gold.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pycaret.datasets import get_data\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import Pipeline, FeatureUnion\n",
    "from category_encoders import LeaveOneOutEncoder, CatBoostEncoder, TargetEncoder, JamesSteinEncoder\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor\n",
    "from catboost import CatBoostClassifier, CatBoostRegressor, Pool\n",
    "from lightgbm import LGBMClassifier, LGBMRegressor\n",
    "from xgboost import XGBClassifier, XGBRegressor\n",
    "from sklearn.model_selection import cross_val_score, cross_val_predict\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "from sklearn.metrics import roc_auc_score, r2_score, mean_absolute_error\n",
    "from sklearn.linear_model import Lasso, Ridge\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.dummy import DummyRegressor\n",
    "from tqdm import tqdm\n",
    "from sklearn.experimental import enable_iterative_imputer\n",
    "from sklearn.impute import SimpleImputer, IterativeImputer, KNNImputer, MissingIndicator\n",
    "from shap import TreeExplainer\n",
    "from scipy.interpolate import interp1d\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_preds_shaps(df, features, target, ix_trn):\n",
    "  model = LGBMRegressor().fit(df.loc[ix_trn, features], df.loc[ix_trn, target])\n",
    "  preds = pd.Series(model.predict(df[features]), index=df.index)\n",
    "  shap_explainer = TreeExplainer(model)\n",
    "  shaps = pd.DataFrame(\n",
    "    data=shap_explainer.shap_values(df[features]),\n",
    "    index=df.index,\n",
    "    columns=features)\n",
    "  return preds, shaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_feature_contributions(y_true, y_pred, shap_values):\n",
    "  \"\"\"Compute prediction contribution and error contribution for each feature.\"\"\"\n",
    "\n",
    "  prediction_contribution = shap_values.abs().mean().rename(\"prediction_contribution\")\n",
    "\n",
    "  abs_error = (y_true - y_pred).abs()\n",
    "  y_pred_wo_feature = shap_values.apply(lambda feature: y_pred - feature)\n",
    "  abs_error_wo_feature = y_pred_wo_feature.apply(lambda feature: (y_true - feature).abs())\n",
    "  error_contribution = abs_error_wo_feature.apply(lambda feature: abs_error - feature).mean().rename(\"error_contribution\")\n",
    "      \n",
    "  return prediction_contribution, error_contribution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Toy Example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 Generate data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "shap_values = pd.DataFrame([\n",
    "  [65,  12,   8],\n",
    "  [72,   3, -22],\n",
    "  [78, -23,  15]], columns=[\"job\", \"age\", \"nationality\"])\n",
    "y_pred = shap_values.sum(axis=1).rename(\"income (predicted)\")\n",
    "y_true = pd.Series([74, 68, 58], name=\"income (true)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 Compute Prediction Contribution and Error Contribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "abs_error = (y_true - y_pred).abs().rename(\"absolute error\")\n",
    "\n",
    "prediction_contribution = shap_values.abs().mean().round(2).rename(\"prediction_contribution\").to_frame().T\n",
    "\n",
    "y_pred_wo_feature = shap_values.apply(lambda feature: y_pred - feature)\n",
    "abs_error_wo_feature = y_pred_wo_feature.apply(lambda feature: (y_true - feature).abs())\n",
    "error_diff = abs_error_wo_feature.apply(lambda feature: abs_error - feature)\n",
    "error_contribution = error_diff.mean().round(2).rename(\"error_contribution\").to_frame().T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3 Display items"
   ]
  },
  {
   "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>income (true)</th>\n",
       "      <th>income (predicted)</th>\n",
       "      <th>absolute error</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>74</td>\n",
       "      <td>85</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>68</td>\n",
       "      <td>53</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>58</td>\n",
       "      <td>70</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   income (true)  income (predicted)  absolute error\n",
       "0             74                  85              11\n",
       "1             68                  53              15\n",
       "2             58                  70              12"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([y_true, y_pred, abs_error], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>65</td>\n",
       "      <td>12</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>72</td>\n",
       "      <td>3</td>\n",
       "      <td>-22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78</td>\n",
       "      <td>-23</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   job  age  nationality\n",
       "0   65   12            8\n",
       "1   72    3          -22\n",
       "2   78  -23           15"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shap_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>prediction_contribution</th>\n",
       "      <td>71.67</td>\n",
       "      <td>12.67</td>\n",
       "      <td>15.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           job    age  nationality\n",
       "prediction_contribution  71.67  12.67         15.0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_contribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20</td>\n",
       "      <td>73</td>\n",
       "      <td>77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-19</td>\n",
       "      <td>50</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-8</td>\n",
       "      <td>93</td>\n",
       "      <td>55</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   job  age  nationality\n",
       "0   20   73           77\n",
       "1  -19   50           75\n",
       "2   -8   93           55"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_wo_feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>54</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>87</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>66</td>\n",
       "      <td>35</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   job  age  nationality\n",
       "0   54    1            3\n",
       "1   87   18            7\n",
       "2   66   35            3"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abs_error_wo_feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-43</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-72</td>\n",
       "      <td>-3</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-54</td>\n",
       "      <td>-23</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   job  age  nationality\n",
       "0  -43   10            8\n",
       "1  -72   -3            8\n",
       "2  -54  -23            9"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>job</th>\n",
       "      <th>age</th>\n",
       "      <th>nationality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>error_contribution</th>\n",
       "      <td>-56.33</td>\n",
       "      <td>-5.33</td>\n",
       "      <td>8.33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      job   age  nationality\n",
       "error_contribution -56.33 -5.33         8.33"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_contribution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Real dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = get_data(\"gold\", verbose=False)\n",
    "df *= 100\n",
    "\n",
    "target = \"Gold_T+22\"\n",
    "\n",
    "features = [f for f in df.columns if f != target]\n",
    "argsort = np.argsort([f.split(\"T-\")[0] + \"T-\" + \"{:2.1f}\".format(float(f.split(\"T-\")[1])).zfill(4) for f in features])\n",
    "features = [features[arg] for arg in argsort]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    2558.000000\n",
       "mean        0.322176\n",
       "std         4.419812\n",
       "min       -14.910681\n",
       "25%        -2.361938\n",
       "50%         0.144122\n",
       "75%         3.153245\n",
       "max        19.472259\n",
       "Name: Gold_T+22, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[target].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.hist(df[target], bins=\"fd\")\n",
    "ax.grid(); ax.set_axisbelow(True)\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "ax.set_xlabel(\"Gold Return (%) - 22 Days Ahead\", fontsize=12)\n",
    "\n",
    "fig.savefig(\"target_hist.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>3M Libor_T-1</th>\n",
       "      <th>3M Libor_T-7</th>\n",
       "      <th>3M Libor_T-14</th>\n",
       "      <th>3M Libor_T-22</th>\n",
       "      <th>...</th>\n",
       "      <th>Yuan_T-7</th>\n",
       "      <th>Yuan_T-14</th>\n",
       "      <th>Yuan_T-22</th>\n",
