Predicting Facebook Comment Counts (Decision Tree vs Neural Net).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "bb462f48",
   "metadata": {},
   "source": [
    "This project is an attempt to compare the effectiveness of classical machine learning techniques and neural networks (using real-life data). The data we use is Facebook posts: we predict the amoount of comments a post will get in the next 10/24 hours based on factors such as total follower count and post category. In the end, we find that using a combination of the two was the most successful: a decision tree classifier to decide between 0 comments or at least 1 comment, and a neural network with embeddings to predict the amount of comments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "80f2e28f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.preprocessing import MinMaxScaler, MaxAbsScaler, StandardScaler, OrdinalEncoder, OneHotEncoder\n",
    "from sklearn.linear_model import LinearRegression, LogisticRegression\n",
    "from sklearn.compose import ColumnTransformer\n",
    "import numpy as np\n",
    "from numpy import *\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense,Dropout,Normalization,Embedding,Input,Flatten,Concatenate"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1d087a4",
   "metadata": {},
   "source": [
    "The source: https://archive.ics.uci.edu/ml/datasets/Facebook+Comment+Volume+Dataset had many different training and test data sets. We use only one of each for computational speed, since this project is more exploratory in nature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "19c2d740",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "train_data_v1 = pd.read_csv(\"FB_Dataset\\Training\\Features_Variant_1.csv\", header = None)\n",
    "train_data_v1.index = np.arange(1, len(train_data_v1) + 1)\n",
    "train_data_v1.columns = np.arange(1, len(train_data_v1.columns) + 1)\n",
    "train_data_v2 = pd.read_csv(\"FB_Dataset\\Training\\Features_Variant_2.csv\", header = None)\n",
    "train_data_v2.index = np.arange(1, len(train_data_v2) + 1)\n",
    "train_data_v2.columns = np.arange(1, len(train_data_v2.columns) + 1)\n",
    "train_data_v3 = pd.read_csv(\"FB_Dataset\\Training\\Features_Variant_3.csv\", header = None)\n",
    "train_data_v3.index = np.arange(1, len(train_data_v3) + 1)\n",
    "train_data_v3.columns = np.arange(1, len(train_data_v3.columns) + 1)\n",
    "train_data_v4 = pd.read_csv(\"FB_Dataset\\Training\\Features_Variant_4.csv\", header = None)\n",
    "train_data_v4.index = np.arange(1, len(train_data_v4) + 1)\n",
    "train_data_v4.columns = np.arange(1, len(train_data_v4.columns) + 1)\n",
    "train_data_v5 = pd.read_csv(\"FB_Dataset\\Training\\Features_Variant_5.csv\", header = None)\n",
    "train_data_v5.index = np.arange(1, len(train_data_v5) + 1)\n",
    "train_data_v5.columns = np.arange(1, len(train_data_v5.columns) + 1)\n",
    "test_data_c1 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_1.csv\", header = None)\n",
    "test_data_c1.index = np.arange(1, len(test_data_c1) + 1)\n",
    "test_data_c1.columns = np.arange(1, len(test_data_c1.columns) + 1)\n",
    "test_data_c2 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_2.csv\", header = None)\n",
    "test_data_c2.index = np.arange(1, len(test_data_c2) + 1)\n",
    "test_data_c2.columns = np.arange(1, len(test_data_c2.columns) + 1)\n",
    "test_data_c3 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_3.csv\", header = None)\n",
    "test_data_c3.index = np.arange(1, len(test_data_c3) + 1)\n",
    "test_data_c3.columns = np.arange(1, len(test_data_c3.columns) + 1)\n",
    "test_data_c4 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_4.csv\", header = None)\n",
    "test_data_c4.index = np.arange(1, len(test_data_c4) + 1)\n",
    "test_data_c4.columns = np.arange(1, len(test_data_c4.columns) + 1)\n",
    "test_data_c5 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_5.csv\", header = None)\n",
    "test_data_c5.index = np.arange(1, len(test_data_c5) + 1)\n",
    "test_data_c5.columns = np.arange(1, len(test_data_c5.columns) + 1)\n",
    "test_data_c6 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_6.csv\", header = None)\n",
    "test_data_c6.index = np.arange(1, len(test_data_c6) + 1)\n",
    "test_data_c6.columns = np.arange(1, len(test_data_c6.columns) + 1)\n",
    "test_data_c7 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_7.csv\", header = None)\n",
    "test_data_c7.index = np.arange(1, len(test_data_c7) + 1)\n",
    "test_data_c7.columns = np.arange(1, len(test_data_c7.columns) + 1)\n",
    "test_data_c8 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_8.csv\", header = None)\n",
    "test_data_c8.index = np.arange(1, len(test_data_c8) + 1)\n",
    "test_data_c8.columns = np.arange(1, len(test_data_c8.columns) + 1)\n",
    "test_data_c9 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_9.csv\", header = None)\n",
    "test_data_c9.index = np.arange(1, len(test_data_c9) + 1)\n",
    "test_data_c9.columns = np.arange(1, len(test_data_c9.columns) + 1)\n",
    "test_data_c10 = pd.read_csv(\"FB_Dataset\\Testing\\TestSet\\Test_Case_10.csv\", header = None)\n",
    "test_data_c10.index = np.arange(1, len(test_data_c10) + 1)\n",
    "test_data_c10.columns = np.arange(1, len(test_data_c10.columns) + 1)\n",
    "test_data_features = pd.read_csv(\"FB_Dataset\\Testing\\Features_TestSet.csv\", header = None)\n",
    "test_data_features.index = np.arange(1, len(test_data_features) + 1)\n",
    "test_data_features.columns = np.arange(1, len(test_data_features.columns) + 1)\n",
    "#print(train_data_v1)\n",
    "#print(test_data_c7)\n",
    "#print(test_data_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99dbcdad",
   "metadata": {},
   "source": [
    "The normalization and error functions we will use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fc64b9c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize(x):\n",
    "    return (x-x.min())/(x.max()-x.min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "60c2cb57",
   "metadata": {},
   "outputs": [],
   "source": [
    "def error(a, b):\n",
    "    return mean_squared_error(a, b, squared = False)\n",
    "    #return mean_absolute_error(a, b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "23b8c366",
   "metadata": {},
   "outputs": [],
   "source": [
    "def maketree(r, c):\n",
    "    return DecisionTreeRegressor(random_state = r, ccp_alpha = c)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2bad3ce3",
   "metadata": {},
   "source": [
    "The data set contains some examples with comments after 10 hours, and some with comments after 24 hours. We create a new variable, C/H, that records the amount of comments each hour for the first X hours."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "997a5760",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            1   2       3   4    5       6           7      8           9   \\\n",
      "1       634995   0     463   1  0.0   806.0   11.291045    1.0   70.495138   \n",
      "2       634995   0     463   1  0.0   806.0   11.291045    1.0   70.495138   \n",
      "3       634995   0     463   1  0.0   806.0   11.291045    1.0   70.495138   \n",
      "4       634995   0     463   1  0.0   806.0   11.291045    1.0   70.495138   \n",
      "5       634995   0     463   1  0.0   806.0   11.291045    1.0   70.495138   \n",
      "...        ...  ..     ...  ..  ...     ...         ...    ...         ...   \n",
      "40945  7170111  70  497000   9  0.0  1881.0  497.200000  269.0  502.318385   \n",
      "40946  7170111  70  497000   9  0.0  1881.0  497.200000  269.0  502.318385   \n",
      "40947  7170111  70  497000   9  0.0  1881.0  497.200000  269.0  502.318385   \n",
      "40948  7170111  70  497000   9  0.0  1881.0  497.200000  269.0  502.318385   \n",
      "40949  7170111  70  497000   9  0.0  1881.0  497.200000  269.0  502.318385   \n",
      "\n",
      "        10  ...  44  45  46  47  48  49  50  51  52  53  \n",
      "1      0.0  ...   0   0   0   0   0   0   0   0   0   1  \n",
      "2      0.0  ...   1   0   0   0   0   0   0   0   1   0  \n",
      "3      0.0  ...   0   1   0   0   0   0   0   0   0   1  \n",
      "4      0.0  ...   0   1   0   0   1   0   0   0   0   0  \n",
      "5      0.0  ...   0   0   0   0   0   0   1   0   0   0  \n",
      "...    ...  ...  ..  ..  ..  ..  ..  ..  ..  ..  ..  ..  \n",
      "40945  0.0  ...   0   0   0   0   0   0   0   0   1   0  \n",
      "40946  0.0  ...   0   0   0   0   0   0   0   0   1   0  \n",
      "40947  0.0  ...   0   0   0   0   0   0   0   1   0   0  \n",
      "40948  0.0  ...   0   0   0   0   0   0   0   0   1   0  \n",
      "40949  0.0  ...   0   0   0   0   0   0   0   0   1   0  \n",
      "\n",
      "[40949 rows x 52 columns]\n",
      "1        0.000000\n",
      "2        0.000000\n",
      "3        0.000000\n",
      "4        0.000000\n",
      "5        0.000000\n",
      "           ...   \n",
      "40945    0.500000\n",
      "40946    0.666667\n",
      "40947    3.000000\n",
      "40948    3.111111\n",
      "40949    1.222222\n",
      "Name: C/H, Length: 40949, dtype: float64\n",
      "1        0\n",
      "2        0\n",
      "3        0\n",
      "4        0\n",
      "5        0\n",
      "        ..\n",
      "40945    1\n",
      "40946    1\n",
      "40947    1\n",
      "40948    1\n",
      "40949    1\n",
      "Name: C/H, Length: 40949, dtype: int32\n",
      "          1       2       3   4    5      6          7     8           9   \\\n",
      "1    1340759   33935   49392   9  0.0  768.0  77.200000  36.5  132.861432   \n",
      "2     161455       0    5978   4  3.0   75.0  22.714286  15.0   23.389340   \n",
      "3      92385      88   34785  70  0.0    0.0   0.000000   0.0    0.000000   \n",
      "4     395272       0    3406  17  0.0   22.0   4.000000   1.0    6.599663   \n",
      "5    4349418       0  614043   9  0.0  958.0  65.974194  24.0  126.960748   \n",
      "..       ...     ...     ...  ..  ...    ...        ...   ...         ...   \n",
      "96     83580       0    6366  20  0.0   17.0   3.000000   0.0    6.271629   \n",
      "97   1045111       0   34768  16  0.0  510.0  17.890909   2.0   81.850123   \n",
      "98   2289651  185932  119581   9  0.0  328.0  56.355932  32.0   68.962226   \n",
      "99   1957593       0  105153  26  0.0  332.0  16.112676   4.0   46.689429   \n",
      "100   410589       0   35208   9  0.0   69.0   6.189655   1.0   12.211437   \n",
      "\n",
      "      10  ...  44  45  46  47  48  49  50  51  52  53  \n",
      "1    0.0  ...   1   0   0   0   0   0   0   1   0   0  \n",
      "2    3.0  ...   0   1   0   0   0   0   0   0   0   1  \n",
      "3    0.0  ...   1   0   0   0   0   0   0   0   0   1  \n",
      "4    0.0  ...   0   0   1   1   0   0   0   0   0   0  \n",
      "5    0.0  ...   0   0   1   1   0   0   0   0   0   0  \n",
      "..   ...  ...  ..  ..  ..  ..  ..  ..  ..  ..  ..  ..  \n",
      "96   0.0  ...   0   0   1   0   0   0   0   0   0   1  \n",
      "97   0.0  ...   0   1   0   0   0   0   0   0   0   1  \n",
      "98   0.0  ...   0   0   0   0   1   0   0   0   0   0  \n",
      "99   0.0  ...   0   1   0   0   0   0   0   0   1   0  \n",
      "100  0.0  ...   0   1   0   1   0   0   0   0   0   0  \n",
      "\n",
      "[100 rows x 52 columns]\n",
      "1      1.590909\n",
      "2      0.083333\n",
      "3      0.000000\n",
      "4      0.000000\n",
      "5      0.000000\n",
      "         ...   \n",
      "96     0.000000\n",
      "97     0.000000\n",
      "98     3.500000\n",
      "99     0.458333\n",
      "100    0.000000\n",
      "Name: C/H, Length: 100, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "train_data_v1[\"C/H\"] = train_data_v1[54]/train_data_v1[39]\n",
    "t1X = train_data_v1.copy()\n",
    "t1X = t1X.drop(columns = 54)\n",
    "t1X = t1X.drop(columns = 39)\n",
    "t1y = t1X['C/H']\n",
    "t1X = t1X.drop(columns = 'C/H')\n",
    "t1y1 = (t1y > 0).astype(int)\n",
    "\n",
    "\n",
    "test_data_c1[\"C/H\"] = test_data_c1[54]/test_data_c1[39]\n",
    "c1X = test_data_c1.copy()\n",
    "c1X = c1X.drop(columns = 54)\n",
    "c1X = c1X.drop(columns = 39)\n",
    "c1y = c1X['C/H']\n",
    "c1X = c1X.drop(columns = 'C/H')\n",
    "c1y1 = (c1y > 0).astype(int)\n",
    "\n",
    "print(t1X)\n",
    "print(t1y)\n",
    "print(t1y1)\n",
    "print(c1X)\n",
    "print(c1y)\n",
    "\n",
    "#print(len(t1X))\n",
    "#print(sum(np.isfinite(t1X)))\n",
    "#print(len(t1y))\n",
    "#print(sum(np.isfinite(t1y)))\n",
    "#print(len(c1X))\n",
    "#print(sum(np.isfinite(c1X)))\n",
    "#print(len(c1y))\n",
    "#print(sum(np.isfinite(c1y)))\n",
