8jIbjXZ85LHZHI/P2Dw+kiRJkua7KWnQantafRbYQDPD1AOrgN7Gp0567/tBxm2iqs4GzgZYtmxZLV++fHObbGJkZIR+2x110kVbvK9ua4/Y8rzMRqMdH3lsNsfjMzaPjyRJkqT5buCPHCbZjmYmwycC+/XMXLiO/r2mFrbLW8YZt65rubCdPXGsOEmSJEmSJM0RA23QSrI18EngacBzq+qanpDVNONe9doNuL6q1nfFLUmyfZ+4e9k4ZtZqYFvgSX3iAL65xYWQJEmSJEnSrDawBq0kC4CPAc8GDq2qK/qErQB2TrJ313Y7AAe367rjtgZe0BW3FfBC4NKquqdNvpimgeuIns85Eri2qtZMqlCSJEmSJEmadQY5htbf0zRAvQ24I8nTu9bd0D56uAJYBZyb5ESaRwxPphnz6u2d4Kq6Ksn5wLvaXl9rgOOAJXQ1XlXVT5O8Ezg5ye3AlTSNXvsChw6wbNNq8ShjcK0948BpzokkSZIkSdLsM8gGrQPa5Sntq9ubgVOr6v4kBwFnAe8FtqNp4Nqnqn7Qs83RNI1jpwE7AlcD+1fVlT1xp9DMjPha4DHAdcDhVXXhIAolSdJc4k0TSZIkzQUDa9CqqsXjjFsHHNO+xoq7C3hd+xor7j6aRq/TxpVRSZIkSZIkDbWBz3IoSZIkSZIkTSUbtCRJkiRJkjRUbNCSJEmSJEnSULFBS5IkSZIkSUPFBi1JkiRJkiQNFRu0JEmSJEmSNFRs0JIkSZIkSdJQsUFLkiRJkiRJQ8UGLUmSJEmSJA0VG7QkSZIkSZI0VGzQkiRJkjRnJHl+kk8m+X6Su5Jcl+T0JA/viVuY5INJbk5yR5LLkizts7/tkpyZ5MZ2f6uSPKtP3IIkJydZm+TuJFcnOWwqyypJ85kNWpIkSZLmkhOA+4A3AvsD7wOOAz6fZAFAkgAr2vWvBg4DtgZWJnlcz/7OAV4OvAk4CLgRuCTJnj1xbwVOBd4DHABcAVyQ5LmDLZ4kCWCrmc7AMLvmh7dx1EkXzXQ2JGlOaS8k3gAsA/YAHgIsqaq1PXELgTOB32tjVgF/VlXX9MRtR3ORcSSwI3AV8Iaq+lJP3IL2c18BPAa4DnhLVX1ykOWTJE25g6vqpq73lydZB3wUWA58ETgE2AvYt6pWAiRZBawBXg+8pk3bA3gJcExVfbhNuxxYDbyl3Q9JHk3TkHZGVZ3Vfu7KJLsCZwCfm7LSStI8ZQ8tSdJssytwOHAL8OV+Ad5ZlySNpqcxq+Or7XLndnkI8KNOY1a73W3AhcChXdsdAvwcOL8rbgNwHrBfkm3b5P2AbYBzez73XGBpkiUTK40kaTQ2aEmSZpsvVdWiqnoucMEoMZ07639QVZ+oqovbtAU0d9aBB91Z/7Oq+kBVfYGmsex6mjvrnbgH3VmvqpVV9QpgJc2ddUnScNu7XX6rXe4OXNsnbjWwS5KHdcWtqao7+8RtQ3MTphN3D/CdPnEAu00w35KkUfjIoSRpVqmq+8cR1vfOepLOnfXXdMVtcmc9yXnASUm2rap7GPvO+oeSLKmqNRMulCRpxiTZmeYmxmVV9bU2eSdgbZ/wde1yIbC+jbtljLidupa3VlVtJq5f/o4FjgVYtGgRIyMjo4WOatFD4PilG7Z4u24T+dy5YP369fO27JPlsZs4j91g2KAlSRpGY91Zf2mSh1XVesZ3Z30147uzboOWJA2ZtqfVZ4ENwNHdq4DexqdOeu/7QcZtoqrOBs4GWLZsWS1fvnxzm2zi3R/7LO+4ZnKXdmuP2PLPnQtGRkaYyDGXx24yPHaDYYOWJGkYzYo764O4qz7aHbrJ3mnfUnPxLqF3P6eWx3fqeGwHo50UZAXwRGDvqrqha/U6+p/bF7bLW7ridhkjbl3XcmGS9NQlvXGSpAGxQUuSNIxmxZ31QdxVH+0O3XTPojsX78x793NqeXynjsd28pJsDXwSeBrwO70z4NL0wH1On013A65ve/l24p6XZPue3r67AfeysWfvamBb4Ek8uLdvZ+ysb060LJKk/mzQkiQNI++sD9jiPg1oa884cAZyIkmTk2QB8DHg2cCBVXVFn7AVwNFJ9q6qy9vtdgAOBj7eE/dm4AXAR9u4rYAXApe24zACXEzTwHVEG99xJHCt4zBK0uDZoCVJGkbeWZckjebvaRqg3gbckeTpXetuaB89XAGsAs5NciLNjZCTaXrmvr0TXFVXJTkfeFfb62sNcBywhKbxqhP30yTvBE5OcjtwJU2j1740k5VIkgZswUxnQJKkCVgB7JykMw179531FT1xW9Nc2HTiNndnvZt31iVp+BzQLk+habTqfr0MHphR9yDg88B7gU8D9wH7VNUPevZ3NPBh4DTgIuDxwP5VdWVP3CltzGuBS4BnAodX1YWDLJwkqWEPLUnSrJPk+e0/f7NdHpDkJuCm9tEQ76xLkvqqqsXjjFsHHNO+xoq7C3hd+xor7j6aBq3TxpVRSdKk2KA1RPqNbwKOcSJpTrqg5/172+XlwPKquj/JQcBZ7brtaBq4Rruz/jaaC4wdgasZ/c76epo7648BrmMa7qxf88Pbpn0AeEmSJGnY2aAlSZp1qmrM2QXbGO+sS5IkSfOUY2hJkiRJkiRpqNhDS5IkSZLmuX7Dmzi0iaTZzB5akiRJkiRJGio2aEmSJEmSJGmo2KAlSZIkSZKkoWKDliRJkiRJkoaKDVqSJEmSJEkaKjZoSZIkSZIkaajYoCVJkiRJkqShstVMZ0CSJM1Oi0+6qG/62jMOnOacSJIkSQ9mDy1JkiRJkiQNFRu0JEmSJEmSNFR85HAO6PdIiI+DSJIkSZKkucoeWpIkSZIkSRoqNmhJkiRJkiRpqPjIoSRJ2iLOfihJkqSZZoOWJEmSJGkT3sCQNJv5yKEkSZIkSZKGij205ijvpkiSJEmSpLnKBi1JkjQQ/W6meCNFkiRJU8FHDiVJkiRJkjRU7KE1z/gooiRpOlnvSJIkaSrYQ0uSJEmSJElDxR5akiRp2tlzS5KGl+dwSbPBnGnQSvJ44J3A7wIBLgP+tKqun9GMDQkH8pUk6xJJ0uRZl0jS9JgTDVpJtge+CNwD/CFQwGnAyiS/XlV3zGT+JEmzn3XJ7OANFknDzLpEkqbPnGjQAl4OPBH41ar6DkCS/wb+F3gF8DczmLehZVdiSfOMdcksZX0kaYjM67rEmxKSptNcadA6BLiiU2kAVNWaJP8BHMocrzim22gXFr2OX7qB5VObFUkaJOuSIbO5+uj4pRs4qo3xgkrSNLEukaRpMlcatHYHPtsnfTXwgmnOi7qMt/FrIrw4kTRg1iVz2FTWR/1YR0nzlnVJD68HJE2VudKgtRNwS5/0dcDCfhskORY4tn27Psl1E/jcRwI3T2C7eeE1U3x88tdTtedp4d/O2Dw+YxvP8XnCdGRkjrEumWOmuh4ay5DXUePl3+7UmS3H1rpky1mXTKMBnWvn5bEbEI/dxM2nYzdldclcadCCZsDFXhk1uOps4OzJfGCSr1XVssnsYy7z+IzOYzM2j8/YPD5TyrpkDvHYTi2P79Tx2A4965Ih4rGbOI/dxHnsBmPBTGdgQG6huRvSayH975BIktTLukSSNFnWJZI0TeZKg9ZqmufVe+0GfHOa8yJJGk7WJZKkybIukaRpMlcatFYAT0/yxE5CksXAM9t1U2VSXYPnAY/P6Dw2Y/P4jM3jMzWsS+Yej+3U8vhOHY/t8LIuGT4eu4nz2E2cx24AUtXvEe/hkuShwNXAXcCf0zy3/lbg4cCvV9X6GcyeJGkIWJdIkibLukSSps+c6KFVVXcA+wLfBv4J+BiwBtjXSkOSNB7WJZKkybIukaTpMyd6aEmSJEmSJGn+mBM9tKZTkscn+ZcktyX5WZJPJdllpvM1EUkel+TdSVYluTNJtc/498YtTPLBJDcnuSPJZUmW9onbLsmZSW5Mcle732f1iVuQ5OQka5Pcnf+fvbsPk6Sq7zZ+f5E3AQmLyhpBsquoCTwETNZEIpEFjaAiqAhGMAZ4BCVGTUAURA0qRqIYTTSoxNcICiFqXB4NKIEBYxajIgirYtBdEUUBl7flZWHhPH9UtVvbWzPbM9sz091zf66rrpo+9evqOqemz+k6VXUquTrJIeNs4zFJfpBkdZLrkryqL5nfgCQvTvK5JD+p83JdkncleURX3Jwrm/qz909ySZJf1J9/Y5J/TbJrV9ycLJ+Wbbmw/n6d1pVu+cxRGaG2pN+Gof4dJYNYPw27JM9NcnmSVfX3+1tJ9msst2zVF3OlLYnHLFM2DG3qAJedxzvDoJTi1OMEbAX8L3At8ALgYOAa4EfA1rO9fVPIz2Lgl8CXgYuo7vFf0BUT4GvAjcBLgQOAy4BbgZ26Ys8BbgeOAZ4JfJ5q/IA9u+LeCawGXg/sC3wEeAh4blfcMXX6O+u40+rXx81A2VwB/CtwBLAP8Fd13q4ANpnLZVN//kuB9wAvrsvnz6ie6nMn8FtzvXxayuqm+vt12lz/bjmNXlsyDeUz0PXvKE2DWD8N+wS8EngAeB/wJ8D+wBuBAy1bpz7/r82ZtgSPWTam7Aa6TR3wsvN4ZwimWd+AYZqA1wEPArs00hYCa4DjZ3v7ppCfTRp/v4L2xuHgOn3fRtpvACuBf2yk7VHHHdVI2xS4DljSSNuh/oK+retz/hP4btd7bwY+1RX38bqC2Gyay+bRLWkvr/O431wumwnK7Ml1Pk+wfH79mdsBv6Bq4LoPGOd8+czViRFrS6ahfAa2/h2laRDrp2GfgAVUByd/NUGMZevUl2kutSV4zLIxZTewbeqgl9045enxzoBN3nI4OQcBV5RSru8klFKWA1+n+mceKqWUh3oIOwj4eSnl0sb77gAuYN08H0R1RvK8Rtwa4Fxg/yRb1Mn7A5sDZ3d9ztnA7kkW1q/3Ah7dEvdp4JHA3j1s+5SVUm5pSf5mPd+xns/JspnAr+r5A/Xc8oF3A8tKKZ9tWWb5zF0j1Zb024DXv6NkEOunYXc01dnxD08QY9mqX+ZMW+Ixy9QNeJs60GU3Do93BowdWpOzG9Vlvd2WAbu2pI+CifK8c5JtGnHLSyn3tMRtDuzSiFsNXN8SB2vLcbd63v3Z3XEzaZ96/v16PufLJsnDkmye5IlUl8P+gqpihjlePkn2pjoD9hfjhMzp8pnj5mJbsrEGpf4dCQNcPw27vYEfAH+a5EdJ1iS5PsmrGzGWrfrFtmRd/q7q3aC0qUNRdh7vDDY7tCZne+C2lvSVwLwZ3paZMlGeYW2+NxS3fWN+e6mvidxAHC3r7I6bEUl2BN4OXFxK+VZjG+Z62XyDqtL9IfC7VJcu39zYjjlZPkk2o2rwziilXDdO2JwtH83JtmTKBqz+HXoDXj8Nu8cCT6Qac+V04NnAV4EPJnldHWPZql9sS9bl76oeDFibOixl5/HOALNDa/K6/7mgGgxuVIXe8jwdcYwTO6PqnvUvUo1JcFRzEXO8bKgGR3wacDjVAIlfzdqnzszl8nkj8HCqQRrHM5fLR3OvLZmSAax/R8Eg10/DbhPgEcArSyn/XEq5pJRyHHAhcHKSYNmqv9z3a/m7agMGsE0dlrLzeGeA2aE1ObfR3ts5j/Ye11GwkvHzDGvzvaG4lY35vPpH3YbiaFnn9l3Lp1WSLYElwOOB/UspNzYWz+myASilfL+U8o16DJZnAtsAJzW2Y86VT6rHZZ8CvAXYIsl2SbarF3deP4w5Wj4C5mZbMmkDWv8OtSGon4ZdZ2yVr3alfwWYD/wmlq36x7ZkXf6umsCAtqlDUXYe7ww2O7QmZxlr71dt2hX43gxvy0yZKM83lFJWNeIWJtmqJe5+1t4HvAzYAnhCSxysLcfO/b/dn90dN23q2zI+B/wB1WNSr+kKmbNl06aUcjtVXjr3gM/V8nk8sCXVII23NSaoHr97G7A7c7d8NDfbkkkZ4Pp32A16/TTslo2T3jkweQjLVv1jW7Iuf1eNY4Db1IEvu24e7wygyTwSca5PwF9RXaL5+EbaAqonFpww29u3kXkb7xG4L6jT92mkbUt1FvIDjbQ967g/b6RtSjXY4AWNtM6jSP+m63MuBq5pvN4MuAX4RFfcR+vP3nyay2MT4F+B+4BnjhMzJ8tmgjKbD6wCPjKXywfYDljcMhWqJ48spjqzMyfLx2m025I+lc/A1r/DPg16/TTsE/C8ujxe3JV+EfDT4v+uUx+nudqW4DHLZMtrYNvUQS+7ccrK450Bm2Z9A4ZpAram6j29huoxnAcBVwM/BraZ7e2bYp5eXE8fqr9kx9Wv96mXbwL8N/BT4E+pHiU6RnUZ4+O61nUu1dndV1BdjvlvdeX5e11xp9fpx1P9eP4Q1VnL53fFvapOP62Oe3v9+tUzUC6d8jiN6p7p5rTTXC6b+vO/QHXLysHAvsArqZ7sdDvwpLlePuOUWQFOa7y2fOboxAi2JX0un4Guf0dxGqT6aZgnqiuxLqE6mHgV1aDwZ9Xle6Rl69Tn/7c51ZbgMctUy22g29QBLzuPd4ZgmvUNGLYJ2Jnqks07gbuAf6frDMEwTXUF1zaNNWK2Bz5efzHvAf4T2KNlXQ8H/p7qUab3UT0RYnFL3MOANwM/oeqB/i5dZzMbsa+keqLEauB/gb+YoXJZMUHZnDqXy6b+7DcC364r9HuA66iemrWgK25Ols8427POAaPlM7cnRqwt6XPZDHz9O2rToNVPwzxRnZn/J+CXVLeRfBc43LJ1mo5pLrUlE7QLY40Yf1etv10D36YOcNl5vDMEU+qCkCRJkiRJkoaCg8JLkiRJkiRpqNihJUmSJEmSpKFih5YkSZIkSZKGih1akiRJkiRJGip2aEmSJEmSJGmo2KElSZIkSZKkoWKHliRJkiRJkoaKHVqSJEmSJEkaKnZoSZIkSZIkaajYoaWBlWRxkpLk1NnelumS5JN1HhdM0/pXJFkxifixJGU6tkWShlFdR491pQ1sXTnZel+S5jrrebVJcmT9v3HkbG+LxmeHliRJkiRJkobKprO9AZIkSUPm5cBWs70R43jmbG+AJI2AQa7nJdXs0JIkSZqEUsoNs70N4yml/Gi2t0GSht0g1/OS1vKWQw2FJE9NcnGSVUlWJjknyaNb4v5vkm8muTvJnUkuT3JwS9y4Y1e1LUuySZJXJvlWktvq9f84yWeT7Nb1/k2SHJPkG/X2rkry30leNHEW89okP0yyOsmPkrxunMBHJ/lAkp8kuT/JTfU2r5eXCT5szyQX1dt2W5J/S7LzOLFbJnlDkmuS3FWX63VJPpFkx14/U5I2pDl2YpK96zp8VV3P/V2Sh9VxL0/y3ST31vXl0S3r2jbJaUl+kOS+JL9K8u9J9hjnsw9N8p069mdJ3pvk4ePErje2SpLHJnl7kv9Jcktdl1+f5IwkjxhvHUk2S/K2uk5fnWRZkpdOrQTbx1ZptGsLJ9HWbJnkxCRX1vvgziRX12W6WVfsC5J8rW4j7q7b4f/bss5fj0eS5OC6Tb2nzvsJdUySHF9v4311eRw4zjY+Jsk/pmqPVyf5ZZKzkyycavlJml7W8xtfzzfW/4i6DrypztM3kzxrnNjfTfK5xnb/sC67rbvixh3DeLxlqY7T/j3JjfW6b0py2Tj77ClJzq/r60479K4k20wh/39Tb8+Lx1n+pnr5SxtpRydZ0tgXtyb5YpJFPX7mpMunXta3fGtdXqGlYfBU4A3AxcCHgacDhwOPT/JHpZQCkOQDwF8CPwE+AmwOHAr8e5ITSil/vxHb8HfA64HvAp8C7gceB+wHfAlYVm9DgM8ChwHfq2MBngd8LslflVL+oWX9ZwB7A/8P+Eq93e9PsrqU8uFOUKpOvCuAxwNfBT4DPInqsujnJXl6KeWHE2Ukye8CXwMeDpwPrAAWA/8F3Nbylk8DLwa+DpwFFGAh8CLgY8DPJvo8SZqCP6Sq979MVe88p35Nkl8Cbwa+CFwO/CnwsSQ/KqVcVsc8ql72O8ClVPX0I4FDgD9J8qxSytLOh9U/uj9GVQd+HLiXqo578iS2+RnAXwP/Cfw3VV35h8AJwDPq+vmBlvedC/xevY2bAi8FPpPk9lLKf0zi83vRa1uzVZ2Pp1G1Zf9c5+e3qfbDGcDtdewJ9etbWNs+vhD4aJI9SimvbdmOF1HdGvkFqrbnRcAZSe6h2mcvqrdxE+BlwOeT7FpKub6xjU8ExoDHAP8BfI6qXT4M2D/J07xaTRpo1vMbV89vRlWPbwv8KzCvXu+XkywqpXy3E5jkj4GLgIfVsTdS/fY/haqs9iml3DeVjUjyFKp6/G6q/XUTsAOwZ709H2/EvpCqLO4H/h34BVW5nATsm+QZpZT7J/HxZwOnAkcA/9ay/HBgVb1dHf8EXEVVdr8Cfgt4AVW7sbiUcsUkPr8n05BvNZVSnJwGcqKqaEs9HdJI34SqISnAXnXaPvXrq4FtGrGPoepweQB4fCP9k3X8gpbPXW8ZsBL4JvCwrthNgd9ovH5l/d5/asYCWwPfAFYDj235rP8F5jfSd6m3+bquz/tEHf+WrvSX1+mXdKWvAFZ0pV1ex76oK/1TnfJupP0G8BDw+ZZy2hLYerb/T5ycnEZn6qr3n9dI35rqR/I9VD/Ef6ux7Pfr+CWNtM/WaS/tWv8uwB3ANY203wDurNMXNtIfQXWyogBjXesZa9aVddoObXUi1UFZAV7Wtg6qg6Jmu9Vpzy6aYhm21fuTbWv+vo7/EJCuZfOBTeu/n1C//2fAYxox21C1xwV4RiP9yDptNfCURvqOwH1UnWTfAx7ZWHZI/Z5/7NqOpfV6ntGVvle9Tf9vtv+fnZyc1p+s5/tWzxeqjvzNGul/Xqd/pJG2CXA91e/5fbrW87E6/q0t++fUCfbdqY20Tnvxuy3xzbr8UfU++DGNY6F62Yn1Ol4/hbK4om4LtutK36Ne5790pS9sWcfvAHcBF3elH1mv48iNKJ9pybfT2slbDjUMLiulfK7zopTyEPAv9cvO5aF/Xs/fUkpZ1Yj9BfAeqo6nIzZyO1aXUh5sJpRS1pRS7mgkvZrqzM9fN2NLKXcD76C6aqzt1sN3llJ+2Yi/nupsx5M6lzAn2ZzqDNUvgHd3bce/UJ1t2DfJ48bLQJLfAv4Y+GYp5fNdi98CPNiVVoBQncVad0Ep99X5kqR+u6SU8qXOi7qu+RLVlaUfKaX8pLHs28CPqH68ds7aHwZ8qZTy2eZK67r1n4H/k+T/1MkHUx3UfKSUsrwRexfwzl43uJRy8zh14pn1vPU2EOBNXe3WZVQHKz3d/jBJvbQ1mwKvAG4F3lDqX92N9/yylLKmfnkEVfv6d3V724lZBfxN/fLPWd/ZpZTvNOJ/Vm/HbwDvKqX8qhH7Baqz2nt0EpL8HtXVY/9cSrm8a/uWUp2Nf06S39hAeUiaPdbzG1/Pn1DWvSLsHGBN13r3pjr58O/1566zXVSdQW319GS1HSs06/KXU+2DN5RSft4V+l6qq3z/dAqfew7V8VX3bYdHNJY3t2l5VxyllO9TXeX3x/XxVj9NV75V85ZDDYPvtKR1bnPbrp7vWc/HWmLHumKm4jzgVUmupLqk9WvAN0rj8tD6Fo3/A9wAvKm6+3AdnTG/frtl/RvK4131+7YElpZSVrfEX0aVxz2Bn46Tj9+t51/rXlBKuSHJDVS3E3bS7kxyIXB43VH27/V7r+zu3JOkPrq6Je0XG1j2h/XfT6U6I71N2zgWVGdioapTr2VtR8l69SJVJ0vPkhxKdaXunlS3fzRPHP7mOG8br/5vHddwI/Xa1jyC6mqFuzawvj3r+VjLsku7Ypp63r+llIeS3AI8tpHc2dc7jbOPf5Oq7J8IfKtluaTZZz2/cfX87aWUFc2EUsqa+nbN7RrJe9bzse4VlFJ+meQHwB5JHtFDnd/mfOB1wDeSfBa4BPhaKeXmrrjOvtu70dHY9ADtx0gbci7VVWJHAB+FXw8B81Kq/5mLm8FJdqHqyNuXql3p7sB6JNWVgv0yXflWzQ4tDYM7WtI6Z4cfVs+3Be4rpdzZEvuLRsxUvZZqbK6jWHsm584kHwVOKdV95/Oormb6LdaemW6zdUtar3kE+GVLLPSWz87Z6u5GpuOXNDq0aodSXUp9ONWZBIBbk7wPOL2+Yk6S+qmtLl+zgWWd3zTb1/N96mk8nbp4onpxvPp2PUlOpLp69maqMZ1+RnUbHVRtwhZt7+u6yrdjDdPz4J5e2ppOeXSfSW4zbrtUSrkjyWra26Sp7N/mQPSdfXxwPY2nrb2VNBis5zeunm9bZ2e9D2u87uX4YY86btIdWqWUpUmeSTUe1zHAXwAlyaXA60op19ahnX32usl+xgY+/5YkX6UaA2vH+orffYCdgPc3T8DXYy/+D9VJm4uprgBeRXU75guoyqF1H26Eacm31rJDS6PiTuAJSbZt6dSa34jp6HTCPIz1rffju76c93Tg9FRPA9wXOA44nup79LrG+r9eStl7SrmYWGf988dZ3pbPbp3Gb4cNrOPX6kukTwJOSvIkqoF8X0PVsXcv8L4JPk+SZlqnDnxnKeXNPcRPVC+OV9+uo75N781UnUB7lFJubSybz8QnOQbN7fX8sRMF1Zrt0joHS/XtflswcZs0VZ11HlNK+eg0rF/SYLOe791kjx8mdYwEUEoZA8ZSPTHxj6jGPnwFcFGSJ9fHEp31P7E0HvDRJ+dQPVTgpVQPKWm93RD4K6qr1w7vvlU1yR/SuLV9ApMtn+nMt5ies3/SbLiqnj+jZdk+XTGw9gf7js3AJJuwgcqslHJDKeVTVJ1aq4CD6vS7gB9Q3bM/HY9gvY7qLNDTxrm/u5P3tsu0OzpPPFmvw63uqJvw0udSyg9LKR8Cnl0nHTThFkvSzPsm1fh/T+sxvlNn/nHLsl5PTjyK6ofs0uZBTu3pPa5jUFxHdZZ+rx7asqvqea9tb7/8Tz3vdR9LGi3W8727qp6vV08n2YHqdrcfN243vL2e79gdDzxlog8qpdxdSvlqKeVVVA8jeSzVgP4wvfX2v1M9ZfHwJFtQjad1XSml+5bzJ9TzC5qJSR5O9dTBXtxez3stH9uraWaHlkZFZ5D4t9VjWQG/rqjfQHX57Wca8d+u5y/vWs9rgcc3E5JskWTfls/cjursc3MQxA9QXdZ8ZpItu9+QZLd6myatHjfrPKr78/+6a71HUFXEY6WUGyZYx0+oxg/4gyTdg9O/g66zDUkenaRtwMrO2Zz1BoCUpNlUD07+b8AzkxzXvTzJJkmat6gsoerAOTbJwkbcNlS3UPTiZqr68PfqH8addfwm8LeTz8XsqQd8/2eqg7f3pGtAyCQ71FcqQNWuPgicmOTRjZitgbfVL/+FPiulfIPqIOGoJM/vXp5ksyTTcaW0pAFgPT8p/0X1hL0XJvmjrmWnUY3P26ynr6PqHHp+ku06iUkeT3WctI4ke3ceKtKlc7zTOVb4BNWFAH9X3/rXvZ7tkkzYYTaeeqD+f6fqUDqB6hjt7JbQzjHSrzsg6zbubxn/7pVukyofpjHfqnjLoUZCKWUsyYeobgO8NskXqAb5O4yqgnpDKeVHjbd8gWpMrP+bZCfgGqpKcE+qwdWbjeDDgUuSXA98g2rA9e2pxu3YjGogwo4PUV1q+2fA4iSXUN2b/ptUA7LvSfVI8fHGsNqQN9TbdnrdyXYl8CTghcCv6vxvyGuoGrd/TXI+1VNWFlOdafguaweOp077ZpLvUg1oeRPwmPrzHgT+YYr5kKTpdBzVWeczk7yCqvNjFdVVqHtRtQtbApRSbk/yV1SPL/92knOpfoC/iOpx7rtu6MPqgcs/THWy4TtJvkTVThwIXA48ua+5m35vofrB/yqqpz59heo2iycB+1Od1Li9lHJ9kjcBfwdcU7cp91O1EQuBM1ueqtUvh1MNPL8kydeorkJYQzWO5R8DK3GgXWmUWc/3oN7uo6nG/Lokyb9Sjf21D1U5fZvG09NLKavrY6rXA1cm+SJVPl8IfJX1n9b+eqqOxUuoOs4epGo/nkZVLt+s13tzfQL+PGBZki8D/0s1ztnj6+35FFW7MxXnUN1qeGr9+jMtMR+mGg/580nOo7oV9Y+p2qsxquOhCU22fGYg33OeHVoaJa+m6nR5FVUj91D9+rhSyuebgaWUe5I8C3g/VeX1R1QV2V5UZ2qaHVp3U40h9cw69tFUnUdXAX9fSvlKY70FeFmS/6C6d/xgYCuqsUW+TzVQ4jVTzWBdKf4h8Faq2/32o/rRfg7wN22Pom1Zx9VJnkF1AHIQ1cHHJVSPjP1UV/gKqobhmVQHMdvXeflP4N31WXJJGiillF8l2YtqfMNDqa7GLVSd8l+neipTM/7jSVZRPfnoaKo6/lyq8VLu6fFjT6K6FeHlVO3Rz4B/ojrz2/Zk2oFVt5H7Uo03cjhVm3o/sJyq7bi7Efvu+oTP8VQHCptQtXfvKqX88zRu44/qs9qvp2prj6F6WtTPqG4naTuYkTQirOd7V0q5rL46663A86gGRb8BeBfwt6WU7jsuTqaq8/+c6tjlh/X8Rtbv0PoQ1ThRf0h1vPAg1fHDG4F/qo+NOtuxJMnvU52gfybw3Pq9N1CdJP/kRmTzq1QXDOwA/Hcp5cfdAaWUK5McQHVXyqF1Hseoxt7q9Uo9mFz5THe+57w0/sckSZIkSZKkgecYWpIkSZIkSRoqdmhJkgZKkp2SfCDJ0iT3JClJFnTFLKjT26btumK3TPKeJDclubdeb9vTfjZJcnKSFUnuS3J1kkOmN7eSJEmSpsIxtCRJg2YXqgc6fJvqqZzPniD2XVRPL2q6q+v1x6jGjDiRasDSVwMXJdmrlHJVI+4dVOPxnFJ/9p8C5yc5sJTy5allRdo4SRbTw0C1VE+5HZvObZEk9Z/1/Fr1Scm/6iH09lLK+6d1YzQU7NCSJA2ay0sp8wHqJxdN1KH141LKFeMtTLIH1aDWR5dSPlGnXUb1VKO3Uz0YgSQ7UHVmnV5KOaN++6VJdgFOB+zQ0mxZDPxNj7Fj07cZkqRpshjr+Y7t6K0sfkL1cC/NcQ4KDzzqUY8qCxYsmPT77r77brbeeuv+b5DGZZnPLMt75vWrzL/97W/fWkp5dB82aVbVHVr/DCwspaxopC+geuLaMaWUj07w/rcAbwG2K6Xc00h/G9XTiratH8H8Z8C/AE8qpfxvI+4o4OPA4zf0FNHutsTvz+xzHwwG98Psm+o+GJW2ZJi0HZfMle+Q+Rw9cyWv5nNi09mWeIUWsGDBAr71rW9N+n1jY2MsXry4/xukcVnmM8vynnn9KvMkP9n4rRkK70ryYeBu4DLglFLKNY3luwHLm51ZtWXA5lS3Ny6r41YD17fEAexK1YE2ru62xO/P7HMfDAb3w+yb6j6YQ23JwGg7Lpkr3yHzOXrmSl7N58Smsy2xQ0uSNIxWAx8BvgLcAvw28Cbgv5P8QSnl+3Xc9sBtLe9f2Vjemd9e1r9suTtuHUmOBY4FmD9/PmNjY79etmrVqnVea+a5DwaD+2H2uQ8kSaPIDi1J0tAppdwEvKqR9LUkF1JdUXUK8LI6PUDbvfVped1LXPd2nAWcBbBo0aLSPGs1V87WDTL3wWBwP8w+94EkaRRtMtsbIElSP5RSfgr8F/DURvJK2q+umtdY3pnPS9LdgdUdJ0mSJGkA2KElSRol3VdaLQMWJtmqK25X4H7Wjpm1DNgCeEJLHMD3+rydkiRJkjaCHVqSpJGQZGfg6cA3GslLgM2AQxtxmwIvAb5SSlldJ19I1cF1RNdqXwZcu6EnHEqSJEmaWY6hJUkaOEleXP/5+/X8OUluAW4ppVyW5L1UJ2WWUg0K/2TgZOAh4G876ymlXJXkPOD9STajelLhccBCGp1XpZSbk7wPODnJXcCVVJ1e+wEHT19OJUmSJE2FHVozZMFJX2pNX3H682Z4SyRpKJzf9frMen4ZsJjqFsHjgCOBRwC3ApcAbyulXNf13qOAdwKnAdsBVwMHlFKu7Io7BVgFvA54DHAdcFgp5YKNzs0c0dbW2c5JkiQNj/H6Lj55wNYzvCUbZoeWJGnglFI29HTBjwMf73Fd9wLH19NEcQ9SdXqd1uNmjjxPxkiSJGlQOYaWJEmSJEmShoodWpIkSZIkSRoqdmhJkiRJkiRpqNihJUmSJEmSpKHioPCSJGlSxhssXpIkSZopdmhJkqRp45MSJUmSNB2mtUMryYXA/sA7SylvbqTPA94DvAB4OLAU+OtSyjVd798SeAfwMmA74CrgjaWUy7viNgHeCLwSeAxwHfD2UsrnpiNfkiRJkqTR5MkYaThM2xhaSV4K7NGSHmAJcADwGuAQYDPg0iQ7dYV/DDgGeCtwIHATcFGSPbvi3gGcCnwQeA5wBXB+kuf2KTuSJEmSJEkaENPSoZVkO+B9wPEtiw8C9gb+rJTy2VLKhXXaJsAbGuvYAzic6sqtfy6l/CdwGHAD8PZG3A7A64HTSylnlFIuLaW8ErgUOH068idJkiRJkqTZM11XaL0bWFZK+WzLsoOAn5dSLu0klFLuAC4ADu6KewA4rxG3BjgX2D/JFnXy/sDmwNldn3M2sHuShRuZF0mSJEmSJA2Qvo+hlWRv4OW03G5Y2w24tiV9GfDyJNuUUlbVcctLKfe0xG0O7FL/vRuwGri+JQ5gV2D5ZPMhSZIkSRptPrlXGl597dBKshnwEeCMUsp144RtD6xoSV9Zz+cBq+q42yaI274xv72UUjYQ172txwLHAsyfP5+xsbFxNnd8q1at6vl9J+y+pjV9Kp87l02mzLXxLO+ZZ5lLkiRJ0ob1+wqtN1I9tfCdE8QE6O586qRPZ9w6SilnAWcBLFq0qCxevHii8FZjY2P0+r4jx3tSxhGT/9y5bDJlro1nec88y1ySJEmSNqxvY2gl2Rk4BXgLsEWS7erB4Wm8fhjVlVNtV03Nq+edq7I2FLeyMZ9XPz1xojhJkiRJWkeSpyf5SpKbk9yZ5MokR3fFzEvy0SS3Jrk7ycVJdm9Z15ZJ3pPkpiT3Jlma5BkzlxtJmjv6OSj844EtqQZjv60xQfUUwtuA3Vk77lW3XYEb6vGzqOMWJtmqJe5+1o6ZtQzYAnhCSxzA96aSGUmSJEmjLcnvAhcDmwHHAIcA3wQ+luS4OibAEuAA4DV1zGbApUl26lrlx+r1vBU4ELgJuCjJntOeGUmaY/rZoXUVsG/LBFUn175UnVBLgB2T7NN5Y5JtgefXyzqWUDUUhzbiNgVeAnyllLK6Tr6QqoPriK7teRlwbSnFAeElSZIktflT4GHA80spXyylfLWU8krgG1QPuoLq6et7A39WSvlsKeXCOm0T4A2dFSXZAzgc+OtSyj+XUv4TOAy4AXj7jOVIkuaIvo2hVUq5HRjrTq/vBPxJKWWsfr0EWAqcneREqiu3TqYa8+rdjfVdleQ84P31YPPLgeOAhTQ6r0opNyd5H3BykruAK6k6vfYDDu5X/iRJkiSNnM2BB4B7u9JvZ+0QJgcBPy+lXNpZWEq5I8kFVMcbr23EPQCc14hbk+Rc4KQkWzROys953U8XPGH3NRx50pdYcfrzZmmLJA2bfl6h1ZNSykNUl99+FTgT+ALwILBvKeWnXeFHAZ8ATgO+BDwOOKCUcmVX3Cl1zOuAi4CnA4eVUi6YrnxIkiRJGnqfrOf/mOSx9bi/xwDPBN5XL9sNuLblvcuAnZNs04hbXkq5pyVuc2CXvm65JM1x/X7K4XpKKes9bbCUshI4up4meu+9wPH1NFHcg1QdWqdNfUslSdJs6z5jD3i2XtK0KaVcm2Qx1Un2v6iTHwBeVUo5t369PbCi5e2dh0/NA1bVcbdNENf2wCuSHAscCzB//nzGxsbWWb5q1ar10kbBCbuvWef1/IdXaTOd1+7tmEg/tm1U92ebuZLXUcvneN+JQczntHdoaWL+cJckDYK29kiSRl2SJwKfo7qK6lVUtx4eDHw4yX2llHOohkYpbW9ved1L3DpKKWcBZwEsWrSoLF68eJ3lY2NjdKeNgiNbbjl87zWbsuKIxbO6HRPpx7aN6v5sM1fyOmr5HO878ckDth64fNqhJUmSJGmu+luqK7IOLKU8UKf9Z5JHAv+Q5LNUV1i1XV3VGWOrc1XWSmDnCeJWtiyTJE3RjI+hJUmSJEkDYnfg6kZnVsf/AI8EdqC6emu3lvfuCtxQSllVv14GLEyyVUvc/VRPfJck9YkdWpIkSZLmql8AeybZvCv9D4H7qK6qWgLsmGSfzsIk2wLPr5d1LAE2Aw5txG1K9QT2r/iEQ0nqL285lCRpDnGsLElaxweB84ELkpxJNYbWQcBLgfeVUu5PsgRYCpyd5ESqWwxPphob692dFZVSrkpyHvD+JJsBy4HjgIXAETOYJ0maE7xCS5IkSdKcVEr5N+C5wBbAR6kGiN8beDVwYh3zEHAg8FXgTKonIj4I7FtK+WnXKo8CPkH19PUvAY8DDiilXDntmZGkOcYrtCRJkiTNWaWU/wD+YwMxK4Gj62miuHuB4+tJkjSN7NCSJEmSJI2MttvrV5z+vFnYEknTyVsOJUmSJEmSNFTs0JIkSZIkSdJQ8ZZDSZIkSZKmyFscpdlhh5YkSZIkaWC1dRjB5DqNxluHpOFlh5YkSZIkaVrYkSRpujiGliRJkiRJkoaKV2hJkqQZ5xl7SZIkbQw7tCRJkiRJ2gBPxkiDxVsOJUmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDpa8dWkn2T3JJkl8kWZ3kxiT/mmTXrrh5ST6a5NYkdye5OMnuLevbMsl7ktyU5N4kS5M8oyVukyQnJ1mR5L4kVyc5pJ95kyTNjCQ7JflAXeffk6QkWdASZ1siSZIkzVH9vkJre+DbwF8CzwZOBnYDrkjyWwBJAiwBDgBeAxwCbAZcmmSnrvV9DDgGeCtwIHATcFGSPbvi3gGcCnwQeA5wBXB+kuf2N3uSpBmwC3AYcBvwtbYA2xJJkiRpbtu0nysrpXwW+GwzLcn/AD8AXgy8FzgI2BvYr5RyaR2zFFgOvAF4bZ22B3A4cHQp5RN12mXAMuDt9XpIsgPweuD0UsoZ9cdemmQX4HTgy/3MoyRp2l1eSpkPkOQVVCdIutmWSJIkSXPYTIyh9at6/kA9Pwj4eecABKCUcgdwAXBw430H1e85rxG3BjgX2D/JFnXy/sDmwNldn3s2sHuShX3KhyRpBpRSHuohzLZEkiRJmsOmpUMrycOSbJ7kicBHgF9QHTxAdQvitS1vWwbsnGSbRtzyUso9LXGbU92S0olbDVzfEgewK5KkUWNbIkmSJM1hfb3lsOEbwO/Xf19PdUvIzfXr7YEVLe9ZWc/nAavquNsmiNu+Mb+9lFI2ELeOJMcCxwLMnz+fsbGxcbIyvlWrVvX8vhN2X9PzeqeyLXPFZMpcG8/ynnmWec8Gvi0Z1H05mfZoUEy1HAd1H8w17ofZ5z7YsHq8xJOA3wMeAn4IvKGUckm9fB7wHuAFwMOBpcBfl1Ku6VrPllRjMr4M2A64CnhjKeXymcjHXLPgpC/N9iaMa7xt++QBW8/wlkija7o6tP4M2BZ4PNWYJF9NsncpZQUQoPuAgTq9+3U/49ZRSjkLOAtg0aJFZfHixROFtxobG6PX9x05icp2xRGT35a5YjJlro1nec88y7xnA9+WDOq+nEx7NCim2i4O6j6Ya9wPs899MLEkr6R6IMgHqTqjNgH2BLaql3ceRLKQ6kEkt1E9/OrSJHuWUm5srO5jwPOAE4EfA6+mehDJXqWUq2YiP8NukDupJA2WaenQKqV8v/7zG0n+g+os+knAq6jOdred6Z5Xzztn0lcCO08Qt7Ixn5ckXWfWu+MkSaPDtkSStNGSLADeD5xYSnl/Y9FFjb/7+iCSUTBep9OK0583w1siaS6b9kHhSym3U9122BmnZBnVWCXddgVuKKWsasQtTLJVS9z9rB3nZBmwBfCEljiA70154yVJg8q2RJLUD0dT3WL44Qli+v0gEklSH0x7h1aS+cBvAz+qk5YAOybZpxGzLfD8ehmNuM2AQxtxmwIvAb5SSlldJ19IdVByRNdHvwy4tpSyvH+5kSQNCNsSSVI/7A38APjTJD9KsibJ9Ule3Yjp94NIJEl90NdbDpN8AbgS+C5wJ/Ak4K+BNcB767AlVIMonp3kRNbegx7g3Z11lVKuSnIe8P4km1Fd0nsc1b3rRzTibk7yPuDkJHfVn/8SYD/WPWMiSRoSSV5c/9l5wMhzktwC3FJKuQzbEklSfzy2nt4DvInqJPyhwAeTbFpK+Qf6/yASqSfe2ilNrN9jaF0BHAacQHUW4qfAGPCuekB4SikPJTkQOAM4E9iS6qBk31LKT7vWdxTwTuA0qqeEXA0cUEq5sivuFKpG5HXAY4DrgMNKKRf0N3uSpBlyftfrM+v5ZcBi2xJJUp9sAjwCOLKU8vk67ZJ6bK2Tk/wj0/yAkQ09fX0Qn1I53hNz27az16frzn/4cD6Jd7Imsz8nU86DaBD/d6fDqOVzvP+7QcxnXzu0Sil/B/xdD3Erqe5XP3oDcfcCx9fTRHEPUh2onNbzxkqSBlYpZcIf/3WMbYkkaWP9Cngi8NWu9K8ABwC/Sf8fRLKODT19fRCfUjneE3Pbnkrb69N1T9h9De+9ZlqeWTZQPnnA1j3vz8mU8yAaxP/d6TBq+Rzv/24y/7szZdrH0JIkSZKkAbVsnPTOiZWH6P+DSCRJfWCHliRJkqS56gv1fP+u9P2BG0spv6D/DyKRJPXB6F/TKUmSJEntvgxcCnwkyaOAHwMvBp5NNQYj9PlBJJKk/rBDS5KkETTek5EkSWuVUkqSFwDvAt5GNd7VD4AjSimfqWOm40EkkqSNZIeWJEmSpDmrlHIn8Op6Gi+mrw8ikSRtPMfQkiRJkiRJ0lDxCi1JkjTQxrt9csXpz5vhLZEkaXo4VIA0eV6hJUmSJEmSpKFih5YkSZIkSZKGih1akiRJkiRJGiqOoSVJkoZS23gjjqslSZI0N9ihNQ0c0E+SJEmSJGn62KE1gHyakyRJkiSpjceLUsUxtCRJkiRJkjRU7NCSJEmSJEnSULFDS5IkSZIkSUPFDi1JkiRJkiQNFQeFlyRJkiStx6e3SxpkdmhJkiRJkjaaHWCSZpK3HEqSJEmSJGmoeIWWJEmSJEkz4Jqf3cGRXskm9YVXaEmSJEmSJGmoeIWWJEkaGc3xW07Yfc2vz4KvOP15s7VJkiRJmgZ969BK8mLgpcAiYAfgBuDzwN+WUu5qxM0D3gO8AHg4sBT461LKNV3r2xJ4B/AyYDvgKuCNpZTLu+I2Ad4IvBJ4DHAd8PZSyuf6lTdJkiRJkgZZ26D8ntDRKOvnLYevBx4E3gQcAHwIOA74at3pRJIAS+rlrwEOATYDLk2yU9f6PgYcA7wVOBC4CbgoyZ5dce8ATgU+CDwHuAI4P8lz+5g3SZIkSZIkDYh+3nL4/FLKLY3XlyVZCXwKWAxcAhwE7A3sV0q5FCDJUmA58AbgtXXaHsDhwNGllE/UaZcBy4C31+shyQ5UHWmnl1LOqD/30iS7AKcDX+5j/iRJkiRJkjQA+naFVldnVsc36/mO9fwg4Oedzqz6fXcAFwAHN953EPAAcF4jbg1wLrB/ki3q5P2BzYGzuz73bGD3JAunlhtJkiRJkiQNqul+yuE+9fz79Xw34NqWuGXAzkm2acQtL6Xc0xK3ObBLI241cH1LHMCuU9xuSZIkSXNMkguTlCSndaXPS/LRJLcmuTvJxUl2b3n/lknek+SmJPcmWZrkGTOXA0maO6atQyvJjlS3B15cSvlWnbw9cFtL+Mp6Pq/HuO0b89tLKWUDcZIkSZI0riQvBfZoSZ+OcYAlSRupn2No/Vp9pdUXgTXAUc1FQHfnUye9+3U/49q28VjgWID58+czNja2obesZ9WqVa3vO2H3NZNeVy+mso2jZrwy1/SwvGeeZS5J0sxLsh3wPuCvgc90Le7rOMDSoGp7SiL4pEQNrr53aCXZkuoMxuOBfUopNzYWr6T9qqnOlVm3NeJ2niBuZWM+L0m6rtLqjltPKeUs4CyARYsWlcWLF48XOq6xsTHa3nfkOBXBxlpxxPqfNdeMV+aaHpb3zLPMpenhj3RJG/BuYFkp5bNJ2jq01hsHOElnHODXNuLWGwc4ybnASUm2KKWsntZcSNIc0tcOrSSbAZ8D/gB4Vinlmq6QZcCzW966K3BDKWVVI+6FSbbqGkdrV+B+1o6ZtQzYAngC646j1Rk763tTzYskSZKk0Zdkb+DltNxuWJtoHOCXJ9mmPo7pZRzgZUgzyBM6GmV969BKsglwDvBM4HmllCtawpYARyXZp5RyWf2+bYHns+6lvUuAtwGHAp+q4zYFXgJ8pXFm40KqDq4j6viOlwHXllKW9yl7kiRJkkZMfUL+I8AZpZTrxgnbHljRkt4cB3gVvY8D3L0NEw6FMpvDEUzXUCpt5j98Zj9vtgxKPiczdM5U///mylAao5bP8f4PBjGf/bxC65+oOqDeCdyd5GmNZTfWtx4uAZYCZyc5karCP5lqzKt3d4JLKVclOQ94f93ILAeOAxZSdV514m5O8j7g5CR3AVdSdXrtR3X5ryRJkiSN543Aw6mOYcYzreP7bmgolNkcjmC6hlJpc8Lua3jvNdMyxPNAGZR8tg1nM97+nurQN3NlKI1Ry+d4/wefPGDrgctnP79Jz6nnp9RT09uAU0spDyU5EDgDOBPYkqqDa99Syk+73nMUVcNyGrAdcDVwQCnlyq64U6jOiLwOeAxwHXBYKeWCfmRKkqRBN97tBJKk8SXZmepY4hXAFkm2aCzeoh4o/i76Pw6wNOv87aBR0LcOrVLKgh7jVgJH19NEcfcCx9fTRHEPUnV6ndbThkqSJElS9RCrLYGzW5a9vp6eQv/HAZYk9cEms70BkiRJkjQLrgL2bZmg6uTal6oTagmwY5J9Om9sjAO8pLG+JcBmVMOwdOLaxgGWJPXB7N+8K0mSJEkzrJRyOzDWnZ4E4CellLH6dV/HAZYk9YdXaEmSJEnSOEopDwEHAl+lGgf4C8CDjD8O8CeohkP5EvA42scBliRtJK/QkiRJkqRaKWW9pxL2exxgSdLG8wotSZIkSZIkDRU7tCRJQynJ4iSlZbq9K25eko8muTXJ3UkuTrJ7y/q2TPKeJDcluTfJ0iTPmLEMSZIkSeqZtxxKkobda4FvNl6v6fyRamTfJVQD8r6GtQP5Xppkz1LKjY33fQx4HnAi8GPg1cBFSfYqpVw1rTmQJEmSNCl2aEmSht33SylXjLPsIGBvYL9SyqUASZZSPXnqDVSdYSTZAzgcOLqU8ok67TJgGfD2ej2SJEmSBoS3HEqSRtlBwM87nVkApZQ7gAuAg7viHgDOa8StAc4F9k+yxcxsriRJkqReeIXWEFlw0pda01ec/rwZ3hJJGijnJHkUcDtwEXBSKeWGetluwLUt71kGvDzJNqWUVXXc8lLKPS1xmwO71H9rxLS1rbarkiRJg88OLUnSsLoDeC9wGXAn8BTgTcDSJE8ppdwMbA+saHnvyno+D1hVx902Qdz2/dtsSZIkSRvLDi1J0lAqpXwH+E4j6bIklwP/QzU21puBAKXl7Wl53UvcuguTY4FjAebPn8/Y2Nivl61atWqd19PphN3XbDhoDpr/8KmVzUztt7liJr8Lauc+kCSNIju0JEkjo5RyZZIfAk+tk1bSfnXVvHp+WyNu5wniVrYso5RyFnAWwKJFi8rixYt/vWxsbIzm6+l05Di3pM91J+y+hvdeM/mfOiuOWNz/jZnDZvK7oHbuA0nSKHJQeEnSqGlebbWManysbrsCN9TjZ3XiFibZqiXufuD66dhQSZIkSVNjh5YkaWQkWQQ8CfhGnbQE2DHJPo2YbYHn18toxG0GHNqI2xR4CfCVUsrqad50SZIkSZPgLYeSpKGU5BxgOXAl1RMOnwKcDPwM+EAdtgRYCpyd5ESqWwxPprqK692ddZVSrkpyHvD+JJvV6z0OWAgcMRP5kSRJktQ7O7QkScPqWuClwGuArYBfAJ8H/qaUcitAKeWhJAcCZwBnAltSdXDtW0r5adf6jgLeCZwGbAdcDRxQSrly+rMiSZI0mBZMYqzOFac/bxq3RFqXHVqSpKFUSnkX8K4e4lYCR9fTRHH3AsfXkyRJkqQBZoeWJElSD8Y7Q+3ZaEmSpJnnoPCSJEmSJEkaKnZoSZIkSZIkaajYoSVJkiRJkqSh4hhakiRJkiRpKLWNcen4lnODV2hJkiRJkiRpqPS1QyvJTkk+kGRpknuSlCQLWuLmJflokluT3J3k4iS7t8RtmeQ9SW5Kcm+93me0xG2S5OQkK5Lcl+TqJIf0M2+SJEmSJEkaDP2+5XAX4DDg28DXgGd3ByQJsARYCLwGuA04Gbg0yZ6llBsb4R8DngecCPwYeDVwUZK9SilXNeLeAbweOKX+7D8Fzk9yYCnly33NoSRJkiSNmLbbtiRpkPW7Q+vyUsp8gCSvoKVDCzgI2BvYr5RyaR27FFgOvAF4bZ22B3A4cHQp5RN12mXAMuDt9XpIsgNVZ9bppZQz6s+4NMkuwOmAHVqSJKlnHtRJc0eSFwMvBRYBOwA3AJ8H/raUclcjbh7wHuAFwMOBpcBfl1Ku6VrfllQn218GbAdcBbyxlHL5NGdFkuacvt5yWEp5qIewg4Cfdzqz6vfdAVwAHNwV9wBwXiNuDXAusH+SLerk/YHNgbO7PudsYPckCyebD0mSJElzwuuBB4E3AQcAHwKOA76aZBNY5w6TA6juMDkE2IzqJPpOXev7GHAM8FbgQOAmqjtM9pz2nEjSHDMbTzncDbi2JX0Z8PIk25RSVtVxy0sp97TEbU51e+OyOm41cH1LHMCuVFd/SZIkSVLT80sptzReX5ZkJfApYDFwCX2+w0TS1HgFtbrNRofW9sCKlvSV9XwesKqOu22CuO0b89tLKWUDcetIcixwLMD8+fMZGxvrYdPXtWrVqtb3nbD7mkmva2NMZduH1Xhlrulhec88y1ySpJnT1ZnV8c16vmM9b73DJEnnDpPXNuLWu8MkybnASUm2KKWs7ncepGHW1km14vTnzcKWaBjNRodWgO7Op076dMato5RyFnAWwKJFi8rixYsnCm81NjZG2/uOnOme42vuXi9pVCuB8cpc08PynnmWuSRJs26fev79et7vO0wkTcArsdSr2ejQWkn7VVPz6vltjbidJ4hb2ZjPS5Kuq7S64yRJkiRpXEl2pLo98OJSyrfq5H7fYdL9mRPeOTJTV2/P9F0m3eY/fPa3YSaMej6b/6vN/91rfnZHa/wJu0//dky3UbvDYrz/z0HM52x0aC2j/emHuwI31Gc3OnEvTLJV11mOXYH7WTtm1jJgC+AJrDuO1q71/Hv92vBu1/zsjpm/GkuSNGd5xlKSpk+SbYAvAmuAo5qLmMU7R2bq6u3ZPq45Yfc1vPea2Tg8nVkjn8/G3UMn7P4g7/2vzuuZzfOKIxbP2GeN2h0W49UFnzxg64HLZ1+fctijJcCOSTqX8pJkW+D59bJm3GbAoY24TYGXAF9p3H9+IVUH1xFdn/My4NpSigPCS5IkSRpXki2pjj8eD+xfSrmxsXgyd5hMFOedI5LUR33vJk3y4vrP36/nz0lyC3BLKeUyqoZiKXB2khOpGoCTqc5cvLuznlLKVUnOA96fZDOqp4gcByyk0XlVSrk5yfuAk5PcBVxJ1em1H9UgjZIkSZLUqj7W+BzwB8CzSinXdIX0+w4TSVIfTMd1f+d3vT6znl8GLC6lPJTkQOCMetmWVB1c+5ZSftr13qOAdwKnAdsBVwMHlFKu7Io7heq+9dcBjwGuAw4rpVzQlxxJkiRJGjlJNgHOAZ4JPK+UckVL2BLgqCT71Cfom3eYfKYr7m1Ud5h8qo5ru8NE0jQbb5iGUX142lzV9w6tUsqE94jXMSuBo+tporh7gePraaK4B6k6vU7rfUslSZI2no8cl4baP1F1QL0TuDvJ0xrLbqxvPezrHSaSpP4Y4dHoJEmSZodnhqWh8Zx6fko9Nb0NOHWa7jCRJG0kO7QkSZIkzUmllAU9xvX1DhNJ0sabjaccSpIkSZIkSVPmFVqSJEmSJGnkOe7laPEKLUmSJEmSJA0Vr9CSJEmSJElzkg9yGV5eoSVJkiRJkqShYoeWJEmSJEmShoq3HI4oL5uUJEmSJKm/uo+1T9h9DYtnZ1PmPK/QkiRJkiRJ0lCxQ0uSJEmSJElDxVsOJUmSJEmSGsYbxkeDww4tSZIGkD+iRpNjXEqSpMnwt8P47NCSJEkaQP6AlSRpOEz2RKRteX/YoTXH+ONYkqTB4xV5kiTNbf4WmDw7tCRJkoZI2w9eT0xJkqS5xg4tSZIkSZKkGeLVWP2xyWxvgCRJkiRJkjQZXqElSZI05BwjU5IkzTV2aAlwPA5JkiRpLvBWJ2k0eAxvh5YkSdLI8sotSZI0qp1fdmhpXP4IliRJkiRpOEzmCsxRuFrTQeElSZIkSZI0VEbmCq0kjwPeB/wJEOBi4K9KKTfM6oZJkobGbLQlo3B2TMNnpm898KpvzSUel0jSzBiJDq0kWwGXAKuBPwcKcBpwaZLfLaXcPZvbN2omc/DlD1VJw8K2RNKwG+832icP2HqGt2Tusi2RpJkzEh1awDHA44Enl1KuB0jyXeB/gVcCfz+L2yZJGg62JZrT+nEVVds6Tth9DeP95PQkmUaQbYkkzZBR6dA6CLii02gAlFKWJ/k6cDA2HLPGH6qShohtidRiUG6LHdUnNGnkDFRbMijfX0maDqPSobUb8MWW9GXAoTO8LZqiyZ4Z9oetpD6zLZGGzGQP1tt+J0zn+F6OHTYnzUpbYseVpLkopZTZ3oaNluR+4O9LKSd1pZ8GnFRKWa/jLsmxwLH1yycD103hox8F3DqF92nqLPOZZXnPvH6V+W+VUh7dh/XMGdPQlvj9mX3ug8Hgfph9U90HtiWTNE3HJXPlO2Q+R89cyav5nNi0tSWjcoUWVAMudsu4waWcBZy1MR+Y5FullEUbsw5NjmU+syzvmWeZz7q+tSXuy9nnPhgM7ofZ5z6YcX09Lpkr+898jp65klfzOXs2me0N6JPbgO1b0ufVyyRJ2hDbEknSxrItkaQZMiodWsuo7lfvtivwvRneFknScLItkSRtLNsSSZoho9KhtQR4WpLHdxKSLACeXi+bLht1y6KmxDKfWZb3zLPMZ0+/2xL35exzHwwG98Pscx/MnOk4Lpkr+898jp65klfzOUtGZVD4rYGrgXuBN1Pdt/4O4BHA75ZSVs3i5kmShoBtiSRpY9mWSNLMGYkrtEopdwP7AT8EPg2cAywH9rPRkCT1wrZEkrSxbEskaeaMxBVakiRJkiRJmjtG4gqtmZTkcUn+LckdSe5M8vkkO8/2dg2qJC9O8rkkP0lyb5LrkrwrySO64uYl+WiSW5PcneTiJLu3rG/LJO9JclO9vqVJntESt0mSk5OsSHJfkquTHDKdeR1kSS5MUpKc1pVuufdRkucmuTzJqrp++FaS/RrLLe8RZxsx+5Isruu77un22d62UZRkpyQfqOupe+qyXtAS11P9p8nrZR8kWTDO96Ik2W52tlwTGYT2pN/f7+n4fZPkmCQ/SLI61XHGq8aJe0GS79Tr+0mSNyd5WIbgWGVj81gv2z/JJUl+Ua/nxiT/mmTXUcrnOPEDdxzUp33a0++dYc/nBpVSnHqcgK2A/wWuBV4AHAxcA/wI2Hq2t28QJ+AK4F+BI4B9gL8Cbq/TN6ljAnwNuBF4KXAAcBlwK7BT1/rOqd9/DPBM4PNUYxTs2RX3TmA18HpgX+AjwEPAc2e7TGZhH7wUuIlqDIfTGumWe3/L+ZXAA8D7gD8B9gfeCBxoec+NyTZiMCZgcV3fvQZ4WmNaNNvbNopTXd6/BL4MXFSX/YKumJ7rP6dp2wcL6vS/7fpePA142GznwWm9fToQ7Um/v9/9/n1Tr+ehOn5f4LT69XFdcfsDD1INaL0vcDxwH/B3DPixSj/yWC9/KfAe4MV1Pv+M6omcdwK/NSr5bPkfHrjjoD7u08Vs4PfOKORzg/XUTFWIozABr6sLe5dG2kJgDXD8bG/fIE7Ao1vSXl5/+farXx9cv963EfMbwErgHxtpe9RxRzXSNgWuA5Y00naov2Rv6/rc/wS+O9tlMsPlvx3wi7oC667ILff+lfOCusL/qwliLO8Rn2wjBmNi7Q+8Z832tsyFifqAr/77FbQf8PZU/zlN6z5YUKe/Yra316mnfToQ7Uk/v9/9/n1Tv/dm4FNdcR+nOljfrJH2HeCyrri3AvcDu7XkeyCOVfqYx8eMs3+fXG/vCaOYTwbwOKifeaWH3zujkM/x8taZvOVwcg4CriilXN9JKKUsB75O9c+iLqWUW1qSv1nPd6znBwE/L6Vc2njfHcAFrFuuB1FdAXNeI24NcC6wf5It6uT9gc2Bs7s+92xg9yQLp5abofRuYFkp5bMtyyz3/jma6ozDhyeIsbxHn22E5pxSykM9hPVa/2kKetwHGi4D0Z70+fvd7983ewGPbon7NPBIYG+obt0E9hwnbjPgD1ryNCjHKv3K43Na8gjwq3r+wIjmcxCPg6Z7n865fNqhNTm7UV36220ZsGtLutrtU8+/X88nKtedk2zTiFteSrmnJW5zYJdG3Grg+pY4mCP7KsneVGeY/mKcEMu9f/YGfgD8aZIfJVmT5Pokr27EWN6jzzZisJyT5MEkv0rymTiW2Wzqtf7T9HtX3UbdkWRJ2zgqGgjD1J7M1u+b3ep592f3FFd3EN5De3kOyrFK3/OYatywzZM8kepWsV9QdVp01jMq+RzU46Dp+L+d6PfOKOWzlR1ak7M9cFtL+kpg3gxvy1BKsiPwduDiUsq36uSJyhXWlu2G4rZvzG8v9fWKE8SNrCSbUTVSZ5RSrhsnzHLvn8cCT6Qam+B04NnAV4EPJnldHWN5jz7biMFwB/Beqttj9gPeATwLWJpkh9ncsDms1/pP02c11e+CV1KNUfJ6YHfgv5P8zmxumFoNU3syW79vOvPudfYa10lb5/fSgB2rTEcev0FVH/wQ+F2q2ypv7nH7hyKfA34c1M+89vJ7ZxTyOaFNNxSg9XTvPKgGW9MG1D3AX6S6//+o5iJ6K9d+x42yNwIPpxqEbzyWe/9sAjwCOLKU8vk67ZJUTwI6Ock/YnnPFe6TWVZK+Q7VeAwdlyW5HPgf4LXAm2dlw+Y266tZVkq5CWg+XeprSS6kOlt+CvCyWdkwTWRYvjOz9fum87ottte4ddY5gMcqfc8j1WDw2wKPp+rY/mqSvUspKya5XYOcz0E+DupbXnv8vTP0+dwQr9CanPF6CefR3quoWpItgSVUlef+pZQbG4tXMn65wtqy3VDcysZ8XpLuL0F33EiqLzM9BXgLsEWS7bL2kdyd1w/Dcu+nzhgEX+1K/wowH/hNLO+5wDZiQJVSrqQ6G/3U2d6WOarX+k8zqJTyU+C/8HsxiIapPZmt3zfjXZG+fY9xUA0avhIG9lilr3kEKKV8v5TyjXpcqWcC2wAn9bj9A5/PITgO6vs+bWr5vTOS+WyyQ2tylrH2Ps+mXYHvzfC2DI36ss/PUQ26+NxSyjVdIROV6w2llFWNuIVJtmqJu5+19/IuA7YAntASB6O/rx4PbEk1uN5tjQmqMzG3Ud1mYLn3z7Jx0juV/UNY3nOBbcRgG++soqZfr/WfZp7fi8E0TO3JbP2+6fz26v7snuLqq+i3Ar43wMcqfctjy7ZTSrm93qbO+EijkM9BPw6a1n3aCWNtvT7K+QTs0JqsJcDTkjy+k1AX9tPrZeqSZBPgHKozAAeXUq5oCVsC7Jhkn8b7tgWez7rluoTqaQeHNuI2BV4CfKWUsrpOvpDqi3dE1+e8DLi2HmRulF1FNT5G9wRV5b4vVaVkuffPF+r5/l3p+wM3llJ+geU9F9hGDKgki4AnUY0dopnXa/2nGVRfyfB0/F4MomFqT2br981S4NZx4lZSPRGSUsoNwNXjxD1Qf96gHqv0K4//0ZInkswHfhv40Qjl8yoG+zhouvdp9++dkcznOkopTj1OwNZUX4BrqB5zeVC9A34MbDPb2zeIE/Ahqh7i04CndU071TGbAP8N/BT4U6pOgLH6n/1xXes7l6pn/RVUDc+/AfcBv9cVd3qdfjywuN6Oh4Dnz3aZzOK+KMBpjdeWe//KNsAlVLcevopqUPiz6jI/0vKeG5NtxGBMVAcmpwEvohok9QSqH1U3AI+a7e0bxQl4cT112vzj6tf71Mt7rv+cpm0fvBd4H3AY1QHdq4CfALcDT57t7Xdab38OTHvSz+93v3/f1P/HD9V1/mKqwdwfAl7dFffcOv0jddxf1+t/DwN+rNKPPNbLv0B1G97BdR3wSqondN8OPGlU8jnB//HAHAf1cZ9u8PfOKORzg3XUTFaIozABO1NdknoncBfw78CC2d6uQZ2AFXUF0jad2ojbHvh4/eW6B/hPYI+W9T0c+HuqR8zeR9X7vLgl7mFUA+H9hOpJHt8FXjzb5THL+2Kditxy73v5bgv8E/BLqjMX3wUOt7zn1mQbMfsTcHL9XbiD6uzeT6k6mH9ztrdtVKcJ2vmxRkxP9Z/T9OwD4Gjgm1QHK2vq9uUz2Jk1sNOgtCf9/H5Px+8bqo6ZH9Zx/wv8xThxL6LqFFxNdcD/1vpzVkyQx1NHIY/1sjcC36bqwLoHuI6qA2FB1zqGOp8b+D8emOOgPu3Tnn7vDHs+NzSlXoEkSZIkSZI0FBxDS5IkSZIkSUPFDi1JkiRJkiQNFTu0JEmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDxQ4taUAkOTJJSXLkbG/LdEoylqRM4/pLkrFJxK9IsmK6tkeaaUkW19+DU2f4c0+tP3fxTH6upodtUt/Wb5skzYCZaPumu74YdbYrfVu/7UqDHVqSJEmSpDlttk4ISZq6TWd7AyRJkiRJGnAvB7aa7Y2QtJYdWpIkSZIkTaCUcsNsb4OkdXnL4QhqXi6b5BlJLklyZ5Ll9fIFST6Z5KYk9yf5SZJ/TPKocdb3giRfS3JXkruTfDPJ/22J+/V90UkOTvKtJPfU6z+hjkmS45P8MMl9SZYlObBlXTsm+ack1ye5N8mtSa5M8p4plkmp72dekORzSW5LsirJRUn2GOc9T0jyiSQ31uV0Y5Izkzy6K25Bvf5PJtk9yf+r1z/le6eT7J9kaV1+N9efu94ZoSSbJXl9kmvqcrotyYVJ/rgldtz7uduWJdkyyRvqdd9V/w9dV5fJjl2xWyQ5McnV9TbfkeTiJPtMkMfNkryt/v9YXf8vvHSc2En9z46zjn2TfL1Rph9Lsn2v75eGUeP//u4kt9T/9+t9b+q4Lyf5VV03fy/JSUnWO/GV5NH1em6p1/tf433Xs4H2qI6xTbJN2uCy2CZJ066rzu78j26o/fi/dT18d/29vDzJwS1xvx5nMclxSb5f17vLk7w5Xe1NJhiXcaJlLbEvTHJekh836qWLkzyre53ApfXLv6nXX5p10Xj1Vqp28QN1XXB/XTd8MsmCltgV9fSIut64qS6Hb3Zv02TEdsV2Za62K6UUpxGbgMVAAb4K3A/8P+DvgDOBJwO3AA8BnwfeBVxUx/8IeHTXuk6ol90MfBD4e2B5nfaPXbFH1ulLgLuBs4H3AzfU6ccB/wjcCHwYOAu4p97GXRrr2RpYUad/Hji9/uyvAvdOsUwKcHW9Lf9d5/szwBrgTuB3u+L3Au4AVgPnA++u8/VQXU7bN2IX1Ov/r3pdlwLvAT47yW3slN8XgPuA84D3AtfU6ed2xQf4Yr1sWb2N/1xvwxrg0K74seor3/rZ6y2r893J13uBM4DP1eWydyNuS+DyOvZ/6n3+z/X/zBrgRW2fVa9rOdX/5VnAXXX6c7riJ/s/W4CxrrRnAw/U/28fo/o+XAtcCfwcWDHb31snp35NrG0DLmrUYX9b102d+mLrRvxf1t+vm4GP19/3b9Sxn+9a9zbA9zrfs3q9/1Z/Tud7ubhlW9Zrj+rltkm2SW2fvd4ybJOcnKZ9YvLtxwfq9BVU9fEHgV/Wacd3rfvUOv3/1d/bj9R103V1+qfGiV/csp3rLWts+6ldsT+gqm8/UX9fPwncDjzYrA/q93+Ste3bqZ2pETPG+nXTo+vvfgG+Un/G5+o64hbgSV3xK4CfAUvr8vwH4F/qOuF+uur/Sew72xXblTnZrsz6BjhNw05dW6EX4PCuZZ0G6c+60t9ap3+8kfaE+ovxM+AxjfRtqCrMAjyjkd6ppFYDT2mk71hXWrdTHQg9srHsELoORICD6rTXtuTtkb2WQ9f7OuXx8a70F9XplzfSNgd+AqwEfqcr/tA6/oONtAWN9b9pI/Zbp/zuB57WSN8S+H5dye3YSP9z1v7o2LSR/jtUB2+3A49opI/RYyUP/EanUm2J3ZJ1f8y8q96Ok7riHk3VaN8CPLz7s6ga220a6ft08jPV/9nGvh5rvN6EqjFZAzy1kb4p8J91/IqZ+n46OU33xLptwMu7lv1jnf72+vVuVPX8FcBvNOJCdWBSgBc30t9Oe+fR0Y3PXDzOthzesq22Seum2ya1LMM2yclpRiYm1350viNXd313HkNVTz8APL6Rfmodfw/w2430Lak6CQqwX0v84pbtXG8Z43doLWx5/3yqkxnXj5P/U7vfUy9fr96i6igrwFu60l9ep1/Slb6CtZ0dmzXSO3XoR6a472xXbFfmZLsy6xvgNA07dW1l/M2u9J3r9Ctb3rMl8AvgXmDzOq3zJWr7Ef+CetnHGmlHdqc1ll08zhd1E6qDjcsaaZ2Dh2P6WCaFqmHdqWVZpxHduX7dqfhfP866vgXc2ni9oI7/OY2GaQrb2Cm/T7Ys+5t62fMbaZfUaeudyaE6c7BOeTO5Sn7b+v3nbGCbNwFuA64dZ/lf1us5sPuzaP+Bshz41VT/Zxv7eqzx+hl12r+2rOOPRr2Sd5p7U6MN+D6QrmWPpDqY+HH9unOAsqhlPdtS/dj7t0baj+v3P7IrNlRnwcc7wPhmy/ptk2yTyjifvc4ybJOcnGZkYnLtx8fr2INa1vNXdHXysLYT6sMt8c9l/ZMYnfjFLfHrLWMDnVEt6+i0fwt6XUdL3bR5/Z2/CdiiJf479foe10hb0f25dfqmVO3Ct6e472xXbFfmZLvioPCj7Vtdr/es52PdgaWU+5JcARxMddnjNRPFs/Ye8z1bll3dkvaLtmWllIeS3AI8tpF8WR3/T0n+BLgQ+K9Syg9b1jsZPyml3NiS/l/AU4E9qC7T/cM6/f+k/bG9DwcemeRRpZRbG+lXl1Ie2MhthKrx6/azer5dI21P4LZSyndb4seA19Uxn57sBpRS7kxyIXB4kscB/w58jaqyfbAR+uR6m34yTlk9sZ7/NtUl5k3j5XPnxus96/lYyza2/c+22aOef61l2RVUZzOkUfT1Uv+a6Sil/CrJD4CnJHkEVX1XgIPSMnYU1Y+o3wao4xdS1QO/6lpvSfLfVN/FNt3tEdgm2Sb1yDZJmnG9tB971ovGWt7fSduzZdl/TZC2R8uyjZLkMcDJwHOAx1F1EjT9JlUn01T8dr2+paWU1S3LL6Mqgz2BnzbSby+lrPOZpZQ1SX7JuvXqZNmu9Mh2ZXTYoTXabu56vW09/+U48b/oihs3vpRyR5LVjZimO1vS1mxg2WZd6/4j4B3AgVSXvlI3om8qpXxhnO3fkFvGSe/kr5OXzuB5f76B9W0NNCv57vKeqjta0jrl97BG2rbA/46zju59ORWHAm8GDqe6rxzg1iTvA04vpTzE2rLag4l/hGzdnVBKGS+fzYdVTPZ/ts1v1PP19k998Hprd7o0Inqp87anurrqLROsp/P9Hfe71LXeNm3vsU1qZ5vUzjZJmjm91E/bAveVUtrq0Ym+C+utu+5cuG+c+CmrB8T+H2Anqk6d/6Cq0x6iuhprH2CLjfiIqdYJbfUNVHXOw8ZZ1gvblcmxXRkBPuVwtJWu150GZ/448fO74saNT/IbVA1AWyO20Uopy0spLwMeBfwB1WWoOwDnJ1k0xdU+epz08fL9J6WUTDD9pHuzp7hdU3Unve9LqBpvkrQ1lOtVkqWUVaWUk0opO1OdFfgLqobynVRnRJrrP2cDZfW2yWXt1yb7P9um05js0L0gySZU/2PSKOqlzruTamDcLSf4/i5sxEPLd6lrvW3a6kfbpHa2SbZJ0mzrtf3YMknbgfZE34X11l1f8bUlLXUE7R08vXZi/F+qq7JOKaU8o5TyulLKW0spp1LdVrmx+lEn9JPtCrYrTXOhXbFDa265qp4/o3tBki2oLj+9j+ppIxPGU53RaMZMi1LKmlLKN0spb6eqWB4GPG+Kq/utJDu1pD+9nnduPfmfev60KX7OTLkKmJfk/7Qsa9s/t9fz7sfQbg08aaIPKqX8sJTyIaonaEA1pgxUPwbuAp46TuOxsa6q573+z7bp7Nf1Hu9LtY+9UlWj6o+SpJmQ5JFUl8UvL6XcRVXfPYzqVoQJ1WfhlwO/U6+nud5QjdMwGVfVc9ukddkm2SZJs62X9uOqetFk6+S9W9I634fmbeC31/MdWd9TWtLaPKGeX9BMrPO2V0t851azXuuP66i+809LsnnL8k7ZtN36Ph1sVyq2K2uNfLtih9YcUkq5gepe7t9P8pKuxa+nuof83FLK/XXaZ6gq9hOT/LrHv64UOr3Q/9Lv7Uzyf+p7mbt1eqjvneKqN2Xtdnc+60VUZ9u/VpcPVPdQ/xQ4KckftGzfw5P8YXf6LOiU/buaFWySJwHHUvXWf7ER/+16/vJGbIDT6LpMNsmjx7nqYJ19UEpZQ/W4+ycBp7VV9En+MMlWk8jXr03hf7bN16nGRnhRkl8ftCfZlOoWImlU/Q7wZ11pf0M1NsbZ9eszqer5f0rym90rSDI/ye80ks6u3/83XaFHMf74Wa1sk2yT6rltErZJGji9tB+d7/zbmt+pJDsAb6C6teozLet+eZLfbsRvSTXIO411w9o64mX1VSad+BdQ3S7Yi049+vSu9L+i/faxlfW8rRNtPaUaN+s8qu/+XzeXJTkC+D2qwbtvaHn7dLBdqdiuMHfalZHurVOr46juIf9MkkOBH1JVtvtTnXl/YyewlHJ9kjcBfwdck+R8qsesvpBqYOAzSymXTcM2Pgt4b5LLqXqjb6NqWA+kugx00gP/1b4LPDvJ14HLgd+iund6FdUTKoCqcarL5j+AK5J8herR7ptSPeVjH2ApcMAUt6Nf/gV4MVW5fCfJfwDzgJdQ/eA4omtcg09Q/cB4e5I9qR7X+3Sqy1CvZt2GfUfgm0m+SzWg4U1Uj2F+IdUB5T80Yt8KLAJOAl6Y5GtUPwh2qtOfRFUZ3zPFfPb8P9umlPJgkldRDdR4WZLPUo0H8Dyq/+ebprhd0qD7CvDPSZ5PNQbFXlQHAd+jqtcppVyT5DXAB4EfJvky1Y+ieVQDne5NNb5W59aMd1PVO69J8rtUj6R+EvD8+vM6ZzZ7ZZtkm2SbZJukwdNL+zGW5ENU34lrk3yB6ql/h1Hd+vSGUsqPWtZ9CfCNJOdS3UZ1ENX38l9KKZc04v4b+CbwJ8B/pXrwyBPr11+mejLihnya6jv5wST7AjdS1QNPA77E+lfYXkf1HfzTJHdTD1ReSjl9gs94A1U9fHr9GVfW+Xkh8Cuq8pkptiu2K3OvXSkD8KhFp/5ObPiRs4+nqiB+QfVPfgPVwcwO48S/iOpLtorqi/ptWh5fztpHsR7ZsuyTtDyitl62gsajRKkOFP6BqnJZWX/m/wIfoPHY20mWSaF6esQC4PNUByR3UzXYe47znp3rcvkR1WPcb6N6usQHgKc24hYwziNoJ7mNE5Vf6zKqgYvfCCyjuhz1duAiYJ9xPmNRXQ731mV7NlXlPca6j7LdjupM3OVUleDq+v/kc8Aftqx3U+DVVE/SuLNe/4+pzgC9HNi0EbvOZ3Wtp3XZZP5nO/u6JX0/qh9H91IdhH6caqDHdf7/nJyGfaLRBtT/91+v67tb6//7R7e8Zy/g/Pr7fn/9XbuC6kfczl2xj67Xc2u93q/Xn3kqU3iM+mS+33W8bZJtkm2Sk9M0TEyy/aB6qMgxVPXwPVT18teAF7Ws+9R63Yupxir6Qf1dXkF14mTTlvfsQHWVV6eOvJTqaqNfr6tt27vW8RTgq/U67qjrpKe2raOO/6M6D6vq5aWxbLw6YYe6DriBtW3ovwALW2LH/Y5vzPcf25VOvO3KHGtXUmdeGmlJCnBZKWXxbG+LJGlus02SNIiSLKbqNHpbqQZO7+e6T6XqQNi3lDLWz3XLdkVzl2NoSZIkSZIkaajYoSVJkiRJkqSh4qDwGkr1E0727CH030spV03rxkygvrx6g/p9WbckaebYJkmS+sl2ReqNY2hpKCX5JPDnPYQeVUr55PRuzfjq+9k3qJSS6d4WSdL0sE2SJPWT7YrUGzu0gEc96lFlwYIFPcfffffdbL311tO3QZoU98dgcX8Mhm9/+9u3llIePdvbMZdMpi3xe9Jflmf/WJb9NezlaVsy8yZ7XNIx7P9rGzLq+QPzOCpGPY9Tyd90tiXecggsWLCAb33rWz3Hj42NsXjx4unbIE2K+2OwuD8GQ5KfzPY2zDWTaUv8nvSX5dk/lmV/DXt52pbMvMkel3QM+//ahox6/sA8jopRz+NU8jedbYmDwkuSJEmSJGmo2KElSZIkSZKkoWKHliRJkiRJkoaKHVqSJEmSJEkaKnZoSZIkSZIkaajYoSVJkiRJkqShsulsb8AoWnDSl9ZLW3H682ZhSyRJw8q2RJLUq2t+dgdH2m5ImmO8QkuSJEmSJElDxQ4tSZIkSZIkDRU7tCRJkiRJkjRU7NCSJEmSJEnSULFDS5IkSZIkSUPFDi1JkiRJkiQNFTu0JEmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDxQ4tSZIkSZIkDRU7tCRJkiRJkjRU7NCSJEmSJEnSULFDS5IkSZIkSUPFDi1JkiRJkiQNFTu0JEmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDxQ4tSZIkSUMhyU5JPpBkaZJ7kpQkC7piFtTpbdN2XbFbJnlPkpuS3Fuv9xktn7tJkpOTrEhyX5KrkxwyzjYek+QHSVYnuS7Jq/pZBpKkih1akiRJkobFLsBhwG3A1zYQ+y5gr67prq6YjwHHAG8FDgRuAi5KsmdX3DuAU4EPAs8BrgDOT/LcZlCSY4CPAJ8DDgDOB85MclyvGZQk9WbT2d4ASZIkSerR5aWU+QBJXgE8e4LYH5dSrhhvYZI9gMOBo0spn6jTLgOWAW8HDqrTdgBeD5xeSjmjfvulSXYBTge+XMdtCrwT+HQp5ZRG3GOBdyT5aCnlgalkWpK0Pq/QkiRJkjQUSikP9XF1BwEPAOc11r8GOBfYP8kWdfL+wObA2V3vPxvYPcnC+vVewKNb4j4NPBLYu4/bLklznh1akiRJkkbRu5KsSXJHkiVJdu9avhuwvJRyT1f6MqoOrF0acauB61viAHZtxAFcu4E4SVIfeMuhJEmSpFGymmocq68AtwC/DbwJ+O8kf1BK+X4dtz3VWFzdVjaWd+a3l1JKD3G0rLM7bh1JjgWOBZg/fz5jY2PtuZrA/IfDCbuvWS99KusaRKtWrRqZvIzHPI6GUc/joOXPDi1JkiRJI6OUchPQfLLg15JcSHWl1CnAy+r0AN2dVJ307te9xjFO7ETbexZwFsCiRYvK4sWLJ/N2AD5wzhd57zXrH9qtOGLy6xpEY2NjTKVchol5HA2jnsdBy5+3HEqSJEkaaaWUnwL/BTy1kbyS9qum5jWWd+bzknR3YLXF0bLO7buWS5L6wA4tSdKMSLJTkg8kWZrkniQlyYKWuHlJPprk1iR3J7m4ZdwTkmyZ5D1Jbkpyb73eZ7TEbZLk5CQrktyX5Ookh4yzjcck+UGS1UmuS/KqtjhJ0lDqvtJqGbAwyVZdcbsC97N2zKxlwBbAE1riAL7XiIO1Y2mNFydJ6gM7tCRJM2UX4DCqsUW+1hZQn/1eAhwAvAY4BNiM6rHnO3WFfww4BngrcCBwE3BRkj274t4BnAp8EHgOcAVwfpLndn32MVRjrnyu/vzzgTOTHDf5rEqSBkmSnYGnA99oJC+hamMObcRtCrwE+EopZXWdfCFVB9cRXat9GXBtKWV5/XopcOs4cSuBr298TiRJHY6hJUmaKZeXUuYDJHkF8OyWmIOoHmu+Xynl0jp2KbAceAPw2jptD+Bw4OhSyifqtMuozo6/vV4PSXYAXg+cXko5o/6MS5PsApwOfLmO2xR4J/DpUsopjbjHAu9I8tFSygN9KwlJ0pQleXH95+/X8+ckuQW4pZRyWZL3Up24X0o1KPyTgZOBh4C/7aynlHJVkvOA9yfZjKqtOQ5YSKNTqpRyc5L3AScnuQu4kqrTaz/g4EbcA0neQnUy5GfAxXXM0cBrSin397koJGlOs0NLkjQjSikP9RB2EPDzTmdW/b47klxAddDw2kbcA8B5jbg1Sc4FTkqyRX1mfX+qR6+f3fU5ZwMfT7KwPrO+F/DolrhPA0dRdbJdiiRpEJzf9frMen4ZsJjq5MZxwJHAI6iumroEeFsp5bqu9x5FdULjNGA74GrggFLKlV1xpwCrgNcBjwGuAw4rpVzQDCqlfDhJAU4ATgRuAP6ylHImkqS+skNLkjRIdgOubUlfBrw8yTallFV13PJSyj0tcZtT3d64rI5bzdpxUJpxUI1rspy14510f3Yzzg4tSRoApZTuwdm7l38c+HiP67oXOL6eJop7kKrT67Qe1vkRqlvYJUnTyA4tSdIg2R5Y0ZLeeTLUPKoz5NtTjcU1Xtz2jfntpZTuR6i3xdGyzvGeWAVAkmOBYwHmz5/P2NhYW9h6Vq1atcHYE3Zfs15ar+ufa3opT/XGsuwvy1OSpOljh5YkaZB0P4GqmT7dcYwTO65SylnAWQCLFi0qixcv7ul9Y2NjbCj2yJO+tF7aiiN6W/9c00t5qjeWZX9ZnpIkTR+fcihJGiQrab8aal49v63HuJWN+bz66YkbiqNlndt3LZckSZI0AHrq0EqyU5IPJFma5J4kJcmClrh5ST6a5NYkdye5OMnuLXFbJnlPkpuS3Fuv9xktcZskOTnJiiT3Jbk6ySHjbOMxSX6QZHWS65K8qpe8SZIGSmfcq267AjfU42d14hYm2aol7n7Wjpm1DNgCeEJLHMD3GnG0fHZ3nCRJkqQB0OsVWrsAh1GdGf9aW0B99nsJcADwGuAQYDOqx57v1BX+MeAY4K3AgcBNwEVJ9uyKewdwKvBB4DnAFcD5SZ7b9dnHUA28+Ln688+nelzucT3mT5I0GJYAOybZp5OQZFvg+fWyZtxmwKGNuE2pHqP+lfoJhwAXUnVwHcG6XgZcWz/hEKpHu986TtxK4OsbkSdJkiRJfdbrGFqXl1LmAyR5BfDslpiDqB5rvl/ncetJllI9PeoN1I9aT7IHcDhwdCnlE3XaZVRnx99er4ckOwCvB04vpZxRf8alSXYBTge+XMdtSvWo3U+XUk5pxD0WeEeSj5ZSHugxn5KkaZTkxfWfv1/Pn5PkFuCWUsplVB1VS4Gzk5xIdSLlZKoxrt7dWU8p5aok5wHvT7IZVVtzHLCQRqdUKeXmJO8DTk5yF3AlVafXfsDBjbgHkryF6mTIz4CL65ijgdeUUu7vc1FIkiRJ2gg9dWiVUh7qIewg4Oedzqz6fXckuYDqoOG1jbgHgPMacWuSnAuclGSL+sz6/lSPXj+763POBj6eZGF9Zn0v4NEtcZ8GjqLqZPNR65I0GM7ven1mPb8MWFxKeSjJgcAZ9bItqTq49i2l/LTrvUdRndA4DdgOuBo4oJRyZVfcKVRPRnwd8BjgOuCwUsoFzaBSyoeTFOAE4ETgBuAvSylnIkmSJGmg9PMph7sB17akLwNenmSbeuyT3YDlpZR7WuI2p7q9sTOGymrWjoPSjINqXJPlrB3vpPuzm3F2aEnSACildA/O3hazkurKqKM3EHcvcHw9TRT3IFWn12k9fPZHqG5hlyRJkjTA+tmhtT2woiW982SoeVRnyLdn7VOq2uK2b8xvL6V0P0K9LY6WdY73xCoAkhwLHAswf/58xsbG2sJarVq1asL4E3Zfs17aZNavydnQ/tDMcn9IkiRJkqZbPzu0AnR3PnXSpzuOcWLHVUo5CzgLYNGiRWXx4sU9v3dsbIyJ4o886Uvrpa04ovf1a3I2tD80s9wfkiRJkqTp1utTDnuxkvaroebV89t6jFvZmM+rn564oTha1rl913JJkiRJkiSNgH52aHXGveq2K3BDPX5WJ25hkq1a4u5n7ZhZy4AtgCe0xAF8rxFHy2d3x0mSJEmSJGkE9LNDawmwY5J9OglJtgWeXy9rxm0GHNqI25TqMepfqZ9wCHAhVQfXEazrZcC19RMOoXr61a3jxK0Evr4ReZIkSZIkSdKA6XkMrSQvrv/8/Xr+nCS3ALeUUi6j6qhaCpyd5ESqWwxPphrj6t2d9ZRSrkpyHvD+JJtRPanwOGAhjU6pUsrNSd4HnJzkLuBKqk6v/YCDG3EPJHkLcGaSnwEX1zFHA68ppdzfc2lIkiRJkiRp4E1mUPjzu16fWc8vAxaXUh5KciBwRr1sS6oOrn1LKT/teu9RwDupHqG+HXA1cEAp5cquuFOonoz4OuAxwHXAYaWUC5pBpZQPJynACcCJwA3AX5ZSzkSSJEmSJEkjpecOrVJK9+DsbTErqa6MOnoDcfcCx9fTRHEPUnV6ndbDZ38E+MiG4iRJkiRJkjTc+jmGliRJkiRJkjTt7NCSJEmSJEnSULFDS5IkSZIkSUPFDi1JkiRJkiQNFTu0JEmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDxQ4tSZIkSZIkDRU7tCRJkiRJkjRU7NCSJEmSNBSS7JTkA0mWJrknSUmyoCVuXpKPJrk1yd1JLk6ye0vclknek+SmJPfW631GS9wmSU5OsiLJfUmuTnLIONt4TJIfJFmd5Lokr+pL5iVJ67BDS5IkSdKw2AU4DLgN+FpbQJIAS4ADgNcAhwCbAZcm2akr/GPAMcBbgQOBm4CLkuzZFfcO4FTgg8BzgCuA85M8t+uzjwE+Anyu/vzzgTOTHDf5rEqSJrLpbG+AJEmSJPXo8lLKfIAkrwCe3RJzELA3sF8p5dI6dimwHHgD8No6bQ/gcODoUson6rTLgGXA2+v1kGQH4PXA6aWUM+rPuDTJLsDpwJfruE2BdwKfLqWc0oh7LPCOJB8tpTzQt5KQpDnOK7QkSZIkDYVSykM9hB0E/LzTmVW/7w7gAuDgrrgHgPMacWuAc4H9k2xRJ+8PbA6c3fU5ZwO7J1lYv94LeHRL3KeBR1J1skmS+sQOLUmSJEmjZDfg2pb0ZcDOSbZpxC0vpdzTErc51e2NnbjVwPUtcQC7NuJo+ezuOElSH3jLoSRJkqRRsj2woiV9ZT2fB6yq426bIG77xvz2UkrpIY6WdXbHrSPJscCxAPPnz2dsbKwtbELzHw4n7L5mvfSprGsQrVq1amTyMh7zOBpGPY+Dlj87tCRJkiSNkgDdnU+d9OmOY5zYcZVSzgLOAli0aFFZvHjxZN4OwAfO+SLvvWb9Q7sVR0x+XYNobGyMqZTLMDGPo2HU8zho+fOWQ0mSJEmjZCXtV0PNq+e39Ri3sjGfVz89cUNxtKxz+67lkqQ+sENLkiRJ0ihZxtrxrJp2BW4opaxqxC1MslVL3P2sHTNrGbAF8ISWOIDvNeJo+ezuOElSH9ihJUmSJGmULAF2TLJPJyHJtsDz62XNuM2AQxtxmwIvAb5SSlldJ19I1cF1RNfnvAy4tpSyvH69FLh1nLiVwNc3Ik+SpC6OoSVJkiRpaCR5cf3n79fz5yS5BbillHIZVUfVUuDsJCdS3WJ4MtUYV+/urKeUclWS84D3J9kMWA4cByyk0SlVSrk5yfuAk5PcBVxJ1em1H3BwI+6BJG8BzkzyM+DiOuZo4DWllPv7XBSSNKfZoSVJkiRpmJzf9frMen4ZsLiU8lCSA4Ez6mVbUnVw7VtK+WnXe48C3gmcBmwHXA0cUEq5sivuFKonI74OeAxwHXBYKeWCZlAp5cNJCnACcCJwA/CXpZQzkST1lR1akiRJkoZGKaV7cPa2mJVUV0YdvYG4e4Hj62miuAepOr1O6+GzPwJ8ZENxkqSN4xhakiRJkiRJGip2aEmSJEmSJGmo2KElSZIkSZKkoWKHliRJkiRJkoaKHVqSJEmSJEkaKnZoSZIkSZIkaajYoSVJkiRJkqShYoeWJEmSJEmShoodWpIkSZIkSRoqdmhJkiRJkiRpqNihJUmSJEmSpKFih5YkSZIkSZKGih1akiRJkiRJGip2aEmSBkqSxUlKy3R7V9y8JB9NcmuSu5NcnGT3lvVtmeQ9SW5Kcm+SpUme0RK3SZKTk6xIcl+Sq5McMo1ZlSRJkjRFdmhJkgbVa4G9GtOzOguSBFgCHAC8BjgE2Ay4NMlOXev5GHAM8FbgQOAm4KIke3bFvQM4Ffgg8BzgCuD8JM/tZ6YkSZIkbbxNZ3sDJEkax/dLKVeMs+wgYG9gv1LKpQBJlgLLgTdQdYaRZA/gcODoUson6rTLgGXA2+v1kGQH4PXA6aWUM+rPuDTJLsDpwJf7nz1JkiRJU+UVWpKkYXQQ8PNOZxZAKeUO4ALg4K64B4DzGnFrgHOB/ZNsUSfvD2wOnN31OWcDuydZ2PccSJIkSZqyvnZoOe6JJKmPzknyYJJfJflMkp0by3YDrm15zzJg5yTbNOKWl1LuaYnbHNilEbcauL4lDmDXqWZCkiRJUv9N1y2HrwW+2Xi9pvNHY9yThVTjntwGnEx1a8eepZQbG+/7GPA84ETgx8CrqcY92auUclUj7h1Ut4qcAnwb+FOqcU8OLKV4m4gkDZc7gPcClwF3Ak8B3gQsTfKUUsrNwPbAipb3rqzn84BVddxtE8Rt35jfXkopG4hbR5JjgWMB5s+fz9jY2ET5+rVVq1ZtMPaE3desl9br+ueaXspTvbEs+8vylCRp+kxXh5bjnkiSpqSU8h3gO42ky5JcDvwPVRvxZiBAd+cTdXr3637GdW/rWcBZAIsWLSqLFy+eKPzXxsbG2FDskSd9ab20FUf0tv65ppfyVG8sy/6yPCVJmj6zMYaW455IkiallHIl8EPgqXXSStqvmppXz2/rMW5lYz6vvop4ojhJkiRJA2C6rtA6J8mjgNuBi4CTSik31MsmGvfk5Um2KaWsordxT5bR27gnyzcuO5KkAdC8imoZ8OyWmF2BG+p2pBP3wiRbdbUnuwL3s7btWAZsATyBdduTzthZ39v4zW+3oOVKLEmSJEkT63eH1siPewIbHg/BcU9mluNTDBb3h6ZDkkXAk4B/rZOWAEcl2aeUclkdsy3wfOAzjbcuAd4GHAp8qo7bFHgJ8JVSyuo67kKqDq4j6viOlwHXllI8MSJJkiQNkL52aM2FcU9gw+MhOO7JzHJ8isHi/tDGSnIO1ZW1V1Jd6fsUqoeH/Az4QB22BFgKnJ3kRNY+YCTAuzvrKqVcleQ84P1JNqvXexzVg0mOaMTdnOR9wMlJ7qo/+yXAfqx7O7wkSZKkATBdtxz+WinlyiRTHfdk5wni1hv3pOsqLcc9kaThdC3wUqon4W4F/AL4PPA3pZRbAUopDyU5EDgDOBPYkqqDa99Syk+71ncU8E7gNGA74GrggHpcrqZTqK4Qfh3wGOA64LBSygX9zqAkSZKkjTPtHVq1kRz3RJLUf6WUdwHv6iFuJXB0PU0Udy9wfD1NFPcgVafXaT1vrCRJkqRZMe1POWyMe/KNOmkJsGOSfRoxnXFPljTeugTYjGrck07chsY9aXLcE0mSJEmSpBHU1yu0HPdEkiRJkiRJ063ftxw67okkSZIkSZKmVb+fcui4J5IkSZIkSZpW0z6GliRJkiTNpCSLk5SW6fauuHlJPprk1iR3J7k4ye4t69syyXuS3JTk3iRLkzyjJW6TJCcnWZHkviRXJzlkGrMqSXOWHVqSJEmSRtVrgb0a07M6C5KEanzfA6iGTDmE6qFUlybZqWs9HwOOAd4KHAjcBFyUZM+uuHcApwIfBJ4DXAGcn+S5/cyUJKn/Y2hJkiRJ0qD4finlinGWHQTsDexXSrkUIMlSqodRvYGqM4wkewCHA0eXUj5Rp10GLAPeXq+HJDsArwdOL6WcUX/GpUl2AU4Hvtz/7EnS3OUVWpIkSZLmooOAn3c6swBKKXcAF7Du09IPAh4AzmvErQHOBfZPskWdvD+wOXB21+ecDeyeZGHfcyBJc5gdWpIkSZJG1TlJHkzyqySfSbJzY9luVE9p77YM2DnJNo245aWUe1riNgd2acStBq5viQPYdaqZkCStz1sOJUmSJI2aO4D3ApcBdwJPAd4ELE3ylFLKzcD2wIqW966s5/OAVXXcbRPEbd+Y315KKRuIW0eSY4FjAebPn8/Y2NhE+Wo1/+Fwwu5r1kufyroG0apVq0YmL+Mxj6Nh1PM4aPmzQ0uSJEnSSCmlfAf4TiPpsiSXA/9DNTbWm4EA3Z1P1Ondr/sZ172tZwFnASxatKgsXrx4ovBWHzjni7z3mvUP7VYcMfl1DaKxsTGmUi7DxDyOhlHP46Dlz1sOJUmSJI28UsqVwA+Bp9ZJK2m/ampePb+tx7iVjfm8+umJE8VJkvrADi1JkiRJc0XzKqplVONeddsVuKGUsqoRtzDJVi1x97N2zKxlwBbAE1riAL63EdstSepih5YkSZKkkZdkEfAk4Bt10hJgxyT7NGK2BZ5fL6MRtxlwaCNuU+AlwFdKKavr5AupOriO6ProlwHXllKW9y83kiTH0JIkSZI0UpKcAywHrgRupxoU/mTgZ8AH6rAlwFLg7CQnUt1ieDLVVVzv7qyrlHJVkvOA9yfZrF7vccBCGp1XpZSbk7wPODnJXfVnvwTYDzh42jIrSXOUHVqSJEmSRs21wEuB1wBbAb8APg/8TSnlVoBSykNJDgTOAM4EtqTq4Nq3lPLTrvUdBbwTOA3YDrgaOKAel6vpFKonI74OeAxwHXBYKeWCfmdQkuY6O7QkSZIkjZRSyruAd/UQtxI4up4mirsXOL6eJop7kKrT67SeN1aSNCWOoSVJkiRJkqShYoeWJEmSJEmShoodWpIkSZIkSRoqdmhJkiRJkiRpqNihJUmSJEmSpKHiUw5nyIKTvtSavuL0583wlkiSJEmSJA03r9CSJEmSJEnSULFDS5IkSZIkSUPFDi1JkiRJkiQNFTu0JEmSJEmSNFTs0JIkSZIkSdJQsUNLkiRJkiRJQ8UOLUmSJEmSJA0VO7QkSZIkSZI0VOzQkiRJkiRJ0lCxQ0uSJEmSJElDxQ4tSZIkSZIkDRU7tCRJkiRJkjRUNp3tDRhmC0760mxvgiRJkiRJ0pzjFVqSJEmSJEkaKnZoSZIkSZIkaajYoSVJkiRJkqShYoeWJEmSJEmShoodWpIkSZIkSRoqdmhJkiRJkiRpqNihJUmSJEmSpKFih5YkSZIkSZKGih1akiRJkiRJGip2aEmSJEmSJGmo2KElSZIkSZKkoTIyHVpJHpfk35LckeTOJJ9PsvNsb5ckaXjYlkiSNpZtiSTNjJHo0EqyFXAJ8NvAnwN/BjwRuDTJ1rO5bZKk4WBbIknaWLYlkjRzNp3tDeiTY4DHA08upVwPkOS7wP8CrwT+fha3TZI0HGxLJEkby7ZEkmbIqHRoHQRc0Wk0AEopy5N8HTiYAW44Fpz0pfXSVpz+vFnYEkma8wa+LWlrM8B2Q5IGyMC3JZI0KkalQ2s34Ist6cuAQ2d4WzaaByySNCuGti0Zr91oY1siSdNqaNsSSRo2o9KhtT1wW0v6SmBe2xuSHAscW79cleS6SXzeo4BbJ7WFfZC/m+lPHBqzsj80LvfHYPit2d6AITSTbcmsfU9GtC2x3ukfy7K/hr08bUsmb6aPSzpa/9dGqM4f9u9SL8zjaBj1PE4lf9PWloxKhxZAaUnLuMGlnAWcNZUPSvKtUsqiqbxX/ef+GCzuDw25GWlL/J70l+XZP5Zlf1mec9aMHZf8euUj/r826vkD8zgqRj2Pg5a/kXjKIdVZkO1b0ufRfoZEkqRutiWSpI1lWyJJM2RUOrSWUd2v3m1X4HszvC2SpOFkWyJJ2li2JZI0Q0alQ2sJ8LQkj+8kJFkAPL1e1m8bdUmw+s79MVjcHxpWM9mW+D3pL8uzfyzL/rI8556ZPi7pGPX/tVHPH5jHUTHqeRyo/KWUtlu8h0uSrYGrgXuBN1Pdt/4O4BHA75ZSVs3i5kmShoBtiSRpY9mWSNLMGYkrtEopdwP7AT8EPg2cAywH9rPRkCT1wrZEkrSxbEskaeaMxBVakiRJkiRJmjtG4gqtmZDkcUn+LckdSe5M8vkkO8/2do26JIuTlJbp9q64eUk+muTWJHcnuTjJ7rO02SMjyU5JPpBkaZJ76rJf0BLXU/kn2TLJe5LclOTeer3PmJHMSLPMdmRq+l0PzWVJXpzkc0l+UtfB1yV5V5JHdMVZlj1Isn+SS5L8IsnqJDcm+dcku3bFWZ6aFsPYrvRSDyVZMM7v/5Jku671Ddxvy34fvwxoHscm2EcX1jFDsx/7/Vuj1/wk2STJyUlWJLkvydVJDpmtPCZ5ZpKzk/yo3u4fJflQkh1a1jfevt1zJvNoh1YPkmwFXAL8NvDnwJ8BTwQuTXWfvKbfa4G9GtOzOguShGqQzQOA1wCHAJtR7Z+dZn5TR8ouwGFUj5n+WlvAJMv/Y8AxwFuBA4GbgIu6Kz5p1NiObJR+10Nz2euBB4E3UZXVh4DjgK8m2QQsy0naHvg28JfAs4GTqZ5ud0WS3wLLU9NniNuVDdZDDe9i3d//ewF3dcUM8m/Lfh2/DGIe/4L1983x9bLuhx8Mw36crWOedwCnAh8EngNcAZyf5LkbnaP1bTCPwKuARwKnUeXzXcBBVO3aNi3xn2T9ffvDrpjpzWMpxWkDE/A6qop3l0baQmANcPxsb98oT8BiqsE0nzVBzMF1zL6NtN8AVgL/ONt5GOYJ2KTx9yvqcl4wlfIH9qjjjmqkbQpcByyZ7bw6OU3nZDuyUWXXt3pork/Ao1vSXl6X3X6WZV/K+Ml1+Z1geTpN5zSs7UqP9dCC+vUrNrCugfxt2c/jl0HN4zh5+hiwGth+2PZjP39r9JofYIe6vN7W9Tn/CXx3lvLY9v18Rh17dFd6AU7bwGdOex69Qqs3BwFXlFKu7ySUUpYDX6f6x9bsOgj4eSnl0k5CKeUO4ALcPxullPJQD2G9lv9BwAPAeY24NcC5wP5JtujLRkuDyXZkivpcD81ppZRbWpK/Wc93rOeW5cb5VT1/oJ5bnpouQ9mu9FgP9WqYf1uO1O/nJA8HDgUuKKWsnOTbZz2Ps3TMsz+wOXB21+ecDeyeZOFk8zGRXvLY5+8nzEAe7dDqzW7AtS3py4BdW9LVf+ckeTDJr5J8JuuODzDR/tl5nMsj1T+9lv9uwPJSyj0tcZtTXQYrjar/3969hVpRxXEc//5Ru1AanpculplRVpJRlEUaWgpBJYYIBqlgD4U9FSRm9RBR5ItRvXSDSMjsJfJSFJWlaZldFE+dREmylC5qJ9OwLOHfw382jXP2OWf2OXv23nPO7wODOrNmnLX+e681s/aaNWpHiqV2oO+mJH/uSP5UWdbIzIaY2UlmdhHwAvALceMCKk8pzkBqV7L1UMWTZnbcYo6wNVXmKmr1a8t63L+0eh4rZgHDgeVVtpU9jhX1jtl4YvTSd1XSQet8j7v7fgIstJhD8qjFnJI3ZLYXnkd1aOXTRjxrmtUJjGzwuQw2fwDLiGGRNxHP4E4HNqcmp+spPqAYFS1v+feWrq3O5yXSStSOFEvtQB+Y2SjgMeADd/8yWa2yrN0W4oJ9FzCBeGxqf7JN5SlFGRDtSjf10DGic/ge4EZi3q3LgU/N7NLU7q16bVnP+5dWzWPWfGA/8E5qXdnjmFXvmLUBhzx5Bq+HdE1j8bKGp4nOrFWZza8Sc6lNB+4m5t760MymptIUnseh/T3AIJINAoA1/CwGGXffBmxLrdpgZh8DnxMTLT5CxEHxaZ685a84yWCnz39xVL/UKPkleTUx386C9CZUlrWaB4wAxhI3bO+b2WR334PKU4pV6s9Wd/WQu/9MTE5dsdHizXkdwMPA3MohaMEyqPP9S0vmMc3MziE6NZ5JHq0Dyh/HKuods5bOt5kNBVYSjxpOSscWwN3npf650cxWEyPYHgcmVw5DwXnUCK18fqd67+FIqve+SoHcfSvxK+g1yapOuo8PKEZFy1v+vaWr9Xl7kTJRO1IstQM1MLNTiDc1jQVudvd9qc0qyxq5+w533+LuK4FpwOnAg8lmlacUpdTtSi/1UBfuvhfYxP/X/1Cia8t+3L+UIY9ziX6Fao8bnqDkcax3zDqBkcnbE3tK13DJG0eXEx2Vt7t7e2/7uPsR4G26xrbQPKpDK58O4vnPrMuAbxt8LhLSvb09xedHd/+zYWc1OOUt/w7gguQ109l0/9D12WqRgUTtSLHUDuRkZsOAN4CJwC3u/nUmicqyH9z9ENGeVeZIUXlKUUrbruSoh7rdlRNHe5Tt2rIv9y9lyON8YLu7b8+ZvqxxrHfMOoCTgQurpIPmfo+fB+YAd7j7uhr2qxbbQvOoDq181gDXmdnYygozGwNMSrZJA5nZ1cDFxJwVEDEYZWZTUmlGADNQfBohb/mvAYYRb0CppBtKVJbvufuxxpyuSFOoHSmW2oEckl9cVxCjiGa6+2dVkqks+8HMzgQuAXYnq1SeUpRStis566Fq+40m8rYltbo015b9uH9p6Twm+RpPjtFZSfoyx7HeMXuX6OC6M/P/zAW+Sd5a2nBmVpn/bYG7r6phvxHArZwY28LzaF3n55IsMzsN2A78RTzz7MTkfsOBCfp1rThmtgL4HtgKHAKuBJYAR4Gr3P1g0jBuAs4DFhHDPZcQE7NekQxtlT4ys9nJX6cRz8HfCxwADrj7hlrK38xeJ17fuoiI60LgNuD6ZCi2yICkdqR/6lkPDWZm9hxRfk8Ab2U273P3fSrL/MzsTeL6pB04TNys3g+cBUx0910qTylKWduVnPXQMmLgxWairh9HfG/OAK51952p47XctWW9719aMY+pc3s2OZ9z3f3XzLZSxbEZ9zxmthS4D3iI+LzMISbRn+nua5uQx8XAUuBl4KXM7gfcfXdynAeIeH4E/AScT8whOQ6Y5u4bG5ZHd9eSYwFGE0NjDwNHiFn+xzT7vAb6QlQS7cTbQv4F9gIvAmdn0rUlX7xOorFYR1QsTc9D2RfiAqnasr7W8gdOBZ4iXmn+N9GDP7XZedSipRGL2pF+lV3d6qHBvAB7eijLR1WWNZfnYuAr4ob1KLCTeKPXmEw6laeWQpYytit56iHgLuALosPgeHLd+BowrsrxWu7ast73L62Yx+S8hhGdIWu72V6qONbzWiNvfoAhRIf0D8RbIduB2c3KI7C+hzSvpI4zA/gEOJh8xn8jRqZNbHQeNUJLRERERERERERKRXNoiYiIiIiIiIhIqahDS0RERERERERESkUdWiIiIiIiIiIiUirq0BIRERERERERkVJRh5aIiIiIiIiIiJSKOrRERERERERERKRU1KElIiIiIiIiIiKlog4tEREREREREREplf8AG67WK4Cl0u8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x1080 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## (optional)\n",
    "train_df.hist(bins=50, figsize=(20, 15));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## (optional)\n",
    "train_df.plot(\n",
    "    kind=\"scatter\",\n",
    "    x=\"longitude\",\n",
    "    y=\"latitude\",\n",
    "    alpha=0.4,\n",
    "    s=train_df[\"population\"] / 100,\n",
    "    figsize=(10, 7),\n",
    "    c=\"median_house_value\",\n",
    "    cmap=plt.get_cmap(\"jet\"),\n",
    "    colorbar=True,\n",
    "    sharex=False,\n",
    ");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What all transformations we need to apply on the dataset? \n",
    "\n",
    "Here is what we see from the EDA. \n",
    "\n",
    "- Some missing values in `total_bedrooms` column\n",
    "- Scales are quite different across columns. \n",
    "- Categorical variable `ocean_proximity`\n",
    "\n",
    "Read about [preprocessing techniques implemented in `scikit-learn`](https://scikit-learn.org/stable/modules/preprocessing.html).  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We are droping the categorical variable ocean_proximity for now. We'll come back to it in a bit.\n",
    "X_train = train_df.drop(columns=[\"median_house_value\", \"ocean_proximity\"])\n",
    "y_train = train_df[\"median_house_value\"]\n",
    "\n",
    "X_test = test_df.drop(columns=[\"median_house_value\", \"ocean_proximity\"])\n",
    "y_test = test_df[\"median_house_value\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Let's first run our baseline model `DummyRegressor`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "results_dict = {}  # dictionary to store our results for different models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "def mean_std_cross_val_scores(model, X_train, y_train, **kwargs):\n",
    "    \"\"\"\n",
    "    Returns mean and std of cross validation\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    model :\n",
    "        scikit-learn model\n",
    "    X_train : numpy array or pandas DataFrame\n",
    "        X in the training data\n",
    "    y_train :\n",
    "        y in the training data\n",
    "\n",
    "    Returns\n",
    "    ----------\n",
    "        pandas Series with mean scores from cross_validation\n",
    "    \"\"\"\n",
    "\n",
    "    scores = cross_validate(model, X_train, y_train, **kwargs)\n",
    "\n",
    "    mean_scores = pd.DataFrame(scores).mean()\n",
    "    std_scores = pd.DataFrame(scores).std()\n",
    "    out_col = []\n",
    "\n",
    "    for i in range(len(mean_scores)):\n",
    "        out_col.append((f\"%0.3f (+/- %0.3f)\" % (mean_scores[i], std_scores[i])))\n",
    "\n",
    "    return pd.Series(data=out_col, index=mean_scores.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "dummy = DummyRegressor(strategy=\"median\")\n",
    "results_dict[\"dummy\"] = mean_std_cross_val_scores(\n",
    "    dummy, X_train, y_train, return_train_score=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>dummy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fit_time</th>\n",
       "      <td>0.001 (+/- 0.001)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>score_time</th>\n",
       "      <td>0.000 (+/- 0.000)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>test_score</th>\n",
       "      <td>-0.055 (+/- 0.012)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_score</th>\n",
       "      <td>-0.055 (+/- 0.001)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          dummy\n",
       "fit_time      0.001 (+/- 0.001)\n",
       "score_time    0.000 (+/- 0.000)\n",
       "test_score   -0.055 (+/- 0.012)\n",
       "train_score  -0.055 (+/- 0.001)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(results_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Imputation "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "knn = KNeighborsRegressor()\n",
    "# knn.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What's the problem? \n",
    "\n",
    "```\n",
    "ValueError: Input contains NaN, infinity or a value too large for dtype('float64').\n",
    "```\n",
    "\n",
    "- The classifier is not able to deal with missing values (NaNs).\n",
    "- What are possible ways to deal with the problem? \n",
    "    - Delete the rows? \n",
    "    - Replace them with some reasonable values?    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- `SimpleImputer` is a transformer in `sklearn` to deal with this problem. For example, \n",
    "    - You can impute missing values in categorical columns with the most frequent value.\n",
    "    - You can impute the missing values in numeric columns with the mean or median of the column.    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20248</th>\n",
       "      <td>-119.23</td>\n",
       "      <td>34.25</td>\n",
       "      <td>28.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>8.0000</td>\n",
       "      <td>2.888889</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>3.222222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12649</th>\n",
       "      <td>-121.47</td>\n",
       "      <td>38.51</td>\n",
       "      <td>52.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.6250</td>\n",
       "      <td>2.222222</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>8.222222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3125</th>\n",
       "      <td>-117.76</td>\n",
       "      <td>35.22</td>\n",
       "      <td>4.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.6250</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12138</th>\n",
       "      <td>-117.22</td>\n",
       "      <td>33.87</td>\n",
       "      <td>16.0</td>\n",
       "      <td>56.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>2.6250</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>2.785714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8219</th>\n",
       "      <td>-118.21</td>\n",
       "      <td>33.79</td>\n",