       "      <th>Gold_T+22</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00</td>\n",
       "      <td>-1.80</td>\n",
       "      <td>-2.90</td>\n",
       "      <td>-4.07</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.46</td>\n",
       "      <td>-1.47</td>\n",
       "      <td>-3.34</td>\n",
       "      <td>-4.51</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.04</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.21</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2556</th>\n",
       "      <td>-0.23</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>-6.73</td>\n",
       "      <td>...</td>\n",
       "      <td>1.25</td>\n",
       "      <td>1.88</td>\n",
       "      <td>3.98</td>\n",
       "      <td>-2.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2557</th>\n",
       "      <td>-0.29</td>\n",
       "      <td>-0.83</td>\n",
       "      <td>-1.93</td>\n",
       "      <td>-5.03</td>\n",
       "      <td>...</td>\n",
       "      <td>1.21</td>\n",
       "      <td>2.22</td>\n",
       "      <td>3.46</td>\n",
       "      <td>-2.71</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2558 rows × 121 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      3M Libor_T-1  3M Libor_T-7  3M Libor_T-14  3M Libor_T-22  ...  Yuan_T-7  \\\n",
       "0             0.00         -1.80          -2.90          -4.07  ...     -0.01   \n",
       "1            -0.46         -1.47          -3.34          -4.51  ...     -0.01   \n",
       "...            ...           ...            ...            ...  ...       ...   \n",
       "2556         -0.23          0.02          -1.18          -6.73  ...      1.25   \n",
       "2557         -0.29         -0.83          -1.93          -5.03  ...      1.21   \n",
       "\n",
       "      Yuan_T-14  Yuan_T-22  Gold_T+22  \n",
       "0         -0.02      -0.00       0.46  \n",
       "1         -0.04      -0.01      -0.21  \n",
       "...         ...        ...        ...  \n",
       "2556       1.88       3.98      -2.03  \n",
       "2557       2.22       3.46      -2.71  \n",
       "\n",
       "[2558 rows x 121 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.max_columns', 8)\n",
    "pd.set_option(\"display.max_rows\", 5)\n",
    "\n",
    "df[features + [target]].round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3M Libor            AUD                 Brent 1             Brent 2             \n",
      "Corn                Crude Palm Oil      DJ Index            Dollex              \n",
      "EM Bond             EMEquity ETF        Euro Curr           Gold                \n",
      "INR                 Light Crude         MR                  MSCI APAC ex J      \n",
      "MSCI EM             Nifty               Nikkei Ind          Palladium           \n",
      "Platinum            S&P 500             Shanghai Index      Silver              \n",
      "US Bond ETF         Wheat 1             Wheat 2             White Sugar         \n",
      "Yuan                "
     ]
    }
   ],
   "source": [
    "for enum, f in enumerate(sorted(set([f.split(\"_\")[0] for f in features]))):\n",
    "  print(f.ljust(20), end=\"\")\n",
    "  if enum % 4 == 3:\n",
    "    print(\"\\n\", end=\"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Train/Validation/Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(123)\n",
    "fold = np.random.choice([\"trn\", \"val\", \"tst\"], p=[1/3, 1/3, 1/3], size=len(df))\n",
    "\n",
    "ix_trn = fold == \"trn\"\n",
    "ix_val = fold == \"val\"\n",
    "ix_tst = fold == \"tst\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert (ix_trn.astype(int) + ix_val.astype(int) + ix_tst.astype(int) == 1).all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7. Single Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.1 Compute Predictions and Shap Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds, shaps = get_preds_shaps(df=df, features=features, target=target, ix_trn=ix_trn)\n",
    "\n",
    "prediction_contribution_trn, error_contribution_trn = get_feature_contributions(\n",
    "  y_true=df.loc[ix_trn, target], \n",
    "  y_pred=preds.loc[ix_trn], \n",
    "  shap_values=shaps.loc[ix_trn, :]\n",
    ")\n",
    "\n",
    "prediction_contribution_val, error_contribution_val = get_feature_contributions(\n",
    "  y_true=df.loc[ix_val, target], \n",
    "  y_pred=preds.loc[ix_val], \n",
    "  shap_values=shaps.loc[ix_val, :]\n",
    ")\n",
    "\n",
    "contributions_trn = pd.concat([prediction_contribution_trn, error_contribution_trn], axis=1)\n",
    "contributions_val = pd.concat([prediction_contribution_val, error_contribution_val], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.2 Write/Read Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "contributions_trn.to_csv(\"contributions_trn.csv\", index=True)\n",
    "contributions_val.to_csv(\"contributions_val.csv\", index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "contributions_trn = pd.read_csv(\"contributions_trn.csv\", index_col=0)\n",
    "contributions_val = pd.read_csv(\"contributions_val.csv\", index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.3 Prediction vs. Error Contribution (Training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>prediction_contribution</th>\n",
       "      <th>error_contribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-1</th>\n",
       "      <td>0.042951</td>\n",
       "      <td>-0.013922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-7</th>\n",
       "      <td>0.090338</td>\n",
       "      <td>-0.039541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yuan_T-14</th>\n",
       "      <td>0.066477</td>\n",
       "      <td>-0.028666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yuan_T-22</th>\n",
       "      <td>0.144774</td>\n",
       "      <td>-0.072751</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>120 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              prediction_contribution  error_contribution\n",
       "3M Libor_T-1                 0.042951           -0.013922\n",
       "3M Libor_T-7                 0.090338           -0.039541\n",
       "...                               ...                 ...\n",
       "Yuan_T-14                    0.066477           -0.028666\n",
       "Yuan_T-22                    0.144774           -0.072751\n",
       "\n",
       "[120 rows x 2 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "contributions_trn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.scatter(contributions_trn[\"prediction_contribution\"], contributions_trn[\"error_contribution\"], s=80, color=\"orange\", edgecolors=\"black\")\n",
    "ax.hlines(y=0, xmin=0, xmax=contributions_trn[\"prediction_contribution\"].max(), ls=\"--\", color=\"black\")\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "ax.grid(); ax.set_axisbelow(True)\n",
    "ax.set_title(\"Train Dataset\", fontsize=12)\n",
    "ax.set_xlabel(\"Prediction Contribution\", fontsize=12)\n",
    "ax.set_ylabel(\"Error Contribution\", fontsize=12)\n",
    "\n",
    "fig.savefig(\"single_trn.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>prediction_contribution</th>\n",
       "      <th>error_contribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>INR_T-22</th>\n",
       "      <td>0.431543</td>\n",
       "      <td>-0.314891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EM Bond_T-22</th>\n",
       "      <td>0.354269</td>\n",
       "      <td>-0.234833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-22</th>\n",
       "      <td>0.349576</td>\n",
       "      <td>-0.270460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Platinum_T-14</th>\n",
       "      <td>0.298751</td>\n",
       "      <td>-0.209640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Shanghai Index_T-22</th>\n",
       "      <td>0.214553</td>\n",
       "      <td>-0.130805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>White Sugar_T-22</th>\n",
       "      <td>0.207254</td>\n",
       "      <td>-0.122756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-14</th>\n",
       "      <td>0.195259</td>\n",
       "      <td>-0.127183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EMEquity ETF_T-7</th>\n",
       "      <td>0.186585</td>\n",
       "      <td>-0.097444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gold_T-22</th>\n",
       "      <td>0.185994</td>\n",
       "      <td>-0.126613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brent 2_T-14</th>\n",
       "      <td>0.180668</td>\n",
       "      <td>-0.088183</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     prediction_contribution  error_contribution\n",
       "INR_T-22                            0.431543           -0.314891\n",
       "EM Bond_T-22                        0.354269           -0.234833\n",
       "3M Libor_T-22                       0.349576           -0.270460\n",