    "#print(np.average(c1y))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "addae0c7",
   "metadata": {},
   "source": [
    "Trying to classify no comments vs comments, using a decision tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "767ee713",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1      1\n",
      "2      1\n",
      "3      0\n",
      "4      0\n",
      "5      0\n",
      "      ..\n",
      "96     0\n",
      "97     0\n",
      "98     1\n",
      "99     1\n",
      "100    0\n",
      "Name: C/H, Length: 100, dtype: int32\n",
      "[1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 1 0 1 1 0 1 1\n",
      " 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 0 0 0 0 1 1 1 0 1 0 1 1 1\n",
      " 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 0 1 1 0]\n",
      "0.4277088614985673 0.41231056256176607\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "treebool = DecisionTreeClassifier(random_state = 42, ccp_alpha = 0.01)\n",
    "treebool.fit(t1X, t1y1)\n",
    "t1y1_pred = treebool.predict(t1X)\n",
    "c1y1_pred = treebool.predict(c1X)\n",
    "print(c1y1)\n",
    "print(c1y1_pred)\n",
    "print(error(t1y1, t1y1_pred), error(c1y1, c1y1_pred))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = c1y1, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "plt.show()\n",
    "\n",
    "#try setting different threshholds (not 0 but like 1 or 10 or whatever)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07edd2f1",
   "metadata": {},
   "source": [
    "Based on the results here, it seems like this decision tree is relatively successful. Given that this serves as an initial \"weeder\" in the model, these results are more than adequate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "da961bfd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13       0.125000\n",
       "22       0.208333\n",
       "27       0.083333\n",
       "32       0.625000\n",
       "39       0.083333\n",
       "           ...   \n",
       "40945    0.500000\n",
       "40946    0.666667\n",
       "40947    3.000000\n",
       "40948    3.111111\n",
       "40949    1.222222\n",
       "Name: C/H, Length: 18370, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t1X2 = t1X.copy()\n",
    "t1X2['com'] = t1y1\n",
    "t1X2 = t1X2[t1X2[\"com\"] == 1]\n",
    "t1y2 = t1y.copy()\n",
    "t1y2 = t1y2[t1y > 0]\n",
    "t1X2 = t1X2.drop(columns = 'com')\n",
    "t1y2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3007e04e",
   "metadata": {},
   "source": [
    "# Classical Machine Learning Techniques"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba13dd39",
   "metadata": {},
   "source": [
    "Now, we attempt to run a REGRESSION decision tree on those the posts that were predicted to have comments\n",
    "The graph created below is using the training data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6c436e58",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7556871793956498\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "treecom = maketree(42, 0.005)\n",
    "treecom.fit(t1X2, t1y2)\n",
    "t1y2_pred = treecom.predict(t1X2)\n",
    "print(error(t1y2, t1y2_pred))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = t1y2, x = np.arange(1, len(t1y2)+1), label = \"actual\")\n",
    "sns.lineplot(y = t1y2_pred, x = np.arange(1, len(t1y2)+1), label = \"predict\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9eddb41c",
   "metadata": {},
   "source": [
    "Running it on test data..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b9fe98de",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 2 28  1  2 24  2  3  1  3 45  1  5  1  2  9  6  2  2  9  4 16  4]\n",
      "11.250661456859085\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "c1y1_pred = treebool.predict(c1X)\n",
    "c1y1_pred[c1y1_pred > 0] = treecom.predict(c1X[c1y1_pred > 0])\n",
    "print(c1y1_pred[c1y1_pred > 0])\n",
    "print(error(c1y, c1y1_pred))\n",
    "#print(error(np.zeros(len(c1y)), c1y))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = c1y, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa462c2d",
   "metadata": {},
   "source": [
    "Next, we try another classical machine learning technique: linear regression."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "44f728c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.7733672393837498\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lr = LinearRegression().fit(t1X2, t1y2)\n",
    "t1y2_pred = lr.predict(t1X2)\n",
    "print(error(t1y2, t1y2_pred))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = t1y2, x = np.arange(1, len(t1y2)+1), label = \"actual\")\n",
    "sns.lineplot(y = t1y2_pred, x = np.arange(1, len(t1y2)+1), label = \"predict\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "09bf79cc",
   "metadata": {},
   "source": [
    "Even for training data, it is clearly much less accurate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "af652824",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14.107385614553609\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "c1y1_pred = treecom.predict(c1X)\n",
    "c1y1_pred[c1y1_pred > 0] = lr.predict(c1X[c1y1_pred > 0])\n",
    "#print(c1y1_pred[c1y1_pred > 0])\n",
    "print(error(c1y, c1y1_pred))\n",
    "#print(error(np.zeros(len(c1y)), c1y))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = c1y, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0372057",
   "metadata": {},
   "source": [
    "The test results are very unimpressive, although the model does seem to have the right \"idea\" for some of the points."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9357ad17",
   "metadata": {},
   "source": [
    "# Neural Network"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2eebc6b9",
   "metadata": {},
   "source": [
    "Now, we use a neural network (with ReLu as the activation function) to classify the posts that we think will get at least 1 comment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6a2c1141",
   "metadata": {},
   "outputs": [],
   "source": [
    "def nn(X_train, X_test, y_train, y_test):\n",
    "    rows, cols = X_train.shape\n",
    "    adam = tf.keras.optimizers.Adam(learning_rate=0.05)\n",
    "    BATCH_SIZE = 256\n",
    "    net = Sequential()\n",
    "    net.add(Dense(20, input_dim = cols, activation='relu'))\n",
    "    net.add(Dropout(0.2))\n",
    "    net.add(Dense(30,  activation='relu'))\n",
    "    net.add(Dropout(0.2))\n",
    "    net.add(Dense(40,  activation='relu'))\n",
    "    net.add(Dropout(0.2))\n",
    "    net.add(Dense(20,  activation='relu'))\n",
    "    net.add(Dense(1))\n",
    "    #net.compile(loss='mean_absolute_error', optimizer='adam')\n",
    "    net.compile(loss='mean_squared_error', optimizer='adam')\n",
    "    net.fit(X_train, y_train, epochs = 100, batch_size = BATCH_SIZE)\n",
    "    y_train_pred = net.predict(X_train).flatten()\n",
    "    y_test_pred = net.predict(X_test).flatten()\n",
    "    \n",
    "    print(\"train error  \", \"  test error\")\n",
    "    print(error(y_train, y_train_pred), error(y_test, y_test_pred))\n",
    "    print(y_train.shape)\n",
    "    print(y_train_pred.shape)\n",
    "    \n",
    "    trainl = len(y_train)\n",
    "    testl = len(y_test)\n",
    "    sns.set()\n",
    "    plot1 = plt.figure(1)\n",
    "    sns.lineplot(y = y_train, x = np.arange(1, trainl+1), label = \"Train actual\")\n",
    "    sns.lineplot(y = y_train_pred, x = np.arange(1, trainl+1), label = \"Train predict\")\n",
    "    plot2 = plt.figure(2)\n",
    "    sns.lineplot(y = y_test, x = np.arange(1, testl+1), label = \"Test actual\")\n",
    "    sns.lineplot(y = y_test_pred, x = np.arange(1, testl+1), label = \"Test predict\")\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "    return y_test_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2c96360f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100, 52)\n",
      "(100,)\n",
      "Epoch 1/100\n",
      "72/72 [==============================] - 1s 4ms/step - loss: 5.2888\n",
      "Epoch 2/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 4.6881\n",
      "Epoch 3/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 4.4267\n",
      "Epoch 4/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 4.2287\n",
      "Epoch 5/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 4.0190\n",
      "Epoch 6/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 3.7190\n",
      "Epoch 7/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 3.5236\n",
      "Epoch 8/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 3.2681\n",
      "Epoch 9/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 3.1607\n",
      "Epoch 10/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 3.0612\n",
      "Epoch 11/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.9952\n",
      "Epoch 12/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.9922\n",
      "Epoch 13/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.8354\n",
      "Epoch 14/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.7990\n",
      "Epoch 15/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.6977\n",
      "Epoch 16/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.6849\n",
      "Epoch 17/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.6370\n",
      "Epoch 18/100\n",
      "72/72 [==============================] - 1s 7ms/step - loss: 2.5839\n",
      "Epoch 19/100\n",
      "72/72 [==============================] - 1s 10ms/step - loss: 2.5636\n",
      "Epoch 20/100\n",
      "72/72 [==============================] - 1s 8ms/step - loss: 2.5338\n",
      "Epoch 21/100\n",
      "72/72 [==============================] - 1s 10ms/step - loss: 2.5084\n",
      "Epoch 22/100\n",
      "72/72 [==============================] - 1s 8ms/step - loss: 2.4277\n",
      "Epoch 23/100\n",
      "72/72 [==============================] - 0s 7ms/step - loss: 2.4976\n",
      "Epoch 24/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.3718\n",
      "Epoch 25/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.3127\n",
      "Epoch 26/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.4614\n",
      "Epoch 27/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.3289\n",
      "Epoch 28/100\n",
      "72/72 [==============================] - 0s 7ms/step - loss: 2.2326\n",
      "Epoch 29/100\n",
      "72/72 [==============================] - 0s 7ms/step - loss: 2.1871\n",
      "Epoch 30/100\n",
      "72/72 [==============================] - 0s 7ms/step - loss: 2.3236\n",
      "Epoch 31/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 2.2296\n",
      "Epoch 32/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.1604\n",
      "Epoch 33/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.3388\n",
      "Epoch 34/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.1637\n",
      "Epoch 35/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.1798\n",
      "Epoch 36/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.1125\n",
      "Epoch 37/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.2020\n",
      "Epoch 38/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.1526\n",
      "Epoch 39/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.1831\n",
      "Epoch 40/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0678\n",
      "Epoch 41/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0191\n",
      "Epoch 42/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0759\n",
      "Epoch 43/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.1394\n",
      "Epoch 44/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0673\n",
      "Epoch 45/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0196\n",
      "Epoch 46/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9683\n",
      "Epoch 47/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.0429\n",
      "Epoch 48/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.9866\n",
      "Epoch 49/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.1081\n",
      "Epoch 50/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.9499\n",
      "Epoch 51/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.9163\n",
      "Epoch 52/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.9680\n",
      "Epoch 53/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 2.0563\n",
      "Epoch 54/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9471\n",
      "Epoch 55/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9963\n",
      "Epoch 56/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9226\n",
      "Epoch 57/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9277\n",