       "      <td>33.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>96.0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>4.5938</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>2.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4591</th>\n",
       "      <td>-118.28</td>\n",
       "      <td>34.06</td>\n",
       "      <td>42.0</td>\n",
       "      <td>2472.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3795.0</td>\n",
       "      <td>1179.0</td>\n",
       "      <td>1.2254</td>\n",
       "      <td>2.096692</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.218830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19485</th>\n",
       "      <td>-120.98</td>\n",
       "      <td>37.66</td>\n",
       "      <td>10.0</td>\n",
       "      <td>934.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>401.0</td>\n",
       "      <td>255.0</td>\n",
       "      <td>0.9336</td>\n",
       "      <td>3.662745</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.572549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6962</th>\n",
       "      <td>-118.05</td>\n",
       "      <td>33.99</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1619.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>886.0</td>\n",
       "      <td>357.0</td>\n",
       "      <td>3.7328</td>\n",
       "      <td>4.535014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.481793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14970</th>\n",
       "      <td>-117.01</td>\n",
       "      <td>32.74</td>\n",
       "      <td>31.0</td>\n",
       "      <td>3473.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2098.0</td>\n",
       "      <td>677.0</td>\n",
       "      <td>2.6973</td>\n",
       "      <td>5.129985</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.098966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7763</th>\n",
       "      <td>-118.10</td>\n",
       "      <td>33.91</td>\n",
       "      <td>36.0</td>\n",
       "      <td>726.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>490.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>3.6389</td>\n",
       "      <td>5.584615</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.769231</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>18576 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "20248    -119.23     34.25                28.0         26.0             3.0   \n",
       "12649    -121.47     38.51                52.0         20.0             4.0   \n",
       "3125     -117.76     35.22                 4.0         18.0             3.0   \n",
       "12138    -117.22     33.87                16.0         56.0             7.0   \n",
       "8219     -118.21     33.79                33.0         32.0            18.0   \n",
       "...          ...       ...                 ...          ...             ...   \n",
       "4591     -118.28     34.06                42.0       2472.0             NaN   \n",
       "19485    -120.98     37.66                10.0        934.0             NaN   \n",
       "6962     -118.05     33.99                38.0       1619.0             NaN   \n",
       "14970    -117.01     32.74                31.0       3473.0             NaN   \n",
       "7763     -118.10     33.91                36.0        726.0             NaN   \n",
       "\n",
       "       population  households  median_income  rooms_per_household  \\\n",
       "20248        29.0         9.0         8.0000             2.888889   \n",
       "12649        74.0         9.0         3.6250             2.222222   \n",
       "3125          8.0         6.0         1.6250             3.000000   \n",
       "12138        39.0        14.0         2.6250             4.000000   \n",
       "8219         96.0        36.0         4.5938             0.888889   \n",
       "...           ...         ...            ...                  ...   \n",
       "4591       3795.0      1179.0         1.2254             2.096692   \n",
       "19485       401.0       255.0         0.9336             3.662745   \n",
       "6962        886.0       357.0         3.7328             4.535014   \n",
       "14970      2098.0       677.0         2.6973             5.129985   \n",
       "7763        490.0       130.0         3.6389             5.584615   \n",
       "\n",
       "       bedrooms_per_household  population_per_household  \n",
       "20248                0.333333                  3.222222  \n",
       "12649                0.444444                  8.222222  \n",
       "3125                 0.500000                  1.333333  \n",
       "12138                0.500000                  2.785714  \n",
       "8219                 0.500000                  2.666667  \n",
       "...                       ...                       ...  \n",
       "4591                      NaN                  3.218830  \n",
       "19485                     NaN                  1.572549  \n",
       "6962                      NaN                  2.481793  \n",
       "14970                     NaN                  3.098966  \n",
       "7763                      NaN                  3.769231  \n",
       "\n",
       "[18576 rows x 11 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.sort_values(\"bedrooms_per_household\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2064, 11)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape\n",
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "imputer = SimpleImputer(strategy=\"median\")\n",
    "imputer.fit(X_train)\n",
    "X_train_imp = imputer.transform(X_train)\n",
    "X_test_imp = imputer.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- Let's check whether the NaN values have been replaced or not\n",
    "- Note that `imputer.transform` returns an `numpy` array and not a dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scaling "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "- This problem affects a large number of ML methods.\n",
    "- A number of approaches to this problem. We are going to look into two most popular ones.  \n",
    "\n",
    "| Approach | What it does | How to update $X$ (but see below!) | sklearn implementation | \n",
    "|---------|------------|-----------------------|----------------|\n",
    "| normalization | sets range to $[0,1]$   | `X -= np.min(X,axis=0)`<br>`X /= np.max(X,axis=0)`  | [`MinMaxScaler()`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.MinMaxScaler.html)\n",
    "| standardization | sets sample mean to $0$, s.d. to $1$   | `X -= np.mean(X,axis=0)`<br>`X /=  np.std(X,axis=0)` | [`StandardScaler()`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html#sklearn.preprocessing.StandardScaler) |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "There are all sorts of articles on this; see, e.g. [here](http://www.dataminingblog.com/standardization-vs-normalization/) and [here](https://medium.com/@rrfd/standardize-or-normalize-examples-in-python-e3f174b65dfc)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# [source](https://amueller.github.io/COMS4995-s19/slides/aml-05-preprocessing/#8)\n",
    "mglearn.plots.plot_scaling()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.908140</td>\n",
       "      <td>-0.743917</td>\n",
       "      <td>-0.526078</td>\n",
       "      <td>0.143120</td>\n",
       "      <td>0.235339</td>\n",
       "      <td>1.026092</td>\n",
       "      <td>0.266135</td>\n",
       "      <td>-0.389736</td>\n",
       "      <td>-0.210591</td>\n",
       "      <td>-0.083813</td>\n",
       "      <td>0.126398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.002057</td>\n",
       "      <td>1.083123</td>\n",
       "      <td>-0.923283</td>\n",
       "      <td>-1.049510</td>\n",
       "      <td>-0.969959</td>\n",
       "      <td>-1.206672</td>\n",
       "      <td>-1.253312</td>\n",
       "      <td>-0.198924</td>\n",
       "      <td>4.726412</td>\n",
       "      <td>11.166631</td>\n",
       "      <td>-0.050132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.218207</td>\n",
       "      <td>-1.352930</td>\n",
       "      <td>1.380504</td>\n",
       "      <td>0.329670</td>\n",
       "      <td>0.549764</td>\n",
       "      <td>0.024896</td>\n",
       "      <td>0.542873</td>\n",
       "      <td>-0.635239</td>\n",
       "      <td>-0.273606</td>\n",
       "      <td>-0.025391</td>\n",
       "      <td>-0.099240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.128188</td>\n",
       "      <td>-0.753286</td>\n",
       "      <td>-0.843842</td>\n",
       "      <td>-0.459154</td>\n",
       "      <td>-0.626949</td>\n",
       "      <td>-0.463877</td>\n",
       "      <td>-0.561467</td>\n",
       "      <td>0.714077</td>\n",
       "      <td>0.122307</td>\n",
       "      <td>-0.280310</td>\n",
       "      <td>0.010183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.168196</td>\n",
       "      <td>-1.287344</td>\n",
       "      <td>-0.843842</td>\n",
       "      <td>1.343085</td>\n",
       "      <td>2.210026</td>\n",
       "      <td>4.269688</td>\n",
       "      <td>2.500924</td>\n",
       "      <td>-1.059242</td>\n",
       "      <td>-0.640266</td>\n",
       "      <td>-0.190617</td>\n",
       "      <td>0.126808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18571</th>\n",
       "      <td>0.733102</td>\n",
       "      <td>-0.804818</td>\n",
       "      <td>0.586095</td>\n",
       "      <td>-0.875337</td>\n",
       "      <td>-0.243446</td>\n",
       "      <td>-0.822136</td>\n",
       "      <td>-0.966131</td>\n",
       "      <td>-0.118182</td>\n",
       "      <td>0.063110</td>\n",
       "      <td>-0.099558</td>\n",
       "      <td>0.071541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18572</th>\n",
       "      <td>1.163195</td>\n",
       "      <td>-1.057793</td>\n",
       "      <td>-1.161606</td>\n",
       "      <td>0.940194</td>\n",
       "      <td>0.609314</td>\n",
       "      <td>0.882438</td>\n",
       "      <td>0.728235</td>\n",
       "      <td>0.357500</td>\n",
       "      <td>0.235096</td>\n",
       "      <td>-0.163397</td>\n",
       "      <td>0.007458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18573</th>\n",
       "      <td>-1.097293</td>\n",
       "      <td>0.797355</td>\n",
       "      <td>-1.876574</td>\n",
       "      <td>0.695434</td>\n",
       "      <td>0.433046</td>\n",
       "      <td>0.881563</td>\n",
       "      <td>0.514155</td>\n",
       "      <td>0.934269</td>\n",
       "      <td>0.211892</td>\n",
       "      <td>-0.135305</td>\n",
       "      <td>0.044029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18574</th>\n",
       "      <td>-1.437367</td>\n",
       "      <td>1.008167</td>\n",
       "      <td>1.221622</td>\n",
       "      <td>-0.499947</td>\n",
       "      <td>-0.484029</td>\n",
       "      <td>-0.759944</td>\n",
       "      <td>-0.454427</td>\n",
       "      <td>0.006578</td>\n",
       "      <td>-0.273382</td>\n",
       "      <td>-0.149822</td>\n",
       "      <td>-0.132875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18575</th>\n",
       "      <td>0.242996</td>\n",
       "      <td>0.272667</td>\n",
       "      <td>-0.684960</td>\n",
       "      <td>-0.332190</td>\n",
       "      <td>-0.353018</td>\n",
       "      <td>-0.164307</td>\n",
       "      <td>-0.396991</td>\n",
       "      <td>-0.711754</td>\n",
       "      <td>0.025998</td>\n",
       "      <td>0.042957</td>\n",
       "      <td>0.051269</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>18576 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "0       0.908140 -0.743917           -0.526078     0.143120        0.235339   \n",
       "1      -0.002057  1.083123           -0.923283    -1.049510       -0.969959   \n",
       "2       1.218207 -1.352930            1.380504     0.329670        0.549764   \n",
       "3       1.128188 -0.753286           -0.843842    -0.459154       -0.626949   \n",
       "4       1.168196 -1.287344           -0.843842     1.343085        2.210026   \n",
       "...          ...       ...                 ...          ...             ...   \n",
       "18571   0.733102 -0.804818            0.586095    -0.875337       -0.243446   \n",
       "18572   1.163195 -1.057793           -1.161606     0.940194        0.609314   \n",
       "18573  -1.097293  0.797355           -1.876574     0.695434        0.433046   \n",
       "18574  -1.437367  1.008167            1.221622    -0.499947       -0.484029   \n",
       "18575   0.242996  0.272667           -0.684960    -0.332190       -0.353018   \n",
       "\n",
       "       population  households  median_income  rooms_per_household  \\\n",
       "0        1.026092    0.266135      -0.389736            -0.210591   \n",
       "1       -1.206672   -1.253312      -0.198924             4.726412   \n",
       "2        0.024896    0.542873      -0.635239            -0.273606   \n",
       "3       -0.463877   -0.561467       0.714077             0.122307   \n",
       "4        4.269688    2.500924      -1.059242            -0.640266   \n",
       "...           ...         ...            ...                  ...   \n",
       "18571   -0.822136   -0.966131      -0.118182             0.063110   \n",
       "18572    0.882438    0.728235       0.357500             0.235096   \n",
       "18573    0.881563    0.514155       0.934269             0.211892   \n",