       "Platinum_T-14                       0.298751           -0.209640\n",
       "Shanghai Index_T-22                 0.214553           -0.130805\n",
       "White Sugar_T-22                    0.207254           -0.122756\n",
       "3M Libor_T-14                       0.195259           -0.127183\n",
       "EMEquity ETF_T-7                    0.186585           -0.097444\n",
       "Gold_T-22                           0.185994           -0.126613\n",
       "Brent 2_T-14                        0.180668           -0.088183"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.max_rows\", 10)\n",
    "\n",
    "contributions_trn.sort_values(\"prediction_contribution\", ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7.4 Prediction vs. Error Contribution (Validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.scatter(contributions_val[\"prediction_contribution\"], contributions_val[\"error_contribution\"], s=80, color=\"orange\", edgecolors=\"black\")\n",
    "ax.hlines(y=0, xmin=0, xmax=contributions_val[\"prediction_contribution\"].max(), ls=\"--\", color=\"black\")\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "\n",
    "ax.grid()\n",
    "ax.set_axisbelow(True)\n",
    "ax.set_title(\"Validation Dataset\", fontsize=12)\n",
    "ax.set_xlabel(\"Prediction Contribution\", fontsize=12)\n",
    "ax.set_ylabel(\"Error Contribution\", fontsize=12)\n",
    "\n",
    "fig.savefig(\"single_val.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>prediction_contribution</th>\n",
       "      <th>error_contribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>INR_T-22</th>\n",
       "      <td>0.3997</td>\n",
       "      <td>-0.0982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gold_T-22</th>\n",
       "      <td>0.1998</td>\n",
       "      <td>-0.0882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-22</th>\n",
       "      <td>0.3347</td>\n",
       "      <td>-0.0730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Platinum_T-14</th>\n",
       "      <td>0.3117</td>\n",
       "      <td>-0.0599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EM Bond_T-22</th>\n",
       "      <td>0.3559</td>\n",
       "      <td>-0.0552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Shanghai Index_T-22</th>\n",
       "      <td>0.2263</td>\n",
       "      <td>-0.0401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Corn_T-22</th>\n",
       "      <td>0.1876</td>\n",
       "      <td>-0.0283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DJ Index_T-22</th>\n",
       "      <td>0.1426</td>\n",
       "      <td>-0.0283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brent 1_T-22</th>\n",
       "      <td>0.1046</td>\n",
       "      <td>-0.0273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3M Libor_T-14</th>\n",
       "      <td>0.1977</td>\n",
       "      <td>-0.0257</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     prediction_contribution  error_contribution\n",
       "INR_T-22                              0.3997             -0.0982\n",
       "Gold_T-22                             0.1998             -0.0882\n",
       "3M Libor_T-22                         0.3347             -0.0730\n",
       "Platinum_T-14                         0.3117             -0.0599\n",
       "EM Bond_T-22                          0.3559             -0.0552\n",
       "Shanghai Index_T-22                   0.2263             -0.0401\n",
       "Corn_T-22                             0.1876             -0.0283\n",
       "DJ Index_T-22                         0.1426             -0.0283\n",
       "Brent 1_T-22                          0.1046             -0.0273\n",
       "3M Libor_T-14                         0.1977             -0.0257"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.max_rows\", 10)\n",
    "\n",
    "contributions_val.sort_values(\"error_contribution\", ascending=True).round(4).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>prediction_contribution</th>\n",
       "      <th>error_contribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>US Bond ETF_T-1</th>\n",
       "      <td>0.0926</td>\n",
       "      <td>0.0130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSCI APAC ex J_T-1</th>\n",
       "      <td>0.1057</td>\n",
       "      <td>0.0076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSCI EM_T-1</th>\n",
       "      <td>0.0241</td>\n",
       "      <td>0.0047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nikkei Ind_T-1</th>\n",
       "      <td>0.0269</td>\n",
       "      <td>0.0038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Corn_T-1</th>\n",
       "      <td>0.0380</td>\n",
       "      <td>0.0033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brent 1_T-1</th>\n",
       "      <td>0.0264</td>\n",
       "      <td>0.0029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brent 2_T-1</th>\n",
       "      <td>0.0258</td>\n",
       "      <td>0.0029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DJ Index_T-1</th>\n",
       "      <td>0.0256</td>\n",
       "      <td>0.0019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Euro Curr_T-1</th>\n",
       "      <td>0.0297</td>\n",
       "      <td>0.0014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>INR_T-7</th>\n",
       "      <td>0.0371</td>\n",
       "      <td>0.0013</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    prediction_contribution  error_contribution\n",
       "US Bond ETF_T-1                      0.0926              0.0130\n",
       "MSCI APAC ex J_T-1                   0.1057              0.0076\n",
       "MSCI EM_T-1                          0.0241              0.0047\n",
       "Nikkei Ind_T-1                       0.0269              0.0038\n",
       "Corn_T-1                             0.0380              0.0033\n",
       "Brent 1_T-1                          0.0264              0.0029\n",
       "Brent 2_T-1                          0.0258              0.0029\n",
       "DJ Index_T-1                         0.0256              0.0019\n",
       "Euro Curr_T-1                        0.0297              0.0014\n",
       "INR_T-7                              0.0371              0.0013"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.max_rows\", 10)\n",
    "\n",
    "contributions_val.sort_values(\"error_contribution\", ascending=False).round(4).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8. Recursive Feature Elimination (RFE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.1. RFE - Error Contribution\n",
    "At each step, we remove the feature with the highest error contribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 120/120 [04:09<00:00,  2.08s/it]\n"
     ]
    }
   ],
   "source": [
    "rfe_error = pd.DataFrame(dtype=float)\n",
    "features_curr = features.copy()\n",
    "feature_drop = None\n",
    "\n",
    "for iteration in tqdm(range(len(features))):\n",
    "  \n",
    "  preds, shaps = get_preds_shaps(df=df, features=features_curr, target=target, ix_trn=ix_trn)\n",
    "  prediction_contribution, error_contribution = get_feature_contributions(\n",
    "    y_true=df.loc[ix_val, target], \n",
    "    y_pred=preds.loc[ix_val], \n",
    "    shap_values=shaps.loc[ix_val, :]\n",
    "  )\n",
    "\n",
    "  rfe_error.loc[iteration, \"feature_drop\"] = feature_drop\n",
    "  rfe_error.loc[iteration, \"n_features\"] = len(features_curr)\n",
    "  rfe_error.loc[iteration, \"contrib\"] = error_contribution.max()\n",
    "  rfe_error.loc[iteration, \"mae_trn\"] = mean_absolute_error(df.loc[ix_trn, target], preds.loc[ix_trn])\n",
    "  rfe_error.loc[iteration, \"mae_val\"] = mean_absolute_error(df.loc[ix_val, target], preds.loc[ix_val])\n",
    "  rfe_error.loc[iteration, \"mae_tst\"] = mean_absolute_error(df.loc[ix_tst, target], preds.loc[ix_tst])\n",
    "  rfe_error.loc[iteration, \"r2_trn\"] = r2_score(df.loc[ix_trn, target], preds.loc[ix_trn])\n",
    "  rfe_error.loc[iteration, \"r2_val\"] = r2_score(df.loc[ix_val, target], preds.loc[ix_val])\n",
    "  rfe_error.loc[iteration, \"r2_tst\"] = r2_score(df.loc[ix_tst, target], preds.loc[ix_tst])\n",
    "  \n",
    "  feature_drop = error_contribution.idxmax()\n",
    "  features_curr.remove(feature_drop)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.2. RFE - Prediction Contribution\n",
    "At each step, we remove the feature with the smallest prediction contribution (i.e. the less important feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 120/120 [04:10<00:00,  2.08s/it]\n"
     ]
    }
   ],
   "source": [
    "rfe_prediction = pd.DataFrame(dtype=float)\n",
    "features_curr = features.copy()\n",
    "feature_drop = None\n",
    "\n",
    "for iteration in tqdm(range(len(features))):\n",
    "  \n",
    "  preds, shaps = get_preds_shaps(df=df, features=features_curr, target=target, ix_trn=ix_trn)\n",