      "Epoch 58/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9150\n",
      "Epoch 59/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9352\n",
      "Epoch 60/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 2.0293\n",
      "Epoch 61/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9942\n",
      "Epoch 62/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8650\n",
      "Epoch 63/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9054\n",
      "Epoch 64/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8105\n",
      "Epoch 65/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8608\n",
      "Epoch 66/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8314\n",
      "Epoch 67/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9176\n",
      "Epoch 68/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8630\n",
      "Epoch 69/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8713\n",
      "Epoch 70/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7389\n",
      "Epoch 71/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8270\n",
      "Epoch 72/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9194\n",
      "Epoch 73/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7905\n",
      "Epoch 74/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7291\n",
      "Epoch 75/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7834\n",
      "Epoch 76/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7528\n",
      "Epoch 77/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.9046\n",
      "Epoch 78/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8771\n",
      "Epoch 79/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8321\n",
      "Epoch 80/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.9606\n",
      "Epoch 81/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.7085\n",
      "Epoch 82/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 1.7124\n",
      "Epoch 83/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 1.8001\n",
      "Epoch 84/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 1.7138\n",
      "Epoch 85/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 1.7244\n",
      "Epoch 86/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.8030\n",
      "Epoch 87/100\n",
      "72/72 [==============================] - 0s 6ms/step - loss: 1.7322\n",
      "Epoch 88/100\n",
      "72/72 [==============================] - 0s 5ms/step - loss: 1.7757\n",
      "Epoch 89/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7390\n",
      "Epoch 90/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8393\n",
      "Epoch 91/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7732\n",
      "Epoch 92/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7765\n",
      "Epoch 93/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.6451\n",
      "Epoch 94/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7557\n",
      "Epoch 95/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7491\n",
      "Epoch 96/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7655\n",
      "Epoch 97/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.8740\n",
      "Epoch 98/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.7514\n",
      "Epoch 99/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.6835\n",
      "Epoch 100/100\n",
      "72/72 [==============================] - 0s 4ms/step - loss: 1.6961\n",
      "train error     test error\n",
      "1.1864372615186933 12.11449471450528\n",
      "(18370,)\n",
      "(18370,)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "t = [(\"norm1\", MinMaxScaler(), slice(0,3)), (\"cat1\", OrdinalEncoder(), slice(3,4)), (\"norm2\", MinMaxScaler(), slice(4,37))]\n",
    "transformer = ColumnTransformer(transformers = t)\n",
    "t1X3 = transformer.fit_transform(t1X2.iloc[:, 0:37])\n",
    "t1X3 = np.concatenate((t1X3, t1X2.iloc[:, 37:]), axis = 1)\n",
    "#print(t1X3)\n",
    "#print(t1X3.shape)\n",
    "\n",
    "c1y1_pred = treecom.predict(c1X)\n",
    "c1X3 = c1X.copy()\n",
    "c1X3 = c1X3[c1y1_pred > 0]\n",
    "c1X3 = np.concatenate((transformer.fit_transform(c1X3.iloc[:, 0:37]), c1X3.iloc[:, 37:]), axis = 1)\n",
    "c1y3 = c1y[c1y1_pred > 0]\n",
    "print(c1X3.shape)\n",
    "print(c1y3.shape)\n",
    "\n",
    "c1y1_pred[c1y1_pred > 0] = nn(t1X3, c1X3, t1y2, c1y3)\n",
    "#print(error(c1y, c1y1_pred))\n",
    "#sns.set()\n",
    "#plt.figure()\n",
    "#sns.lineplot(y = c1y, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "#sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12d31392",
   "metadata": {},
   "source": [
    "Results are not bad, slightly worse than the decision tree regressor with this sample. Like the decision tree, it seems to predict too \"vanilla\" (predictions are too similar, not enough highs and lows)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "678166a4",
   "metadata": {},
   "source": [
    "# Neural Network with Embedding"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2645bfd",
   "metadata": {},
   "source": [
    "Many of the variables in the data set are categorical (with many categories). Previously, we used simple regression on them, assuming that there is some \"ordering\" among the categories. This is not ideal.\n",
    "    \n",
    "    \n",
    "Pure one-hot encoding would result in greatly increased computations, which are not ideal either. So, we use embeddings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "760f80b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def embedding(num_data1, cat_data, num_data2, target, nd1, ohe, nd2, y_test):\n",
    "    no_of_unique_cat  = len(np.unique(cat_data))\n",
    "    oh = OneHotEncoder()\n",
    "    one_hot_encoded_cat_data = oh.fit_transform(cat_data).toarray()\n",
    "    print(one_hot_encoded_cat_data)\n",
    "    embedding_size = min(np.ceil((no_of_unique_cat)/2), 50 )\n",
    "    embedding_size = int(embedding_size)\n",
    "    \n",
    "    \n",
    "    inp_cat_data = Input(shape=(no_of_unique_cat,))\n",
    "    inp_num_data1 = Input(shape=(num_data1.shape[1],))\n",
    "    inp_num_data2 = Input(shape=(num_data2.shape[1],))\n",
    "    \n",
    "    BATCH_SIZE = 256\n",
    "    \n",
    "    emb = Embedding(input_dim=no_of_unique_cat, output_dim=embedding_size)(inp_cat_data) \n",
    "    flatten = Flatten()(emb)\n",
    "    conc = Concatenate()([inp_num_data1, flatten, inp_num_data2])\n",
    "    dense1 = Dense(30, activation=tf.nn.relu, )(conc)\n",
    "    dropout1 = Dropout(0.2,) (dense1)\n",
    "    dense2 = Dense(25, activation=tf.nn.relu, )(dropout1)\n",
    "    dropout2 = Dropout(0.2,) (dense2)\n",
    "    dense3 = Dense(20, activation=tf.nn.relu, )(dropout2)\n",
    "    dropout3 = Dropout(0.2,) (dense3)\n",
    "    dense4 = Dense(15, activation=tf.nn.relu, )(dropout3)\n",
    "    out = Dense(1, activation=None)(dense4)\n",
    "    model = tf.keras.Model(inputs=[inp_num_data1, inp_cat_data, inp_num_data2], outputs=out)\n",
    "    \n",
    "    #model = Sequential()\n",
    "\n",
    "    #model.compile(optimizer='adam', loss='mean_absolute_error')\n",
    "    model.compile(optimizer='adam', loss='mean_squared_error')\n",
    "    \n",
    "    #print(num_data1.shape)\n",
    "    #print(one_hot_encoded_cat_data.shape)\n",
    "    #print(num_data2.shape)\n",
    "    \n",
    "    #print(cat_data)\n",
    "    #print(ohe)\n",
    "    one_hot_encoded_cat_data2 = oh.transform(ohe).toarray()\n",
    "\n",
    "    #print(nd1.shape)\n",
    "    #print(one_hot_encoded_cat_data2.shape)\n",
    "    #print(nd2.shape)\n",
    "    \n",
    "    \n",
    "    model.fit([num_data1, one_hot_encoded_cat_data, num_data2], target, epochs = 100, batch_size=BATCH_SIZE)\n",
    "    \n",
    "    y_train_pred = model.predict([num_data1, one_hot_encoded_cat_data, num_data2]).flatten()\n",
    "    y_test_pred = model.predict([nd1, one_hot_encoded_cat_data2, nd2]).flatten()\n",
    "    \n",
    "    print(\"train error  \", \"  test error\")\n",
    "    print(error(target, y_train_pred), error(y_test, y_test_pred))\n",
    "    \n",
    "    trainl = len(target)\n",
    "    testl = len(y_test)\n",
    "    sns.set()\n",
    "    plot1 = plt.figure(1)\n",
    "    sns.lineplot(y = target, x = np.arange(1, trainl+1), label = \"Train actual\")\n",
    "    sns.lineplot(y = y_train_pred, x = np.arange(1, trainl+1), label = \"Train predict\")\n",
    "    plot2 = plt.figure(2)\n",
    "    sns.lineplot(y = y_test, x = np.arange(1, testl+1), label = \"Test actual\")\n",
    "    sns.lineplot(y = y_test_pred, x = np.arange(1, testl+1), label = \"Test predict\")\n",
    "    plt.show()\n",
    "    return y_test_pred\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "612b38aa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1. 0. 0. ... 0. 0. 0.]\n",
      " [1. 0. 0. ... 0. 0. 0.]\n",
      " [1. 0. 0. ... 0. 0. 0.]\n",
      " ...\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " [0. 0. 0. ... 0. 0. 0.]]\n",
      "Epoch 1/100\n",
      "72/72 [==============================] - 2s 14ms/step - loss: 4.7710\n",
      "Epoch 2/100\n",
      "72/72 [==============================] - 1s 15ms/step - loss: 4.2865\n",
      "Epoch 3/100\n",
      "72/72 [==============================] - 1s 17ms/step - loss: 3.7083\n",
      "Epoch 4/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 3.1509\n",
      "Epoch 5/100\n",
      "72/72 [==============================] - 1s 17ms/step - loss: 2.9279\n",
      "Epoch 6/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 2.7726\n",
      "Epoch 7/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.6636\n",
      "Epoch 8/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.6901\n",
      "Epoch 9/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 2.5579\n",
      "Epoch 10/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 2.4701\n",
      "Epoch 11/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 2.3961\n",
      "Epoch 12/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.3176\n",
      "Epoch 13/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.2997\n",
      "Epoch 14/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 2.3263\n",
      "Epoch 15/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 2.2708\n",
      "Epoch 16/100\n",
      "72/72 [==============================] - 1s 17ms/step - loss: 2.1526\n",
      "Epoch 17/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.1810\n",
      "Epoch 18/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 2.0099\n",
      "Epoch 19/100\n",
      "72/72 [==============================] - 1s 18ms/step - loss: 2.1365\n",
      "Epoch 20/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 2.0455\n",
      "Epoch 21/100\n",
      "72/72 [==============================] - 2s 27ms/step - loss: 2.0282\n",
      "Epoch 22/100\n",
      "72/72 [==============================] - 2s 30ms/step - loss: 1.9850\n",
      "Epoch 23/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 2.0421\n",
      "Epoch 24/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.9705\n",
      "Epoch 25/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.9262\n",
      "Epoch 26/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.9777\n",
      "Epoch 27/100\n",
      "72/72 [==============================] - 2s 29ms/step - loss: 2.0140\n",
      "Epoch 28/100\n",
      "72/72 [==============================] - 3s 40ms/step - loss: 1.8911\n",
      "Epoch 29/100\n",
      "72/72 [==============================] - 3s 45ms/step - loss: 1.8802\n",
      "Epoch 30/100\n",
      "72/72 [==============================] - 3s 43ms/step - loss: 1.9931\n",
      "Epoch 31/100\n",
      "72/72 [==============================] - 3s 38ms/step - loss: 1.9123\n",
      "Epoch 32/100\n",
      "72/72 [==============================] - 2s 32ms/step - loss: 1.7850\n",
      "Epoch 33/100\n",
      "72/72 [==============================] - 2s 31ms/step - loss: 1.8937\n",
      "Epoch 34/100\n",
      "72/72 [==============================] - 2s 30ms/step - loss: 1.8510\n",
      "Epoch 35/100\n",
      "72/72 [==============================] - 2s 33ms/step - loss: 1.8022\n",
      "Epoch 36/100\n",
      "72/72 [==============================] - 2s 29ms/step - loss: 1.7837\n",
      "Epoch 37/100\n",
      "72/72 [==============================] - 2s 28ms/step - loss: 1.8113\n",
      "Epoch 38/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.7441\n",
      "Epoch 39/100\n",
      "72/72 [==============================] - 2s 27ms/step - loss: 1.8346\n",
      "Epoch 40/100\n",
      "72/72 [==============================] - 2s 28ms/step - loss: 1.7208\n",