       "18574   -0.759944   -0.454427       0.006578            -0.273382   \n",
       "18575   -0.164307   -0.396991      -0.711754             0.025998   \n",
       "\n",
       "       bedrooms_per_household  population_per_household  \n",
       "0                   -0.083813                  0.126398  \n",
       "1                   11.166631                 -0.050132  \n",
       "2                   -0.025391                 -0.099240  \n",
       "3                   -0.280310                  0.010183  \n",
       "4                   -0.190617                  0.126808  \n",
       "...                       ...                       ...  \n",
       "18571               -0.099558                  0.071541  \n",
       "18572               -0.163397                  0.007458  \n",
       "18573               -0.135305                  0.044029  \n",
       "18574               -0.149822                 -0.132875  \n",
       "18575                0.042957                  0.051269  \n",
       "\n",
       "[18576 rows x 11 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train_imp)\n",
    "X_test_scaled = scaler.transform(X_test_imp)\n",
    "pd.DataFrame(X_train_scaled, columns=X_train.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8090877831586284"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsRegressor()\n",
    "knn.fit(X_train_scaled, y_train)\n",
    "knn.score(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "- Big difference in the KNN training performance after scaling the data. \n",
    "- But we saw last week that training score doesn't tell us much. We should look at the cross-validation score."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### ❓❓ Questions for class discussion "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    },
    "tags": []
   },
   "source": [
    "#### True/False on scaling and imputation\n",
    "\n",
    "1. `StandardScaler` ensures a fixed range (i.e., minimum and maximum values) for the features. \n",
    "2. `StandardScaler` calculates mean and standard deviation for each feature separately. \n",
    "3. In general, it's a good idea to apply scaling on numeric features before training $k$-NN or SVM RBF models. \n",
    "4. After normalization, all numerical features in training and test set will have values between 0 and 1.\n",
    "5. It is OK to modify features before splitting into training and testing set if the modification is row-wise.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```{admonition} True or False\n",
    "1. False\n",
    "2. True\n",
    "3. True\n",
    "4. False\n",
    "5. True (but we don't recommend it)\n",
    "```    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### More True/False on scaling and imputation\n",
    "\n",
    "4. The transformers such as `StandardScaler` or `SimpleImputer` in `scikit-learn` return a dataframe with transformed features.\n",
    "5. The transformed feature values might be hard to interpret for humans.\n",
    "6. After applying `SimpleImputer` The transformed data has a different shape than the original data.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Consider a toy data with the following two columns. If you apply `StandardScaler` on this data, both columns A and B will end up being identical. True or False?  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>-2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>40</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    A  B\n",
       "0  10 -2\n",
       "1  20 -1\n",
       "2  30  0\n",
       "3  40  1\n",
       "4  50  2"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ex_test_cols = np.array([[10, -2], [20, -1], [30, 0], [40, 1], [50, 2]])\n",
    "ex_test_df = pd.DataFrame(data=ex_test_cols, columns=[\"A\", \"B\"])\n",
    "ex_test_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "Let's create some synthetic data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.datasets import make_blobs, make_classification\n",
    "\n",
    "# make synthetic data\n",
    "X, y = make_blobs(n_samples=100, centers=3, random_state=12, cluster_std=4)\n",
    "# split it into training and test sets\n",
    "X_train_toy, X_test_toy, y_train_toy, y_test_toy = train_test_split(\n",
    "    X, y, random_state=5, test_size=0.2\n",
    ")\n",
    "plt.scatter(X_train_toy[:, 0], X_train_toy[:, 1], label=\"Training set\", s=60)\n",
    "plt.scatter(\n",
    "    X_test_toy[:, 0], X_test_toy[:, 1], color=mglearn.cm2(1), label=\"Test set\", s=60\n",
    ")\n",
    "plt.legend(loc=\"upper right\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "Let's transform the data using `StandardScaler` and examine how the data looks like. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "train_transformed = scaler.fit_transform(X_train_toy)\n",
    "test_transformed = scaler.transform(X_test_toy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_original_scaled(X_train_toy, X_test_toy, train_transformed, test_transformed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training score: 0.75\n",
      "Test score: 0.55\n"
     ]
    }
   ],
   "source": [
    "knn = KNeighborsClassifier()\n",
    "knn.fit(train_transformed, y_train_toy)\n",
    "print(f\"Training score: {knn.score(train_transformed, y_train_toy):.2f}\")\n",
    "print(f\"Test score: {knn.score(test_transformed, y_test_toy):.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "#### Bad methodology 1: Scaling the data separately (for class discussion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training score: 0.75\n",
      "Test score: 0.50\n"
     ]
    }
   ],
   "source": [
    "# DO NOT DO THIS! For illustration purposes only.\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train_toy)\n",
    "train_scaled = scaler.transform(X_train_toy)\n",
    "\n",
    "scaler = StandardScaler()  # Creating a separate object for scaling test data\n",
    "scaler.fit(X_test_toy)  # Calling fit on the test data\n",
    "test_scaled = scaler.transform(\n",
    "    X_test_toy\n",
    ")  # Transforming the test data using the scaler fit on test data\n",
    "\n",
    "knn = KNeighborsClassifier()\n",
    "knn.fit(train_scaled, y_train_toy)\n",
    "print(f\"Training score: {knn.score(train_scaled, y_train_toy):.2f}\")\n",
    "print(f\"Test score: {knn.score(test_scaled, y_test_toy):.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "- Is anything wrong in methodology 1? If yes, what is it?  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_original_scaled(\n",
    "    X_train_toy,\n",
    "    X_test_toy,\n",
    "    train_scaled,\n",
    "    test_scaled,\n",
    "    title_transformed=\"Improperly transformed\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Bad methodology 2: Scaling the data together (for class discussion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((80, 2), (20, 2))"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_toy.shape, X_test_toy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100, 2)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# join the train and test sets back together\n",
    "XX = np.vstack((X_train_toy, X_test_toy))\n",
    "XX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(XX)\n",
    "XX_scaled = scaler.transform(XX)\n",
    "XX_train = XX_scaled[:80]\n",
    "XX_test = XX_scaled[80:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training score: 0.75\n",
      "Test score: 0.55\n"
     ]
    }
   ],
   "source": [
    "knn = KNeighborsClassifier()\n",
    "knn.fit(XX_train, y_train_toy)\n",
    "print(f\"Training score: {knn.score(XX_train, y_train_toy):.2f}\")  # Misleading score\n",
    "print(f\"Test score: {knn.score(XX_test, y_test_toy):.2f}\")  # Misleading score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "- Is anything wrong in methodology 2? If yes, what is it? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_original_scaled(\n",
    "    X_train_toy,\n",
    "    X_test_toy,\n",
    "    XX_train,\n",
    "    XX_test,\n",
    "    title_transformed=\"Improperly transformed\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not a big difference in the transformed data but if the test set is large it might influence the mean and standard deviation significantly. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Methodology 3: Cross validation with already preprocessed data (for class discussion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "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>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_score</th>\n",
       "      <th>train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000452</td>\n",
       "      <td>0.000802</td>\n",
       "      <td>0.6875</td>\n",
       "      <td>0.671875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000250</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>0.7500</td>\n",
       "      <td>0.671875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000252</td>\n",
       "      <td>0.000595</td>\n",
       "      <td>0.6875</td>\n",
       "      <td>0.734375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000242</td>\n",
       "      <td>0.000624</td>\n",
       "      <td>0.6250</td>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000245</td>\n",
       "      <td>0.000616</td>\n",
       "      <td>0.5000</td>\n",
       "      <td>0.687500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fit_time  score_time  test_score  train_score\n",
       "0  0.000452    0.000802      0.6875     0.671875\n",
       "1  0.000250    0.000600      0.7500     0.671875\n",
       "2  0.000252    0.000595      0.6875     0.734375\n",
       "3  0.000242    0.000624      0.6250     0.750000\n",
       "4  0.000245    0.000616      0.5000     0.687500"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsClassifier()\n",
    "\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train_toy)\n",
    "X_train_scaled = scaler.transform(X_train_toy)\n",
    "X_test_scaled = scaler.transform(X_test_toy)\n",
    "scores = cross_validate(knn, X_train_scaled, y_train_toy, return_train_score=True)\n",
    "pd.DataFrame(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Is there anything wrong in methodology 3? Are we breaking the golden rule here?  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_improper_processing(\"kNN\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_proper_processing(\"kNN\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "<br><br><br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Feature transformations and the golden rule [[video]()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### How to carry out cross-validation? \n",
    "\n",
    "- Last week we saw that cross validation is a better way to get a realistic assessment of the model. \n",
    "- Let's try cross-validation with transformed data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    },
    "tags": []
   },
   "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>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_score</th>\n",
       "      <th>train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.009345</td>\n",
       "      <td>0.182786</td>\n",
       "      <td>0.710905</td>\n",
       "      <td>0.803734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.009117</td>\n",
       "      <td>0.165151</td>\n",
       "      <td>0.706893</td>\n",
       "      <td>0.803212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.009148</td>\n",
       "      <td>0.174530</td>\n",
       "      <td>0.711039</td>\n",
       "      <td>0.803030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.009180</td>\n",
       "      <td>0.177301</td>\n",
       "      <td>0.695769</td>\n",
       "      <td>0.806275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.009134</td>\n",
       "      <td>0.130101</td>\n",
       "      <td>0.697941</td>\n",
       "      <td>0.805146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fit_time  score_time  test_score  train_score\n",
       "0  0.009345    0.182786    0.710905     0.803734\n",
       "1  0.009117    0.165151    0.706893     0.803212\n",
       "2  0.009148    0.174530    0.711039     0.803030\n",
       "3  0.009180    0.177301    0.695769     0.806275\n",
       "4  0.009134    0.130101    0.697941     0.805146"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsRegressor()\n",
    "\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train_imp)\n",
    "X_train_scaled = scaler.transform(X_train_imp)\n",
    "X_test_scaled = scaler.transform(X_test_imp)\n",
    "scores = cross_validate(knn, X_train_scaled, y_train, return_train_score=True)\n",
    "pd.DataFrame(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- Do you see any problem here? \n",
    "- Are we applying `fit_transform` on train portion and `transform` on validation portion in each fold?  \n",
    "    - Here you might be allowing information from the validation set to **leak** into the training step."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- You need to apply the **SAME** preprocessing steps to train/validation.\n",
    "- With many different transformations and cross validation the code gets unwieldy very quickly. \n",
    "- Likely to make mistakes and \"leak\" information.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "- In these examples our test accuracies look fine, but our methodology is flawed.\n",