    "  prediction_contribution, error_contribution = get_feature_contributions(\n",
    "    y_true=df.loc[ix_val, target], \n",
    "    y_pred=preds.loc[ix_val], \n",
    "    shap_values=shaps.loc[ix_val, :]\n",
    "  )\n",
    "\n",
    "  rfe_prediction.loc[iteration, \"feature_drop\"] = feature_drop\n",
    "  rfe_prediction.loc[iteration, \"n_features\"] = len(features_curr)\n",
    "  rfe_prediction.loc[iteration, \"contrib\"] = prediction_contribution.min()\n",
    "  rfe_prediction.loc[iteration, \"mae_trn\"] = mean_absolute_error(df.loc[ix_trn, target], preds.loc[ix_trn])\n",
    "  rfe_prediction.loc[iteration, \"mae_val\"] = mean_absolute_error(df.loc[ix_val, target], preds.loc[ix_val])\n",
    "  rfe_prediction.loc[iteration, \"mae_tst\"] = mean_absolute_error(df.loc[ix_tst, target], preds.loc[ix_tst])\n",
    "  rfe_prediction.loc[iteration, \"r2_trn\"] = r2_score(df.loc[ix_trn, target], preds.loc[ix_trn])\n",
    "  rfe_prediction.loc[iteration, \"r2_val\"] = r2_score(df.loc[ix_val, target], preds.loc[ix_val])\n",
    "  rfe_prediction.loc[iteration, \"r2_tst\"] = r2_score(df.loc[ix_tst, target], preds.loc[ix_tst])\n",
    "  \n",
    "  feature_drop = prediction_contribution.idxmin()\n",
    "  features_curr.remove(feature_drop)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.3 Write Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfe_error.to_csv(\"gold_rfe_error.csv\", index=False)\n",
    "rfe_prediction.to_csv(\"gold_rfe_prediction.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.4 Read Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfe_error = pd.read_csv(\"gold_rfe_error.csv\")\n",
    "rfe_prediction = pd.read_csv(\"gold_rfe_prediction.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.5 Plot RFE on Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhcAAAEcCAYAAAB09pYWAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABoNUlEQVR4nO3dd3hUVfrA8e+bQnqoIZAIhCqidERdGyiKiq4oKC5rLyj2vlixrN11Fbv+7KKIYMWOgN1VkCZgobcEQycEUt/fH2cmM5lMkklImZD38zz34ZZz7z0nIZk3p4qqYowxxhhTUyLqOwPGGGOM2btYcGGMMcaYGmXBhTHGGGNqlAUXxhhjjKlRFlwYY4wxpkZZcGGMMcaYGmXBhWnURERFpItn/xkRuS2UtNV4zz9F5PPq5tMYYxoSCy5MgyYin4rIXUHOnywiWSISFeqzVPUSVb27BvKU4QlESt6tqhNV9dg9fXY577tZRFaISI6IrBWRt0K871wR+bY28mSMadwsuDAN3SvAmSIiAefPAiaqamE95KnOiMg5uLIOUdVEYADwZf3myhjT2FlwYRq694CWwOHeEyLSHDgReFVEBorIDyKyVUQyReQJEWkS7EEi8rKI/Nvv+AbPPetF5PyAtMNEZK6IbBeRNSJyh9/lrz3/bvXUJhwSWEsgIn8TkZ9FZJvn37/5XZslIneLyHciskNEPheRVuWU/0DgM1VdBqCqWar6nN+zmorIC55yrBORf4tIpIjsBzwDHOLJ49Zyv8LGGFNFFlyYBk1VdwGTgbP9Tp8O/Kaq84Ei4BqgFXAIcDRwaWXPFZHjgOuBY4CuwJCAJDs972wGDAPGishwz7UjPP82U9VEVf0h4NktgI+ACbjA6BHgIxFp6ZdsNHAe0Bpo4slLMD8CZ3sCoQEiEhlw/WWgEOgC9AWOBS5U1SXAJcAPnjw2K/eLYYwxVWTBhdkbvAKMFJFYz/HZnnOo6hxV/VFVC1V1JfAscGQIzzwdeElVf1XVncAd/hdVdZaqLlTVYlVdALwZ4nPBBSN/quprnny9CfwGnOSX5iVV/cMveOoT7EGq+jpwBTAU+Ar4S0T+BSAiqcAJwNWqulNV/wL+C5wRYj6NMaZaQu7sZky4UtVvRWQjMFxEfgYGAqcCiEg3XM3AACAe939+TgiPTQtIt8r/oogcBNwPHICrWYgB3g4xy2mBz/Mcp/sdZ/nt5wKJ5T1MVScCE0UkGhju2Z8HbAGigUy/LikRwJoQ82mMMdViNRdmb/EqrsbiTFwfhA2e80/jagW6qmoycDMQ2PkzmEygnd9x+4DrbwAfAO1UtSmu/4L3uZUtNbwe6BBwrj2wLoR8lUtVC1T1bWABLuhZA+QBrVS1mWdLVtX9Q8ynMcZUiwUXZm/xKq5fxEV4mkQ8koDtQI6IdAfGhvi8ycC5ItJDROKB8QHXk4DNqrpbRAbi+kh4ZQPFQKdynv0x0E1ERotIlIiMAnoA00LMWwlPR9FhIpIkIhEicjywP/A/Vc0EPgf+IyLJnuudRcTbfLMB2Ke8Dq7GGFNdFlyYvYKnP8X3QAKuRsHretwH/w7geSCkOSBU9RPgUWAGsNTzr79LgbtEZAdwOy4Y8d6bC9wDfOcZpXJwwLM34UazXAdsAm4ETlTVjaHkLcB2XG3MamAr8CAwVlW9I1POxjXbLMY1k0wB2nquzQAWAVmeZiVjjKkRomo1o8YYY4ypOVZzYYwxxpgaZcGFMcYYY2qUBRfGGGOMqVEWXBhjjDGmRu01k2i1atVKMzIyqn3/zp07SUhIqLkM1SMrS3iysoSnxl6WOXPmbFTVlFrKkmmk9prgIiMjg9mzZ1f7/lmzZjFo0KCay1A9srKEJytLeGrsZRGRwNlijdlj1ixijDHGmBpVZ8GFiLzuWfZ5u4j8ISIXVpC2k4hM8yw3vVFEHqyrfBpjjDFmz9RlzcV9QIZnfYe/A/8Wkf6BiTxTEX+Bmz2wDbAP8Hod5tMYY4wxe6DOggtVXaSqed5Dz9Y5SNJzgfWq+ohnmejdniWtjTHGGNMA1GmfCxF5SkRycatUZuIWcAp0MLBSRD7xNInMEpGedZlPY4wxxlRfna8tIiKRwCHAIOABVS0IuP45MBjXdPIlcBVuJcvuqpofkHYMMAYgNTW1/6RJk6qdr5ycHBITE6t9fzixsoQnK0t4auxlGTx48BxVHVBLWTKNVL0tXCYizwCLVXVCwPn3gWRVHew5Ftxqj0eo6vzynjdgwAC1oaiOlSU8WVnCU0MuS/5fC3nlqeUce3IaHXp0ZtYPC6ozFNWCC1Pj6nMoahTB+1wswPXHMMYYU4FFH03kom7D6bBkIB89+lR9Z8eYEnUSXIhIaxE5Q0QSRSRSRIYC/8A1ewR6HThYRIZ4mlCuBjYCS+oir8YY0xAUFsLyxZklx9HJbesxN8aUVlc1F4rrN7EW2AI8DFytqh+ISHsRyRGR9gCq+jtwJvCMJ+3JwN8D+1sYY0xjNmkSJEetLzk+/Ni0esyNMaXVyfTfqpoNHFnOtdVAYsC5d4B36iBrxhjT4BQXw333wVtn+2ou4lq0hTVb6y9Txvix6b+NMaaBee89WLwY0pr7ai6Is5oLEz4suDDGmAZEFe69F2Kid9MicYs7KVEQ06p+M2aMHwsujDGmAVmxAubMgbbNfE0ixLUBsV/nJnzY/0ZjjGlAMj0xRakmkVgbKWLCiwUXxhjTgGRnu39L1VzEW38LE14suDDGmAbEG1xYZ04Tziy4MMaYBmTjRvdvqZoLaxYxYcaCC2OMaUBKai6a+dVcWLOICTMWXBhjTANSUnPR3GouTPiy4MIYYxoQq7kwDYEFF8YY04AEHS1iNRcmzFhwYYwxDcjGjW52zpZJm90JiYTYlPrNlDEBLLgwxpgGJDsb2jTN8p2Itdk5Tfix/5HGGNNA5Oa6zea4MOHOggtjjGkgvCNFSgUX1pnThCELLowxpoGwzpymoaiz4EJEXheRTBHZLiJ/iMiFIdzzpYioiETVRR6NMSac2dTfpqGoy5qL+4AMVU0G/g78W0T6l5dYRP4JRNdV5owxJtwFnfo7zmouTPips+BCVRepap730LN1DpZWRJoC44Eb6yh7xhgT9qzmwjQUddrnQkSeEpFc4DcgE/i4nKT3Ak8DWeVcN8aYRidonwuruTBhSFS1bl8oEgkcAgwCHlDVgoDrA4D/AwYA+wArgGhVLQzyrDHAGIDU1NT+kyZNqna+cnJySExMrPb94cTKEp6sLOGpIZXl4Ye78dFHaWx8pmXJJFrfpU6lILIFUL2yDB48eI6qDqjxzJpGrc6Di5IXizwDLFbVCX7nIoAfgRtU9SsRyaCC4MLfgAEDdPbs2dXOz6xZsxg0aFC17w8nVpbwZGUJTw2pLKecAh9PyyPvlVh3QiLhjPySSbSqUxYRseDC1Lj6HIoaRdk+F8m4Gou3RCQL+Nlzfq2IHF6XmTPGmHCTnQ3pzdf5TtjsnCZM1ckQTxFpDRwFTAN2AUOAf3g2f9sA/95J7YCfgP5Adu3n1BhjwtfGjdC7w3zfiaQu9ZcZYypQVyGvAmOBtcAW4GHgalX9QETai0iOiLRXJ8u74QsoNqhqfh3l1RhjwlJ2Ngzs/JPvRMuB9ZcZYypQJzUXqpoNHFnOtdVA0B5IqroSkNrLmTHGNAyFhbB5MxzY6WffyZYH1l+GjKmANdYZY0wDsHkziBQzoKNfx/UWFlyY8GTBhTHGNADZ2dC1zZ80S9jmTsSkQEKH+s2UMeWw4MIYYxqA7OwgTSJircYmPFlwYYwxDcDGjQHBhTWJmDBmwYUxxjQANlLENCQWXBhjTAOwKbuAvh3m+k7YSBETxiy4MMaYBqBJ7kJim7iFpbcVdYDYlHrOkTHls+DCGGMagBb4+ltsEWsSMeHNggtjjGkA2jbxBRd5idYkYsKbBRfGGNMANI1aVbIf1+aAesyJMZWrNLgQkUgROV9EYuoiQ8YYY0orLoZo2VFy3CqtaT3mxpjKVRpcqGoR8Iiq5tVBfowxxgTYsAESY3zBRXxyUj3mxpjKhdos8qGInFSrOTHGGBPUqlWQGJvjOxEVdK1HY8JGqKuixgJTROQHYA1uCXUAVPXs2siYMcYYZ/Vq6Bbrq7kgymouTHgLNbj41bMZY4ypY6tXK0mt/YKLaAsuTHgLKbhQ1TtrOyPGGGOCW78mn+i0QgCKNJrISOtfb8JbyENRRWSQiLwoIp95/h1clReJyOsikiki20XkDxG5sJx054jIHE+6tSLyoIiEWsNijDF7nb/W+2otisT6W5jwF1Jw4QkEJgNZwDtAJvCmiFxUhXfdB2SoajLwd+DfItI/SLp44GqgFXAQcDRwfRXeY4wxe5Ut2b7gQq2/hWkAQq0RuBE4RlXne0+IyFvAVOD5UB6gqov8Dz1bZ2BOQLqn/Q7XichEoEq1JMYYszfZvskXXETEWHBhwp+oauWJRDYBbVS1wO9cDLBeVVuG/DKRp4BzgThgLnCEquZUcs97wG+qOi7ItTHAGIDU1NT+kyZNCjUrZeTk5JCYuHdUN1pZwpOVJTyFe1l27Yrk7qsi+f6OQwHYFr0fc1OeCpq2OmUZPHjwHFUdsMcZNcafqla6Ae8DjwPxnuMEYALwYSj3BzwrEjgMuBWIriTt+cBaoFVlz+3fv7/uiZkzZ+7R/eHEyhKerCzhKdzLsmiR6rE9P1WdiNu+HFJu2uqUBZitVfw9bpttlW2hdui8BOgFbBORDcBWoDdwcTWCmSJV/RbYBxhbXjoRGY7rp3G8qm6s6nuMMWZvsHo1JMXZHBemYam0z4WIRALHA0NxnSzTcM0ha2vg3Z3LeedxuL4cw1R14R6+xxhjGqzVq212TtPwVGVtkd2qulZVf6pqYCEirUXkDBFJ9CyENhT4B/BlkLRHAROBEar6U1XeY4wxe5tVqyAp1ibQMg1LXa0torgmkLXAFuBh4GpV/UBE2otIjoi096S9DWgKfOw5nyMin+zBu40xpsFavdqCC9Pw1MnaIqqaDRxZzrXVQKLfsQ07NcYYj9WrYb821ixiGhZbW8QYY8LY6tWQlGEdOk3DEmqHzk7AGFXNq/0sGWOMASgqgrVrA0aLWLOIaQBC7dB5LFBc+9kxxhjjlZkJhYU2WsQ0PKF26PwvcKeIRNdmZowxxvisXu3+tQ6dpqEJtc/FFUAb4FoRyaZ0h8725d5ljDGm2tascf+WCi6sz4VpAEINLs6s1VwYY4wpY90692+pZhGruTANQEjBhap+VdsZMcYYU5o3uChdc2F9Lkz4q7DPhYi8H3B8Z8Dxz7WRKWOMMW6kCNhoEdPwVFZzETih1RXAeL/j7jWbHWOMMV4lzSIx/qNF6i64+OWXX4ZGRUWNV9U2hD4AwOz9ioBvCwsLL+rfv39+sASh9rnwkoBjDZrKGGPMHlu7FqIiC4ht4pliSCIgMrZO3v3LL78MjYmJeSIjIyM/Li5uS0REhP2+NwAUFxfLqlWrDtu6detY4LFgaaoaidp/LmOMqQPFxbB+fZCRIhL4N17tiIqKGp+RkZGfkJCwywIL4y8iIkLT0tJyIiMjzy0vTWU1F9Eich6+GosYETm/CvcbY4yphuxsKCiAxOT6GSmiqm3i4uK21NkLTYPSpEmTAlVtUd71yoKD/wH+C5P9BJwVcN0YY0wNC4ORIhFWY2HKI64GrdzWjwqDC1UdVMP5McYYE4KgI0VsAi3TQFjvX2OMCUNBR4rYMNRGR0T6//rrrzEAo0ePbn/DDTe0rc5z4uPj+y5evLhJzeaufBZcGGNMGCppFomzCbSCSU9P7xkbG9svPj6+b6tWrXqPGDEiY9u2bSWfaSNGjMiIjo7uFx8f39e7Pf/8880D7/VuZ599drWWspg2bVpSRERE//j4+L4JCQl9MzIyDnjsscda1lQ5/b3xxhurH3rooczK0g0cOHDfRx55pJX/udzc3Lk9evQIOmy0NtRZcCEir4tIpohsF5E/ROTCCtJeIyJZnrQvikhMXeXTGGPCQUmziC1aVq5Jkyb9mZubO/eXX35Z9Ouvv8bfeuutpf6qHzt2bFZubu5c73bRRRdtCbzXu7366qurq5uPlJSUgtzc3Lk7duyYe/fdd6+97rrrMubMmVNmzHBBQUF1X9Hg1GXNxX1AhqomA38H/i0i/QMTichQYBxwNNAB6ATcGZjOGGP2ZkHXFbE+F0G1b9++cPDgwdsXLlwYV5/5iIiI4KyzztqalJRUuGDBgrgJEya07NevX/cLLrigXbNmzfpcd911abt27ZIxY8bs07Zt254tW7bsPXr06PY5OTkl44tvu+221JSUlF6tW7fu9eijj5aqARkxYkTGlVdemeY9fv3115t17969R2JiYt927dodMGXKlOQrrrgifc6cOYk33XRTe/8aGf/mlU2bNkWecsopGc2bN++dlpbW88Ybb2xbVFQEwIQJE1r2799/3zFjxuyTnJzcJz09vefkyZOTq/y1CDWhiHQXkdtE5Em/416h3q+qi1Q1z3vo2ToHSXoO8IIn/RbgbuDcUN9jjDF7A6u5CN2yZcuiZ8yYkdyxY8e8ylPXnqKiIl599dVmO3bsiOzXr18uwIIFCxI6deqUl52dPe/ee+/NvPzyy/dZunRp7Lx58xYvXbp0YVZWVpN//etfaQBTpkxJfuqpp9p8/PHHfyxduvTXL7/8stwP9ZkzZ8aPHTs247777lu7bdu2uV9//fXvnTt3zn/88cfX9e/fP+e+++5bXV6NzIUXXthu+/btkcuXL184c+bM3ydPntxywoQJJc0o8+fPT9h33313b968ed6VV16Zdfnll2cUFxdX6WsR0jwVInIa8CTwDjAauAxIBO4HhoT6MhF5ChcoxAFzgY+DJNsf8F/TZD6QKiItVXVTwPPGAGMAUlNTmTVrVqhZKSMnJ2eP7g8nVpbwZGUJT+FaltWrDwOiSvW5WLE2m1XbZpV7T7iWpbaMHj26i4iQm5sbcfDBB+946KGH1vtff/bZZ9u89NJLrQGioqJ0y5Yt8/3vjYyMLBlqe8cdd6y97rrrNlYnH9nZ2dFJSUl9IiIiaNu2bf5TTz21onfv3nlfffVVYkpKSv4tt9zyF0BkZKS+8cYbrWbPnr04NTW1CODmm2/OPOecczo9+eST6956660Wo0aN2njggQfuBrjnnnvWT5s2LehcEs8//3zK6aefvumUU07ZDtCxY8cCoNJ2l8LCQqZNm9bixx9/XNy8efPi5s2b51922WVZb775ZstrrrlmI0BaWlq+92tx6aWXbho3blz7tWvXRrVv374w1K9JqJNg3QUco6rzRWSU59x8oHeoLwJQ1UtF5ArgEGAQECzKTAS2+R1795OAUsGFqj4HPAcwYMAAHTRoUFWyU8qsWbPYk/vDiZUlPFlZwlM4lmX7dsjNdfvNEnzNIh279qJj90Hl3heOZalNb7zxxtLhw4fv+OijjxLPO++8TllZWVGtWrUq8l6/+OKLsyZMmLC+onsre0d8fHxf7/78+fMXde3atUynyJSUlIINGzYsCHZ/27ZtSz7wMzMzo3bv3h1x8MEH7+efpqioSACysrKi+/Xrt9N7Pti7vNatWxc9dOjQbeVdL09mZmZUYWGh+D+7Y8eO+Rs2bIj2L493PykpqRhg+/btkUDIwUWozSKtAe8XTv3+rfIEK6papKrfAvsAY4MkyQH8q4K8+5X+JzDGmL2Bt78FQGoLm+eiMsOGDcsZNWrUpquvvnqfmn62f6fPij7syyMiJZ+Tbdq0KYyNjS1esGDBoh07dszzbrm5uXMBUlNTC9asWVMyXHTp0qXlDh1NT08vWLZsWdCFZvzfGaht27aFUVFR+ueff5Y8e+XKlU1SU1NrtLdpqMHFHErPzAlwBm7GzuqKInifi0WUrhHpDWwIbBIxxpi9lX9w0aqZDUUNxU033bThu+++S/7hhx/qtVNnRSIjIznjjDM2XnbZZe3WrVsXBbBixYroqVOnJgOcfvrpmydPntxqzpw5sTt27Ii49dZb08p71kUXXZQ9efLklu+//35SUVERK1asiJ47d24sQEpKSuHy5cuDjrKMiorihBNO2DJu3Lj0LVu2RPzxxx9NnnzyydQzzjijRj9jQw0ursSN7vgKSBCRz3AdLa8J5WYRaS0iZ4hIoohEekaE/AP4MkjyV4ELRKSHiDQDbgVeDjGfxhjT4Hk7cwK0SLIOnaFIS0srPPXUUzfdcccd5X4g+zvjjDO6+s9zccwxxwT7Y7fGPfnkk2s7deqUd9BBB+2XmJjY9+ijj+62ZMmSWIDTTz99+5gxYzYMHTp0386dOx8wePDg7eU9Z/DgwblPPvnkyhtuuKFdcnJy3yOPPHLf5cuXNwG4+uqrN0ybNq15cnJyn3PPPbdd4L3/93//tzo+Pr64U6dOPY844ojuI0aM2HzVVVdVq79JeUQ1tJYNEYkHTsQND10DTFPVnIrvKrk3BZiCq4WIAFYBE1T1eRFpDywGeqjqak/6a4F/4Tp+TgUu8RtpEtSAAQN09uzZIZUlmL2prdLKEp6sLOEpHMtyzz1w661uf/mzh9Mx8Vt3MOQraH1EufdVpywiMkdVBwSenz9//srevXvX6AeO2bvMnz+/Ve/evTOCXQt1tMgEVb0SmBxw/lFVvbqy+1U1GziynGurcZ04/c89AjwSSt6MMWZv419zkVh/C5cZU22hNoucW875wH4Yxhhj9pB/n4u4aOvQaRqeCmsuROR8bzq/fa9OgFWZGWNMDdqwAX780XccE5ED3sGV9djnQoQyMyrXFlXm1NW7TO2orFnEWzPRhNK1FApswM2maYwxpgYUFcGZZ0J2tjtOTYUorFnENDwVNouo6mBVHQzc7933bEep6j9U9ceK7jfGGBO6e+6B6dPdvgi89kohUrTLc1UgKqHe8mZK69Kly/7Tpk1LArj22mvTTj755I419exx48a1GTVqVIeael59CHWGzttFJGggoqpVm3DcGGNMGYsWwR13+I5vuQWOGbzTjbMDV2shEuzWOhGOTRXPPPNMiyeeeCJ1+fLlsQkJCUX77bffrltuuSVz6NChIY1kDGbEiBEZ6enp+eXN7Om1dOnSRdV9h79p06YlXXDBBR39Z/i8//77s2ri2fUp1A6dhbg5y4Ntxhhj9tCXX4J3ZoC//Q3GjwcK/ee4sCYRf3fccUfqzTff3O6GG27IzMrKmr927dqFl1xyyV/vvPNOs9p8b2NaNn1PhBpcdMR14PRuhwIf4lk0zBhjzJ5Zs8a3P2wYREUBBTZSJJhNmzZFPvjgg2kPP/zw6nPOOWdrcnJycUxMjI4ePXrbs88+u3bXrl1y/vnnt2vdunWv1q1b9zr//PPb7dq1S8DVFKSmpvYaP358aosWLXqnpKT0euyxx1oCPPzww63ef//9Fk8//XSb+Pj4vkcddVQXgPT09J633HJLm27duvVISEjoV1BQQHp6es/33nuv5JuSl5cnw4YN65SQkNC3R48e+/nPFOq/3Dn4lk7fvn17xMiRI7tmZ2dHeyfzWrlyZXRgM8vEiRObdunSZf+kpKQ+AwcO3PeXX34pmfY7PT295+23357arVu3HklJSX2GDRvWKTc3t/6quDxCCi5UdVXA9iOuM+e/ajd7xhjTOKz2Wxi7nXdOxUK/2n2bnbPEzJkzE/Lz8yPOOuusLcGu33TTTW3nzJmTMHfu3MXz5s1bPHfu3IRx48a19V7ftGlT9LZt2yIzMzMXPPHEE6vGjRvXPjs7O/L666/fePLJJ28eO3ZsVm5u7twZM2Ys9d4zderUFh9//PGfmzdvnhsdHV3mndOnT282cuTILZs3b543cuTIzSNHjuySl5dX4Yd8cnJy8ZQpU/5MSUkp8K5fkpGRUapqZMGCBTEXXnhhp4ceemjNxo0b5x977LFbhw8f3mX37t0lz3733XdbfP75538uXbp04ZIlS+KeeOKJVmXfVrdCrbkIJhlIqamMGGNMY+Zfc9G+vWenwEaKBJOdnR3VrFmzwmAf8uACgZtvvjkzPT29MC0trfDWW29dP2XKlJbe61FRUfrQQw+tj4mJ0VGjRm2Li4srXrBgQdBFwLwuueSSDV26dClITEwMOq31/vvvn3veeedtiYmJ0fHjx2/Iz8+XmTNn7nEP3Ndee63F4MGDt51yyinbY2Ji9M4779ywe/fuiOnTp5f8hxg7duyGjIyMgtTU1KJjjz1227x58+p9fZVQZ+h8jdIroMYDRwCv10amjDGmsfGvuSgJLnb/5TsZY3/LeaWkpBRu3bo1qqCggGABRnZ2dpPOnTuXLBnRqVOn/L/++qskYdOmTUsFJnFxccU7duyo8I/tDh06VNjZIi0trWTF1MjISO8Kp8GjnypYv359dLt27Uo9u23btvn+z05LSyvJW3x8fHFmZuYev3dPhVpzsRRY5rf9CIxW1StqK2PGGNNYFBTAes/YBBFIT/dc2J3pSxTXps7zFa4GDx68Mzo6uvj1119vHux6SkpK/rJly0r6OKxYsaJJ69atQ+qJKeWMyKloGXOA9evXlyxhXlRUxIYNG6LbtWtXABAbG1u8c+fOks9b/0CnsuempaWVWoa9uLiYzMzMJt5nh6tQ+1zcGbA9rKrTaztzxhjTGKxf7xsp0qYNNPF+lOzyG5EY17bMfY1Vy5Yti2688cb1119/ffvXXnut2Y4dOyLy8vJk8uTJyZdccsk+p5xyyub777+/7fr166MyMzOj7rnnnrYjRowIaUnx1q1bF6xYsSLocuUVWbRoUfwrr7zSrKCggLvvvju1SZMmOnjw4J0A++23365XXnmlRWFhIVOmTEn++eefSzrQpKWlFW7bti1q06ZNkcGee+aZZ26eOXNm0/fffz8pLy9P7rjjjtQmTZrokCFDqj3cti6U2ywSZLrvoFT1xZrLjjHGND5BO3MC7PKruYi1mgt/d95554Y2bdoUPPDAA23HjBnTMSEhofiAAw7Yecstt2QeeuihuZdeemlk7969ewAMGzZsy/33359Z2TMBxo4du/G0007rnJSU1Oeggw7aMX369GWh3DdkyJCtkydPbnHppZd2bN++fd5bb721LCYmRgEeffTR1eeff37Hpk2btj7mmGO2HnPMMSUdUfv27bv7pJNO2ty5c+eeRUVFLFy4sNT8Gb1798579tlnV1x77bXtzzzzzOju3bvveu+99/6MjY0NbUnzelJRn4tQFiVTwIILY4zZA0E7cwLs9qu5sOCijLFjx24eO3bs5mDXXn755TXAmsDzJ5544g7/CasA1q1bt9C737Nnz7zffvttcXnXg5175JFHKpxw64gjjsitaNKtt99+e6X/ceDzzj777K1nn3321mD3BuatsrzUlXKDC8+038YYY2pZuTUXu61ZxDRMoU7/jYg0B04C0oF1wIeqGnSMsTHGmNCVW3Oxyzp0moYppA6dInIIbpTIJUAv4GJgmed8KPfHiMgLIrJKRHaIyDwROb6ctCIi/xaRdSKyTURmicj+IZbHGGManKA1F8UFkLfRcyA2FNU0KKEORX0UuFRV/+ZZDfVQYCwwIcT7o3BtX0cCTYFbgckikhEk7WnA+cDhQAvgB+C1EN9jjDENTtCaC/85LmJbQ0TIFc3G1LtQg4tuwOSAc1OALqHcrKo7VfUOVV2pqsWqOg1YAfQPkrwj8K2qLlfVItxEXT1CzKcxxjQ4QWsubKSIacBCDYX/BM4A3vA7dxquqaTKRCQVF7AE6z07CThdRLrhApBzgE/Lec4YPIunpaamMmvWrOpkB4CcnJw9uj+cWFnCk5UlPNV3WXbtimTLlsMBiI4uZvHir/ntN2i5+wd6etJs3tWEBSHksb7LYoxXqMHF1cA0EbkSWAVkAF2BE6v6QhGJBiYCr6jqb0GSZALfAr8DRbjmlKOCPUtVnwOeAxgwYIAOGjSoqtkpMWvWLPbk/nBiZQlPVpbwVN9lWbLEt9+uXQRHHeXJy9I/4Se32yJ9fwYdPKjSZ9V3WYzxCim4UNXvRaQzMAxIwy23/rGqBh1fXB4RicD1n8gHLi8n2e3AgUA7IAs4E5ghIvuram5V3meMMeGu/JEiNseFabhCXhVVVbeo6uuq+iDwPW5V1JCJm7D9BSAVGKGq5c2L3gd4S1XXqmqhqr4MNMf6XRhj9kI2x4XZG4U6FPVNEfmbZ/88XF+JRSJyQRXe9TSwH3CSqu6qIN3PwGkikioiESJyFhCNWzzNGGP2KiHNcWE1F2Wkp6f3jI2N7RcfH9+3VatWvUeMGJGxbdu2ks+0ESNGZERHR/eLj4/v692ef/755oH3erezzz67fflvK9+0adOSIiIi+vs/Kz4+vu/06dP3eLn1hizUPhdH4zpWAlwLDAG2Au/haiMqJCIdcHNj5AFZfqvOXQx8AywGeqjqauABoDUwD0jABRUjVHVriHk1xpgGI7SaCwsugpk0adKfw4cP37F69eqoY445ptutt97a9vHHH1/nvT527NisCRMmBJ0O23tvTeQjJSWlIHBK8WCKi4tRVSIjfWuUlbdsfHmqmr6+hNos0kRV80UkHWihqt+p6iJcE0elVHWVqoqqxqpqot82UVVXe/ZXe9LuVtXLVLWtqiaraj9VDTpaxBhjGrrQ+lxYs0hF2rdvXzh48ODtCxcujKvvvAQaOHDgvldccUV6v379usfHx/dbsmRJjIj0v++++1I6dOhwQEZGRk+A//znP63at29/QNOmTfscddRRXVauXOm/LHuZ9OEu1OBinojcBNwGfATgCTS211bGjDGmMVi1yrdfUnOhCrtt6u9QLVu2LHrGjBnJHTt2zKvvvAQzZcqUFs8999zKHTt2/NK1a9d8gA8//LDZTz/9tOT333//9YMPPkj697//nf7GG28sz8rKmt+uXbu8kSNHdvJ/hn/6+ilF1YQaXFwA9ATicAEGwCG4IaXGGGOq4dNPYamnN1lkJHTo4LlQsB2KdnsuxENUYr3kL9yNHj26S0JCQt8uXbr0atmyZeFDDz1Uqgnk2WefbZOUlNQnKSmpT/PmzXsH3uu9lpSU1Oc///lPq+rmIzs7O9r/WUlJSX22b99e8vk6atSoTQMGDNgdHR2Ndxn2cePGZaWmphYlJibq66+/3mLUqFGbDjvssNy4uDidMGHCunnz5iX8/vvvTbzP8E9f3XzWpVCHoi4DRgecm4KbpdMYY0wV7doFl13mOz7zTEhK8l70r7VoC75+asbPG2+8sXT48OE7Pvroo8TzzjuvU1ZWVlSrVq2KvNcvvvjicvtceO+t7B3x8fF9vfvz589f5K158FdZn4t27dqVuadjx44l57Kyspr07dt3q/e4adOmxc2aNStatWpV9L777psfmL4hqMqqqOcD/8DNc7EeN5Pmi6raIKIoY4wJJ/fdB8uXu/3mzeHBB/0uWmfOKhk2bFjOqFGjNl199dX7TJ8+vVozR5cnNzd37p4+Q4IEh/7n2rRpk79q1aoY7/H27dsjtm7dGtmhQ4eCYOkbglCHoj4I/At4B7gBmApcjxvZYYwxpgqWLoUH/H573ncftG7tl8Am0Kqym266acN3332X/MMPP4Rdp87KjB49evNbb73V8vvvv4/btWuXXHXVVem9e/fe6a21aIhCrbk4F+inqmu9J0TkI+AX4MZayJcxxuy1Jk2CfM/HxkEHwUUXBSTYHdAsYiqVlpZWeOqpp26644470j777LNKay/OOOOMrhERESU174ceeuj2L774olq1HtnZ2dH+zScATz311Ipzzz13ayj3Dx8+fMcff/yxftSoUZ23b98e1a9fv5zJkycvr05ewkWowcUOzxZ4zkaLGGNMFfnPbXHWWRARWIdsNReVWrdu3cLAcxMnTiz5yk6dOnVlVe6trhNPPHFHcXHxnPKu//TTT78HnlPVMulvvPHG7BtvvDE72DOCpQ935QYXIuI/DOZR4B0RuR9Yi1v34wbgv7WaO2OM2Qut9+timJ4eJIFN/W0auIpqLpYCCvj3IhkckOYo4ImazpQxxuzN/IOLtLQgCWzqb9PAlRtcqGrIi5oZY4wJXaZf7NA2WMXELr/oI1xGi7wh/evsXaMbXjOAKa3aAYSI7O8ZRWKMMSZEhYWwYYPvuE1g7KAKOX59+RI71km+jKlJVQouRKSViFwpInOABdgy6MYYUyUbNrj4Adzw0zJrUO3KhCLPwtFNWkCT5nWaPxOaLl267D9t2rQkgGuvvTbt5JNPrrEocNy4cW1GjRrVofKU4avS0SIiEg2chFsV9ThgDW4irQNV9ZfazZ4xxuxdKu1vkbPUt5/YudbzE7IwbKp45plnWjzxxBOpy5cvj01ISCjab7/9dt1yyy2ZQ4cOzanuM0eMGJGRnp6eX97Mnl5Lly5dVN13+Js2bVrSBRdc0NF/hs/7778/q6J7GoIKay5E5EkgE3gSWAUcqapdgG24USPGGGOqoNL+Fjv8goukLrWen4bqjjvuSL355pvb3XDDDZlZWVnz165du/CSSy7565133mlWm+8tKCioPJGptFnkEtyIkTuA21T1x1rPkTHG7MUqrbmw4KJSmzZtinzwwQfTHn744dXnnHPO1uTk5OKYmBgdPXr0tmeffXbtrl275Pzzz2/XunXrXq1bt+51/vnnt9u1a5eAqylITU3tNX78+NQWLVr0TklJ6fXYY4+1BHj44Ydbvf/++y2efvrpNvHx8X2POuqoLgDp6ek9b7nlljbdunXrkZCQ0K+goID09PSe7733nnc1GPLy8mTYsGGdEhIS+vbo0WM//5lCRaT/r7/+WjK994gRIzKuvPLKtO3bt0eMHDmyq3cSrvj4+L4rV66MDmxmmThxYtMuXbrsn5SU1GfgwIH7/vLLL7Hea+np6T1vv/321G7duvVISkrqM2zYsE65ubn1Pld4ZcFFZ+Bx3JwWWSIyVURGhHCfMcaYICpvFvGbJDLRgotgZs6cmZCfnx9x1llnbQl2/aabbmo7Z86chLlz5y6eN2/e4rlz5yaMGzeupJ5o06ZN0du2bYvMzMxc8MQTT6waN25c++zs7Mjrr79+48knn7x57NixWbm5uXNnzJhREulNnTq1xccff/zn5s2b50aX6SgD06dPbzZy5Mgtmzdvnjdy5MjNI0eO7JKXl1fhh3xycnLxlClT/kxJSSnIzc2dm5ubOzcjI6NU1ciCBQtiLrzwwk4PPfTQmo0bN84/9thjtw4fPrzL7t27S5797rvvtvj888//XLp06cIlS5bEPfHEE9Ve4bWmVBgkqOpKVb3L0xRyLLAZeAFIAe4RkZA6dIpIjIi8ICKrRGSHiMwTkeMrSN9JRKZ50m60USnGmL1F1WouwqjPRRjJzs6OatasWWGwD3lwgcDNN9+cmZ6eXpiWllZ46623rp8yZUpL7/WoqCh96KGH1sfExOioUaO2xcXFFS9YsCA26MM8Lrnkkg1dunQpKG/J8/333z/3vPPO2xITE6Pjx4/fkJ+fLzNnzkzYo4ICr732WovBgwdvO+WUU7bHxMTonXfeuWH37t0R06dPT/SmGTt27IaMjIyC1NTUomOPPXbbvHnz6n19lZBrIFT1G1W9CGgD/BM3S+e8EG+PwnUEPRJoCtwKTBaRjMCEItIE+AKY4XnXPsDroebTGGPCWYV9LlQDOnRazUUwKSkphVu3bo0qr/9DdnZ2k86dO+d5jzt16pT/119/lUQiTZs2LRWYxMXFFe/YsaPCz0P/FUqDSUtLK1lkLDIyktTU1II1a9YEj36qYP369dH+S7ZHRkbStm3bfP9np6WlleQtPj6+eOfOnZF7+t49VeXmDVXdrapvqupxQEaI9+xU1Ts8NSHFqjoNWAEEm5TlXGC9qj7iuW+3qi4Iks4YYxqcCmsu8jZCgWfJpqhEiG2NKWvw4ME7o6Oji19//fWg43RTUlLyly1bVtLHYcWKFU1at24dUk/M8pY2F5GgNRZe69evb+LdLyoqYsOGDdHt2rUrAIiNjS3euXNnyeetf6BT2XPT0tIK1qxZU/Ls4uJiMjMzm3ifHa5CXbgsKFWtcKhOeUQkFegGBBvKczCwUkQ+AQ4EfgWuUNUyC82IyBhgDEBqaiqzZs2qTnYAyMnJ2aP7w4mVJTxZWcJTXZdl1aq/Ae6zYuXK79m507eqdnL+Yvp59nfQhjlffVWlZ+9N35eKtGzZsujGG29cf/3117ePiorS4cOHb2/SpIm+//77STNmzEg+5ZRTNt9///1tDz/88J0iwj333NN2xIgRm0J5duvWrQtWrFgRU3nK0hYtWhT/yiuvNBs9evTWe+65J7VJkyY6ePDgnQD77bffrldeeaVF//7917333nvJP//8c1Lfvn13glvNddu2bVGbNm2KbNmyZVHgc88888zNBx98cI/3338/6bjjjsu55557Wjdp0kSHDBlS7eG2dWGPgovq8MybMRF4RVV/C5JkH9waJn8HvgSuAt4Xke6qWmpte1V9DngOYMCAATpo0KBq52vWrFnsyf3hxMoSnqws4akuy1JQAFu3un0ROOWUvxHl/1t4xVrY6HaT2vZm0OFVy9fe9H2pzJ133rmhTZs2BQ888EDbMWPGdExISCg+4IADdt5yyy2Zhx56aO6ll14a2bt37x4Aw4YN23L//fdnVvZMgLFjx2487bTTOiclJfU56KCDdkyfPj2kZdiHDBmydfLkyS0uvfTSju3bt8976623lsXExCjAo48+uvr888/v2LRp09bHHHPM1mOOOaakI2rfvn13n3TSSZs7d+7cs6ioiIULF5b6o7t37955zz777Iprr722/ZlnnhndvXv3Xe+9996fsbGxFdZ41DdRrbv8iUgE8AaQDJysqmWqdUTkfSBZVQd7jgXYChyhqvPLe/aAAQN09uzZ1c7b3vRDaWUJT1aW8FSXZVm7Ftq1c/upqZAVOFXSgjvg1zvdfo9/QZ/7q/T86pRFROao6oDA8/Pnz1/Zu3fvjVV6mGlU5s+f36p3794Zwa7V2ZBST5DwApAKjAgWWHgswM2tYYwxe5Wqzc5pnTlNwxVSs4iItACuB/oAif7XVPWIEN/1NLAfMERVd1WQ7nXgOhEZAswErsRVFC4J8T3GGBOWKh+G6lcDb8NQTQMWap+LN4AYYDKQW9WXiEgH4GIgDzcZl/fSxcA3wGKgh6quVtXfReRM4BmgNfAL8PfA/hbGGNPQVDr1t9VcmL1EqMHF34AUVc2rNGUQqroKqGimssDakHeAd6rzLmOMCVcV1lzkb3VDUQEiYiA+va6yVZ7i4uJiiYiIsGZqU4anv2ZxeddD7XOxADeKwxhjTDVVGFzkBDSJSP2usiAiWbt27apw1krTeOXn50eLyLbyrodaczED+FREXgJK9W9W1Rf3IH/GGNNoVBhcbP/Ttx8GS60XFhbeuXLlyicyMjKIi4vbbTUYxqu4uFjWr1+fWFRUNKG8NKEGF4fjllg/JuC8AhZcGGNMCCrsc5H9rW+/Wc86yU9F+vXr99kvv/xy+bJly8arahtswUrjUwR8W1xc/HR5CUIKLrxzThhjjKm+CmsuNszw7aceVSf5qUy/fv0+Az6r73yYhqfKM3R65qso6ZypquV26DDGGOPk5kJ2ttuPiIDW/suG7MqC7Z7R9hFNoNXf6jx/xtSkkKq5RCRdRN4VkU1AIVDgtxljjKnE1Km+/W7dKD3t94aZvv1Wh0BUva+YbcweCbUN7RkgHzgayAH6AR8Al9RSvowxZq/y/PO+/XPOCbgYhk0ixuyJqsxz0V5Vd4qIqup8EbkA+B54vpJ7jTGmUfv9d/jmG7cfGWnBhdn7hVpzUYRrDgHYKiIpwE6g3md5McaYcPfCC779E08MGCmycxXkLHf7kfHQcmCd5s2Y2hBqcPE/4ATP/mfAW7gZNKu/DKkxxjQC+fnwyiu+44suCkjg398i5TCIbFIn+TKmNoXaLHIWvkDkauA6IAl4tOazZIwxe48PP4S//nL76ekwdGhAgiy/JpE21iRi9g6hznOx1W9/F/Dv2sqQMcbsTV57zbd/3nkBo0QKd0Lmp75j629h9hKhDkWNEZF7RGS5dy5xETlWRC6v3ewZY0zDtWMHfOoXO5x1VkCCJQ9Dnmfyi7g0aN63zvJmTG0Ktc/Ff4EDgH/ipvwGWASMrY1MGWPM3mDaNMjzrCXdq5eb36JE7npY/KDvuNfdEFHleQ2NCUuh/k8+BejiGYpaDKCq60TERosYY0w53n7bt3/aaQEXF9wGRbluv1kv6Bg4PtWYhivUmot8AgIRz3DUTTWeI2OM2Qvk5MAnn/iOR470u7h1ISx/yXfc92GIiKyzvBlT20INLt4GXhGRjgAi0hZ4ApgUys2ePhsviMgqEdkhIvNE5PgQ7vtSRFRErK7QGNOgfPQR7N7t9g84ALp397u46i1KWpjbHg9tAxecNqZhCzW4uBlYASwEmgF/AuuBO0O8PwpYAxwJNAVuBSaLSEZ5N4jIP4HoEJ9fswpzYcVE38Q21aQKS5e6ce7GmMalwiYR7yJlAB3OqJP8GFOXQgouVDVfVa9R1UQgFUjyHIf0samqO1X1DlVdqarFqjoNF6z0D5ZeRJoC44EbQytGDVKFr06CH86Ej3vBlgXlJl2yBE45Bc49F/74o/S1wkJ3rWtXOO44KLa1Y41pNHbuhI8/9h2XDS5+9+0nd8eYvY2oavkXRdpX9gBVXV3ll4qkAquAPqr6W5DrTwJLgXdxQUi0qhYGSTcGGAOQmpraf9KkkFppgsrJySExMZE2uZ/SfesDJed3RbZlTsozFEYkl0q/YEFTbrnlAHJyXOVKVFQxp566jjPPXEVSUiETJnTh3Xf3KUl/990LOeywuumi4i0LwObNTZg6NZ327XMZOnRDnby/JvmXpaGzsoSn2ijLvHlNueYaN6y0ffudvPLKz76LWsQRmccT4VlU+ps2H1IUUTPvr05ZBg8ePEdVB9RIBozxqCy4KMY39FSCJFFVrVIvJBGJBj4BlqnqxUGuDwD+DxgA7EMFwYW/AQMG6OzZVZ+NfMYLr5EiP/Leb8Npm9GPfyZ3Jy5iY6k0qwuOZVrOxxRrJLm5ELljIcv+9yOvfXMGObuTSqVNSnIz8E2ZUvo9hx4K335b5exVy6xZsxg0aBDLl8OQIbBihTs/bRoMG1Y3eagp3rKAa1566CGIjYVrroGIUBv1woR/WRo6K0vFHnsMrr7a7Z97Lrzk13eTnOXwQWe3H5sKp2bV2HurUxYRseDC1LjKOkrOB+KAV4DXcf0sqk1EIoDXcKNPykzA5bn+FHCVqhaKBItnapAWk7b1Xrq3/Y39ez3NiuyOJYHFph0taJm0GYD20Z+z9vMHuO+Dm0lvsZZf7z+cZvtt4/heH3DR6x/SuTN8/7175I4dZQMLgO++gx9+gLQ0mDgRjjgCDjus9oq2eDEccwys9/uO3XUXnHAC1PaXtbZMmAC33ur2k5ODrNFgTJiYO9e33zdwXqxSTSL71kl+jKlrFf7tp6p9gZFAC+A74GPgDKCJqhapalGoLxIXKbyA67MxQlULgiRLxtVYvCUiWYC3LnGtiBwe6rtCtmEW3du6VpmICKVzqq8D5/nPv8i/37ul5Pj6YQ8TH7OTS4c8RbOEbQCc1G8as2f8xrffwltvwX77lX58796l21ovv9ydu+UWF1y8+aY7v2ABnHkmDBoEhx/u+mp4l2eujuzsJgwaVDqwAPjpJ5g+vfrPrW/+QZv/QlDGhJt583z7ffoEXPQPLpIsuDB7p0orllX1V1W9AcgAHgFOBDJFpF8V3/U0sB9wkmd9kmC2AWlAH8/mXYm1P25l1pqVOpgnf/uSXzefUOr0gs0nsc9BJ5OVcifZuzoB0CJxC6/d9gxXDXuuVNp9dj2LCJx+Ovz6q/sAHDLENT9Mmwa33eZL+8svsM3FJajC2We7qtOBA11txldfuaaT996Do4/2BR9VoQoPPtidbM+MwomJLj9e//asClNQ4NI2FNnZLjjy+u47WF3l3j7G1L68PFi0yHfcu3dAAqu5MI1AVVqtu+KGkh4CzAW2hHqjiHQALsYFDFkikuPZ/iki7T377dXJ8m6A5yOSDaGOTKkSES676ygOuPwjfkp5Cfa9BrpdTq8xr/Lkk/DEk5GkHHZ1SfJTO40jISqgU+aKV6DQxUoRETBiBHzxhQss9tkHevYMsgqiR2Gha5v1Tg/sr6AARo92TQElr1rhajcOPRS+/tqdU3VrF7z0kvsAfuopmD27hbd4fPABvPCCb7Gkr7+Ggw6CuDjo0AGWLSv9XlW45x5o0QIGDHC1LB99BJ9/DjNmwJaQv+s16/PPywZDkyeHfn9RETzzDFxwgWueMqa2LF7sfrYBOnWCpk0DEuyw4MLs/SrscyEiLYB/AOfgllh/DTiiqiNEVHUVwTuEegXt3qyqKyu5r8bkRmdA/3PLXuh0Hiy4HQq2QrA+pflbYM0U6Bi4IpHPrbe6gKO42PWDePhhOPlk0JyVHNL1B0SUdu2EE84cSEFsZ666yv2CArjqKvfBeNFF8Pe/u9oRgKOOgn/9yzWfeJtQoqJKd3K8/noYPNjtn3OOCzLAVwOwZg1ceaULHsC95/LL3YcwwJw5bvOXlgazZ0PbtqXPr1gBq1a5fiRRQf5X5ea6802alPtlqpD/sD6vSZNcGSuzcqWrJfJ+nV56CS67DO6913XArS3Fxe5r3L59/fdz2bTJlffPP2H8eOgfdBC4qQkVNomANYuYRqGymov1uI6X7wGXAT8CXUTkKO9Wy/mrf9GJ0PWS0uci42C/G3zHS58te19xIcwbB1Nbc1jCDfzvf64245NP3AJGMz9cxq8P9uLNy0fzxmX/5IG/j+bInb0YMuBXvvkGDjnE96hrr3Uf2t7AAmBoz48YGnkEPWOfLDlXWOibsOuAA+Duu33px42D6CBTkn38sasVyMtz/T68gUV51q93AQ+4Jp7x4927OnVygczf/152To/Fi6FLF/cX3M8/l31mZYqK4LPPyp6fM8d9WPoLnLDss8/c19u/D4sqPPGEO7/cb5607dt9zVZ7ShVOPBEyMuDCC2vmmdVRXAwvvgj77guPPAIffuia3Pw/AE3NqrAzZ8EO2OXpDBURDYkd6yxfxtSlyoKLLCAWuAjXGTNw+79azV246HaF+0Xg1fEs6H4teGclz/4O1vutq5y3CWYeB4sfcMspL3mYAR2+Y9gwiPQM3M3YeguJMTtKv6coF2ZfRovmyhdfuM6dXvPn+/YP7beeyVeczhHdv+HJcy/n3CNfLfUXUnR0Ma+/DjExvnNdurjA5tpr4eWXXZOL1zXXuOYW/2lCRo92/UfGjHG1Lccds4uRA9+mW9vfefttF4QcdpgbgeLfvvzJJ6WbclRdbUhmppsK+d57y355p01zq0UecYR77ubNpa///LP7yxugTRs46STftbfecv8WFLhamORk1wy1ebOb2Oz0090IHnBf+wMP9N27cqWb4Cw7G/7zH0hJcU1Z/jMrBvPnn/Dqq65G6txzg3cuXbXKt67ESy/B2rUVP7O2XHmlawra5Neat20bHHts2YnfTM2osOZih98XPbGzrYJq9l6qulds/fv31z0xc+bMihP8fIXqRFTfSlDdutid++Y0d24iqm9Eqf7xtOqSR1Xf6+A7791mDPU9a9Oc0te+HuHu9x6vmKiqqtu2qQ4YoOo+ot126aWqRd+eU+r+ojdiVTfN1j//VH3+edUXXvip0vKuX6+akFD62d7t8stVi4oCbvjxItWJ6Nbnk7Vdy1Vl7hHx7cfEqC5a5G77+OPS6eLjVXNzfY+dPVs1NrZ0mpgY1eee831fbrvNd+3cc1XffNN33Lq16v/9n+qxx5Z+Rs+eqvvv7zveZx/VH39ULS5Wfe019w7vtebNS98bGak6aVLp4u/Yofr006oHHxz8a/bzz6XTT51a+vp//hPC/7Ea9sEHpfPQoYNqs2a+4/btVTdurN6z67ostam6ZSkuVp0/X3XnTt+5oiLVpCTf13jt2oCbVkz0/ex+dXJ1s1yu6pQFmK1h8Dvctr1ra2DTENWjfv+Fw6fCsT9AU8+Y0z4PQJxn1XkthJ/Hwi9Xw85Vfjd6GtszP4ONns4O827yXd5nOBw+Bfa9ynful+tgVxbJiYV8+qmvavXoo+HR22YTsar0n8oRuhu+PoUuefdx4ZA32bdD5X8mt23rmkr8RUbCf//rah5KTVBVlA+rXLVG0/jtXHfCf0ouRUW5TqSbNvn+SsvLg7POcn+t3xgwgXturm847IYNMHy4b3Enr7w8uPhi12GzqEj48EPftRNOcDUX3kkI//rLNTt8/nnpZyxc6KtRiYmB9993HVlFXPPPG2/4+kEEdlItKnI1Nzfd5PL64IPQsSOMHQs//hj0y1kqj+BGBvkLnDw2K8vVKPToAe3auVqT4493z1m2zNUwpaW5pqbApp9QZGeXbo45+WTXPPXRRxAf786tXu0bPVQZVVcbVBTy4PO93/XXu5EgHTr4+gStWOGrKWvVyn0PS7H+FqaxqO/opqa2Wq+5KM/ONaof9SxbUzG5qerqqarfjvadm3mC6vJX/Wo7IlS3ev7Ez9+u+k5a2ee83UKLPj1UN350gRbNHa/6SX/ftc8Pc+8JuCfvzeaquVnuuZnTVae2Uf38UNWclaWzvlO1c2dVUG3ZUnX69HLKmDWr1PN3vhinLROzNSFB9bPPfMkWLlRt0kSD1mb4bxdeqJqfr3rYYb5zTZuq3n+/6r77+s41aaKakZFTqkZhyxb3rmnTytY4gOrpp6tGRJQ+9+KLwYv1xBOl0110kWqPHsHz7L9FRakOG6Z62mm+c4ceWvrZxx9f9r7XX/9RVVXffVe1VavK3+PdEhNVX301pP+Nqur+oj7lFN/9bduqbtrku/7WW75r0dGqy5aVvr+oSHX5cvccr+uvd+mPPlo1L8/387J2rfteNmTV+dkvLnY/M/7fp2uvVX38cd/xMccEufGbUb6fpaUv7HHeA1nNhW3hstV7Bmpqq7fgQlU1b6vqrJPcL4xPB6r++axq/jZ3besi1YlSNmiYiOoP55V+zoo3g6cLtr0ZrbrtD9W109x+4PUfL3T5mtrGd25qG9VNs0u9ct061VdeUc3MrKB8c/9V5vlTx9+uc+eWTfrYYxr0A9L/wzY1VfXOO33HERGqn3zi7t+4sXSAEfjh72/bNtW771Zt0cIFJ94P4ClT3IcmqI4dW/G37qWXVE84wTW1qKpu2KB6wAHB39+hgytfdrZL+9dfvmtRUa7pxCs1tez95567XC+/PPizQ9mOPlp1xgzV7dtVv/5adeLE4M0azz1X+j7v19aruNgFQ97rZ5zhuzZzpmrXru78sGGqBQWq33xT+nm33qo6Y8ZMveYad7zffqWbBhqa6vzsr11b+ffrhhv8bshZpZr5peqH+/p+jv76tsbK4GXBhW3hstV7Bmpqq9fgwqtwd/Dz/n0zSmokmrtfOP6Ki1XnXKv6TlvVt5JczUZ5wcWca333bV+q+ttjqv8b45dGVGccV/a+SfGq6z+vWrk+6h08//k7gib/4APVv/1NS37J7rOPCwRat/ad86/VuPfe0vevWKHapo3velyc6n33ub+YgykqUt0d8KVfutR9EPv/9R2qXbvcB/cll6j27q06cKD7wA72/t69ffn86CN3bv16DfphI1Jc6jg93X2tVq5UXbJE9cYbXaAErg/Jiy/6PujL21q39r1X1fUB8O/DUl5w9f33pZ9zww2u1ifw+TfdpNq3b+lzkZGqQ4dmljoX2Ecl1K/z1q1Vv6+mVedn/6OPSn89gn1v3njDk3jNB8H/wNiVXaPlULXgwrbw2eo9AzW1hUVwUZ5tv6lOaeV+obzXQfXnK11AUJniYtWc1arrPlX97XHVebeo/niB6vzxqoXlfNLOPCF4MDIpzq8Go7VrhgnFznWla0v8O6sueaTCW3/4QXXCBNXVq93xBRdomV/AAweqFhaWvXfRIveX83HHrdcVK0LLan249lpfWa65xp2bNs13rk8fFxwFlnvECNXNm8s+Ly/PVzOi6mpDLrqobFNP4HbBBaqTJ5eu9TnggIprFEaMqPiZVdnOOiv0r1lengsW4+NdIPRtzf8BXyXV+dm/915f2S+9VPWFF1SHDPEFdi1a+NUqfXtG2Z/H9zJqtAxeFlzYFi5bvWegprawDi5UVfO2qO5YUb0/pati229aNDEyYKTKcapbl5RuIpk/PrTnLX3Rd8/0o9yIGO/xu/uUH+QEETh6ITpa9ddfK74n3Ecl+P8F26uXO3fXXb5zV15ZtkZg3Lggo3EqsXSp6sUXu1EukZFuNEywphfvFh+vunhxxc/844/SfWS826mnqg4aVPb8+ee75wZ7X8uWZYPEX35RfeAB1ays0ucC+7WMGlW1r0VNq87/Mf/vqXdkk6qrQZs9u3SAqJ8f5vuZ+bifGyWy4Zs9zXZQFlzYFi6bjRapK02aQWJG7U/VmLwv6xJO8R1HxsKAJ6Bpd+hzv+/8bw/DrszKn5fpN39H2+Og4zlumWiA3LWwcmLIWRsyxDdSAdw8EfvvH/LtYemII3wzki5Y4Eav+I8U6dfPjSpo0sTNP/L003DffVVfLr5zZzcHyM6dkJPj3vXrr27SsmCefrrsQnqBunZ1o2yuuspN8/7f/7rJxqZOdevatG7tS9uuHTz+uBs549Wnj5t3BNxoof/5rf6TlQVHHulmkT3+ePcxnJvrRvt4Z5/1+vLLshOvhTv/eWf81w6JiXGzn7Zq5Zc4d51v/9A34Ij3oHUtLolsTBiw4GIvtCrpHGjeDyQS+j8OSZ3dhYwzoVkvt1+4ExbeWfGDigsh6wvfcdpxEBUH+17tO7fkASgObXxiXJxvIbcTTig7FLYhSkyEgw/2Hc+cWTa4OPBA92H79