      "Epoch 41/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.8022\n",
      "Epoch 42/100\n",
      "72/72 [==============================] - 2s 27ms/step - loss: 1.7516\n",
      "Epoch 43/100\n",
      "72/72 [==============================] - 2s 27ms/step - loss: 1.7127\n",
      "Epoch 44/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.8278\n",
      "Epoch 45/100\n",
      "72/72 [==============================] - 2s 32ms/step - loss: 1.8047\n",
      "Epoch 46/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.8106\n",
      "Epoch 47/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.6894\n",
      "Epoch 48/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.7169\n",
      "Epoch 49/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.7701\n",
      "Epoch 50/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.7235\n",
      "Epoch 51/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.8470\n",
      "Epoch 52/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.8955\n",
      "Epoch 53/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.6240\n",
      "Epoch 54/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.6577\n",
      "Epoch 55/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.6662\n",
      "Epoch 56/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.5439\n",
      "Epoch 57/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.6148\n",
      "Epoch 58/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.6805\n",
      "Epoch 59/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.7842\n",
      "Epoch 60/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.8110\n",
      "Epoch 61/100\n",
      "72/72 [==============================] - 2s 21ms/step - loss: 1.6119\n",
      "Epoch 62/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.7042\n",
      "Epoch 63/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.6167\n",
      "Epoch 64/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.6024\n",
      "Epoch 65/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.5081\n",
      "Epoch 66/100\n",
      "72/72 [==============================] - 1s 19ms/step - loss: 1.6281\n",
      "Epoch 67/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.6489\n",
      "Epoch 68/100\n",
      "72/72 [==============================] - 2s 21ms/step - loss: 1.6474\n",
      "Epoch 69/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.5904\n",
      "Epoch 70/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.5498\n",
      "Epoch 71/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.4343\n",
      "Epoch 72/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.6290\n",
      "Epoch 73/100\n",
      "72/72 [==============================] - 2s 21ms/step - loss: 1.4657\n",
      "Epoch 74/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.6071\n",
      "Epoch 75/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.5426\n",
      "Epoch 76/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.5451\n",
      "Epoch 77/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.5409\n",
      "Epoch 78/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.4550\n",
      "Epoch 79/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.5945\n",
      "Epoch 80/100\n",
      "72/72 [==============================] - 2s 21ms/step - loss: 1.5119\n",
      "Epoch 81/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.5135\n",
      "Epoch 82/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.4367\n",
      "Epoch 83/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.5576\n",
      "Epoch 84/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.5468\n",
      "Epoch 85/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.6081\n",
      "Epoch 86/100\n",
      "72/72 [==============================] - 2s 24ms/step - loss: 1.6241\n",
      "Epoch 87/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.5147\n",
      "Epoch 88/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.5660\n",
      "Epoch 89/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.6289\n",
      "Epoch 90/100\n",
      "72/72 [==============================] - 2s 22ms/step - loss: 1.5145\n",
      "Epoch 91/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.4845\n",
      "Epoch 92/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.5026\n",
      "Epoch 93/100\n",
      "72/72 [==============================] - 2s 26ms/step - loss: 1.5293\n",
      "Epoch 94/100\n",
      "72/72 [==============================] - 2s 23ms/step - loss: 1.5076\n",
      "Epoch 95/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.5901\n",
      "Epoch 96/100\n",
      "72/72 [==============================] - 1s 20ms/step - loss: 1.4656\n",
      "Epoch 97/100\n",
      "72/72 [==============================] - 2s 21ms/step - loss: 1.5112\n",
      "Epoch 98/100\n",
      "72/72 [==============================] - 2s 25ms/step - loss: 1.5385\n",
      "Epoch 99/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "72/72 [==============================] - 2s 27ms/step - loss: 1.4951\n",
      "Epoch 100/100\n",
      "72/72 [==============================] - 2s 27ms/step - loss: 1.4242\n",
      "train error     test error\n",
      "1.0584410135243587 10.411996597990912\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD7CAYAAABjVUMJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAABKvklEQVR4nO2deXwU5f3HPzN7JNncCRtAQDxrq62K0gpi4Yf+CGdEEBXxotiKVrEe1SKirfqjIh4IKlpPimJRECwgIiACQhAIN8gRIAnk3px7ZO95fn9sdrPH7O7M7swe2efdVyU78xzfeWae5/sc3+f7MIQQAgqFQqGkHGy8BaBQKBRKfKAKgEKhUFIUqgAoFAolRaEKgEKhUFIUqgAoFAolRaEKgEKhUFIUqgAoFAolRVHGWwCxtLaawHHity4UFmahudkog0TSkgxyUhmlIxnkTAYZgeSQMx4ysiyD/PxM3ntJpwA4jkSkANxxk4FkkJPKKB3JIGcyyAgkh5yJJCOdAqJQKJQUhSoACoVCSVGSbgqIQqEkDoQQtLbqYLNZAMR3aqOxkQXHcXGVIRzyychArU5Hfr4WDMMIjkUVAIVCiRijsR0Mw6Bnz75gmPhOKCiVLByOxFYAcslICIe2tiYYje3Izs4THI9OAVEolIgxm43Izs6Le+Of6jAMi+zsfJjN4iyM6FujUCgRw3FOKBR0IiERUCiU4DinqDhUAVC6LYRzwvD+VNgOb4i3KN0aMXPOFPmI5D1Q1U3pvjhsAABr2Uqof1McZ2EocvPqqy/j4MGDcDjsqK4+hwsuuAgAcNttkzF27M2C0pg6dQoWL/5cFvlWr16FrKxM3Hij+G9x3bo12L9/L5599h+SykQVAIVC6RY89dQzcDg41NXVYsaM6RE15HI1/gBw+PBBXHvtQNnSjwSqACgUSrdn0qQSXH75r1FefgKLFn2IL7/8D/bu3QO9Xo8ePXrgxRdfRkFBIW64YSC2by/DRx/9C01NOpw7dxYNDfUYN2487rvvfp80TSYjXn75Jeh0jWhq0mHgwN9h5sznAADvvvsWtm3bAqVSgZtvnogLL7wI27dvw759ZcjPL8SmTd9hwIBrMWZMCQB48tXpGvHyyy/BaDSgqUmHMWNK8Mc/PihbuVAFQKFQJGHH4TpsP1QnS9o3XNkbQ37TO6o0Bg26Hi+++DKqq8/h7NlKvPfex2BZFi+99Dy+++5b3Hnn3T7hT50qx6JFH8JoNOD222/BxIm3Izs723O/tHQ7Lr30F/i//3sFdrsdd999G06cOI7a2hocPnwQS5Ysg8PhwJ///Ee8/vpC3HDDUFx77UBcd91gbNr0Ha+MGzd+hxEjRmL06HEwGo2YOHEsJk2aHNVzh4IqAAqFkhJcfvmvAQB9+/bDI488jjVrvsbZs1U4evQw+vTpGxD+mmsGQqVSIT+/ADk5OTCZjD4KYMSIUfj55yP48svPUVlZgfb2dpjNHThwYC9uvHEE1Go11Gq1qGmlKVPuwb59Zfj8809RUXEaDocdFos5+ocPAlUAFIpIHNVHwBb0BavJi7coCcWQ30TfS5eTtLQ0AMDx48fwj388i8mTp2D48JugULAgJHAXs1qt9vzNMExAmBUrlmHLls24+eYJmDTpd6ioOA1CCJRKJbwNcurqapGXl+8T1zs9h8Phuf7WW/NRW1uDESNGYejQ/0FZ2W5e2aSCmoFSQmK2OjD3s72ob+mItygJg3nda+j4+qV4i0GJkAMH9mLAgGtxyy2T0K/f+Sgt3R6Re4Y9e3bh5psnorh4NGw2G8rLT4LjOFx11TXYsmUzHA4HLBYLnnxyBnS6RigUCjidLjv93Nw8VFScAQBs27bFk2ZZ2S5MmXIPbrzxf3H2bBV0ukZZ3VvQEQAlJEcqWnCyuh0rt57Gnyf8Jt7iRIYMPShibJY8TUpsuOmmYsya9RTuvfcOAMBll/0KdXW1otO5/fYpeO21l/HZZ58gMzMLv/71lairq0VJyS04fvxnTJt2FziO4Lbb7sT55/fHwIG/w/vvL4JGk4lbbrkVzz//DO67bzKuuea3KCzsAQC4++6peOml55GWloaiol745S8vR21tjaTP7w1D5BxfyEBzszEif9pabTZ0OoMMEklLosm553gj3v36CAZepvUogESTkQ+tNhuNtToYP3kQUKYhe9q/JEvb8P5UAED2A4ujTitZyjKYjPX1VejVq3+MJeInlX0BueF7HyzLoLAwizc8nQKiUCiUFIUqAAqFQklRqAKgUCiUFIUqAAqFQklRqAKgUCiUFIUqAEoKkFSGbhRKzKAKgNKNoX7qKZRQUAVAoVC6Ba+++jKmTp2Cu+++Df/zP4MwdeoUTJ06Bd98s1pwGlOnTpFRwtA88sgD2LevDMeP/4y5c4PvNDcajXjmmb9KkifdCUyhULoFiX4egFB++cvLMXPm5UHvGwx6lJefkCQvqgAoFIok2E/ugP3ENlnSVl02FKpfDIk4vhznAaxbtwalpT+itbUVzc1NGDLk93jkkcexf/9evPvuQjidHC666GI88cTf8MYbr+DMmdMghMOUKfdixIhRsNlseOWVl3D8+DH06nUe2tvbAAD79pXh44/fx9tvv4/y8hOYN++fsFotyMnJxfPPv4Q333wVTU06PPPMX/Hyy69FU6zyKoB77rkHLS0tUCpd2bz44oswmUx4+eWXYbVaMXr0aDz++ONyikChUCgApD8PAAAOHTqIxYs/R3Z2DmbMmI5t235AdnYOzp07ixUr1iIrKwvvvvsWLrvsV5g9+wVYrR344x+n4vLLf42tW38AACxdugLnzp3FfffdGSDzCy88h4cemoEhQ36PVatWYPnyZXjssacwY8b0qBt/QEYFQAhBZWUlfvjhB48CsFgsGDVqFD799FP07t0b06dPx9atWzFs2DC5xKBQKDFC9YshUfXS5Ubq8wAA4IYbhqKgoBCAy8nc3r178D//cxP69euPrCyX/52yst2wWi345pvVYBhXO1hRcQYHDuzFzTdPBAD063c+fvObK33Sbmtr84wsAGDChEkAEJHjumDIpgDOnHG5Op02bRra2tpw++234xe/+AX69++Pfv36AQBKSkqwfv16qgCSAGpISUl2pD4PAAAUiq4mlBDO89udFwBwnBPPPfcSLrvsl1AqWTQ26pCTk4vVq1fBu2YpFAqftF3nCnRZslmtVjQ16cCy0tnuyKYA9Ho9Bg8ejOeeew52ux333nsv/vjHP0Kr1XrCFBUVoaGhQVS6wbzaCUGrzQ4fKAFIJDlza11eINPSlD5yJZKMwejRIwtGuIxBpZTX7RdTqjSToSyDydjYyEKpTBxjQqWShULBev72RqFwyXro0H5ce+21mDTpdrS3t2Hu3O0YPvwmT3ilkgXLMgFpuOO7YVkGu3fvhMVigkqlxqZNG/DAAw9CoWDBMIwn7MCBv8N///sVZs16Dk1NOtx335344INPcN1112HjxvUYOnQYGhrqceTIIY/sDMMgLy8HRUVF2Lt3F667bjDWrv0W+/fvxcMPPwqn08lb7izLivqeZFMAAwYMwIABAzy/J02ahIULF+Laa6/1XCOE+Gg4IVB30LFFr3cdR2e1OjxyJZqMfGi12WhqNgJwHQcgh7xSpJksZRlMRo7jEsYFs9vVstPpksdfLqfTJevw4f+LWbOewpQptwFwnQdQU1PjCe9wcJ42xjsNd3w3HEeQm5uHxx+fgba2NowcOQYDBw7Cvn1lIIR4wk6d+ke8/voruPPOSeA4Dn/+86Po1asPxo+fhFOnTuGOOyaiV6/euPDCiz2yu+M/99xLeO21l/HWWwuQm5uH5557ETk5eejZsxceeuhPeOstXzfnHMcFvKtQ7qBlUwBlZWWw2+0YPHgwANcD9enTBzqdzhNGp9OhqKhILhEoFEoK0rv3eVixYo3PNe/fWm0RPvjg37xxt28vAwDcf//0oPG96d//Ajz77D98rl1zzUBcc81Az+/MzCw8/7zLrt/7PAClUom//W02b7ru+BdffAneffejgPvvvfcxbzyxyDZ2MxgMmDdvHqxWK4xGI1atWoUnnngCFRUVqKqqgtPpxNq1azF06FC5RKBQKBRKCGQbAQwfPhwHDx7ELbfcAo7jMGXKFAwYMABz587FjBkzYLVaMWzYMIwaNUouESgUCkU2xowpwZgxJfEWIypk3Qfw2GOP4bHHHvO5NnjwYKxeLXxrNoVCoVDkIXGW7ykUSlKSZMeKd1sieQ9UAVC6Le4KQeguBtlQKtUwmfRUCcQZQghMJj2USnX4wF5QX0CUbovF5gQQaA4oNYdONyMnU4ULeuXImk8ikp+vRWurDkZjW7xFAcuy4LjEMEkNhpwyKpVq5Odrwwf0jiOLJBRKQhCb8wDeXH4QAPDxzBtjkl8ioVAo0aNH73iLASD591TEAzoFRKFQKCkKVQAUCoWSolAFQKFQKCkKVQAUCoWSolAFQBFGElr5ifQzSKGkHFQBUCgUSopCFQCFQqGkKFQBUCgUSopCFQCFQqGkKFQBUCgUSopCFQCl28LEyBUEhZKsUAVAoVAoKQp1BkehRMnw9KNocfIfuk2hJDJUAVBSAHl3sd2i2dv51x9kzYdCkRo6BUTpxtA1AAolFFQBUCgUSopCFQCl+0IHABRKSKgCoFAolBSFKgCKIJLQGSiFQgkDVQCUkFCXyhRK94UqAEpISDfo+lMdRqHwQxUARRC0EaVQuh+yK4BXXnkFM2fOBACUlpaipKQExcXFmD9/vtxZU1IcqrQolNDIqgB27tyJVatWAQAsFgtmzZqFRYsWYd26dThy5Ai2bt0qZ/YUCoVCCYFsCqCtrQ3z58/Hgw8+CAA4dOgQ+vfvj379+kGpVKKkpATr16+XK3sKhUKhhEE2X0DPP/88Hn/8cdTV1QEAGhsbodVqPfeLiorQ0NAgOt3Cwsidbmm12RHHjSWJJGdOnQEAoE5T+siVSDIGo4c2C7Wdf0spr8EvTf/fYkmGskwGGYHkkDORZJRFASxfvhy9e/fG4MGDsXLlSgAAx3FgvGwKCSE+v4XS3GwEx4k3TdFqs6HTGcIHjDOJJqe+3QwAsFkdHrkSTUY+tNpsNOmMnt9yyOufZiR5JEtZJrqMQHLIGQ8ZWZYJ2nGWRQGsW7cOOp0O48ePR3t7Ozo6OlBTUwOFQuEJo9PpUFRUJEf2FAqFQhGALArgk08+8fy9cuVK7N69Gy+88AKKi4tRVVWFvn37Yu3atbj11lvlyJ5CAUA3sVEo4YjZeQBpaWmYO3cuZsyYAavVimHDhmHUqFGxyp5CoVAofsiuACZOnIiJEycCAAYPHozVq1fLnSVFBrKc7XC21kCR3yfeoggmkXcx1+iM6KPtfqeI2exObN5XgxG/7QsFS/eZJjr0DVEEMbH1Q3QsfzbeYnQL9p5oxHMf7cae443xFkVy1pRW4ssfTmHH4fp4i0IRAFUAFEqMqdaZALhGAd0Ni9UJwDUSoCQ+VAFQKBRKikIVQArx1leHMG3uZlFxEsGSRtdmxrS5m1EmcsokEWSnUBIZqgBSiP3lTfEWISKq6l0bZ3b9LH7nOIVCCQ5VAJRuD0PPM6NQeKEKgEKJE4lsphot3fjRuhVUASQIhOPAWRLbjwlFGrr10kS3frjuB1UACYJtzwqYlsyAs0Mfb1EolJhhsTlgstjjLUbKQhVAgmCv3AcAcJrpKICSOjz97k7MePPHeIuRslAFQKFQ4obRTHv/8YQqAAqFQklRqAKgUOJEt7aU6dYP132gCoBCiTXd2FKmGz9at4QqgESjOxuHxxxalhRKKKgCSBBoux+aHqweSkIXDCkUKaEKIEFoNVoBAM3tljhLkpg8l/c1huu/jrcYKcf27zahdMtP8RaDIhMxOxKSEhqOcw0BHE4nfSlB6G0/G28RUo6rqj4DqgBcsTjeolBkgI4AKGGgy3qJwvMf7cLcpfviLQalG0E7m5SEp9v59Sccfqs+DZB+oqK5TxJLBuiSVnJARwCJBq053Z4+rXtxd9YO9Gsti7co0tOprA+eSs6zJ1KNlFEALVuXwfDBH+ItRlBou586qByunrzK2RFnSeTjWFVrvEWgCCBlpoDati+PtwgUCoWSUKTMCMDNzqP1qKxPXJfLdD8AJRVxODlYbI54i5FypJwC+GDNz3hxcTece6VQkpg5n+7Fn9/YFm8xUo6UUwAUCkU+NpVVRxSvqp6egxEPZFUACxYswJgxYzB27Fh88sknAIDS0lKUlJSguLgY8+fPlzP7JIXOAXV7BNq11jaZsP+kTmZhKKmMbApg9+7d+Omnn7B69Wp89dVX+PTTT3H8+HHMmjULixYtwrp163DkyBFs3bpVLhEkw/Dh/bAd+lbmXLqbsTslWmZ/uAtvrTwcbzEo3ZiQVkAPPvhgyMjvvfde0Hu/+93vsGTJEiiVSjQ0NMDpdEKv16N///7o18+1AaakpATr16/HsGHDIhA9hnBOWH/6AuorR8dbEkoEUNUaGwi1YEg6QiqAkSNHev5euHAhHn30UVGJq1QqLFy4EB9//DFGjRqFxsZGaLVaz/2ioiI0NDSISrOwMEtUeDf+M4xabbbouGLiiKXC62858xGbfq7f3Kw7rtwyepNT1yWDmHx7FGajJoJ44fD/HsR+H2qVAgCg6vw3XDz/+7EsezHP9s32Mz6/I5Ez2meLZdlESiLJGFIBTJgwwfP3v//9b5/fQnn00Ufxpz/9CQ8++CAqKyvBeM1/EkJ8fguhudnocZwWDTqd+EWnSOJEgtz5iEm/3c87qU5ngFabHbOyAAC93oxeXvkLQavNRlNzV1g55PVPU2geNrsTAGC3OwTFizQfKRGS56mzvpu/Yl3HYv1dRkI8ZGRZJmjHWfAagNiG+vTp0zh27BgAICMjA8XFxdi1axd0uq5FLZ1Oh6KiIlHpUijJD52UoiQGsi0CV1dXY/bs2bDZbLDZbPj+++8xefJkVFRUoKqqCk6nE2vXrsXQoUPlEoEiAcnsiC3eU9IOJ4fPNpyAocMWX0EolCCEnAJqa2vz/O10OtHe3u6z0JOXlxc07rBhw3Do0CHccsstUCgUKC4uxtixY1FQUIAZM2bAarVi2LBhGDVqVNQPQenuJKcW2nO8EZv31cBic+KP4y6PtzgUSgAhFcCgQYPAMIyn0b/uuus89xiG8UzxBGPGjBmYMWOGz7XBgwdj9erVkcpLiTHx7kUnM+56Q61jKIlKSAVw/PjxWMlBoYQg2gY0Pg2wwmbEy3nLsM0+GcAVcZGBQglFyDWAyZMn47333sOpU6diJQ+F0m3IbjsJDWvDL038vqeY7jYwSM6ZOkHYz+wB19EebzEkJ6QC+OCDD3DeeefhrbfewoQJE/DKK6+grKyMDmllhJZt94O+0eSG2MywbHoH5m9fi7cokhNSAWRnZ+Pmm2/GggUL8MUXX2DQoEFYvXo1xo8fj1mzZsVKRgolKWG6c5e4m9HUbsaOw3UAAMI5YSldCq6jDZ0XAACcoTlO0slHSAVgs9nwt7/9DRs3boRarcawYcPQ0NCAX/3qV5g4cWKsZEwJ3L1EsfstKBRK9Lz82T589M0xOJwcnDU/w35kIyzbPom3WLITUgEsXLgQRqMR11xzjefaiy++CIPBgB9//FF24SiU7gBDp/USHr3Ja6+G+3119vy7MyEVwJYtW/D666+jsLDQc61nz56YN28eNm3aJLtwcqCAEwy6/4tNFJz15SB2S/iA3RASdjRHFYMvBIRzxluIEHS/9xVSAahUKqSnpwdcz8rKglqtlk0oOXmjYCkeyk5O5ZVscBYDOlbPgfn74F5j44WztTaOuXfPab5o1zwmaMpg/PB+EI520GJFSAXAsiyMRmPAdaPRCIcjec/vvExVH28RUgOHa1jNNZ+NqxisX7tkP70bHctnwV65Lz4CUXj5fZp73xFVALEipAIYN24cZs+ejY6ODs+1jo4OzJ49G8XFxbILl4rQ3o/8uBUS1xLZ8YVScapGD4s1eTtS3Z7uN+MTQEgFcN999yE7OxtDhgzB7bffjkmTJmHIkCHIycnBww8/HCsZk5IanRGNrR3hA1JSD68Rydb98VVClEDaTFYAQIuh+69dhXQFwbIsXnrpJTz44IM4evQoWJbFlVdeSV04C+C5j3YDAD6eeWOcJaEkKkwqdDGTBAYc1HCNxlr1VvSCn2UQ0C1HBCEVgJs+ffqgT58+cstCEYHpv/8HruEUsh9YLGs+ibEtISGEEE1ySt2FFAcvRUbsS25Sxk8YnFYOwt3EI06yv8ngyHYeAEVeuIb4+GdyttYkuKkeRSq+2nZaVHip2kkp3aEYPnkI5o1vhw33O7W7PqXWGhxVAEHgzHqYvvo7OGP32/4dDR3Ln0Xrti9FxhJfocur2/DI/G0wWeyi4yYLlya4Ndqh0/H59p1SjjzsZjgq+J3xUagCCIrj5HZwzVWwHdkY03xJEkw0WmtPCgwZeZdwTWklOqwOnK7RR5xG3AnTJb5A2QRNU+gzNSixJdU2bVMFkDBIO8/4w/4aLFxxSNI0uwtcZy2P3RR38IyU1uAuhjMZC3qybTLIk5gk/kx799MOghaBKcnHp9+diLcICQB/hT1Tq8f5nf9ecW1sJRLD33LXIJc1A0g1x4vxVQWBA7fEV02RQkcAFPEIHCe7F/PiZ03Cj8PJ+fwrOxE+vqvxp1DkgyoAimy0d9pRGzqiWchNLOUhBiKo45i8zycX3djqMuGgCiBhoA2BN/QwlVQkEepAIsgQO6gCoIhHdBcttSqVG0FKLIGLRuxbZogTCwqWeDl1oyQ6VAEEQddm6fw3VvOwndWNp0Gw2p1ojJkcAkg1WzmJSaRT3wzvT0XH2lckSUvBuab8xmQckCS9eJMKXzlVAEFo1rsa3Ba9Nc6SAAuWH8TM93bGW4xuh9zNcNeeDuJ3PbFw1ibWXgQ6/Rc7qAJIAo6fbYu3CJLgqC+H4+xBUXGSe7CRmg2ZhrUhi4n9iJUQgiYJR8qp8PaoAqDEDPPqOTCvny8obF9HFRYULAFrM8gslZwktfaKikeyxe+g9z+4Ryxrtp/B0+/tRFW9NN9MKrw9WRXA22+/jbFjx2Ls2LGYN28eAKC0tBQlJSUoLi7G/PnCGoN4ELdhaBJ0O4RWDMbYFHEeV1hcp3WlGZLXX37XN5QEL1USup6zSBF8h7NcHD3j8l0kZN2OcBws25eAM+h8r8vU6s96/yes3xXfk/H4kE0BlJaWYvv27Vi1ahW+/vprHD16FGvXrsWsWbOwaNEirFu3DkeOHMHWrVvlEoEiAZE2XcRihHLz69LIIFP7GbseXqicErmf2X0Vl7OhHPafN8Pywwcxya++pQNf/hAfD76hkE0BaLVazJw5E2q1GiqVChdffDEqKyvRv39/9OvXD0qlEiUlJVi/fr1cIkiC2B4BsVtBnOKP+UvYZoBw+N/0wz6XhDQLxJZAVksJB8PzFyVapOi9J5CBVkyQzRfQpZde6vm7srIS3377Le6++25otVrP9aKiIjQ0NIhKt7AwKyJ5/GcFtdrskOHValfRqFRdOjJcHAA4M2cqFIV9AdwoOA4AnPEyAw0WR6vNxqFTOihYFhkCZeK7L1QmAOhxoBQlmv2i07CrOmDq/JvpDO9+B0LyZzsnhDUaNdicDM91MbIXFGbB7XDZO577narVSlHpufF/jmDPVZmhBgAoFKzPvTS1b7ULJkNgugQj0w8hP2MolFn5ouUWlkcXSiULOEPL6I1Go/L8rWBIRGULAD20WVAqI2+acnIzwr4bs1kDMwCVSgGtNhtt7rwLs9GcmQYAYBkGWm02OKsSRnR9x0IghICzmKDICGyvIi0XOZDdGVx5eTmmT5+Op59+GgqFApWVlZ57hBDRNtHNzUZJfMvodKEXimw2Vy/ebu/yFxMujhtnc9e8tdA43m6g/eMwILgh7Tga64bg2XdLAQALCoSl733/Vs0uXKGqhk4n/JhKizHwXGMiIF9Ob/L87R9eSJm433GHyQZjmhk9RcQFXJWsudnIm6fd7mrVbHan4PT48I/r/9tsdtnFO52czz33txUsXrB8+iuaMEZzEDXL34Bm3NMRyx0qD28cDnHffo7Od6QYadk26QxQRKEA9O3msO/G0eb6ru02h8+9pmYDTCYrsuDyGqvTGUDsrj1BpPO3EGxHN8G64zNk3jEXbG6vkLLIDcsyQTvOsi4C7927F1OnTsWTTz6JCRMmoFevXtDpuhZddDodPV9YAFeqzmJS5h5Yd6+IKp2h6SdQqDCFDygFSTCUZmS2MZW6CFjGJW9buzFMyPhQYDkXbxGEI/Ncj9vcmWtvlDWfaJFtBFBXV4eHH34Y8+fPx+DBgwEAV111FSoqKlBVVYW+ffti7dq1uPXWW+USodugZlw9RmLtwG2an+CMqfVuZI2kdG2rHI10bLQTCdLIkCB/C8VgTsxT0mJRqj8eqkXvwkxc0ic34J7b+2w0bXty7zsRj2wK4KOPPoLVasXcuXM91yZPnoy5c+dixowZsFqtGDZsGEaNGiWXCNHR+REpiS22+RICS+nnsB/Z4DnwnelsJmx2J25IF3oaV3xpNVqRFlUKXbU41SqlXFjtTny09mdMvulSFOSkx1ucoIQ6Fe+TdS4/Qx/PDDWNKVwDJMMJfHIimwKYPXs2Zs+ezXtv9erVcmUrOUWm8hjl1PXR2o9s4A1R1WDEhTGSxkOE9UOyRlvC+kksRjDpkRkRBMO677+SpicEJsJC2X9Sh7ITOihZYOqFNVBd8b8SS5ZMROKoT0S5e1WA2PkTEw/dCcwDIRwuqOVvhMXy7a4qWDsXHSPlcnWNJLJERvfoITmqj8K45BE4znkdkynBnIWtbFV0CYhpU6LLycOlHQdg/WkZbId8TbBb9BacrhW3gYsQDtayVeDM8T+7Wb6RYhQfCgN89E1i+Vryhh4JyQPRS7dws/yH02g32jD5pkvDBwZ/JR+grpJMnlgh2XywX0JWuxNpKoXoZJyNrk04zvryOMwpJYYSdUvhmda0+/ZMn1pUCoJw0yu+OOtOwrbvv3A2VUoioxRItr6bBIYM0UJHADyYrdH12P2x2MRvDOMjMZqR2BDsNK1/r5fO1zxJ1hoe6YfQGS/YU0eULOk0FXXEeK0sBNG8VcLzw11/7Y4YHSEaQ6gC4MGUoFYWCYOQHrRXLSQS9rjrmgL3JYgm7u1+tAJEaJkVTgPEEf/d5pEg5jMLWQR+N52de1K6ozECVQAxILk/nEDhha1pSNTKEJLsBRgGMc+WgC23RPDtNo8Y2YpJ+LtqM7pGRHqTPaHfGlUAcHkGJJy00z6iZYhr7uKobUzMjUiCiVFhu3e5M8n0cr1ICb84IkezQmkzug6SatZbxEeOIVQBANAvmwnjh/d3XZD4w0+E+m/d81WEMQOlFzR3LmEZStMQBSYif/sWJIc4NaxdbV0qtOxhEPFRdefSogoAAGtM7O3aUmDbv0aytCK1Q5co825AfLsEshQhIYj3c7nXOBLhSMm41hERUAWQYIRaME0mq5VEqIQOZ2JYbYRuCoQ3FNEuhXTFl7Bx8ulJx/ede55PMjMgXyJNNpGn0qgC4EHy9yXRNGNcehVhsuz49g10rP6njNlH/sz1LYm3A3PrgRocOt0SVxnoaQRi6b7lRDeCxZmzDYaEti/m/fS9Ljq9d9aGixdRztF49ur602x1QAGgprYRgPiNZNHgrbiXbT6FIXHqdsnr9yY5pjxiR3IoDaoA4sw/PtkDAHi2IHzY+EwB8VRskXU9l42mJ04ksQK11Z9CBgBtwy4Y8q+PPsE4EvFIUIopkiQgMR4vORQinQLiQ3IroPAfQ3J8LiKQaeJTTKre5c4447e5z+7gsHFPIvnKF1aKidGQyk1gzfNfv0rkOfxooQqAjzhuPAp1Qlq3UxJeHDrdhDO10joUYzhpXHBEi9Fix3++L0dDqwS7mAHYHAQ7j9SHD+iHPN9PZK0jsZlh3fMVWEg3/amwd2BU+kFs2FUZOm9CYAsy7Uo6/xc9TGdeEiQlIymhAGqbYnQKlh8sOBSnH4KSSxw/KYnKm8sP4f+WlAXeiKICKY1e500nQDeO46SZxOMIwQdrf5Ygpfhh3fs1bPvX4Hfq0xHF339Shxqd74bEaw0/YLTmIBT1R0LG3XqgFvM+F7HzOMEb8WhICQVQrRO3c7WqQZozO69RV2Cs5gB+rf9RcByxfnOk9LMTJIPAazFsS5PFnjqZkKVMxX6Hnc7jFAxfTzx8Wm+tPIznPtrtc03V6eVUESb+3pO6kPelMGFOlu82JRSA2I+zqd0qSbYqxuVeQknCzz97nX8lKJSb+1/5QbRcyQF/JRTTkU/04TcQ23l22TsLyUiKF0lKKIDM1hPxFiEsybTJSwih1jKiQVwblhq1e8fhOqwtrRQegefdbIh2kTrC982CwxUq8XmPydjPE6/rfe88Wo+/LBQ+8nYj1wbGQtaANCTeVHBKKADWKa5Hn2lrlkmSaEmMBs3YYfM4u5KdaB45MYpLdj765hhWbjsTNlyo4lj2fZRHn0Y4uhiRcQQPZPuOYv2Tqm40ornd16nayIzDAfHcTTcB8NmGEzB0yG/5pWsz489vbEVDC/8Cv/tRns9bhb/kfBd4nxD8Z1M5TteIO4lNKlJCAYjl6upl8RaBl4TZBwDg0Gl5lSTx+zd64jfC8nGWwPi+x0ieL9p9AIJLQtD29OjKVcgekec/3o2n3i0NG87L6FeEBMTvl7iy/ennBlhsTmw/XBc0jHs03EfZijM8Df3GsnOY8+leUflKBVUAMUGaZkzIWkLiEF3DECx2AhjziKDTHTQIMpkgboFF9Jyl6gA0dfamE/mw8kgwdgifYolVZ8o/F7NVvGnyuUYjVgkY4UVCiuwEjvxlO6NwKCb1R3ahJXEPl5Ya72Yx4kafeL27OGqOK9XncKX6HIyWa+MmA9BVpoYOO6Bx+arPFRJBIPG2fHHnH04KsV9CvDsdcz4tg83OYdz1/aFSSuvGhI4AwuA+Dk4sDAh+nyb8/Fq7o/NAmgSbt06USt0tsEW2H0Uyr6Yi5+kNSXo0KgGQzxhxg4j6JwQV44T91E6fa7HQDU6nfD48qAKQiWvUFeirbBUcPlGtgJyOCHfTxqjbxHW0wVEbrKLHWXlJVAR7T/jarbuTnTZ3Mw6dbkIfRTMuUYbfGUz8/go3391qEL7QH8zEVG+yYdrczTh5rs33hqzfB4M/pX+L2zJ3Bw3xZO66iFK2bP5X0HvT5m7GO6t8zzaW8guUo8hSQwHEYQyXznT1npLZ/Dr31LcRxTOapTF5C1d0HatehHntXAEpJYKCjexD4Hg/IILfqk9j/4l6PJ37DWbkbIgg5ejLpL7T+kVv4n/f7oY/wBeSDJXC+2k0THDlJTZnMaah/srav+2JZC+GnO1HaiiAFKDqo7/i2Ip3JE83VEUKxftr5HJV4FehTCF86yex4g3HVaqzuDtrB36l3yE4jhwNiXtBOZLFzUgJ9k16rwGEGlEfrQj+zSRyZy3pRgBGoxHjxo1DdXU1AKC0tBQlJSUoLi7G/Pnz5czaB0ISofcnLwXOJvRt2ROz/PxL1P9MA0OQHqHwDPjf2TmZDqQ/dOQUvln4Mtr00jhs40PKw+E1rKsRTHfKJ6+USNmuvpz/RZgQwuu7SchIVUByhADnKVqQzUhvWSXnDm7ZFMDBgwdx5513orKyEgBgsVgwa9YsLFq0COvWrcORI0ewdetWubJPOtKY+C24HTrdhGlzN6M9ikZ7+mtbpBMICNoVE7Mg6p2Cwxm6Eil2L8XQ9BPQHZPbHluazoj3pqfEgF+SYL1Ws034iOF8RRN6sNJ6inWjj3KzmPfz/S13LWbnrer8RTz/RNtzd5fsqWrpN4vJpgC+/PJL/P3vf0dRUREA4NChQ+jfvz/69esHpVKJkpISrF+/Xq7sfYi3GRcAtOgt+Grr6aDavEjhdkAXyRxhdM3Apr2uEVpVvXAneOFyTIhFba9yaTF42+Hz+YCX4VQ2//lfviBC0iE+/+B8ZXNEbpQDvhPJtYfwBM82CB/JPZm7Ds/lfS1KioirRFsdFLbInUGmMy7FJuXXf7GyAQsKlqC1Ifhms0iRTQHMmTMHAwcO9PxubGyEVqv1/C4qKkJDQwNf1G7Je6uP4pudVagM18jGqEvXarAGbF8/fla41VLMkKg8eJOJec8guofxllbhUQDCn0Fs7moIcWKYAIq+E29JBD+rX0DnmhdQeFS8JwB/i6poO0AOJ4flW07BbHVgcJrLTUdWe0VUafIRs41gHMf5OAgjhETkMKywMEt0nMqMtIBrWm2252+D37VgTbR3HG8Wrz2Ky/oXYPBvevPGVShYoHMdIi9PEzQdwDU0zhAoj5sePbLh7k/5p82XhlabjWlzNwMA1rw+HplKDr9Tn8L6XQQP3z7AJ36wvHOy06HVZgekGwzvsP7hLF4LiO57rIIF7IBGo4IiJ533mbzl48s7K0vt+ZtlvPo6nd+dkyNd+bGuaxpNWsjn8M8z1POf06h9fufmpMNsdYCk+7qCCJafO+3snPSAMO4UWEXXc4WTOyvLtx4oVfybitzppDGB78WfjM5nZFkGSoVvf1KrzUZOnesp1GqFTxr+YX3i9ciGSt1VdqHese819yJwYLsSTH6Gcd1r47vHMtBqs9FhZOBtZOudVmamq0w1GWqf+yc7v6fMTDXUKt9mtsNB8MhrP+Dzl0YjM13Fm+7GXVX49qezUCgUyOu8lp6hCvuOxRIzBdCrVy/odF0mUjqdzjM9JIbmZiM4kZuzOsw2FPpd0+kCmza+a0Luf/XDKQDAxzNv9Fzz3sDEOTk4nK6NXm1tHdBlBC/2t5cfxFOd2zMP/FyHPtrwCk/X1CWXTmfwOXkq3HPqdAYM1H+PX2QdRS9zO3S6m/D93mos3XgS7zw+NGieBoMlIO1Q5ed9r7FRD2fDKSh7XQoAeOOLAwHh3O+4o8MOY7sFPcPk439Nq82G0WD1KFNvU8qqOgMuSgOOVbZggCc/1z1ThzXsdxAsT//fHR02T+UFgHY9/wJhuPwMegv2/8w//Hd6rW2ES8fYadfvbh7tdqdgeYKl3dHpfiHPWo96dX5AHL3eNfVmtTp80jjbYMD5vnq9K16TASqVOvB6GLlCdSeDyU8ICX6Pc92zmHxHyt7hTSZXmZq9Ns3pdAaQzg+qvKoZ9sp6wKvdfuQ1lxO7mW//iOen/pY33bZ217eiN1oxIs3lBsJitgn+Nr1hWSZoxzlmZqBXXXUVKioqUFVVBafTibVr12Lo0OANjKTEeZTqra685yYPnGoKGW/z/pqwaS8oWBIwjP0wjAnmym2+pzBlcK4dqjdlHAUAbNhzFgBgEOFbRQz241thXj0Hx3/chOc/2o0T/puEfIh82kSMY6+Y7DjmzUJYvs9/tNsn5MiMQ1GLI021cEmlgBM9TSeDhyIIOMFLLqR+k0cqQ5gae/LkzzWnujTAa6kbf8s5wHV41bS5m3m9gzKEX2FHQ8wUQFpaGubOnYsZM2ZgzJgxuOiiizBq1KhYZR93+Ga7fjxYG3tBAKwtrRIULmRFiqL14NpdA+r9+4+hWmfkrQj+M6pS4uRCLaCKmFMPu9IoX89Dw7qUs649iJM5Hvw9rEZbqsu3nMLmvV2dlFBHnx441YTnPtqNumb5jmfNZV099XxWeB5C3pDFGIkFkivlHIhTekfOuJRNmefUMnk7JrJPAW3evNnz9+DBg7F69Wq5s+Qh9kOAWFrBRH+INeP3i3EnHCrT6PPzck8c7RMQzgGGDfU5C3sfYuTYsuMoBoYPJis2Bye8Fos9btT/t8UIwjnAavIAAOcd+TeGqsQdTt9msKJ3YaaoOEI5T9kGABiSFnwk4o+QEskwN4pOyGxzAGyn2XKQ9xPqdTACwkhByu4Ebmozo7E1/Caa6BvX4PS3ncKCgiUoYIP0EiLNWqTuCXA1ECa+kOTdvTE+PI7vQuYh7iGsO/8Di80Bk4XfckWOt9j72GcypCqOAhG93WgxLnkEps8e8/y+Un0OGlacHX2s9i3E2ww5Gi/CsSRlFcDT7+3EzH/9BL1M89xCuNjqmqvvq+iaY7xEJd401ui3mUVJ7KKO2fOulN57AaKprA9lbwp6r71TXtLprjl0PsKkcNYew8z3dmLGm13HAEa0V0HEQyuJOPcHp/jmdUXE55P31+pqEfHFIUcTmjgb11wIOvNGqoTgev5sxoyR6Qd5O5ehDCPlUGopogCCF9zmvcIrkPm7BTD8+2FsO1iL70XEC0/XhzBBUyY6dl2zb2/7RvsWPJD9A3orXHb9YnbPHj/bisZWlwVCyDluniK1l5eC2FyyZLHBfQg1V4dXTt67km1BrFV8IIG7OuVe2M0jbSHv+1fm9btcay83Z4TfbUw46Rf84uHnJqKtFjLL2ZNtkz7RIHvs+B7/rswdGKM5iD4I3tmL1atKDQUQ4iP0P2s0FI6q/YDVhMXfHsfSjcLnGYMR6iMBALvAhtv/Y8nlXD1Nt3uJHSGOq/OXQEgjwRekt6IVlh/eh2XLR2Hjq4KebEbgbHI1kt7TUhYhCkAEfZWuEdflKj4rK+EtlredPADB5skKAU6BrLuXe/6WbBqSENyQdjxAbgCYm/cf5DC+HQkpG6EcpgM3Z5R12dtGwCtL9wkOG6q3/IiP59QoN+dFoOHc9ZIlgWXhn5qPdDzhoyU1FEAISo/IY4kTrPfpW5lDfzzbDwnb+v3xN35mn4xv6jY/K5s/ZG0JmpaQxoZPanejwnUI2E3M8P+8Pq0cHSv/Dse5Lp/qcvRaL1K6FvX4p9siz3D1Dt+dmoJkDxLIevZoxHIEo0B/Ardl7saIjCMB9zJYO65Un/W5JuWEw5SsHbgp42eoW10myJFMZ4QyFxbjDiWH7er0CYnl38b7m1HzxhGQbqjm3GpzdXq8LeTkGBWkvAKYpAl+aEQweitacZ4ivG2wL/ItSjXrfadb3JXLrYQyzL6WGlerz2J4On8DU3HiFCZqduON/E9xOMTB75YOi1+lCzee4eeWjDL8Lee/AIDzOqesOH1DBA0ET/Xw0bVd6SlCVr3Iq1lFXeQ+ZPyxinCWJhQFF9q1t3+Zj844GHWe7hQVneXKdPZipZ6ei3RKNpJauba0MuBasKfxv86a20TlZZL5VLYUORM4kD6KFijA4Rp1aP8aeWzg7s2ZuWsAAGdqi3G6lt9D32Wqrt57uN6J3HPVOYbKgGu3aPbiB8sVAADi1ThmNR/DsEzXCVv+Iwc36YwNgw69CJtyvOeaOwWTxYnwRn6M57/DMwI3rVlLP8cAtbjhbrijO/OcXQqb8BiedlgcUdeGcKOneHvNETuVdG1a10HknEkiP1GdIrhHYVJRWW/AdZ4s5C3pBQWfwnbIBvWVo8OGVTF+05cGXZd0hAmqOdIZG/opmkFQ4HWVLgJHBN8H8XTuWjyZuy6qj+X/lpThP5vKee/leZlBBs0hEdyU+uMlUrDmIrPzQA7rie2ea27vlOI2+gTJIYK5zha9BQsKlmBqJr+L8TwueAPW2NoBZ2eex6paMW3uZiz7nv+9hsTvcVinb497TMZ+FLKBowQnx2Ha3M1YsSX81EI0MGE6In39RrXpXmsFpqWPR5V31+Yz119um/1Ep91o5Z2ms+5bIyh+T0XwTWSh3sa0rK14JGcj0riuDmi6WRciRmSkhAIIBReLImC6dp/WNQXaxzNBFgUvENxL6opvtTsD19kinEgPtvXcrTRNhq79C492LqwJUahipGFAoHKE36/hdmcxIK1rl7PQHm+HlzO6k50+1zf4H2EoAP/cFHZfuX+jrsbzHn/xXTgcrpjrfvLeoR0oOyuyB6832XCgvMvdSN+GbT73T9f6Nk7Xp5fjEmW9LAeQeL6LGJi3iBpRh3hWu4PD42/vwE8/B7fWKdCfxGXKwHVEf2XqlSGyOtcgCsAze9DZKeylaAMAsF51sFfD9sDwUUIVgF+DdUIil8jE74fb//nibwMPMA/WZA4MMz3Fx0Ovb4W102pG+PiC8fqrS/Ie+hO8od0h+Ew9BS2qBeQamvNqNoYNI3ZDkhuHk/MsuPHBmfXg9BFOV0g8wLszs1RU+NeW7cfCrw55Nt5pbMHXdNzMyNkA+4ltYcOJ5WKluxGN3JLlClU1FAh8V9U6IxxemwuJRJOqbis8A888vN3B4WhFC66oXIo/52xCYccZn/tuVx3+sEYdtJ1nfwxifNdYnI1nMHj/c15lRXcCS0Twmsh5HRdptTtR5n+oswicjeGH8GLmYX+fzt8A+9PHr7dB/P4K9hH5m/35E0zSQkXwaZ5LVI28PUi3++lgzM5dhb5K/gaqvilCJ2ICas8Li/fglc/3B71v+vRRmJY9HVn+EiN22qShcz+HSOe5IO0NIfdecGbxvnFUTPQmjA9kb8YbBUt9ruk7bHj+o91gK7qU4/lBviOxhNLfdgeH17282GpNwqbv0nYv8fxdxLTCvOrvnt+OGtd62PVpJ70MFeTVAKmhAEK8Se8poIde38rrmCwYCwqW+PRIbIc38IYjADIYKy5R1vu9T9/hXqQ8nfuN5+8ith2XqIT1WF/KX9EphbQfmePUTtFxtAoDLlTye0cl1kCF43BygS4s/FDY+BWHd6wanThXCgtXhPDC6SeP0N4bX6dAae1qZJU2Pa5WVwpLzIs+bDMWFCwBaRV/ktTjb/B/y4BLKQrGr+4JnV7Sm2wwCrCAcZ8l4d1rBuLvCkIoXLO36a2rbAamVXhG13KeBwykiAJoNQQ3f/OfAtom0kNnutdZvg4v5eGdKgNgevb3mJGzAUp0zTe7X216iPOAxX4AYzQHAmQIN+oI6tIgwkXq9ubQPTB3shcqhY22+DZsPfDqFry/OrStfI+j/+HPX+A1PkK58PYvZXuYc4hDofI67D2r/TT+kCV+WuZqlWsK0XmOf4QTbE2BEIK5YQ9eFwghGKj2mh4RWCSPvbUdjy74MWw4RfkPvNfFrpfwcZmyBmMygo8OvWkQ4FcsHHzHZA607Yo63VCkhAII1amP9jPx7mmcrmnz/O0/DHX3brNY753Hrrj8O1Jd2A6uEyWP94d/vQCviJxeBw0J0guOsPfR2BBaAbjL7HJ1+PMOQrH7mLSmhABAvKYEP9/kW37EYQ2YbvPmxNk2yeVxyRRhxDAuJfk2hElNRms57snqWrwUslvaKUJxqg8s572ezQrf4Q8E3509MuMwb+dAw9p8jsw0W6LfrW7mWYu6wiH9hkBvUkIBqJzBRwDRDhVfzv8C6YwNOUwHLnaENx3U8pgBahXBNxA5Tu4QJY/3dM61aZUAQo8iTMueQgHX1WDnMl1mZ73bhPV+/Kk9K96CRkq27K9ByZP/jSjurZm7Ma1zp/Tx/b6LdB3fvIqnc9cGjftizjJUL5kl+sQ6Oc7V5TgiqiH1RbppB4XDt+7tOBxe6dc0d/WE538Z/WY0b+xn9oiOE2yK9NGc76IVx4fDZ8RuLo2elFAAoea4M5jovYGer2j2zKfz4eC8p4a8ZJFhmlLDhN7tGY4BnUoDADKskS2mSbqmEEH3V+w0njc9FXpc1ekS4SmvtZXGNjO4hlMh42axVuRaanHQPU0UwRSa98JiNDTpvdwdxMD0Mih+ZdButOE/G46FjPJzhdsSj8B5TloFYNn0js/vHoi80e2n9IorwL9TIpISCiBUQ5vFdFWUWzW7kMXwn9saCv91BH+81yAYhqCxtQPT5m5GuzG88rF0iFuk5F0AFvFtem9g09giqxz886+RVRCbg/NRKEKUi8PB4cb04NMbyggsUma+J3xhmxPlyZNg294zuCfzR2gYC45WBJY53/GAfHy64UTXwjivmw6hIsnXmJ2vbILy2Lehs+/899eqakzPDm09Bri+iXCq9tEFP6KpLbBuC1kryBTQqQrtXkQYUnRGxZIaCiAErNeXMzT9BObk888phkJMdWFBPH7hhSwSEpt4hRSQRtQpRE+kg50tB3ynDDIEVMbfmHZgvEa450hA2jLKMLimwIS0ow4Hh7qfvsPAtAqMzHA5wauv8R3BCD1L94d9NR7X4HHt9Xvj9+JHZhz26WTw4Z6yzOFxw8LHmwWf4rdpZ0KGMZrtKD0q7vQyN3dnhZ+GHSrQZDsUsViT8Sc1FIDMLhfC9T98LYJiWzOdHCfZ5jah8D2j9zUNF/nh4NenlYc93yDUonowpPxCup41fKq7fw5slDK/mRW1DARd36VoU0KJtAfHkSCaNXT61Y1GjMo4gDsyf5JEDoHZxo3zFcEty7yp5wrCBxJJijiDk1cBhJsC8kYBAkIAFYR6e4zuq928rwY2Oweoo0pGFOGUXKFT+Ga769PKfSw0rlZX4YFXt0QqWgAPZW9EX0ULqp1SVq5OeQV0PG7RhD8cJlzv1s2o9INQNBUAPa72afRb9FZkC0rBRX1LBwpFhA/Gym1ncKU6sAwGpYXeNHXmzDncmxdiv0WEfL29AsN5XrN115eS5yWGLNaCaRlbwoZziN3RJ4DUGAHIrAC8TQf5+KWqa0ivYhzYufsYXiv4HAPTwrt6iLYztjeKnc2Rwm9nH9mD/MLv0PGeivDz4WJ2gv5SVYcs1hqwkSgamM5T0cQ+cRr494N4e5YNxWjNQWRuezPguo5n7jsUBomOST1WFdnI8x95KyXJXyhiTa3DEW6Doj/jNXs9hgexJjUUgMybAsNNAeV6zWVO1OyBtTWyuchIOBniEA25yOaZu5XqFaj93etKhBSuCtywB7/u/EtcQ+DdUYgW785ivKaA2oxW7D0p/V6NRMe0g38DYjB6CejUAPLsbk4JBSD3pnAxU0BZrNVzOEZ3JZdXAUj7zP0VuqhNXuUi1M7uUOQrot9N6oEQqN3unEU26EL8VdW3hJdVb+hAXXP0RgxJx8/B3WhEgxzrhymhAOTG215cCKyI3makjYmbWC86A/wb26SUo6+iGU/kfouHs+WpaNGSr3CZ+fLt7IwVBMCwdJfn2aKO0PsX/LmwvSxsmFnvh1+gZUUfQUOJNamhABLs4JWR6dIvcAXD5UI4/tVQyjfgVrh9lbGzbhqRfjh8ID/2nRRm3eHNr1US7aL2euUqzopDIY73jISebJsAERKn3v1WLe9hO7GATgFFSsIYRbu4SBW7hdnr0k7z+u2PNUqZ5u7lwt9Ka5xGnFuMXyhrMTkCM8Y/ZfM7NxOD1eZEm7HrnWc7mrHkK3EuRcIxK2912DD9g3h3jQd3Z+3ArRp5HaslIwyR298oD2vWrMG7774Lh8OB++67D3fddZfguM3NRtG+Vg5+8S4uaqcvn5Ia/KXlXlwucBet3FQ78mM6UuvO1JFC/GL666LjsSyDwsIs3nsx3wfQ0NCA+fPnY+XKlVCr1Zg8eTKuu+46XHLJJbLlSRt/SiqxoGBJ+EAxgjb+0qFJk765jvkUUGlpKQYNGoS8vDxoNBqMHDkS69evj7UYFAqFklR09L5G8jRjrgAaGxuh1Wo9v4uKitDQIN0mHD7OOqTY10ihUCjx44IbJ0meZsyngDiOA+NllUMI8fkdjmBzWaEY9vy/UH6uDSzDoLndDJVSgfQ0BVRKFieqWtG7RyZsdicy0pQwmu0ozMmAQsGAZRjUt5igVLBgWQa6lg7k56Sjw+JA355Z6LA4YDLZYOc42B0civI1MFsdyNKoUNdkgtXuRM98DRwcgdXmBMsyaDNYka5WQK1iPUfeZaarcH6vbLQbbSjMdaWvSVdCb7LBZLbDZnciJysNutYOEALkZ6dBrVKgRmeESqlAj7x0VDcawTBAfnY6ehVqUN1ohMPBwWSxQ61SID87HYQQ1OiMOL9nNlr0FuTnpKPNy1NptkYNu8MJjgBpKgVMFjtUShbpagVsdg7GDhtyMtOgVDLgCGAy26FgGfTrmY3G1g4UdJaN1eaE0WyDSqnAhefloL65A1a7E7rWDmRr1LDZncjLTvM8l95oRc/CTFisDlisTpynzYRCwaKl3YJzjQbkZaWhf+8cmMx2VDcaYbM7QUCQplKiKD8DPQs1qKrTQ6VUgCME9c0m5GSmobnNjKsu1aK8ug0ZagUYhkGrwYKsDBUsNifS1QqolAo4nBw06UoADM7UtCMjTQG200ugw0mgVrFwODhkpiuRnqaCJl0JpYJFi94CtUoBi9UBtUoBJ8chI00Fu8MJq80JlZKFQsHiXIMBTo7goj65MFvs0OZrALiOM1QqWWjSVKhuNIBhAKWChdXOwWS2w+5w4rL++WjRW2F3cEhXK5CbpYYmXYWz9Qbk56ShutGIjDQlmtvMyNSokJ+VDn2HDQwD2OxOaPM0aDNaYbM7kZulhoJlQQhBmloBi80J0vkeexZqUFHTjqwMFRQKFmqVAiazHXnZaWAZBjU6I3Kz1GhptyA3Ow0Mw8DYYcMlffNwrsGA3Kw0KJUsGLjS69UjExU17ThPm4VzDQYoWMZT5pp0FXoWapCjUaOiTg+VgkVmhhLVjUakqxWwWh3osDlRlK+BocMGh5Pg0n55YBkGDo4D25l3s94Cm51DVoYKLAucrm7HxX3zkKVRQW+0wmh2oGeBBq0GC9JUCrQZrMjLToOCZdC7RyYOlOuQkaaESqmAgmVgMtuRm5WGjDQlOI6gqd2M3j0yYXdwyEhToqXdglaDBQqWgdXmxHnaLCgUDHoWaFBRq0e70QptvgYWqwMEgErBwuZw4qLzclHbZITNzkHBMmg32dCrUIPmdgvS1UpY7Q5kZaiQlaGG3mSFyeJAUX4GWIZBU5sZrILFVZf2wHk9xLd94Yj5IvCqVatQVlaGOXPmAADeeecdEELwyCOPCIofySIwAGi12dDpgh+8kigkg5xURulIBjmTQUYgOeSMh4yhFoFjPgV0/fXXY+fOnWhpaYHZbMaGDRswdOjQWItBoVAoKU/Mp4B69uyJxx9/HPfeey/sdjsmTZqEK6+8MtZiUCgUSsoTF3fQJSUlKCkpiUfWFAqFQukkNXYCUygUCiUAqgAoFAolRaEKgEKhUFKUpDsSkmUj94gXTdxYkgxyUhmlIxnkTAYZgeSQM9YyhsovLs7gKBQKhRJ/6BQQhUKhpChUAVAoFEqKQhUAhUKhpChUAVAoFEqKQhUAhUKhpChUAVAoFEqKQhUAhUKhpChUAVAoFEqKQhUAhUKhpCgpoQDWrFmDMWPGoLi4GEuXLo15/m+//TbGjh2LsWPHYt68eQCAZ555BsXFxRg/fjzGjx+PjRs3AgCOHTuGiRMnYuTIkXj22WfhcDgAALW1tbjrrrswatQoPPTQQzCZTJLKeM8992Ds2LEeeQ4ePIjS0lKUlJSguLgY8+fP94SNl4zLly/3yDd+/Hhce+21ePHFFxOmLI1GI8aNG4fq6moAkKz89Ho9HnjgAYwePRp33XUXdDqdZDJ+8cUXGDduHEpKSvDMM8/AZrMBcH2zw4cP95Spu97EQkY+OaV6x3KV5datW32+zUGDBmH69OkA4l+WISHdnPr6ejJ8+HDS2tpKTCYTKSkpIeXl5THLf8eOHeSOO+4gVquV2Gw2cu+995INGzaQcePGkYaGhoDwY8eOJfv37yeEEPLMM8+QpUuXEkIIeeCBB8jatWsJIYS8/fbbZN68eZLJyHEcueGGG4jdbvdcM5vNZNiwYeTs2bPEbreTadOmkS1btsRNRn9OnjxJRowYQZqbmxOiLA8cOEDGjRtHrrjiCnLu3DlJy++FF14g//rXvwghhKxatYr85S9/kUTGM2fOkBEjRhCDwUA4jiNPP/00+eSTTwghhEyfPp3s27cvIA25ZeSTkxAi2TuWqyy9aWxsJDfddBOpqKgghMS3LMPR7RXAypUryTPPPOP5/fbbb5O33norZvmfPHnS84ES4nq5ixcvJtdccw25//77ybhx48iCBQuI0+kk1dXV5KabbvKE3bNnD7nnnnuIzWYjAwYM8DTQtbW15MYbb5RMxlOnTpEbbriB3HPPPaSkpIR8+umnZNeuXeTee+/1hFm1ahWZOXNm3GT0Z8qUKWTdunWko6MjIcpy1qxZZM+ePWT48OHk3Llzkpbf8OHDSW1tLSGEELvdTgYMGEBsNlvUMlZXV5Pt27d77n/44Ydkzpw5hBBChgwZQqZPn07GjRtHXnjhBWKxWGIiI5+cUr5jucrSm7/+9a/kgw8+8PyOZ1mGo9tPATU2NkKr1Xp+FxUVoaGhIWb5X3rppbj66qsBAJWVlfj222/x+9//HoMGDcI///lPfPnllygrK8OKFSsCZNVqtWhoaEBrayuysrKgVCp9rkuFXq/H4MGD8c4772Dx4sVYtmwZamtrecstXjJ6U1paCovFgtGjR6OpqSkhynLOnDkYOHCg53ew7y4SubzjKJVKZGVloaWlJWoZ+/TpgyFDhgAAWlpasHTpUtx0000wmUz41a9+haeeegqrVq2CXq/HokWLYiIjn5xSvmO5ytJNZWUldu/ejXvvvRcA4l6W4ej2CoDjODBMlztUQojP71hRXl6OadOm4emnn8ZFF12Ed955B0VFRcjIyMA999yDrVu3BpWVT2Ypn2HAgAGYN28esrOzUVBQgEmTJmHhwoW8ssRLRm+WLVuGP/zhDwCAfv36JVRZugmWvxRyEULAstJV3YaGBtx333249dZbcd111yEzMxMffPABLr74YiiVSkybNg1bt26Nm4xyvmOpy/KLL77AlClToFarASDhytKfbq8AevXq5bOIotPpUFRUFFMZ9u7di6lTp+LJJ5/EhAkTcOLECXz33Xee+4QQKJXKAFmbmppQVFSEgoICGAwGOJ1OWZ6hrKwMO3fu9JGnT58+vOUWLxnd2Gw27NmzBzfeeCMAJFxZugn23UUiV1FREZqamgAADocDJpMJeXl5ksh5+vRpTJ48GRMmTMDDDz8MwLU4uWLFCk8Yd5nGS0Yp37GccgLA999/jzFjxnh+J1pZ+tPtFcD111+PnTt3oqWlBWazGRs2bMDQoUNjln9dXR0efvhhvPbaaxg7diwA10fwz3/+E+3t7bDb7fjiiy8wYsQI9OnTB2lpadi7dy8A4L///S+GDh0KlUqFgQMHYt26dQCAr7/+WtJnMBgMmDdvHqxWK4xGI1atWoUnnngCFRUVqKqqgtPpxNq1azF06NC4yejmxIkTuOCCC6DRaAAkXlm6ueqqqyQrv2HDhuHrr78GAKxbtw4DBw6ESqWKWkaj0Yj7778ff/nLXzBt2jTP9fT0dLz66qs4d+4cCCFYunQpRowYERcZAWnfsZxytrS0wGKxoF+/fp5riVaWAciyspBgrF69mowdO5YUFxeT999/P6Z5v/TSS+Tqq68mN998s+f/n3/+Ofnss8/I6NGjyYgRI8irr77qCX/s2DFy6623kpEjR5InnniCWK1WQggh1dXV5O677yajR48m06ZNI21tbZLKOX/+fDJq1ChSXFxMFi9eTAghpLS0lJSUlJDi4mIyZ84cwnFcXGUkhJBvvvmGPPbYYz7XEqksvRcFpSq/1tZWMn36dDJmzBhyxx13BCw6RirjJ598Qq644gqfb/PNN98khBCyfv16T52ZOXNmzGX0lpMQ6d6xXGVJCCEHDx4kt912W0CYRCjLYNATwSgUCiVF6fZTQBQKhULhhyoACoVCSVGoAqBQKJQUhSoACoVCSVGoAqBQKJQUhSoACoVCSVGoAqBQKJQUhSoACoVCSVH+H+b7ihNsswkxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD7CAYAAACBiVhwAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAABQb0lEQVR4nO2deWBcZb33P2ebLZlszaQtbWmhFKgs5bJcKOVtBbECpai11wtexfUqigXxumBBuYIgKFqvKC6IuOGCCFJKLYtApRTFVqB0YSltuqZJmn32mXOe948zZzKTpGmSTmZ9Pv8kM5mceZ4zZ873+a2PIoQQSCQSiUSSgVroAUgkEomk+JDiIJFIJJJBSHGQSCQSySCkOEgkEolkEFIcJBKJRDIIKQ4SiUQiGYQUB4lEIpEMQi/0AHJFV1cIyxpZycaECdV0dATHeUTFRyXOuxLnDJU570qcM4x93qqqUF9fdci/l404WJYYsTg4r69EKnHelThnqMx5V+KcYXzmLd1KEolEIhmEFAeJRCKRDKJs3EoSiaS4iURCBIPdmGZyXI7f1qZiWda4HLuYGX7eCi6Xh/r6AIqijOq4UhwkEsm4E4mE6Ovroq4ugGG4Rn2jGgm6rpJMVp44DDdvISy6uw8SDPbg99eN6rjj6lYKBoNceuml7N27F4D169ezePFiFi5cyIoVK9Kv27ZtG0uWLOFd73oXN9xwA8nk+KwsJBJJYQgGu6mrC+ByucdFGCRDoygqfn89kcgYspnGYTwAvPLKK1xxxRU0NzcDEI1GWb58OXfffTerV69m8+bNrF27FoAvfvGLfO1rX+Pxxx9HCMEDDzwwXsOSSCQFwDSTGIar0MOoSDRNx7LMUf/fuInDAw88wE033URTUxMAmzZtYvr06UybNg1d11m8eDFr1qxh3759RKNRTjvtNACWLFnCmjVrxmtYEklOuf3+f/H8qy2FHkZJIC2GwjDW8z5u4nDrrbdy5plnph+3tbURCATSj5uammhtbR30fCAQoLW1dbyGJZHklLf29bCnrfIKr0qd73znDj7ykQ/wwQ/+B29/+zl85CMf4CMf+QCPPbZyxMcIBoN85StfOOKxbN26mbvv/v6Y/relZT/vec+iIx7DUOQtIG1ZVpaCCSFQFOWQz4+WCROqR/X6QMA/6vcoBypx3uM5Z0sI3B6jKM9rMY2prU1F18c/c36k7/HlL38FgP379/OZz/w3v/nN70f9XuFwkDfffP2I57V7dzPd3V1jOo6m2f9zuP9VVXXU10PexGHSpEm0t7enH7e3t9PU1DTo+YMHD6ZdUaOhoyM44irBQMBPe3vfqN+j1KnEeY/nnC0hEAKCoVjRnddi+6wtyxr3TKKxZCuZpv165//27t3DnXd+k97eHtxuD9dd90WOP/5EnnhiDb/97a9QVZWjjjqKr371Fr7znTs4eLCdL37x83zzm3dmHfcnP/khGzf+k97eXhobG7n55m/S0DCBJ55Yw69+dS+gMHv22/jMZ67lpz/9EZFIhHvvvYdAoImXXtrIDTf8LwCf/ewn+djHPsmpp57Gd75zOzt2vEVnZyfHHXcc//u/tw4a/6GwLGvQ9aCqyrCL6ryJw5w5c9i5cye7du1i6tSprFq1ive9731MmTIFt9vNxo0bOeOMM3jkkUeYP39+voYlkYwZZzFSqS0bjoTnX21h3abcxmoUBYSA806dzLxTJo/pGLfeehPXXfcljj/+RHbu3MHy5V/gd797iHvu+RE//el91Nc38MMf/h+7dzfzuc99kWXLPjVIGPbu3cPu3c38+Mc/R1VVbrnlazz++F+48MKF3HXXd7n33l/T1DSRW275Kq+++gqf+MRVvPTSRj784Y+zevWjQ45r8+ZN6LrBT35yH5Zlcc01V/HCC89zwgmzxzTPkZA3cXC73dx+++0sW7aMWCzGggULuOiiiwC48847ufHGGwkGg5x00klceeWV+RqWRDJmzJQomFIcyoJwOMy2bVu57bab089FIhF6erqZN+//8elPf5z589/OggUXMGvWCbS07B/yOFOnTuOzn72ORx/9M7t372LLlleZMmUqmzdv4pRT5tDUNBGAr371FoBDCkImp512OjU1tfzpTw+we3cze/fuIRKJ5GDWh2bcxeHpp59O/z537lxWrhwc8DnxxBN58MEHx3soEklOkZbD2Jl3ythX94fiSIvgLMvC5XLzi1/8Nv1cW1srNTW1fO5zX2D79nfzwgvruOWWr6ZdPUPx2mvb+N//vYHLL/8A55//DjRNRQiBrutkhlO7uroG/a+iKAjRfz051eTr1q3lZz/7Cf/xH5dzySWX0d3dnfW68UD2VpJIxoglpDiUE9XV1UydOo3HH18NwD//+XeuvvqTmKbJ5Ze/l7q6Oj70oY9y0UWLeOON19E0DdMcXD/w8ssb+bd/O4P3vGcp06Ydzfr167Asi9mzT2LLls10dBwE4K67vsu6dWuzjlNbW8euXTsRQrB//z62b98OwIYNL3LBBReyaNFlVFdX89JLG8dUuzAaZPsMiWSMSLdS+XHTTd/g29++jd/+9lfousHNN9+Grut8/OOf4nOfuxq32019fT033PC/+P01TJw4iWXLPsVdd/0kfYx3vGMhy5d/kSuv/E8ATjhhNi0t+2lsDHDttf/D5z+/DMsyOfnkU7nkksXs27eXn//8p/zoR3fx8Y9/iscee4Qrrngf06dPT1snixe/l69//QaeeupxdN3glFNOZf/+/ZxxxvidC0WMt22SJ2S20uGpxHmP55y7gzE+/4PnOeP4AFcvOWVc3mOsFNtnfeDALiZNmj6u7yF7Kx2aoc7/4bKVpFupxOkJxfnuH14mGEkUeigVhyUtB0kZI8WhxNnT2sfmnZ3sa5dVuvnGEQWrPIxviSQLKQ4ljikzZgqGtBwk5YwUhxJHBkULhxRmSTkjxaHEkavXwiHPvaSckeJQ4sjVa+Hot9oqL0NGUv5IcShx5Oq1cMgiOEk5I8WhxEmmVq0yYyb/yHhP6VJM+zkcjnvv/Qn33msX2X3kIx8Y9rW33fZ1DhzITUNDWSFd4kjLoXDI3kqly//8z5cBe7OcZcs+ldVPaaT09fXy5puv53pow3K4cf7rXxv46Ef/OyfvJcWhxEmLgylvUPlGWg5jJ/HG8yRe/1tOj+k0rTNOmI9x/LwxHWM0+zl873vf5uDBdr7ylS9kte1uadnP9dd/nqOPnsHOnTuYNGkSX/vaLdTU1HLppRdywglvo6PjID/72a/43e9+wzPPPIlpWpx99jl8+tPXoCgKv/3tr1i58mFqa+vw+/3Mnn0SAOeddybr1m2gt7eHb37zFnbvbsblcvHZz17H1q1bUvtLXMsPf3gPtbV1R3Q+pVupxJGFWIVDWg7lx6233sRnPnMNP//5/XzpSzdw003LAbjnnh+xYsUP+PnPf8PkyVPS+zk0NgYG7ecA8NZb23nve5fym988wPTpx/Dzn/8UgO7ubv7rv67kF7/4LRs2vMjrr2/jnnt+xX333U97eztPPPEXXnttK489tpKf//x+vve9u2lvbxt0/Hvu+TFTp07j/vsf5KabbuGnP72bD33oIzQ2Bvj2t//viIUBpOVQ8ki3UuGQlsPYMY6fN+bV/aE40t5KudrPAWDatKM5/fQzAbj44kv5+tdvSP/tpJNOBuxOq1u3bubjH/8QALFYlIkTJ9HR0cE558zD5/MBcP75Fw7q/vryyxu56aZbATjuuFn85Cf3jXneh0KKQ4kjU1kLhyWttrIiV/s5AGha/61VCCvrsdvtSb2fyfvffwWXX/5BAPr6+tA0jUceeQgQGcca3Brc3huif3OIXbuamTbt6DHN+1BIt1KJI1evhUOe+/IiV/s5AOzZsysdrH7ssUc555xzB73m9NPP4vHHVxMOh0kmk3zlK//Ds8/+lTPPPIvnn3+OYDBILBbjb397ZtD/zplzOk899TgAzc07+Z//WYaiKMOOabRIy6HEsWQhVsGQdQ7lRy72cwDw+2u4996fsHfvXmbOPI7rr//qoPc677z5bN/+Bp/85EewLJOzzz6Xiy++FEVR+I//uIJPfOJK/H4/EycO3jHv4x//FHfc8Q0+/OEr0HWNr371ZhRF4dxz/x9f+MK1fPe7d3HUUVOO6FzI/RxKnIf/toNH1zfzvgXHsmjujGFfW07zHinjOee/bz3AT1dupcqjc9fn5o/Le4yVYvusK2k/Byc99sEHD783dC6Q+zlIhkS6NgqHTAaQlDNSHEocmU5ZOGQygGQoJk8+Km9Ww3gixaHEkZZD4ZCWg6SckeJQ4sgbVOGQVttoUBCi8PGASmSsYWUpDiWOk6Ukb1D5xxFkgTz/h8Pl8tDdfZBkMjHmm5Vk9AghCIV60XXXqP9XprKWONKtVDgyBcG0BKqqDPPqyqa+PkAw2ENnZyuWlZs8/IGoqopVgSndh5u3rruorw+M+rhSHEoc6dooHGbGClie/+FRFAW/vw6/v27c3qPY0nfzxXjNW7qVShznBiUth/wz0HKQSMoJKQ4ljqyQLhyZgiD7K0nKDSkOJY6zj4N0a+QfaTlIyhkpDiWODEgXjizLQZ5/SZlREHF45JFHWLRoEYsWLeKOO+4AYP369SxevJiFCxeyYsWKQgyrJJHN3wpHtuUg3XqS8iLv4hCJRLj11lv59a9/zSOPPMKGDRt4+umnWb58OXfffTerV69m8+bNrF27Nt9DK0mk5VA4pOUgKWfyLg6maWJZFpFIhGQySTKZpLq6munTpzNt2jR0XWfx4sWsWbMm30MrSUzTXrFKccg/MuYgKWfyXudQXV3Ntddey8UXX4zX6+Wss86ira2NQKC/SKOpqYnW1tZ8D60kkXUOhUPWOUjKmbyLw2uvvcaf/vQnnnnmGfx+P1/4whdobm7O2vJOCJH1eCQM15d8KAIB/6heX6yomm38abo6ojmVy7xHw3jN2eXq//rU1vmK7twW23jyQSXOGcZn3nkXh3Xr1jF37lwmTJgAwJIlS7j33nvRNC39mvb2dpqamkZ13Erd7CcWTwIQjSYPO6dymvdIGc85h8Lx9O8HO4JUG8WT/Cc/68phrPMuus1+TjzxRNavX084HEYIwdNPP82cOXPYuXMnu3btwjRNVq1axfz5xbWzVrHi1DmYsggr78iYg6ScybvlcN5557F161aWLFmCYRiccsopLFu2jHnz5rFs2TJisRgLFizgoosuyvfQSpJ0+wxT3pzyjSWzlSRlTEEa733yk5/kk5/8ZNZzc+fOZeXKlYUYTkkjA9KFI9NakOIsKTeKx0kqGROyzqFwZLmVpFtPUmZIcShx0paDvDnlHVkEJylnpDiUONJyKByZgizPv6TckOJQ4qTFwZS9ffKNaQkM3f4KSctBUm5IcShxpFupcFiWwNCkOEjKEykOJY50KxWOTMtBnn9JuSHFocRxWkXLlWv+saRbSVLGSHEocWSdQ+GQloOknJHiUOI4N6WkvDnlnayYg4z5SMoMKQ4ljCUEzj1JWg75R1oOknJGikMJI3v7FBZLyJiDpHyR4lDCOKtVQ1cxLYGQro28YloCXVoOkjJFikMJ46xWHb+31Ib8YlmWrHOQlC1SHEqYTMsh87EkP1hZMQdZoS4pL6Q4lDDWIHGQN6h8IttnSMoZKQ4lzEDLQd6g8ktmKqu02iTlhhSHEsaxFFy6lnosb1D5xLQEqqqgKoo895KyQ4pDCTPQrSQth/xiCVscNE2R515SdkhxKGFkQLqwmJZAUxVUVVoOkvJDikMJI8WhsFgpt5KmSMtBUn5IcShhpFupsGRZDrLIRFJmSHEoYRxLwSUth7zj9LVSFQVNlZaDpPyQ4lDCSMuhcDjnWsYcJOWKFIcSJh1z0GQqa75xxEFVpeUgKU+kOJQwaXEwpFsp35hpy0FFleIgKUOkOJQwAxvvyRtU/nA293EsBynMknJDikMJkw5IG7K3Ur4xB8QcpDBLyg0pDiWMtBwKR1bMQbbPkJQhUhxKGMdSMIxUQFrm2ueNgdlKcg9pSbkhxaGEMQdYDqYpb1D5wjn3Tp2DtBwk5YYUhxJG1jkUjkF1DqaM90jKi4KIw9NPP82SJUu4+OKL+cY3vgHA+vXrWbx4MQsXLmTFihWFGFbJISukC4cp6xwkZU7exWHPnj3cdNNN3H333axcuZKtW7eydu1ali9fzt13383q1avZvHkza9euzffQSo6BdQ7S750/nHMteytJypW8i8OTTz7JJZdcwqRJkzAMgxUrVuD1epk+fTrTpk1D13UWL17MmjVr8j20kmNgtpK0HPJHdoW0Ki0HSdmh5/sNd+3ahWEYXHXVVbS0tPD2t7+dWbNmEQgE0q9pamqitbV1VMedMKF6VK8PBPyjen0x4vO5AGhKzcXncx92XuUw79EyHnPujiYBqK/34fUYhGLJoju3xTaefFCJc4bxmXfexcE0TTZs2MCvf/1rfD4fn/70p/F4PCiKkn6NECLr8Ujo6AiOePUWCPhpb+8b1fGLkZ7eKADBvmjqcWTYeZXLvEfDeM25oyME2Oc+mTSJx5NFdW7lZ105jHXeqqoMu6jOuzg0NjYyd+5cGhoaALjwwgtZs2YNWqp5HEB7eztNTU35HlrJIQPShSMzIC27skrKkbzHHM4//3zWrVtHb28vpmny3HPPcdFFF7Fz50527dqFaZqsWrWK+fPn53toJUe6CE6msuaddCqr3M9BUqbk3XKYM2cOn/jEJ/jABz5AIpFg3rx5XHHFFRx77LEsW7aMWCzGggULuOiii/I9tJJjYJ2DXL3mDzOj8Z4q22dIypC8iwPA0qVLWbp0adZzc+fOZeXKlYUYTsni3JB0TTbeyzeZRXCabJ8hKUNkhXQJYwmBqijomh28l66N/CFjDpJyR4pDCWOaAk2z8+xBupXyySDLQZ57SZkhxaGEMS2Bqio4Wb/yBpU/rAGWgzz3knJDikMJY1kCTVFQZGfQvGMOsBzkuZeUG1IcShhT2JYDIP3eeUZaDpJyR4pDCWOaAi1DHOQNKn+YA+ocpDBLyo1hU1mvuuqqYf/5xz/+cU4HIxkdlmUHpAF0eYPKK9YQdQ5jafsikRQrw4rDu971rvTv3//+97nmmmvGfUCSkWNadiorSMsh3wyMOQAIAVIbJOXCsOLw3ve+N/37L3/5y6zHksJjiWy3krQc8sfAmAP0Z49JJOXAiGMO0lwuPjJvRjLXPr8MZTnI8y8pJ2RAuoQxTStdAGf7vWX7jHyRvdlPv+UgkZQLw7qVuru707+bpklPTw8io4dMXV3deI1LMgIsq9+tpGmqvDnlEUeItQy3kuyvJCknhhWHc845B0VR0oJw9tlnp/+mKArbtm0b39FJhiWzzkG6lfKLtBwk5c6w4vDaa6/laxySMZBpOci20fkl3XhPybAc5PmXlBHDxhwuv/xyfvzjH7N