    "- Implications can be significant in practice!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Pipelines\n",
    "\n",
    "Can we do this in a more elegant and organized way?\n",
    "\n",
    "- YES!! Using [`scikit-learn Pipeline`](https://scikit-learn.org/stable/modules/generated/sklearn.pipeline.Pipeline.html).\n",
    "- [`scikit-learn Pipeline`](https://scikit-learn.org/stable/modules/generated/sklearn.pipeline.Pipeline.html) allows you to define a \"pipeline\" of transformers with a final estimator."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Let's combine the preprocessing and model with pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "### Simple example of a pipeline\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "pipe = Pipeline(\n",
    "    steps=[\n",
    "        (\"imputer\", SimpleImputer(strategy=\"median\")),\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"regressor\", KNeighborsRegressor()),\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Syntax: pass in a list of steps.\n",
    "- The last step should be a **model/classifier/regressor**.\n",
    "- All the earlier steps should be **transformers**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Alternative and more compact syntax: `make_pipeline`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Shorthand for `Pipeline` constructor\n",
    "- Does not permit naming steps\n",
    "- Instead the names of steps are set to lowercase of their types automatically; `StandardScaler()` would be named as `standardscaler`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.pipeline import make_pipeline\n",
    "\n",
    "pipe = make_pipeline(\n",
    "    SimpleImputer(strategy=\"median\"), StandardScaler(), KNeighborsRegressor()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(steps=[('simpleimputer', SimpleImputer(strategy='median')),\n",
       "                ('standardscaler', StandardScaler()),\n",
       "                ('kneighborsregressor', KNeighborsRegressor())])"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Note that we are passing `X_train` and **not** the imputed or scaled data here. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "When you call `fit` on the pipeline, it carries out the following steps:\n",
    "\n",
    "- Fit `SimpleImputer` on `X_train`\n",
    "- Transform `X_train` using the fit `SimpleImputer` to create `X_train_imp`\n",
    "- Fit `StandardScaler` on `X_train_imp`\n",
    "- Transform `X_train_imp` using the fit `StandardScaler` to create `X_train_imp_scaled`\n",
    "- Fit the model (`KNeighborsRegressor` in our case) on `X_train_imp_scaled`\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([122460., 115220., 216940., ..., 240420., 254500.,  60420.])"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.predict(X_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that we are passing original data to `predict` as well. This time the pipeline is carrying out following steps:\n",
    "- Transform `X_train` using the fit `SimpleImputer` to create `X_train_imp`\n",
    "- Transform `X_train_imp` using the fit `StandardScaler` to create `X_train_imp_scaled`\n",
    "- Predict using the fit model (`KNeighborsRegressor` in our case) on `X_train_imp_scaled`. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src='./img/pipeline.png' width=\"800\">\n",
    "    \n",
    "[Source](https://amueller.github.io/COMS4995-s20/slides/aml-04-preprocessing/#18)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Let's try cross-validation with our pipeline "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_score</th>\n",
       "      <th>train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>dummy</th>\n",
       "      <td>0.001 (+/- 0.001)</td>\n",
       "      <td>0.000 (+/- 0.000)</td>\n",
       "      <td>-0.055 (+/- 0.012)</td>\n",
       "      <td>-0.055 (+/- 0.001)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imp + scaling + knn</th>\n",
       "      <td>0.029 (+/- 0.010)</td>\n",
       "      <td>0.173 (+/- 0.014)</td>\n",
       "      <td>0.706 (+/- 0.006)</td>\n",
       "      <td>0.806 (+/- 0.005)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              fit_time         score_time          test_score  \\\n",
       "dummy                0.001 (+/- 0.001)  0.000 (+/- 0.000)  -0.055 (+/- 0.012)   \n",
       "imp + scaling + knn  0.029 (+/- 0.010)  0.173 (+/- 0.014)   0.706 (+/- 0.006)   \n",
       "\n",
       "                            train_score  \n",
       "dummy                -0.055 (+/- 0.001)  \n",
       "imp + scaling + knn   0.806 (+/- 0.005)  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_dict[\"imp + scaling + knn\"] = mean_std_cross_val_scores(\n",
    "    pipe, X_train, y_train, return_train_score=True\n",
    ")\n",
    "pd.DataFrame(results_dict).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Using a `Pipeline` takes care of applying the `fit_transform` on the train portion and only `transform` on the validation portion in each fold.   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Break (5 min)\n",
    "\n",
    "![](img/eva-coffee.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br><br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Categorical features [[video](https://youtu.be/2mJ9rAhMMl0)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Recall that we had dropped the categorical feature `ocean_proximity` feature from the dataframe. But it could potentially be a useful feature in this task. \n",
    "\n",
    "- Let's create our `X_train` and and `X_test` again by keeping the feature in the data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "      <th>ocean_proximity</th>\n",
       "      <th>rooms_per_household</th>\n",
       "      <th>bedrooms_per_household</th>\n",
       "      <th>population_per_household</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19121</th>\n",
       "      <td>-122.64</td>\n",
       "      <td>38.24</td>\n",
       "      <td>40.0</td>\n",
       "      <td>1974.0</td>\n",
       "      <td>410.0</td>\n",
       "      <td>1039.0</td>\n",
       "      <td>398.0</td>\n",
       "      <td>3.7917</td>\n",
       "      <td>151600.0</td>\n",
       "      <td>&lt;1H OCEAN</td>\n",
       "      <td>4.959799</td>\n",
       "      <td>1.030151</td>\n",
       "      <td>2.610553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20019</th>\n",
       "      <td>-119.05</td>\n",
       "      <td>36.09</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3297.0</td>\n",
       "      <td>568.0</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>568.0</td>\n",
       "      <td>4.0217</td>\n",
       "      <td>99200.0</td>\n",
       "      <td>INLAND</td>\n",
       "      <td>5.804577</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.079225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15104</th>\n",
       "      <td>-116.98</td>\n",
       "      <td>32.85</td>\n",
       "      <td>12.0</td>\n",
       "      <td>3570.0</td>\n",
       "      <td>713.0</td>\n",
       "      <td>3321.0</td>\n",
       "      <td>666.0</td>\n",
       "      <td>4.0882</td>\n",
       "      <td>134500.0</td>\n",
       "      <td>&lt;1H OCEAN</td>\n",
       "      <td>5.360360</td>\n",
       "      <td>1.070571</td>\n",
       "      <td>4.986486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3720</th>\n",
       "      <td>-118.42</td>\n",
       "      <td>34.20</td>\n",
       "      <td>27.0</td>\n",
       "      <td>3201.0</td>\n",
       "      <td>970.0</td>\n",
       "      <td>3403.0</td>\n",
       "      <td>948.0</td>\n",
       "      <td>2.2377</td>\n",
       "      <td>231700.0</td>\n",
       "      <td>&lt;1H OCEAN</td>\n",
       "      <td>3.376582</td>\n",
       "      <td>1.023207</td>\n",
       "      <td>3.589662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8938</th>\n",
       "      <td>-118.47</td>\n",
       "      <td>34.01</td>\n",
       "      <td>41.0</td>\n",
       "      <td>2704.0</td>\n",
       "      <td>557.0</td>\n",
       "      <td>1047.0</td>\n",
       "      <td>478.0</td>\n",
       "      <td>4.4211</td>\n",
       "      <td>462900.0</td>\n",
       "      <td>&lt;1H OCEAN</td>\n",
       "      <td>5.656904</td>\n",
       "      <td>1.165272</td>\n",
       "      <td>2.190377</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "19121    -122.64     38.24                40.0       1974.0           410.0   \n",
       "20019    -119.05     36.09                 9.0       3297.0           568.0   \n",
       "15104    -116.98     32.85                12.0       3570.0           713.0   \n",
       "3720     -118.42     34.20                27.0       3201.0           970.0   \n",
       "8938     -118.47     34.01                41.0       2704.0           557.0   \n",
       "\n",
       "       population  households  median_income  median_house_value  \\\n",
       "19121      1039.0       398.0         3.7917            151600.0   \n",
       "20019      1749.0       568.0         4.0217             99200.0   \n",
       "15104      3321.0       666.0         4.0882            134500.0   \n",
       "3720       3403.0       948.0         2.2377            231700.0   \n",
       "8938       1047.0       478.0         4.4211            462900.0   \n",
       "\n",
       "      ocean_proximity  rooms_per_household  bedrooms_per_household  \\\n",
       "19121       <1H OCEAN             4.959799                1.030151   \n",
       "20019          INLAND             5.804577                1.000000   \n",
       "15104       <1H OCEAN             5.360360                1.070571   \n",
       "3720        <1H OCEAN             3.376582                1.023207   \n",
       "8938        <1H OCEAN             5.656904                1.165272   \n",
       "\n",
       "       population_per_household  \n",
       "19121                  2.610553  \n",
       "20019                  3.079225  \n",
       "15104                  4.986486  \n",
       "3720                   3.589662  \n",
       "8938                   2.190377  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = train_df.drop(columns=[\"median_house_value\"])\n",
    "y_train = train_df[\"median_house_value\"]\n",
    "\n",
    "X_test = test_df.drop(columns=[\"median_house_value\"])\n",
    "y_test = test_df[\"median_house_value\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- Let's try to build a `KNeighborRegressor` on this data using our pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pipe.fit(X_train, X_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "- This failed because we have non-numeric data. \n",
    "- Imagine how $k$-NN would calculate distances when you have non-numeric features."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Can we use this feature in the model? \n",
    "- In `scikit-learn`, most algorithms require numeric inputs.\n",
    "- Decision trees could theoretically work with categorical features.  \n",
    "    - However, the sklearn implementation does not support this. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### What are the options? \n",
    "\n",
    "- Drop the column (not recommended)\n",
    "    - If you know that the column is not relevant to the target in any way you may drop it. \n",
    "- We can transform categorical features to numeric ones so that we can use them in the model.     \n",
    "    - [Ordinal encoding](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OrdinalEncoder.html) (occasionally recommended)\n",
    "    - One-hot encoding (recommended in most cases) (this lecture)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>language</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>English</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Vietnamese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>English</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Mandarin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>English</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>English</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Mandarin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>English</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Vietnamese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Mandarin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>French</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Spanish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Mandarin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Hindi</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      language\n",
       "0      English\n",
       "1   Vietnamese\n",
       "2      English\n",
       "3     Mandarin\n",
       "4      English\n",
       "5      English\n",
       "6     Mandarin\n",
       "7      English\n",
       "8   Vietnamese\n",
       "9     Mandarin\n",
       "10      French\n",
       "11     Spanish\n",
       "12    Mandarin\n",
       "13       Hindi"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_toy = pd.DataFrame(\n",
    "    {\n",
    "        \"language\": [\n",
    "            \"English\",\n",
    "            \"Vietnamese\",\n",
    "            \"English\",\n",
    "            \"Mandarin\",\n",
    "            \"English\",\n",
    "            \"English\",\n",
    "            \"Mandarin\",\n",
    "            \"English\",\n",
    "            \"Vietnamese\",\n",