ts/cMklZZ9RFZGREBvr9lu1covcffihC2CGDnVDW//9bzfdeSiOPBIefdTdc/XVbmI0cEHDG29AQoKb2fWZZ1xgeOmlLnAZPXoVX3zhZiH1mjbNt//EE74hmXPnuunvJ050Xwdwz/UOJ9640QVL9eGPP2DGjNbs3Bn6Pbm5vuHBERFuhtpyaTHs8gsuvEPXjdnLWXCxFyqMSIShP8FpO6CL32QHEZFubg6vZf9Xety9v4LtMPsyt3YKuF+KTT2/RbuOhehkt7/9d1j3fsh5GzfO/XKeNq3664yEm6OP9u2/+KJvtdbYWF/tQfPm0LRpQY2/W8R9wD/0kFvAbtEiVwtRE44+2s25sWqVCwa977vkErjoohW0ahU8uNi50wUg/v77X7dIn9ddd5W+1zv3SV1avdqtSHz33T0YOzb0+3791VfT0rVr6dq4MvI2QrFnTvropm45AWMaAQsu9lYRka6WIVDboZDq+TTUIhdgBMqaDh8dAEv9lpdvP9LXpNOkKXS91Hdt7g3w7Sj45jTY+D8qExcHUrQTZp4AbzeD9zvCpwfCwruhuOY/gGubf3DhP5lXr17BF3FrSFJTyy5S5+/oo33TzC9c6AKRV14pO4W7N/ABV2tx/vmumczriy8o1/r1runEu9Joee680y1EFzhxW0XpvWvJvP027NoV2n2VLkzmz79JJH6f0F5gzF7AgovGRsSti+K1arJrEAfPYms3w4xjIXeNL03aidArYCrHfa92/TkAcpbD6sluddhZJ4TWl2PJw5D5CRRsg50rYfNsWHi7W5Mlb3Olt4eTQw91/QoC9etX93mpa4mJvpV3wTWf/Pe/vuPk5LL3nHceNGtWOrj45puyM7WC65+x//4u7ZVXlp+P336DO+5w/VEeegi+/77ifP/2m1tjx2v3bvjqq4rv8Sqvv0VQuX6z5VpwYRoRCy4aozZDoEkLt5+7Gjb+CPnb4MvBsPg+wBNsxLSCv70JR35Qtjo3LrV07YVX/mb48QJfwBJM3mb47ZHg1zbMgM8Ogu0VzHm98X/wzQh4uznMr6E2gD0QEeGmF7/pptL9dQcOrL881SX/5o3774elS91+s2aun0Ugb5DQoYNrVgBXa/DDD6XT5eTAyJGwdas7fvbZ0qvY+vMPaMAFGhW57baynUi9U3j727LFrax7//2uOWjGjCoGF9bfwjRSDbzS1lRLZBNod6qvSWTVJCjeDdnf+tK0OQYOec0FEeXpcz807wP5W1278tzr3fnMT1yTSteLfWkLc12TR5OmrtaiYLs7n9wdjvgAVr0JC8e7czlL4euT4XhPz8jNc2HZC+4X9c6VsGWe77mL7oWMs9xomHoUHe1WfB0yBG6/3TUn+K88uzc75RS44YayzQqXXALDhrkPYO8H8okn+gIKcF8vb+fIV191H/Br17pOst98A0uW+NIWF7vVa598svR7srPdvf6++AK++87VKgWaPdut+BvIu4qt17ffutVj/cv1/PO+lY3Bai6MKU+d1FyISIyIvCAiq0Rkh4jME5EgFckgIueIyBwR2S4ia0XkQRGxIKimdRjl21/xMix93nfc624Y/GnFgQW4Zeg7ngX7XgH7XQf7XuO79su1sOQR1+Y8/xaYmgJTW8J3/4Df/Xr29bwTkrtCz9vdwnCRnn4i25fAgttosft/8PlB8OeTsPa90oGF1x9PVLHwteeoo9yH0tSppZe835ulpcFPP/kW8gLo3t0tvibiht5GRbm+FuPHl77Xv2nk5Zfh4YfdIm9XX+2+hoFefNEN9/X31FPBm1TuDDIYqrCwdPPKSSdBbKwb7bR0qS/QyctzC8sFBkzFxVDg6RbUsmWQhckC+ddcxFvNhWk86qpZJApYAxwJNAVuBSaLSEaQtPHA1UAr4CDgaOD6OsllY9J6EMR6JjIo2E5JU0j6SXDArSDV+K/R515o2sPtF+XC3OvgvX1c7UJRrutAumqS2wc3LLb9SN/97U6Ffn7NJUv+w/6bby/byVOiXMdUrxWv+GpCvNa8A4sfKHve1IoDDnB9HebNcx/Iixe7lV7B9UdZtswN+xwwoPR9gwdXPvXLmDFu7ghwQcSECb5r27aVrsm46y7fHCJffAEffFD6WQ895Gt+iYpy83b067el5Lq39uLhh11+wfUrGTu2bN579w5h2hr/mos4q7kwjUedBBequlNV71DVlaparKrTgBVA/yBpn1bVb1Q1X1XXAROBIJWbZo9EREG7kaXPRcZC/8eCpw9FZCwcOhni24WWvtddZYOYLhe7JhkAlEg8w/gSOsChb8GQr+GU9TDoE18gU5gDy1/2PWPRfa5PxrxxMPN4KAryZ62pNbGxZT9027cP/ld+8+Zw3HG+42OPhUcegVNPhX32gdNPd0NY//UvX5pHHnE1REce6QKY7GzfO266Cc4805d2+HA3h0dxsQt8/GtOxo93NSwDB/o6EH/yievX4b8U/b33utqRWbNcTYeX/yihctloEdNIiVbU8a62XiqSCqwC+qjqb5WkfQ/4TVXLTLkkImOAMQCpqan9J02aVO085eTkkJiYWHnCBiDUsjTNm0/fTVeXHK9IOp9VSWft8fsjNJ82uZ/SPucNYos2sCsyjRXJF5AbmU67nW/TLG8+G2MP5c+mVwX90y+m6C8O/Ot8otTNbJQX0YJ5rSawK6p0tXLazvfptu1RAHIj9+Gn1q+QvvMdum4v3Sj/V+xgFjevZm1MDWqM/8dCe1YkP/zQik6dcujcOfhsVkVFcM45A1m3rvxJJS677E9GjlzHhg0xXHZZPzZt8rVLxcQUoQr5+a7DRI8e25gwYR6RkcqyZUVceKGLFKKji2nWLJ/sbDcSqmvXHTz99C9ERmpJPj79tA3bt0dz6qnriImpeGrRwzKHEaWupu7bNu9RGNE0xK9K9VTn+zJ48OA5qjqg8pTGhK7OgwsRiQY+AZap6sWVpD0fuAsXhGysKO2AAQN09uzZ1c7XrFmzGDRoULXvDychl0WL4bOBsHmOa6IY+hNE1mBHgeJCN0w1sZOrKamK1VPh+3+wW1oQO3Q6NAsyDWJBjmt2KfBMVhCV6Goxgkk7sfy/HCNj3NwfbYe6zq5eqq6fx+4sSD0KkrpVXg++dRH8dBHEpMBBL0Csbx7oRvl/rAZ98AGcdhrk55c+37OnO3/TTb55RTIz4R//CD68ND7e1WJ4O5bOmjWLSy8dVKrzKLhv9Y8/7sGon4Lt8LYnmIiMhdNza336/+p8X0TEggtT4+q0o6SIRACvAfnA5ZWkHQ7cBwypLLAw1SQRMOhT2Pi9+/CsycACXECR3K1697YfAWlb+N8333FksMAC3PDYTufB74+6Y//AotXfXEDinQhs/bQyt5fy+2MQ0xLanwEdz4aEdm5IbabfEILEztDtMjfHR7APibxN8NUw2LnKHc86Ho6eAdFJoZTYVOLvf3edOVeudP/m5rqp1YM1t7Rt62b9vPNO16zinYo8MtLNxeE/YgXgnHNKT0cfGelG/ezRcGL/JpG49NpfV8iYMFJnwYWICPACkAqcoKrlTsUoIscBzwPDVHVhHWWxcYptBfuUs/pVfYtKQKWSOcJ7jIMNM2Gr3+QDKYfBkdMgKsH9gl//UWjvy9vkRqX8+aTrNKoBU0LmLHOjYDb/AgOfgZVvuOG8CR3dxGQLb/cFFuAmBvv6ZBj0sW/CsVCpujLFpUNsStXu3Ys1bRrC8E+PqCi4+24XYOTmumnJ4+PdLJ6Brr8eMjJcB9FevdzEXcHSVYkNQzWNWF3WXDwN7IeriSh3ol0ROQrXifMUVf2prjJnGqi4VDhhHhTlQ9EuN7IkpqXvr8TD33EjRwq2lP+MnJVung3/WUn9A4vWR7imo0JPf4CVr7tneke9bPoJVr8V/NkbZsLPl8HBL4ReprxN8MM5Lihq0hxO+BXiKxvz6Kcgx3VozV0DB78MrRrJbF7liIhwIz4q6ooQGQmjRpV/vVpsAi3TiNVJcCEiHYCLgTwgS3zVgxcD3wCLgR6quhq4DTdc9WO/dN+oatB5MYwBXF+JyCC1HJFNIOOMyu/vcx/89bUb1rr6bRdIxLdzH85tjnIjTmZf4Zt4zBtYBLP/zRCVBPNvcscrXoUBjwdPq+qCiNVTXDNS3D6w/AXfX735W2Dla9DjX8HvD+bPpyHLs8jJtyNccNLEryNhUZ4rY0wrNzInwm9WqPxt8Nt/IfsbNwNr+xGhv9eUZjUXphGrk+BCVVcBFTU4JvqlHVxBOmNqh0RA6iC39X8cti5ws49GeUYnRMbCwOdcvwtv0NCkuet/sf03V3OhxZB2AvT0DLFd8aqbDEwLXRNJoDXvwsI73LsqsnqqCy6KC9yEYbuz3YRMiZ3LBgfgghGv3LUw5yo45GXfuQW3ullSwQVQHc9yZcnb5Pqo5HuGZv71FcR9AymHhPIVrD2FO2FXFiR1rt98VJUNQzWNmM18aUyg6ERI+VvZ8yKw/zhIHewCin1OhibN3LXe97o+Ga0H+T7sUw51wQVA9vfAwW5fi2HeTbDkwfLz0KQ5FOzwBCY/w87V8MfjvqDAq9Xf4KgvfEHQlvmwNaCb0opXoN0pLr9Fu2Gp30q4uWvcJGfBaBF8/0/X7BQdZAWyulCw3a2Yu+MP6H4d9Hu48nvCRamaC2sWMY2LLVxmTFW1Ogg6neMLLAASM6DN0aVrEVr5BSgbPct0FhfAD+eWDiwi493U6QOegP1ucFOiHz/fjeDx+vPp4NOcb/we/ue3UNwKv1qLiGjf/k9jXBPLug+hYGvF5UvoCNGeZpSdK1yfkT2V/UPwqdsrs3KiCywAfvuPa84JN0W74X8XwmcHwx9Puv4/ALtsdk7TeFnNhTG1pZVfc8LG76HFNfC/i0o3W6T/vcx8GCXaj/D1nVh8v+98Ujdo0c9NpQ7u32a9XWCy6g1fuoNegnk3wq71sPsvWHQ/bFvku97jJmje260yK+KCkaYHQPvTYe278J2nr8rK190cIB39pr4MlaobYeMdLnzQ/0HnC0K/f9mLpY//dxG0GACJHauel4qowrppEJta9Q6wsy93C+sBbPofLPkPHHBL6Q7CVnNhGhkLLoypLcnd3NL2+ZshbxMt8n6CzFd81ztfBAc+Vf4EY/sMh5/HumYUf30fdE0c0c1g6TPu3PybIWs67Mp0x7Gt3eJ0Egnf/8Od+/2x0qNgulzoJjjrEGSYRIdRkPmpb1r1n8dCy4PcInNeRXmu30hsimsOkggXRK2e4uZMaXucZ7iu36J4P10MsW0g7XjY9DNs+BI2zPL1SYmMd/0TBj4DRJTtq1KwDb460a2E23Kg6yPjnXm1cBds/A5aHuyatsD11Vj+kvsaNmnm6acypOzXfOF4+PVuNwT5mO9CDzCW/p8vsPDaucLVZHhJpCuzMY2IBRfG1BaJcLUXnnk2um7zW7el7VAY+GzFEyvFtoaUw13HSq9mvV1tB7h1YLYv8VxX90Ht1eEf7gO0w+nw2yOu30Zxnu96yuEusKhI/wmQ/R3s+NNNUPbdGXDs977J1n4831dTEtsG4trClrm++xc/UPaZWgTfng4xLUr3SSixxQ3h/OokaH2k73SLAa5ZRQth22Jfp9rUo+CI913+vjzKfT2S94Njv3Mf6l8Ogu2/l35FyuFw6BslnSxjC7N8edVCWPZcaMHF5jmu1sKr1d9cX5z8zaXTxbUt2+nWmL2c9bkwpjb5dQyNK8r0ne92eWgzNrY7tfRxz9t990U2ccvUtz2u7H0dPWvESAT0fajs9U7nVP7u6CQ4dBJEeIb4bvkF5rkhsfEFK0o3wezOKh1YBOpwBiRkuP2i3HICCz+5a11/C6/e97pgSgI+pDfMgFnH+QILcP9+c5rr2xIYWIAbZvtJH1jngr6OO16AYr85xde84+s3UR4tds/3BmzNesJRn8Pfl0OfByGxiy9tqyCdg43Zy1lwYUxtCvbBEt8O2oY4bUu7Eb7ZPZv1dk0l/mJawuBP4Ngf3PopEdHQ8Rxo3s+XJvVISPdbzjMyDtqfFtr7W/QrHZz8/hhkTad9jt8igf4LwkXEQKfzoetYiG8PiJu99G8TYfCnrpnIq0kLl9dDXoW/r4ARG12wFDhqPb696yzb7VIYthgGPu+aRbyyv/MFFl4bvnT9Rrw6X+imdffmNW+Ta175+lRSd00vfW/+FtfE5G/dxzDzOFjhCXjWTIVtv7r9qAQ4bKr7t0lT6HEDnPS7m/p94POun4kxjYw1ixhTm1oe6P7a1iLfuc4XhV5NHp8OR34EWV9A10vKX9m11cEw6EMoLgr+7D4PuNlCC3Pc+6sytLTbFe7Ddt2H7viHc2i9a4Pv+tFfub/g8za5Ybre6coHqBsd453cLHlfOG62+2Bu3sc1e/iPaAFXU3PA7fDrnb5znc7zlTu5m9u6XOgCn1+u8aWTCBdErX0/IP9XwgBPk1TnC+G7f/hmz/QPQCKiXX7BzVuSfoLb350N357malwyP3Nfu4V3lf76+PdF8eYldbDbjGmErObCmNoUleA+SL0ksmqjJcDNENrnPkjoUHna8oKWpvu5D/Yj3oN+/6na+0XcX+AxLd3xrvVE4AmWWh8JrQ9zNQsdTi+9DopI2VlTEzvCftd7OlUGBBZePW+HtGFuP7pp+V+v7lfDgCdds01kHBwyEQ5/F9r7dVBNObR0zUvrw+H4eaXTgPu+HPSS73jNu26IKbgZS/1nZP12ROlai+7XBc+fMY2YBRfG1Db/ppH0v1dtnZCalLyvG2VS3uiUisSlwoHPlD3fY1zZc3tKIlwQdPi7MPR/boXa8nS7FIavhVPWuWneReDgl1wA0/ki94zAACe2FRw2ya07E5vqKce/IGO0G00CULgD1n8KeZvd5GX+vLUb4PrOBBtGbEwjZ8GFMbWty8UQ3ZRCiYOe4+s7N9XXfqQbheLVvI8b9VIbIqKg3XAXEFUmNsXNaOoVFedqKw56ruIVZdudAiev4sfWr0Ovf7vAxH9Y7u+PuunZC3PccXy70rUtVmthTLksuDCmtjXbH05Zz3dt3nOTVjVkA56ANseQG7mPayoJZcRLOIuMYXdUuq8c/s0lf31VutaizwPQ169Jqfu1FQcvxjRi1qHTmLoQFY9KkFVbG5qYFnDU5/w0axaDWg6o79zUvGY9XU3T0mdLn0/q5mYujYh0/UbyNvmG+xpjyrCaC2OM8RKBA592I3Sa9fKd73W3r7Ns+olunpDyRu4YY6zmwhhjShFxw1DTjoPML1w/izZHVX6fMaaEBRfGGBOMREBaLXVYNWYvVyf1eiISIyIviMgqEdkhIvNEpNwpCkXkGhHJEpHtIvKiiMTURT6NMcYYs+fqqtEwClgDHAk0BW4FJotIRmBCERkKjAOOBjoAnYA7A9MZY4wxJjzVSXChqjtV9Q5VXamqxao6DVgB9A+S/BzgBVVdpKpbgLuBc+sin8YYY4zZc6Kqdf9SkVRgFdBHVX8LuDYfuFdV3/IctwKygVaquikg7RhgDEBqamr/SZMmUV05OTkkJiZW+/5wYmUJT1aW8NTYyzJ48OA5qroXjis29anOO3SKSDQwEXglMLDwSAS2+R1795OAUsGFqj4HPAcwYMAAHTRoULXzNWvWLPbk/nBiZQlPVpbwZGUxpubV6UBtEYkAXgPygcvLSZYD+C/Z6N3fUYtZM8YYY0wNqbPgQkQEeAFIBUaoakE5SRcB/nMk9wY2BDaJGGOMMSY81VmfCxF5BugDDFHVnArSHQe8DBwFrAfeAX5S1QqXXxSRbFw/jupqBWzcg/vDiZUlPFlZwlNjL0sHVbVFUkyNqpPgQkQ6ACuBPKDQ79LFwDfAYqCHqq72pL8W+BcQB0wFLlHVvFrO4+y9pVOTlSU8WVnCk5XFmJpXJx06VXUVUNHyiaW6N6vqI8AjtZopY4wxxtQKW3nHGGOMMTXKgguf5+o7AzXIyhKerCzhycpiTA2rl0m0jDHGGLP3spoLY4wxxtQoCy6MMcYYU6MsuDDGGGNMjdqrgwsRuVxEZotInoi87Hf+YBH5QkQ2i0i2iLwtIm39rouIPCAimzzbA54ZRutVeeXxXDtaRH4TkVwRmemZW8R7LUZEXhSR7SKS5ZlHJKyISIaIfCwiWzx5fEJEojzX+ojIHE/Z5ohIn3rObqVE5AwRWSIiO0VkmYgc7jlf7vcpnIlIVxHZLSKv+50bLSKrPGV8T0Ra1GceK+P5OXjBk+cdIjJPRI73u94gvzcAItJCRN71fC9Wicjo+s6Tadz26uACN8Pnv4EXA843x/WqzgA64NYtecnv+hhgOG7q8V7ASbgJv+pb0PJ4Vo59B7gNaAHMBt7yS3IH0BVX1sHAjZ6ZUMPJU8BfQFvcTK5HApeKSBPgfeB13PftFeB9z/mwJCLHAA8A5+EW3DsCWB7C9ymcPQn87D0Qkf2BZ4GzcFP65+K+h+EsCliD+7/VFLgVmOwJbBvy9wbc9ycf9734J/C053tkTP1Q1b1+w30gv1zB9X7ADr/j74ExfscXAD/WdznKKw8uGPre7zgB2AV09xyvB471u343MKm+yxFQpiXACX7HD+E+vI4F1uEZ2eS5tho4rr7zXEFZvgcuCHK+wu9TuG7AGcBkXJD6uufcvcAbfmk64z7ckuo7v1Us2wJgREP93vjlNR/o5nfuNeD++s6bbY1329trLkJ1BG7BNK/9gfl+x/M958JVqfyq6k5gGbC/iDTH1QaEe3keBc4QkXgRSQeOBz7F5XOBqvqPmV5A+OUfABGJBAYAKSKyVETWepp44qjg+1Q/ua2ciCQDdwGBTWmBZVmG5wOu7nK3Z0QkFZffRTTA742fbkChqv7hdy4cf8ZNI9LogwsR6QXcDtzgdzoR2OZ3vA1IDId+F+UIzC+e4yR8U6sHliepDvJVFV/jfhluB9biqqXfo+KyhaNUIBoYCRyOa+Lpi6uCb2hlAVfL9YKqrg043xDLUkJEooGJwCuq+hsNuzyJuJ8bfw0l72Yv1aiDCxHpAnwCXKWq3/hdygGS/Y6TgZyAv57DSWB+8Rzv8FyDsuXZUQf5ComIROBqKd7BVfG2wvWveICKyxaOdnn+fVxVM1V1I26dnBNoYGXxdJwdAvw3yOUGVRZ/nv9vr+FqWi73nG6w5aFh593spRptcOHpCT4duFtVXwu4vAjXmdOrN6WbTcJNqfyKSAKuDXyRqm4BMgnv8rQA2gNPqGqeqm7CdbA9AZfPXgG1Rr0Ir/yX8Hy91wL+gah3v9zvU51lsGoG4To9rxaRLOB6YISI/ELZsnQCYoA/yj4mfHj+H72Aq2EaoaoFnksN7Xvj7w8gSkS6+p0Lt59x08js1cGFiESJSCwQCUSKSKznXDowA/dh9kyQW18FrhWRdBFJA64DXq6zjJejvPIA7wIHiMgIz/Xbcf0UfvPc+ipwq4g0F5HuwEWEQXm8PH/drwDGesrYDDgH17diFlAEXOkZSuj9S3NGfeQ1RC8BV4hIa0+fl2uAaVT+fQo3z+E+YPt4tmeAj4ChuCaFk0TkcM8H8V3AO6oa7n8tPw3sB5ykqrv8zje0700JT/+Qd4C7RCRBRA4FTsbVzhhTP+q7R2ltbrje7Rqw3QGM9+zn+G9+9wnwILDZsz2I32iFcCuP59oQ4DdctfwsIMPvvhjc8NXtwAbg2vouS5Cy9fHkewuwETc6IdVzrS8wx1O2X4C+9Z3fSsoSjRuWuRXIAiYAsZV9n8J9w2+0iOd4NG7kzk7ccOEW9Z3HSvLfwfMzszvgZ/+fe8H3pgWuj9JOz/dkdH3nybbGvdnCZcYYY4ypUXt1s4gxxhhj6p4FF8YYY4ypURZcGGOMMaZGWXBhjDHGmBplwYUxxhhjapQFF8YYY4ypURZcGGOMMaZGWXBhGgURWSkif3lmk/Seu1BEZpWTPkNEVERy/Lb5wdJWIQ/eZ0btyXOMMSbcWXBhGpNI4Koq3tNMVRM9W+/Kk9cecexn1hgT9uwXlWlMHgKu96xdUm0i0l1EvhCRzSLyu4ic7ndtmIjMFZHtIrJGRO7wu/Vrz79bPTUhh4jIHSLyut/9pWo3RGSWiNwjIt8BuUCnSt5/gogsFpEdIrJORK7fk7IaY0x1WHBhGpPZuDUjqv2B62lW+QJ4A2gNnAE8JSI9PEl2AmcDzYBhuMXYhnuuHeH511sb8kOIrz0LGAMkAdmVvP8F4GJVTQIOILwXeDPG7KUsuDCNze24FUtTQky/UUS2erbrgROBlar6kqoWqupcYCpwGoCqzlLVhaparKoLgDeBI/cwzy+r6iJVLQSOq+j9QAHQQ0SSVXWLqv6yh+82xpgqs+DCNCqq+itu+fNxId7SSlWbebaHcStrHuQXcGwF/gm0ARCRg0Rkpohki8g24BKg1R5me43ffoXvB0YAJwCrROQrETlkD99tjDFVZr3WTWM0Hrd0+3+qce8a4CtVPaac628ATwDHq+puEXkUX3ARbAninUC833GbIGn876vw/ar6M3CyiEQDl+OWrm9XTl6NMaZWWM2FaXRUdSnwFnBlNW6fBnQTkbNEJNqzHSgi+3muJwGbPYHFQGC0373ZQDHQye/cPOAIEWkvIk2Bm6r7fhFpIiL/FJGmqloAbPe8zxhj6pQFF6axugtIqDRVAFXdARyL60i5HsgCHgBiPEkuBe4SkR24/h2T/e7NBe4BvvM0aRysql/gAp0FwBxc8LAn7z8LWCki23FNMv+sahmNMWZPiWqwmlpjjDHGmOqxmgtjjDHG1CgLLowxxhhToyy4MMYYY0yNsuDCGGOMMTXKggtjjDHG1CgLLowxxhhToyy4MMYYY0yNsuDCGGOMMTXq/wF8IeEnnf1J4wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.plot(-rfe_prediction[\"n_features\"], rfe_prediction[\"mae_val\"], lw=3, color=\"blue\", label=\"RFE - Prediction\\nContribution\")\n",
    "ax.plot(-rfe_error[\"n_features\"], rfe_error[\"mae_val\"], lw=3, color=\"orange\", label=\"RFE - Error\\nContribution\")\n",
    "ax.set_title(\"Validation Set\", fontsize=12)\n",
    "ax.set_ylabel(\"Mean Absolute Error\", fontsize=12)\n",
    "ax.set_xlabel(\"N Features\", fontsize=12)\n",
    "ax.set_xticklabels([-int(x) for x in ax.get_xticks()])\n",
    "ax.grid()\n",
    "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=12)\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "\n",
    "fig.savefig(\"rfe_val_all.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.6 Plot RFE on Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_iter_show = len(rfe_error) - 9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.plot(-rfe_prediction.head(max_iter_show)[\"n_features\"], rfe_prediction.head(max_iter_show)[\"mae_trn\"], lw=3, color=\"blue\", label=\"RFE - Prediction\\nContribution\")\n",
    "ax.plot(-rfe_error.head(max_iter_show)[\"n_features\"], rfe_error.head(max_iter_show)[\"mae_trn\"], lw=3, color=\"orange\", label=\"RFE - Error\\nContribution\")\n",
    "ax.set_title(\"Training Set\")\n",
    "ax.set_ylabel(\"Mean Absolute Error\", fontsize=12)\n",
    "ax.set_xlabel(\"N Features\", fontsize=12)\n",
    "ax.set_xticklabels([-int(x) for x in ax.get_xticks()])\n",
    "ax.grid()\n",
    "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=12)\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "\n",
    "fig.savefig(\"rfe_trn.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.7 Plot RFE on Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "idxmin_prediction = rfe_prediction[\"mae_val\"].idxmin()\n",
    "idxmin_error = rfe_error[\"mae_val\"].idxmin()\n",
    "\n",
    "ax.plot(-rfe_prediction.head(max_iter_show)[\"n_features\"], rfe_prediction.head(max_iter_show)[\"mae_val\"], lw=3, color=\"blue\", label=\"RFE - Prediction\\nContribution\")\n",
    "ax.plot(-rfe_error.head(max_iter_show)[\"n_features\"], rfe_error.head(max_iter_show)[\"mae_val\"], lw=3, color=\"orange\", label=\"RFE - Error\\nContribution\")\n",
    "ax.scatter(-rfe_prediction.loc[idxmin_prediction,\"n_features\"], rfe_prediction.loc[idxmin_prediction,\"mae_val\"], s=200, marker=\"o\", facecolors=\"none\", edgecolors=\"red\", lw=3, zorder=100)\n",
    "ax.scatter(-rfe_error.loc[idxmin_error,\"n_features\"], rfe_error.loc[idxmin_error,\"mae_val\"], s=200, marker=\"o\", facecolors=\"none\", edgecolors=\"red\", lw=3, zorder=100)\n",
    "ax.set_title(\"Validation Set\", fontsize=12)\n",
    "ax.set_ylabel(\"Mean Absolute Error\", fontsize=12)\n",
    "ax.set_xlabel(\"N Features\", fontsize=12)\n",
    "ax.set_xticklabels([-int(x) for x in ax.get_xticks()])\n",
    "ax.grid()\n",
    "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=12)\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "\n",
    "fig.savefig(\"rfe_val.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>feature_drop</th>\n",
       "      <th>n_features</th>\n",
       "      <th>contrib</th>\n",
       "      <th>mae_trn</th>\n",
       "      <th>...</th>\n",
       "      <th>mae_tst</th>\n",
       "      <th>r2_trn</th>\n",
       "      <th>r2_val</th>\n",
       "      <th>r2_tst</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>Nikkei Ind_T-22</td>\n",
       "      <td>17.0</td>\n",
       "      <td>-0.034201</td>\n",
       "      <td>0.389733</td>\n",
       "      <td>...</td>\n",
       "      <td>1.940193</td>\n",
       "      <td>0.98585</td>\n",
       "      <td>0.642669</td>\n",
       "      <td>0.605735</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        feature_drop  n_features   contrib   mae_trn  ...   mae_tst   r2_trn  \\\n",
       "103  Nikkei Ind_T-22        17.0 -0.034201  0.389733  ...  1.940193  0.98585   \n",
       "\n",
       "       r2_val    r2_tst  \n",
       "103  0.642669  0.605735  \n",
       "\n",
       "[1 rows x 9 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfe_error.sort_values(\"mae_val\", ascending=True).head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>feature_drop</th>\n",
       "      <th>n_features</th>\n",
       "      <th>contrib</th>\n",
       "      <th>mae_trn</th>\n",
       "      <th>...</th>\n",
       "      <th>mae_tst</th>\n",
       "      <th>r2_trn</th>\n",
       "      <th>r2_val</th>\n",
       "      <th>r2_tst</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>Brent 2_T-14</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.254594</td>\n",
       "      <td>0.372378</td>\n",
       "      <td>...</td>\n",
       "      <td>2.04357</td>\n",
       "      <td>0.986542</td>\n",
       "      <td>0.594444</td>\n",
       "      <td>0.582419</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     feature_drop  n_features   contrib   mae_trn  ...  mae_tst    r2_trn  \\\n",
       "101  Brent 2_T-14        19.0  0.254594  0.372378  ...  2.04357  0.986542   \n",
       "\n",
       "       r2_val    r2_tst  \n",
       "101  0.594444  0.582419  \n",
       "\n",
       "[1 rows x 9 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfe_prediction.sort_values(\"mae_val\", ascending=True).head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 8.8 Plot RFE on Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "idxmin_prediction = rfe_prediction[\"mae_val\"].idxmin()\n",
    "idxmin_error = rfe_error[\"mae_val\"].idxmin()\n",
    "\n",
    "ax.plot(-rfe_prediction.head(max_iter_show)[\"n_features\"], rfe_prediction.head(max_iter_show)[\"mae_tst\"], lw=3, color=\"blue\", label=\"RFE - Prediction\\nContribution\")\n",
    "ax.plot(-rfe_error.head(max_iter_show)[\"n_features\"], rfe_error.head(max_iter_show)[\"mae_tst\"], lw=3, color=\"orange\", label=\"RFE - Error\\nContribution\")\n",
    "ax.scatter(-rfe_prediction.loc[idxmin_prediction,\"n_features\"], rfe_prediction.loc[idxmin_prediction,\"mae_tst\"], s=200, marker=\"o\", facecolors=\"none\", edgecolors=\"red\", lw=3, zorder=100)\n",
    "ax.scatter(-rfe_error.loc[idxmin_error,\"n_features\"], rfe_error.loc[idxmin_error,\"mae_tst\"], s=200, marker=\"o\", facecolors=\"none\", edgecolors=\"red\", lw=3, zorder=100)\n",
    "ax.set_title(\"Test Set\", fontsize=12)\n",
    "ax.set_ylabel(\"Mean Absolute Error\", fontsize=12)\n",
    "ax.set_xlabel(\"N Features\", fontsize=12)\n",
    "ax.set_xticklabels([-int(x) for x in ax.get_xticks()])\n",
    "ax.grid()\n",
    "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=12)\n",
    "ax.tick_params(axis=\"both\", which=\"major\", labelsize=12)\n",
    "\n",
    "fig.savefig(\"rfe_tst.png\", dpi=200, bbox_inches=\"tight\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dev",
   "language": "python",
   "name": "dev"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}