    "            \"Mandarin\",\n",
    "            \"French\",\n",
    "            \"Spanish\",\n",
    "            \"Mandarin\",\n",
    "            \"Hindi\",\n",
    "        ]\n",
    "    }\n",
    ")\n",
    "X_toy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Ordinal encoding (occasionally recommended)\n",
    "\n",
    "- Here we simply assign an integer to each of our unique categorical labels. \n",
    "- We can use sklearn's [`OrdinalEncoder`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OrdinalEncoder.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>language</th>\n",
       "      <th>language_enc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>English</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Vietnamese</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>English</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Mandarin</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>English</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>English</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Mandarin</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>English</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Vietnamese</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Mandarin</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>French</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Spanish</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Mandarin</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Hindi</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      language  language_enc\n",
       "0      English           0.0\n",
       "1   Vietnamese           5.0\n",
       "2      English           0.0\n",
       "3     Mandarin           3.0\n",
       "4      English           0.0\n",
       "5      English           0.0\n",
       "6     Mandarin           3.0\n",
       "7      English           0.0\n",
       "8   Vietnamese           5.0\n",
       "9     Mandarin           3.0\n",
       "10      French           1.0\n",
       "11     Spanish           4.0\n",
       "12    Mandarin           3.0\n",
       "13       Hindi           2.0"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "\n",
    "enc = OrdinalEncoder()\n",
    "enc.fit(X_toy)\n",
    "X_toy_ord = enc.transform(X_toy)\n",
    "df = pd.DataFrame(\n",
    "    data=X_toy_ord,\n",
    "    columns=[\"language_enc\"],\n",
    "    index=X_toy.index,\n",
    ")\n",
    "pd.concat([X_toy, df], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "What's the problem with this approach? \n",
    "- We have imposed ordinality on the categorical data.\n",
    "- For example, imagine when you are calculating distances. Is it fair to say that French and Hindi are closer than French and Spanish? \n",
    "- In general, label encoding is useful if there is ordinality in your data and capturing it is important for your problem, e.g., `[cold, warm, hot]`. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### One-hot encoding (OHE)\n",
    "- Create new binary columns to represent our categories.\n",
    "- If we have $c$ categories in our column.\n",
    "    - We create $c$ new binary columns to represent those categories.\n",
    "- Example: Imagine a language column which has the information on whether you \n",
    "\n",
    "- We can use sklearn's [`OneHotEncoder`](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OneHotEncoder.html) to do so."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```{note}\n",
    "One-hot encoding is called one-hot because only one of the newly created features is 1 for each data point. \n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>language_English</th>\n",
       "      <th>language_French</th>\n",
       "      <th>language_Hindi</th>\n",
       "      <th>language_Mandarin</th>\n",
       "      <th>language_Spanish</th>\n",
       "      <th>language_Vietnamese</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    language_English  language_French  language_Hindi  language_Mandarin  \\\n",
       "0                1.0              0.0             0.0                0.0   \n",
       "1                0.0              0.0             0.0                0.0   \n",
       "2                1.0              0.0             0.0                0.0   \n",
       "3                0.0              0.0             0.0                1.0   \n",
       "4                1.0              0.0             0.0                0.0   \n",
       "5                1.0              0.0             0.0                0.0   \n",
       "6                0.0              0.0             0.0                1.0   \n",
       "7                1.0              0.0             0.0                0.0   \n",
       "8                0.0              0.0             0.0                0.0   \n",
       "9                0.0              0.0             0.0                1.0   \n",
       "10               0.0              1.0             0.0                0.0   \n",
       "11               0.0              0.0             0.0                0.0   \n",
       "12               0.0              0.0             0.0                1.0   \n",
       "13               0.0              0.0             1.0                0.0   \n",
       "\n",
       "    language_Spanish  language_Vietnamese  \n",
       "0                0.0                  0.0  \n",
       "1                0.0                  1.0  \n",
       "2                0.0                  0.0  \n",
       "3                0.0                  0.0  \n",
       "4                0.0                  0.0  \n",
       "5                0.0                  0.0  \n",
       "6                0.0                  0.0  \n",
       "7                0.0                  0.0  \n",
       "8                0.0                  1.0  \n",
       "9                0.0                  0.0  \n",
       "10               0.0                  0.0  \n",
       "11               1.0                  0.0  \n",
       "12               0.0                  0.0  \n",
       "13               0.0                  0.0  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "enc = OneHotEncoder(handle_unknown=\"ignore\", sparse=False)\n",
    "enc.fit(X_toy)\n",
    "X_toy_ohe = enc.transform(X_toy)\n",
    "pd.DataFrame(\n",
    "    data=X_toy_ohe,\n",
    "    columns=enc.get_feature_names_out([\"language\"]),\n",
    "    index=X_toy.index,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Let's do it on our housing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "ohe = OneHotEncoder(sparse=False, dtype=\"int\")\n",
    "ohe.fit(X_train[[\"ocean_proximity\"]])\n",
    "X_imp_ohe_train = ohe.transform(X_train[[\"ocean_proximity\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We can look at the new features created using `categories_` attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['<1H OCEAN', 'INLAND', 'ISLAND', 'NEAR BAY', 'NEAR OCEAN'],\n",
       "       dtype=object)]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ohe.categories_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>ocean_proximity_&lt;1H OCEAN</th>\n",
       "      <th>ocean_proximity_INLAND</th>\n",
       "      <th>ocean_proximity_ISLAND</th>\n",
       "      <th>ocean_proximity_NEAR BAY</th>\n",
       "      <th>ocean_proximity_NEAR OCEAN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6051</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20113</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14289</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13665</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14471</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7763</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15377</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17730</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15725</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19966</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>18576 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       ocean_proximity_<1H OCEAN  ocean_proximity_INLAND  \\\n",
       "6051                           0                       1   \n",
       "20113                          0                       1   \n",
       "14289                          0                       0   \n",
       "13665                          0                       1   \n",
       "14471                          0                       0   \n",
       "...                          ...                     ...   \n",
       "7763                           1                       0   \n",
       "15377                          1                       0   \n",
       "17730                          1                       0   \n",
       "15725                          0                       0   \n",
       "19966                          0                       1   \n",
       "\n",
       "       ocean_proximity_ISLAND  ocean_proximity_NEAR BAY  \\\n",
       "6051                        0                         0   \n",
       "20113                       0                         0   \n",
       "14289                       0                         0   \n",
       "13665                       0                         0   \n",
       "14471                       0                         0   \n",
       "...                       ...                       ...   \n",
       "7763                        0                         0   \n",
       "15377                       0                         0   \n",
       "17730                       0                         0   \n",
       "15725                       0                         1   \n",
       "19966                       0                         0   \n",
       "\n",
       "       ocean_proximity_NEAR OCEAN  \n",
       "6051                            0  \n",
       "20113                           0  \n",
       "14289                           1  \n",
       "13665                           0  \n",
       "14471                           1  \n",
       "...                           ...  \n",
       "7763                            0  \n",
       "15377                           0  \n",
       "17730                           0  \n",
       "15725                           0  \n",
       "19966                           0  \n",
       "\n",
       "[18576 rows x 5 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transformed_ohe = pd.DataFrame(\n",
    "    data=X_imp_ohe_train,\n",
    "    columns=ohe.get_feature_names_out([\"ocean_proximity\"]),\n",
    "    index=X_train.index,\n",
    ")\n",
    "transformed_ohe"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "```{seealso} \n",
    "One-hot encoded variables are also referred to as **dummy variables**. \n",
    "You will often see people using [`get_dummies` method of pandas](https://pandas.pydata.org/docs/reference/api/pandas.get_dummies.html) to convert categorical variables into dummy variables. That said, using `sklearn`'s `OneHotEncoder` has the advantage of making it easy to treat training and test set in a consistent way.  \n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### ❓❓ Questions for class discussion "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "#### True/False: Pipelines and one-hot encoding\n",
    "\n",
    "1. You can \"glue\" together imputation and scaling of numeric features and `scikit-learn` classifier object within a single pipeline.  \n",
    "2. You can \"glue\" together scaling of numeric features, one-hot encoding of categorical features, and `scikit-learn` classifier object within a single pipeline.  \n",
    "3. Pipelines will `fit` and `transform` on the training fold and only `transform` on the validation fold during cross-validation.\n",
    "4. What's the better encoding for weather labels such as \"sunny\", \"overcast\" and \"rainy\"?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "```{admonition} True or False\n",
    "1. True\n",
    "2. False (but we'll find a way)\n",
    "3. True\n",
    "4. One-hot\n",
    "```   \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## What did we learn today? \n",
    "\n",
    "- Motivation for preprocessing\n",
    "- Common preprocessing steps\n",
    "    - Imputation \n",
    "    - Scaling\n",
    "    - One-hot encoding\n",
    "- Golden rule in the context of preprocessing\n",
    "- Building simple supervised machine learning pipelines using `sklearn.pipeline.make_pipeline`. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Problem: Different transformations on different columns\n",
    "- How do we put this together with other columns in the data before fitting the regressor? \n",
    "- Before we fit our regressor, we want to apply different transformations on different columns \n",
    "    - Numeric columns\n",
    "        - imputation \n",
    "        - scaling         \n",
    "    - Categorical columns \n",
    "        - imputation \n",
    "        - one-hot encoding        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "**Coming up: sklearn's [`ColumnTransformer`](https://scikit-learn.org/stable/modules/generated/sklearn.compose.ColumnTransformer.html)!!** "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Python [conda env:cpsc330] *",
   "language": "python",
   "name": "conda-env-cpsc330-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}