9+/Z8jUcyCjKL4KTlkF+cjriKkpmtJGM+kvJhWMvhnnvu4ZlnnuGuu+5i9+7dnHPOObzjHe/gjDPOkNlLRYApBEZqLweZyppfBmaKgbQcJOXFsOLg9/u57LLLuOyyy4jH47zwwgusXLmSm2++mZNPPpnbbrstX+OUDIFlCTTDCUhLccgnWS49GXOQlCHDupXi8Thf/vKXefLJJ3G5XCxYsIDW1lZmz57NkiVL8jVGySHIWr0q0q2UT7ItB7mHt6T8GFYcvv/97xMMBjn99NPTz91888309fXx3HPPjfvgJMMzsPGeKVMp88bAZACQloOkvBhWHJ599lm+853vMGHChPRzEydO5Fvf+hZPPfXUuA9OMjyZ7TNkQDq/WEPFHKQ4S8qIYcXBMAw8Hs+g56urq3G5XOM2KMnIyHRtqKqCacqbU74wZcxBUuYMKw6qqhIMBgc9HwwGSSaT4zYoyciwLCvLcpA3p/xhZXTETRfBSXGWlBHDisOll17KjTfeSDgcTj8XDoe58cYbWbhw4bgPTjI8A10b0q2RP0wx2HKoFLfe9r09/GzV1qxWOpLyY1hx+PCHP4zf72fevHm8//3vZ+nSpcybN4+amhquvvrqfI1RcgiSluhvvCcth7wyVMyhUhICtjR3sn7zAWIJs9BDkYwjw9Y5qKrKLbfcwlVXXcWWLVtQVZVTTz2VpqamfI1PMgxZjfdUBUtW6OaNoWIOlWI5xJNm6qeFR4Yey5ZhxcFhypQpTJkyZbzHIhklliWoMnsI/XE5Pve7peWQR4a0HCrk/CcSVtZPSXki93MoYUxLUJc8iNW1nzqzq2JuTsVAJbfPiCet1E/pVipnpDiUMKYl0LGzxgySFXNzKgaGap9RKec/kRIH56ekPJHiUMJYGeKgK2bF3JyKgaEsh0qx3BIpi0GKQ3kjxaFEEUJgWgIjw3KolJtTMWBZAk2pTMuh360kxaGcKag43HHHHVx//fUArF+/nsWLF7Nw4UJWrFhRyGGVBE7WpCZSloMUh7yStQtfhfVW6ncryZhDOVMwcXjhhRd4+OGHAYhGoyxfvpy7776b1atXs3nzZtauXVuooZUEzsYyabeSkG6lfDJkzKFC6hzSqawyW6msKYg4dHd3s2LFCq666ioANm3axPTp05k2bRq6rrN48WLWrFlTiKGVDM4qVRf2F1UngWkJWbWaJyo75pCyHEwpDuVMQcTha1/7Gtdddx01NTUAtLW1EQgE0n9vamqitbW1EEMrGRwrQRMJAPSUe0lqQ36Q2UoyIF3ujKgILpf88Y9/ZPLkycydO5eHHnoIsBvIZW47KoQY9TakEyZUj+r1gYB/VK8vNnqCMQA8hkj9tJ9vmFCFoWuH/L9Sn/dYGI85K6qC12sQCPjxRmyB9vpcRXV+x2ssjoXkchtFNV+ozOsbxmfeeReH1atX097ezrvf/W56enoIh8Ps27cPTeu/obW3t4+6RUdHR3DEK7dAwE97e9+ojl9sOOJgxZ2fUQAOtPbicQ39sZbDvEfLeM05HjdJJkza2/uIxm2rrbc3WjTndzw/62jMnm9Xd7ho5guVeX3D2Oetqsqwi+q8i8N9992X/v2hhx7ixRdf5Otf/zoLFy5k165dTJ06lVWrVvG+970v30MrKZzVm2Y5biU79lApro1CM3T7jMpwszixBpnKWt7kXRyGwu12c/vtt7Ns2TJisRgLFizgoosuKvSwipq0OKRiDU7soVKCooXGHKLOoVLOvZOlJGMO5U1BxWHJkiUsWbIEgLlz57Jy5cpCDqekcCwENWU5qCmRkJZDfrAy6hycTX8q4dxblkiLoOytVN7ICukSxfmCOqLguJcqZfVaaJyW3ebBXRALoSqVsZ9GpiBIy6G8keJQoqTFwYrbP1MiUQk3qGLAiTmEV91O/NXHUVWlIiyHzDiDLIIrb6Q4lCj9bqVsy6ESblDFgB1zEBCPIKLBitnDO5khDrIIrryR4lCiODciJR1zkG6lfGJZAhf2ORfJWEVaDgm5TWhZI8WhRBkYkFYsGZDOJ6YlMBT7nJOI2ZZDBZSnxzMEoVhTWa1QF8ldLxV6GCWPFIcSxcmpV8yU5SAD0nlloOWgVYjlkBmELtaAdGLbM0SeuAtRIXUn44UUhxJlkFspJRKV0hm0kAghsIRI97MiGUetkJiDYy24XVrRprKKeASEBclYTo/bG4qTrKA4ixSHEsWyBCoWikhZEI7lYJb/DarQOAJsOJZDovIshyqPXrSWAwlbFEQOxUEIwY0/+wdPbdibs2MWO1IcShTTErgcn7fLiyJMFKyKaeFQSBwRSHfCraCAtLPBT5XHKFpxEAm7z5gjErkgkbQIRhK0doVzdsxiR4pDiWJZAgP7i6q47eZZBnLDn3zguI+cLVpJxRwqya1U5dGLNiDtWAy5tByiqUB8XziRs2MWO1IcSpTMbBnFXQWAoZgVkTFTaPoth363UuVYDo44GMW7TajjVsqh5RCLm5zu2gnBgzk7ZrEjxaFEMTMtB49tObiUZEXcoAqNOUAcSMbQKqV9RmoF7fPoRVshnbYYHPdSDojFEnyoah2zoptydsxiR4pDiWJlxBzSlgOmDEjngYGWA5aJrlgVkSnmVEVXeQ1MSxTnYmQcAtKxSARVEWgJGXOQFDm2W2lAzEFJVsTqtdD0t0uPp59zqyZmBaQ5JhL9MQcozlqHcQlIR0IAGFasYpI+pDiUKKZlDXIrGYpZEavXQpPev9vqD066K0SY40kLTVVwG1rqcfHFHdIB6Ry6leLRCABeJU4wkszZcYsZKQ4lylBuJRdmRdygCo0T9NesfsvBo1ZGvCeeNHEZKq6UOBSj5ZC2GHLoVjJjtjvJq8bpC8cP8+ryQIpDiZLtVurPVqqEG1ShSVsOot9ycCnJisgUSyYtDE3F0OxbR7GlsworCak+Y7nMVjIzLYcKSWeV4lCi2HUOKcvB4wSkK8O1UWgG7t8NlZMpFk9aGLqGodu3jqKzHDIEIZcBaTPeLw59ESkOkiJm6IC0dCvlg3RHXLPfveCqEGGOJ62UW8mxHIor5pBlLeQw5mDF7WN5FelWkhQ5me0zZJ1Dfum3HOJgeABwkaiIc59ImBi6iqGnYg5FVusgkv2CkEu3kkiJg6FYhIKRnB23mJHiUKJkts8gs86hAm5QhcYRAcVKoHhrAXBViNUWT1q4dA2XXpwxBxIZq/pcdmXNEJ1osDd3xy1ipDiUKLblYIKqo+huQAak84WZ4VZSvH4ADKVCLAfTSlkOTsyh2NxK42M5ZLqo4uFg7o5bxEhxGCPCsoiu+zVWd0tB3t8SApdqgm6AZgBKqgiuyFZyZUj/LnxxVG8NAIaojJhDIjFQHIrsenOsBZc3pwFpJSO+lIxIcZAMgwh3kdj6V5K7XynI+5umbTkomgtFUUAzpFspTzgpq4oZR/HYloOLREUUIMaTJi5dxaU7RXDFJQ6OtaB4a3MakFYz3EpmLJSz4xYzUhzGiEgVxYh4YXqtmJZjObjsJ3SXdCvlCZGZrWR4QDPQKyRbKeGkshpFajmkBEH11ebUraRZcZxPV8RkQFoyDI4oOCKRb+w9jE2UlDgouqtigqKFxrQECsK2HAw3iu7GEJURc0insupFmsqadCyHmpwGpDUrRkS1swKVeLgirEQpDmPFEYd4YVYRpkilsqYtBwNXhbRwKDSZBYjobjDc6JWSyupUSDsxh2JLZU30i0MueyvpIkFEs12IHiVOOFr+/ZWkOIyRtMVQKLeSaWEoJopmAKBo0nLIF1k1Jnq/5VAM515YJpGn7ibW8ta4HD+RNDEMFU1V0VQl3cK7aEjGQFHtwtBkHCFyMz5DxEnoVViqXjGFcFIcxkjarVQgcbCcCulMy0GKQ16wRIY4GLbloBWJW0mEu0nueJHwjtwnSliWIGmKdDDa0NWi2/BHJKJguO3PBSCZm5u4IRII3YWle1PiUP4tNKQ4jJFCxxxMYbs20jEHrfwD0rsO9HGwu/DBQNMSuJV+t5Kiu9CLxXJIXY9WNPfplk7w2Yk3uHS16OocSMRQDE+6cj0XrqWkaeFWEqB7UFw+KQ7jyQ9+8AMWLVrEokWL+Na3vgXA+vXrWbx4MQsXLmTFihWFGNaoKHS2kuXsIa1lZCuVeSrrjx/ZzJ/+tqPQw0glAziWgwt0d1ocRIEDlSJmi4I1Drn4TvDZiTcYulZ8qazJWEqwc2c5ROOmLQ6GB8Xts9t2R6RbKeesX7+edevW8fDDD/PnP/+ZLVu2sGrVKpYvX87dd9/N6tWr2bx5M2vXrs330EZHKhBdsFRW026foeipmIPuwiiT3kotHSFiicEr0t5wnO6+HFa9jpHMmAO67cJw9nYodBKLs2gxx9Fy6BcHtfjEIRFNu/qcx0dKLJbAoyRRDA+6pwqvkpCWw3gQCAS4/vrrcblcGIbBzJkzaW5uZvr06UybNg1d11m8eDFr1qzJ99BGRVoU4tGcBb1Gg2kJdDKylTQjtdlPcX1ZR4tpWdz8iw38dePerOeTpkUkZtJbBIFAa0BAGt2dbt9dcMstVaA1vm4lLfVTJVlk4kAiZicJpN1KR76YiEbshaDi8qB6qqiqkA1/8i4Os2bN4rTTTgOgubmZv/zlLyiKQiAQSL+mqamJ1tbWfA9tVPRbDCKnlZgjxUrFHBy3kqK70Mtgq8pQJEksYdLZm31OndTBYlixZVkOqToHZ+OfQltuwhGHcXErDbAcDLU46xwMj51iDDmpdUhE7O+65nJiDomK2PBHL9Qbv/nmm3zqU5/iS1/6Epqm0dzcnP6bEMJuCTEKJkyoHtXrAwH/qF4/kL1mzOmJSn21glF7ZMcbLZqqoGNSVVNNQ8DPQX8VQUwMQx92bkc67/EmYtodLxNW9lijVh8AoWiChoYqNG3k65pcz9nnc+FOxRwmTGygd6+fWMqtVN9QRZXXyOn7jYZOLUEMMCPBnM+7M3VDDDRWEwj4qfK6SCStorqmNJHAVV1FXVMD+4Aar0LVEY6vZY8tiv76OqpUQVSJE02YRTXv8RhLQcRh48aNXHPNNSxfvpxFixbx4osv0t7env57e3s7TU1NozpmR0dwxKu2QMBPe3vfqI4/kEQ4aDe8MxN0tLSjxb1HdLxRv3/MXhGF4wKzvY9YHHSShCPxQ84tF/Meb/bs7wGgozucNda9qeeFgJ17uqitco3oeOMx557eaNpy6OxNkogrqMJExaKtvY/qAopDtKsLACsayvm82w/a1kg4FLOPLQThSKJorqlAwE8yGkGYGt199ufT09FN+AjH19XeSTUQt3TClo6GSW93sKjmPZaxqKoy7KI6726llpYWrr76au68804WLVoEwJw5c9i5cye7du3CNE1WrVrF/Pnz8z20USFiIdqTVanf8x+UVsxUJahjPusuNCyEWVxm/mhx3EbBAVsxBqP9j/tChfX3ZtU56G47Y4nUPtJF4lYSiSjCzK3rY5BbSVeLrghuUEA6B26lZNT+fuseL4rLB0AiWv7N9/JuOdx7773EYjFuv/329HOXX345t99+O8uWLSMWi7FgwQIuuuiifA9txAghEPEwBxOTCLi605lL+UQVKZ+31p+tBIBV2r7QUEoEQgPEIfNxoYPS/TEHuxuu4992UfhsMZHRMVTEQii+upwd2yl4y6xziA+RVVZQBgSkcxEPNFON9gyPL92624wGx+T+LiXyLg433ngjN95445B/W7lyZZ5HM0aScRRh0WmlLIcCpLOqqVVhWhScwHSOV4v5xrEYgpHs3jXFJA6WUwRnuFEUJW29uYtgP41kJIhzuxKxMORQHBJD1DkUU1dWYSbBSqYC0vb3IRfZSs7+0W6fDyVmX3suEScaN/G6Cxa2HXdkhfQYcMSg06rOepxPlFR2TGb7DAC1xC0HJwskljCzqm+DGY3OCp2x5FgOaWFOuTCKYQ9vKxYiaKVcKjnedyA+RCprMdU5WE7TPcONoqigu3LiVnLEweWtQnHbbqVK6K8kxWEMOGLQZRUu5uCIgJKRygqgihIXhwwLIdN6CEUTVHl0VEUp+JfSsgQeJZl2JzkujGKIORALcdDyp3/PJUMVwRVT+wyRuolnfS45sBycQjrN7UnHHCqhhYYUh7GQEoOQcBNHL4jloFlOhW4qMyYlEqpV2q2Es8Wh//dQJEG118DvM+gNFfZL6RTBpZu7FUnMQQgLNRnloGmLQ64th6HEIWmKotnbwErdxNPxBt2dkwppxTmG7gEpDpLhcMQgYrmICFdB2nY7Aen+zX5SgelSdytFEmiqkv7dIRRJUOU18PtcBbcc+nfhS61QU5+Bu9CWQzyCgkhbDk6fpZwdPmUluFK7wLkM271ULHEHx3JQ0paDOzcb/iSjxIWOoqooLjtlXbqVJEOSFgfhImy5sArgVtKtgTEH+wtRDpZDU703/Xv6+Wiy33IoAreSm37LQcmMORRwFe1YCp1mNULk3t2ZSFpoqoKm9lsOzvPFgGM5kGHR5SIgrSRjJJSMrEDNsMUhUtoLscMhxWEMOLu/RYSLiGVgFiDnOW05ODGHVEqrVgYxh0kNvvTvDqGIHXOoqXLRV2C3krMLX3+NSb84mGbhxSEk3ESEK+f9leIJKy0I0J/SWqh01nA0SU9GzUvacki5lRTDk5OAtGrGSCj9RZeKy0e1lpCWg2QwzoosKmy3khktQMzBqXNIWw6u1POFF4c9bUG+8/uXhuysOhyWEISiCSYOJQ7RBFWeYrEcrNQWrYMth0K6lRxxCAsXIeEmGc5tBW/CzBaHQlsO9z/5BiseeDn92BoUkHbnpM5Bt+IkM8XB7cOvZwtTOSLFYSzEwySERl1dFRHhKkxAWjjZStlFcFoRuJW27OxkS3MXLR2js6jC0SRCQF21G7ehpWsbnI6s1V6DGp+LaNwsaJaMmdrPwamMLpqAdGrREk5ZDskcN99LJMy0tQD9Ka2FEofmA70c6Ayn99AQGamsQM7cSpoVx1Qz2rW4bHHo6i18+/jxRIrDGLBiISLCYEpjdcEC0joDLIcicis5HVU7R/nlcSwFv9eg2mukH4dj9lyrvAY1qZ5KhcwUsSyBQaJ/hapqCEXDrSQwCxpzsMXAVVVNSLiwcuzujCctjJQgQL/lUIhah6Rp0dYVIZ6wCKVqYNKWg5HbgLQu4piaO/1YcfvwqXG6imBvkfFEisMYSEZCRISLoxqriAoD1YzmfQcwXSQRMKh9RrqtRgHpTH1pBrbdPhyOGFQNEAfHgqjy6Ph99nwL6VoyTSs75gAI3V3wIjjHrTShcQJhyw3x3KeyuoaIORTCijvQGU678JybtJO2mv5cDE9OLAdDJLAyxcHlw6PE6QrGCr7z33gixWEM9IuDj7BwoQorZxuZjxSNJJai9/d2SYmDXgTi0NUXTf0cpeWQsgb8PoNqr54hDv2Wg99nz7OgtQ5WEhXRnxUDCM1V8FTWRKiPuNCYFKglIlwoidxatPGkiWFkxBxSqayFsBz2H+wXPud6s+JRUJSMBZNtORzpZlwu4gg9WxzcIkYi2W+1lCNSHMaAFQsTEQZ11W5MLbXjVJ5dS7pIYqoZraFVHYEtGoXGcSd1jlYcMiyHqgzLwenIascc7DkXMlPEab6WaTmguwoec4iHg4RTFm1YuNGTkZzuUphIWhgZ+2g4vzsN+fJJpjg415kVj4LuSS+Y0rGHI1i4mZZl99HSs91KuhkFxKit41JCisMYEPEwEctFjc+VLqfPZwsNSwh7v2ilXxwURcFUjIJbDknTojeVxTFWt1K1x3YrhYZ0K6UshwKKg7MlaKblgFZ4t5IZCRKx3NRVu7EMH0qOdymMJ6104Rv0F8MlCtAmft/BEIE6D4pCOjAsErF+QYD0Df1IXEvRmImLRH/VNYDLhyJMDMwjjju89GY7t/zynySLrPU5SHEYE2oiQkS48PsMVI8tDvkMStsBURNLze4IaSl6f6C6QHT3xRCAqihjCkhrqoLXrVHtNQhH7S6njule5TXwuDQMXS1oQFqxBlsOQncVPJXVigYJpa5L3E7fr9zFHRLJQ6SyFshymBqopq7anb5BW/FIlmCnb+hHEJSOR8Koir1/dPq4GS00jlQcNu/oZGdLHx09xWeBSHEYA6oZJSJcVPsMNOdLOM7i0NIR4v4n38C0LExLYCgmVqZbCbDUwlsOjok/bWI13cHYqCqGg6kWGYqiUO01EEAomiQYSaAo4HXbMZYan1HQDX+0IcRBMdy4C92VNR4mIlzUVLnQvKmOwTkUh/ghUlnzHXNwMpWmBKqo97vpTMUcRCKGomes8I0jtxxiEfv8OW0zgHRnVp8aH7XrdCAHOu37RmtX/veEORxSHEaJSMbRRBJT86CpKq6q/LTtfvMfz3He9v+jpaUz3fhNDBIHHR2zoBkUjivpuCm1mJZIu5hGgtNcD0j/DEUS6QI4NeVL9vtc9BbQchjSrZTKViqk5aAmwoSFG7/PwPDVACBymM6aOEQqa77rHJxMpaMm2OLQlRFzyHQrpcX7CFxr8Yj9vdYyxSFlOTRV2ZbykeCIQ1tX/tPhD4cUh1HitM4gdbG4U+Iw3v2VPAdfp0EL0bn7DdtyYLDlYKoGhmIXkhUKZyU18yj75jQa11LfEOIQjCTSrTMcbHEonOWgpi2HjKpZ3W0HpAt48jUzSiK1aHH7U833jiCddeW6nXz7dy+lHx/KrRQ/TCqrFekl9OBXMTt2j3ksmTjB6KMaU5ZDbyy1O2M0W7BTbqUjaaGRdMTBM9hyaPRa6UypsRCNJ9PCJi2HciBlITirB0+1/SWMh3JbjToQX/QAANHW3XbMQTERWrY4CMXAUMyCrl67emN43TqTJ9juttEEpTMth6pMcYgm048B261UQHHQh7AcFKOwloOwkhgijqWnXB41dQBH1EJj885OXtvVle6dFB9Q56BrKqqiHNZyMFvfxOrcQ/KtF8c8lkz2HwyhKDB5go8Gv4dYwiQSM7ES0UGuPjgyt1Iimtoi1D3YcmjwWEfkVmrt7BeENikOpY/jPtI89s2vutpHQqjEwuMnDvGESaPVAYDaszdlOQzhVtIM26IooDh09kVpqHHTUGN/MUcTsMu0HPyOOIQTBDOeB/BXuegNJQrmPnPcSlk3ogIXwaWz5VI3rqraWgCifWMTByEE+w4GEcD+jhCWECQH9FYCMAz1sOJgde4FINny2pjGMpB9B0M01XkxdC3jOoumLYdwNMGO/b05cSuZKU+B7iSeQPoc17qOLFuppdO2gCZP8NEq3Uqlj+NW0lMBvxqf3XwvOY7i0LZvP1WqvVKujrZiWhYuxURorqzXCdVIrV4LlxbX2Rujwe+h2mtg6Go6WHg4hBBDWw7RwW6lGp+LpGkRjRemv5I2cItWHMvBxCxAWif0t85QPfZ1WVtbTVxoxMdoOXT2xojE7Lnsaw+lBSAzlRVGtlWo1bHH/tm2E5GIHnEX1/0HQxzVaC/O6v22AHT2xVKWg4dVL+zim7/ZSFTY18wRuZVi9vfd5e0XByc4XWMkicZNIrGxJYEc6AijAKccO4GOnmjRpbNKcRglTr8al8++OP0pcTBzvOtWJt173wLgoD6RAJ1Eogk7tjDQraQW3q3U2Rel3u9GUZS0P3gkROP2uB1x8Lg0NFXpdyt5MiyHAhfC6WKIbCVXqp9Pnivl08Qci9YWhzq/m7BwjdmttLc9mPX7wF3gHFy6SuIwN3uzcy9xzQfC5Ic/eZhPf3ctq/++a0zjSpoWrZ2RQeLQ1RdDxO06hx37ejAtwe6O1GdxBG4lEUvtH+3LEIfUng7Vmr1IGKtr6UBnmAm1HqYEqjCt4iuok+IwSqIh+8vmrrJjDTU+w+7MGhs/n2G83f4iJaf9Oy7FpG33LgzMdJsAB1Fgt1IiadIXTqRN/YaMTJLD4Wyc4oiDk87aF0oQiQ2IOVQ5hXC5z1iKxg+/CtTTlkO/OKiOfztZmC94ItzfdA+gttpNxHJjjXHRsi8V9A3UebIsh4HioOvasJaDSMaxelp5LjgDE5VzGro5aUYDf3r2LTbv7Bj1uA50hrGEYEpKHOqq3ShAV08YYSYQuptdrfa5eKvV/k4eyVahVsI+httblfW84qujyuwFxp6xdKAzzKQJPibW28JTbEFpKQ6jJBa0xcFXY2fjVHkNIsIY1yI4rWc/3fipPfZkAIL7d9iprEO4lYwCBkUdIWjw21kiDTWeEbuVQgPEwfnd8cVmxRwcyyHHtQ5v7e/hsyueY/venmFfp4kESextIx1Up0jKLIzlEOmzx+z229dlXbWbkHDBGMVhb3uQhho3s6bWsbc92L9F6FCWwzDiYHXtR0GwywygNc1ktreNq5ecwpRANT95ZAsHu0d3Q8zMVAI7KF5T5aK31xaEvria3kdkR0sQdNcRuZUcYdEzspUAtIkz8XTvBMSIr/Gs4wpBa2eESQ0+JqZ2Piy2oLQUh1GSCIcwhUJ1la32uqaSUDyoyfH7YP2xNvpcTUyYdgymUFC6dqMpougsB8eFlLYcatx098VHNJ506wzfAHFI5YEPjDlA7ltovPTGQSwhWPvyvmFfZwzsawWoqbRJtUBuJWfR4q22A9Eet04UD+oYm+/taw8xpbGaKYEquoPx9OrYpQ+OOQzXldXqtOMN1E3BNXU21sFmXCLG1UtOxhLwwz9vHlVX18xMJYeGGjehPlscDoZsoZo+yd8flD6SzqyJGHGhoajZ89YmHY8S7WGCGhxTULqrL0YsYTK5wUdNlQu3oaWv9WJBisMoSUbtjqz+6n6Xgql70Kzx6e0ej0ZpoJuE/yg0l4supZ7aSOrmpWdbDmiugsYcnC+J4weu93uwhKA7ePhz43RkHWg5OK6jqiEsh1y7lbbs7ARgw+vtw7qXdBIklWxxSBdfFchyiIdsF0d1yqIFSGoetDEsWpKmRUtHiGkT3BxdZ98Ud7bY4jMoW+kwAenEwT3EhUbjlKPRjpoNQmC2vMHEeh//fenb2HWgj5XPN494bLtbg+lMJYd6v4dI0BaH1l4Lt6Ex7+RJ9ITiWJrriNxKajJKHNeg57VJJwBwkq9jTOLgFL9NavChKApN9V7aRmlFjTdSHEaJGUuJg6//ghG6F5c1Pr7m9l070BSBK3A0AEHPRKaoB4H+XeDSaAa6YmEmC9NCwzGvG2pSbiX/yNNZg0O4lTIFITMgbegaXreWU7dSbzjOrtY+Tjl2ArGEyYbX2g/5WkMkMAeIgxN/UHOwucxYSEaCRCyDGn9/+wjL8I3pumzripA0BWf1Psm0f34XDZPmA7b4DBaH4WMOkdZmDph1zDq6Hq1pJmh6OqX1tFmNnP22iTy5YQ89I1hA7Gzp5ZXtB/m34wNZz9f73YTDtruppSfB9InVHDfVtqBiwjiirqyKGSMx8LMG1PrJ4K7ieHf7EYmDsyXuxHqvjDmUPDG7f021N6Ppnctnt60YB5dC774dANROPRYAUTfVbiEMWQFRIJ29ZCUKs3rt7I1R5dFxp9IdHZEYSTZHX6p/ks/df16zrYjsJoN+nysdxM4FW5ttq+Gy82Ywsd7L86+2HPK1BoPdSo7l4DTlyzciFkq1zsjeztIggTBHt1jY2x7ETYL6zk0o4U7OrdrFzhZbHIZ2Kw2Tgtm9j/1mPcdNqUXRXWhNx2Hu35b+83vOO4ZkUvDYC8NnL1lC8JsnXqemysXic2dk/a3B74bUNb+3O8mMyTVMDVRj6CoRUzsyy8HM3j/aQVFUtImzOFo5MOoGk2CnsboNLW1lN9X7ONgdOWQa+sGe/AuHFIdRoiQjxBU3WmYwMlUgk26tMUoO/uMxDqz+8ZBFXckO2ywPTJsOgHfS9P6xDLAclFSA2iqQ37uzN5oWBOiPPYwkRc+uZTBQVSX93KGsCIDGWg879/fmrF3Flp2dVHl0jplUw7xTJvP6nu5DmvkGiew9helPa1UL5FYSsRARXHhc/TdvZYydWfe1h5jj3o1iJlA8fi7w9AeOhy6CGzpmYIV7cCVD9OiB9HWhHTUb6+Du9JgmNvg479RJPPvyPjqGcaus29TCzpY+3n/+cXjd2QuFer9dgAgQTmrMmOxH11SmT/TTl1DTAente3u44/5/jepGq1lxkupgcQDQJx9PrdVFoq9rxMdzONAZZmKDN733xMR6L6Yl6BhCaDZv28k/fvk9Xnpl+6jf50iQ4jBKtGQ0az9ZIN2ZdSz9lXoOtqO8/DBVe/9Oz/ZNg/7uCrZwUGnAMOwvRMPRs9J/S29w72DYN9DxsGBGQldfLL0SAtsKcBnqiFZWThW0sJJEn/slyf3bqEpZC05H1kzOO3Uybd0RXn1r9OmQAxFCsLW5i9kzGlBVhXNPnoQCrD+E9WCQHNTXKr2fdIEsBzURIaH0b3QDoPlSfb+ioyvQ3Nse5NyqZhR/APfcK2gQXbzNsKucnT0chGXfjIcrgjNTwWijcVr/mI46ERAk97yafu6yecfgJUbkoa8Se2nVoOMEIwkefPYtZk2t5ZyTJg76e73fjRvbioyjc8xkO+5y7FE19MbsjCNLCO5/8g1e39PNzx7dOuKkDcOKDVoIpOcy6XgAJpr7R13Yd6AzzKSG/qB6k5OxNERQuvfvDzPf8zrRf/whr8kmUhxGiW5FMfXstDYntzzc2zuqY1lCsGXV/WiY9Fkeetb/Mct6EEJQG28j5On/QtQ3BQha9ipMGSIgDQV0K/XFsiwHRVFo8HtG1JzMEYfE5idJbHuG6NM/ocawbzqZHVkdzjyhibpqF09t2HPE497fEaarL8bJxzQAtjts9ox61m9uGdIycQ1lOTh1DvFYQbLFdDNCYsB1afjsWpxI7/CpuQPpOdjGDPZhzDoXfea/E3PV8Q7PFvuYJIk8dTfBX1+L1duGoWuH3M+hb99OABqOnpl+Tps4E7VuMrF//AGREq2GGg+fmrKF2uRBYhv/jNWXHe95+G87CEUT/Nc7j88SP4eGGk/a1aoYHprq7PNw7FE1RC2dZDTChtfa2NXax+nHB3hjbw9/+cfIivD0AftHZ6I2zsBSDWYabXSNIGbiEH7lcT5g/ZlpNf3XSdMhah327NzN7PhmIng4mTd55R+56U81EipaHMLRJBtfbx9Vjx6XiCGM7C+hO7VCC/eNThyefW4Ts2Ob6Go6ne0T5tMQ20v3G/9K/z3e20mVEsWqnZJ+TlVVOnU7IDdQHJzH1jgGRUPRBC0dg90UsYRJMJJIB6EdGmrcI4o5BCMJmtxRYhv+jDrhaES4m4l7/wpkp7E66JrKBadPZUtzF/vaR7cy/vvWA3zrt/9Kt0l2spTeNqMeAGEmeG/tZv5H+QVvPPvYoP83lCTWwNVkynJoae3i6hV/495fPsYrT60msW8rZvf+Ufv9R4thRbEGiIO7yl5Bj+a6jMVNZkS2oQDGrLkoqk702Lcz02hjjrEL1zPfJbnjn2AmiDzzU1zaofdz6NvfTK/lYcYx/devoup4LvgUItxLdN0vEUKQ3PUS04Kv8o/48ZiWIPrig+nXb97RwTMv7eMdZ0zl6In+Id+nrrrfrTQxUJcWkGOPqiEmDMx4jIf+toOpgSo+856TOevEJv783M50HGU4DOKHFAdF04nXHs2xelt6N7rDkTzwJskX/8BMo42zWn6fdkXXVbtwGeqgWof25/8EQNVlXyGID++mP5HIU8JJUYnDo48+yiWXXMLChQu5//77x/393mhu5bFHn+bXj78+It+1sJK4SKa7Mjp4Uu2RnerpkbB9Xw/mK6tQFIWjL7yCt71jMZ1WFb0vPJgWq45ddtsM78QZWf8b9U0ChhAHpyhunNxKHT1RbvnlBm645x/c/efNWSKRLoCrGSAOfs+IYg7BSIJzY38DYeF95zKME+fja17LZK1rULzBivQS2/hnFrg249EFT23ce9jjW+EezANv8tcNe/jpyq28trubO377Egc6w2zZ2cnEBh+NtV6S+7cRevCrNO35K6bqZtIbf6J5w7r0cYQQdmtubeC51xGKyr/PrOaTU7fy/tgfOXbHA0Qf+xbhB5YT+v2XSOzcOC7NAoUQeEQ0vfubg7fGab43cnHY3xHiDNcOIv6jUWvt66z2tHcQslx8zL8WpfcA3oXX4Jn/UazW7czqWU/StIacl9K9lwOiIV3N7KA1zsB15ntJ7vgnic1PEv3bL1AbplE1/0P8NfI2zLf+gdm2g95wnHsf28ZRjVUsXTBz0PEdDF2lxmUL1JTJE9LPT6jx2EVwiShtXRHet2AmqqrwoQumcYl/K72PfoueNzYOez5c2FXXh0KdOIupWic93Yc/xyIWIvr0j0m46/hV8Dw8oRYiT/4AYSbsdNa67AZ8nfv3ckzoVfbVnkbVpOmET1zMFKWN19Y+cdj3ygWDl2QForW1lRUrVvDQQw/hcrm4/PLLOfvssznuuOPG7T1Pch9gZs1faN7xTx59+J0sfs/CrIDoQKyo078mWxx8/v623X29fby59Q2q6+qZefyMrMA12BuWP/HiHtatf4UvVW1HO/F8NH8jjcCWxgWc3rmarlefo9olUF/5CwB1R2d/MZTGYyD4TxR3dfbzqRjESGMOwkwg4hEUwzNIaJyN6RXFHn9bV5hv/+4lwjGTd54xlRde3cOdbzRz5gmNnHXGCcST9s2hvtqN1dOK1dOK4quhyZukJxgjkbTYvrebN/b28LYZqeyVDBfBUfFmpkVex3Xme1FrArj+/T9I7NjIUt8/eNY90+7XH+okseWvxLf8NSWAghsb6vn9tjMJLpiZFcB2sOIRYhseJr5pDSRjTExO4OIZ53Pmgvl874+vcMf9/yIWi7N0Zh/hR7+J2fI6ij+A95IvQM10Wn5/MxM23se+qhqmzD4VIbB3fBsYc8B2LQUOPE9ACPTZ5/N87ET++fJ2jqu3WKi/RvTJu9COnoP79Hej+BtRPNUoimp/DpE+RKQXEerCCneBmUANHIsWmJFOPLCEoCcYp67alXXurGQMXbFQB4hDVarmIRkcecD04K7tnKh3EZv5zvRz/ho/f7FO59/ENqa/5zr0phn2cXe/wrHb13K0djGJ1P7SpmWhqSrCMqlOHGSPb86Q3ynXnEsw92wi9sJvQdHwXvx55jZO4/f730lf85uEn/4VD+jvZXJiNx8/pp3Enx8lnoxCPIri9aMdNRttytvQJs5C8dZQ4wbLVDh6cn3/56EoVPmrcUeSzJ5SxUlVnUT//hRi27NcqEXpszyoz97FW6+dyzEXfwR1QAzPtCzcJCFzd7kBeKfNxnxtDcnW7cB0rGgfCIHqrcl6nRCC6N9+gRXqYk31+9kYd/HRc2dhPn8fkcf/D33qKZzt62FXl4GIHYfiruLA3x5kMjBp/lIAjjlvITtef5qG7Y8ROfVUPI1ThnSz5QpFFHLbsAwefvhh/vnPf3LbbbcB8MMf/hAhBJ/97GdH9P8dHcER+3oDAT/t7X0IYZF443l6nn8AT7KP/frRGP46NE1DVRUUAGGhYKEmY6jxIL7QPnbMWMKchZelj9fb3YvywDVEhYGLBM53oU/4CPun4aquRREWWCYdB7vQE31MNMK41STVV3wb1VcHQGd3iJ7fXU+jZlsgrWYN65Kn8sHPfAxd6xeZnr4Iz/1tAxdfPC9LfHZv3Uz9ujuJql6iWjUhy4OJAiiggIskLiuCW0Rxixg6dhBNADGjloS3ESEEerQLd8JeCcUMP0l3HW29CXxEmeQzURNByHCTRIVOm1VH0HQxu7oXJZ7t5kkKlQhuQpaLsOUiKDwIVxUTAg1oVpJkNIS/bwduj5fAB29PC1Vs67PE1/2CHq2BWjWcar2soB93Nq5/uwwR6iS49pdooXa6jYno1bVYmoe4UIlF48RicZqS+6gmwqbEDF6PN3FJzWtUmb2ogWOIY7C/tZta+qhVIyj+RlxvuwDjpHeks48629roffgbeInRUX0cCMG08Faam97OKe/5SNY8Q3+4HhEN4l7wMYwZpwP2BvL3PLqVeDzBZRN28P/YkG7cZ6Ha+36LQ4u5pehEvE1E4yaxeAJhCWKqF81Xi7e2DkOxUONBaju38Na0yzjt4iXpa/ytne0k7r8al2KSdPmJVh2FqXtRhIUikgihYCoaJnaGk6oIrK4W6pNt+D/4PbSq2vQ4vvuHl9na3MXPvnx++jkRC9Hx2+VEY3H6PEcRjFlE4hZuj0GtR2NK5HW2TXkP/77oPUPPra+d8CO3Ypx8Ie7TLrWvFdPisV/+iguSz9Jl+qjXwuCuQp98AhgeFMOD1XcQs+X1/r2hXV6iCQGWiXL592ms7XevbXr0fo5peRKhaiiWCYqKfuy/45pzMe2ilrcevY9Tk6/QpdQR1WvRzTCGFcVEJSE0JimdNE9+J6cs/q8h5yDiEXrv+wxdaj1e3cKX6LavXdVLrxEgqlWBsNDNGJPjzawMn87axClccPpULn/HLOKvPk7sxQfBzE7LjnsmoEU6edN3Gmd96Nr082++tJHAiz9AUwQhpYpowyzmfOCz9MRGv85XVYUJE6oP+feisRza2toIBPqLW5qamti0aXD2zqEYbpJDEQik/JdNlzD57Hew/ve/RNv5d5RYJxYgEAgUBAoWClFhEBUu+qxjOO5tZ/b/P1DfUMUj1vF41CS+STOYevzx9B1sp3PHNqp79qL17sIUKiYqVaqL+skTmTB5ItUnnotv+rSsMT00aynbtv2d0OR/4/jTz+BTJx+VbjSX+brjjr1k0JzMk07mj0+cRoMapM4Vp05PoCsWigCEIKG46FYbiCkewsJFyNQJJnW0RJiGWC+BSA8C6LJqiej2uDyhXurCYbya4KjpU6mun4Dm86N5/ajeauLxBN2vv47WsovGZJjqE87CO/V4XIFpmKFeWvfuYf3zWziqBqbUKEzVE4S6u0iE9uBqe5OE0IjiJqLV4L/gozRluAXE/EWsfP4FplULamYdh9EwGe8xp+JqnNr/mlPO5MG7foCvewfeSCdeJY6OiVtR8Wo6fZ7JvNK4gHj9DOYEqpl9xmSCLz1BaNsLuFSVGdMCtIcn0nj+xfhPOHNQm4RAwM9u143s/uN3qOrbjYVCm1LHhBNPy7oGAOo+dBOq4UXz9T+/MODntBMnse6VfWx8rYm1zVM4WmmjRg1Tq0ZwqSYh4SGCh17TQ2fSS4/lw0Jhht7OMXobR8W60XUX1dU1VHkMPKEelMhePOHtJNCJC51m0UjD8XOyxnTM9EY+H7uMqeYepsY6mRputSvoUbGEvYLRFAs9tYSwUDBR2ew7nffOmJo1t9NOaKK9Jzpgzn6az/44Pc/8Bl+0m0kuBZcHkokkiVCSFquWWWfPHXSe+k+uH3HtTwed8/d88qO8+oPX8Lmh8V0fo/qkc9ONDdOfu5kgtn87sQM7SHTsp/uNN3mrz8PSmYGs1fRZF15A25N7qZ8+C8+0E/FMm42W2kJ1EnDSiV/l+dVrUF55BJGIEtarsdwBNMXeM6VNaeDEeQsOPQf8bHbNpCbawhuxRtq040HVaEx20ZjspErpxkJFKCrb3Kcy+4Ir+MScqVQ79SgXLEWc/z6saIh/bHiNPz/2IlO0TqbGO6lRNI675L+y3juw8O1saZpI87/+jmh5jfrWZrZt3c6cc846xPjGTtFYDj/60Y+IxWJ87nOfA+CBBx5g8+bN3HzzzSP6/7FYDgMxLYtY3N5VKpowGWq/TUNXCdR5B5lzliVQFAY93xeOZ+2j3FjnTReJDYUQAtMSWZbCaOgJxnAZ2qDUTzj0vIUQROMmfeE4uqZSW+1KWySmZdETjFPlNYYd91jo7I3icWn4PINdNA6xuImuK4Pcc1mvSZhZDdy8bj3dNvxQcy4UkViS3lAcr1vH69az6gaEEMSTFqFIImuPAMPQCNRmp6laQtDeHSGZCgbrukpTxnXpzHvg9ZeJqioYmoquqwgBpmmRMC3q/W48ruzrx7QsEklr0PNC2PuE11QNcHVZgt5wnLrqQ/vrhyMaT+LStWHdvJkkTYu6+iqCvfkvFjMti2RS4DLUI3LzCCFS1en2Z+p2aVlW0FCvD0WTHHN0w5iu8ZKxHCZNmsSGDRvSj9vb22lqasrrGDRVxedRh71ZHYpDXcR+X3arjcOhKAq6NvYLrHYMX0ZFUdI3q4FoqpqVnppLRnJct+vwguQ2NKYERmc5FopDnWewPwe3oY1IhFVFSbd6Ho7RXn+HQlNVNNdggVYUZchrTlWVMQsDMEiEDoeuqXjdOuO7We/Q2OfmyI+jKEq6ncZIXz9UnC1XFE220rnnnssLL7xAZ2cnkUiEJ554gvnz5xd6WBKJRFKRFI3lMHHiRK677jquvPJKEokES5cu5dRTTy30sCQSiaQiKRpxAFi8eDGLFy8u9DAkEomk4ikat5JEIpFIigcpDhKJRCIZhBQHiUQikQyiqGIOR8JI86HH+vpyoRLnXYlzhsqcdyXOGcY278P9T9EUwUkkEomkeJBuJYlEIpEMQoqDRCKRSAYhxUEikUgkg5DiIJFIJJJBSHGQSCQSySCkOEgkEolkEFIcJBKJRDIIKQ4SiUQiGYQUB4lEIpEMouLE4dFHH+WSSy5h4cKF3H///YUezrjxgx/8gEWLFrFo0SK+9a1vAbB+/XoWL17MwoULWbFiRYFHOH7ccccdXH/99UBlzPnpp59myZIlXHzxxXzjG98AKmPejzzySPoav+OOO4DynXcwGOTSSy9l7969wKHnuW3bNpYsWcK73vUubrjhBpLJ5KEOeXhEBXHgwAFx/vnni66uLhEKhcTixYvFm2++Wehh5Zznn39e/Od//qeIxWIiHo+LK6+8Ujz66KNiwYIFYvfu3SKRSIiPfexj4tlnny30UHPO+vXrxdlnny2+/OUvi0gkUvZz3r17tzjvvPNES0uLiMfj4oorrhDPPvts2c87HA6Ls846S3R0dIhEIiGWLl0q/vrXv5blvF9++WVx6aWXipNOOkns2bNn2Ot60aJF4qWXXhJCCPGVr3xF3H///WN+34qyHNavX88555xDXV0dPp+Pd73rXaxZs6bQw8o5gUCA66+/HpfLhWEYzJw5k+bmZqZPn860adPQdZ3FixeX3dy7u7tZsWIFV111FQCbNm0q+zk/+eSTXHLJJUyaNAnDMFixYgVer7fs522aJpZlEYlESCaTJJNJqqury3LeDzzwADfddBNNTU3Aoa/rffv2EY1GOe200wBYsmTJEc2/bLqyjoS2tjYCgUD6cVNTE5s2bSrgiMaHWbNmpX9vbm7mL3/5Cx/84AcHzb21tbUQwxs3vva1r3HdddfR0tICDP15l9ucd+3ahWEYXHXVVbS0tPD2t7+dWbNmlf28q6urufbaa7n44ovxer2cddZZZft533rrrVmPDzXPgc8HAoEjmn9FWQ6WZaEo/W1qhRBZj8uNN998k4997GN86UtfYtq0aWU99z/+8Y9MnjyZuXPnpp+rhM/bNE1eeOEFbrvtNv7whz+wadMm9uzZU/bzfu211/jTn/7EM888w3PPPYeqqjQ3N5f9vOHQ13Wur/eKshwmTZrEhg0b0o/b29vTplq5sXHjRq655hqWL1/OokWLePHFF2lvb0//vdzmvnr1atrb23n3u99NT08P4XCYffv2oWla+jXlNmeAxsZG5s6dS0NDAwAXXngha9asKft5r1u3jrlz5zJhwgTAdqHce++9ZT9vsO9jQ32XBz5/8ODBI5p/RVkO5557Li+88AKdnZ1EIhGeeOIJ5s+fX+hh5ZyWlhauvvpq7rzzThYtWgTAnDlz2LlzJ7t27cI0TVatWlVWc7/vvvtYtWoVjzzyCNdccw0XXHABP/vZz8p6zgDnn38+69ato7e3F9M0ee6557jooovKft4nnngi69evJxwOI4Tg6aefLvtr3OFQ85wyZQput5uNGzcCdjbXkcy/oiyHiRMnct1113HllVeSSCRYunQpp556aqGHlXPuvfdeYrEYt99+e/q5yy+/nNtvv51ly5YRi8VYsGABF110UQFHOf643e6yn/OcOXP4xCc+wQc+8AESiQTz5s3jiiuu4Nhjjy3reZ933nls3bqVJUuWYBgGp5xyCsuWLWPevHllPW8Y/rq+8847ufHGGwkGg5x00klceeWVY34fuROcRCKRSAZRUW4liUQikYwMKQ4SiUQiGYQUB4lEIpEMQoqDRCKRSAYhxUEikUgkg5DiIJFIJJJBSHGQSCQSySCkOEgkEolkEP8fUznorJGMHrcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t = [(\"norm1\", MinMaxScaler(), slice(0,3)), (\"cat1\", OrdinalEncoder(), slice(3,4)),(\"norm2\", MinMaxScaler(), slice(4,37))]\n",
    "transformer = ColumnTransformer(transformers = t)\n",
    "t1X4 = transformer.fit_transform(t1X2.iloc[:, 0:37])\n",
    "t1X4 = np.concatenate((t1X4, t1X2.iloc[:, 37:]), axis = 1)\n",
    "#print(t1X3)\n",
    "#print(t1X3.shape)\n",
    "\n",
    "\n",
    "\n",
    "c1y1_pred = treecom.predict(c1X)\n",
    "c1X4 = c1X.copy()\n",
    "c1X4 = c1X4[c1y1_pred > 0]\n",
    "c1X4 = np.concatenate((transformer.fit_transform(c1X4.iloc[:, 0:37]), c1X4.iloc[:, 37:]), axis = 1)\n",
    "c1y4 = c1y[c1y1_pred > 0]\n",
    "#print(t1X4.shape)\n",
    "#print(c1X4.shape)\n",
    "#print(t1y2.shape)\n",
    "#print(c1y4.shape)\n",
    "#print(c1y4.shape)\n",
    "#print(t1X4[:, 0:3])\n",
    "#print(t1X4[:, 3:4])\n",
    "#print(c1X4[:, 3:4])\n",
    "\n",
    "c1y1_pred[c1y1_pred > 0] = embedding(t1X4[:, 0:3], t1X4[:, 3:4], t1X4[:, 4:], t1y2, c1X4[:, 0:3], c1X4[:, 3:4], c1X4[:, 4:], c1y4)\n",
    "#embed_model.layers[1].get_weights()[0]\n",
    "\n",
    "#print(error(c1y, c1y1_pred))\n",
    "#sns.set()\n",
    "#plt.figure()\n",
    "#sns.lineplot(y = c1y, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "#sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f7d9a1d7",
   "metadata": {},
   "source": [
    "These are the best results yet. It still tends to underestimate many of the high-comment posts, but it overall does a pretty good job. The training error is also the lowest out of all the models, save for the decision tree (and decision trees are notorious for being able to achieve extremely low training error)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ab2c434",
   "metadata": {},
   "source": [
    "# EXTRA (log transform)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51cb0d3e",
   "metadata": {},
   "source": [
    "Using log(target) with linear regression (exploration)\n",
    "We hypothesize that the distribution is roughly logarithmic, so we attempt to explore a linear regression with a log transform."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "927f8f0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13       0.125000\n",
      "22       0.208333\n",
      "27       0.083333\n",
      "32       0.625000\n",
      "39       0.083333\n",
      "           ...   \n",
      "40945    0.500000\n",
      "40946    0.666667\n",
      "40947    3.000000\n",
      "40948    3.111111\n",
      "40949    1.222222\n",
      "Name: C/H, Length: 18370, dtype: float64\n",
      "13      -2.079442\n",
      "22      -1.568616\n",
      "27      -2.484907\n",
      "32      -0.470004\n",
      "39      -2.484907\n",
      "           ...   \n",
      "40945   -0.693147\n",
      "40946   -0.405465\n",
      "40947    1.098612\n",
      "40948    1.134980\n",
      "40949    0.200671\n",
      "Name: C/H, Length: 18370, dtype: float64\n",
      "[-1.51030246 -2.1664459  -1.7717223  ...  3.01623298  0.87755756\n",
      "  1.52495971]\n",
      "0.9662858531409079\n",
      "Scaled error\n",
      "61.16281114530331\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scaled plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lrlog = LinearRegression().fit(t1X2, np.log(t1y2))\n",
    "t1ylog_pred = lrlog.predict(t1X2)\n",
    "print(t1y2)\n",
    "print(np.log(t1y2))\n",
    "print(t1ylog_pred)\n",
    "print(error(np.log(t1y2), t1ylog_pred))\n",
    "print(\"Scaled error\")\n",
    "print(error(t1y2, np.e**t1ylog_pred))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = np.log(t1y2), x = np.arange(1, len(t1y2)+1), label = \"actual\")\n",
    "sns.lineplot(y = t1ylog_pred, x = np.arange(1, len(t1y2)+1), label = \"predict\")\n",
    "plt.show()\n",
    "print(\"Scaled plot\")\n",
    "plt.figure()\n",
    "sns.lineplot(y = t1y2, x = np.arange(1, len(t1y2)+1), label = \"actual\")\n",
    "sns.lineplot(y = np.e**t1ylog_pred, x = np.arange(1, len(t1y2)+1), label = \"predict\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "1401aeaa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0194440282795294\n",
      "Scaled error\n",
      "156.93217574606643\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "c1y1_pred = treecom.predict(c1X)\n",
    "c1y1_pred[c1y1_pred > 0] = lrlog.predict(c1X[c1y1_pred > 0])\n",
    "print(error(np.log(c1y, out=np.zeros_like(c1y), where=(c1y!=0)), c1y1_pred))\n",
    "print(\"Scaled error\")\n",
    "print(error(c1y, np.e**c1y1_pred))\n",
    "sns.set()\n",
    "plt.figure()\n",
    "sns.lineplot(y = np.log(c1y, out=np.zeros_like(c1y), where=(c1y!=0)), x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "sns.lineplot(y = c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "plt.show()\n",
    "print(\"Scaled plot\")\n",
    "plt.figure()\n",
    "sns.lineplot(y = c1y, x = np.arange(1, len(c1y1)+1), label = \"actual\")\n",
    "sns.lineplot(y = np.e**c1y1_pred, x = np.arange(1, len(c1y1)+1), label = \"predict\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51abfced",
   "metadata": {},
   "source": [
    "We see that the model fits pretty well. However, using the logarithm does mess up our error (and the overal predictions) since linear changes in the logarithm have multiplicative effects on the original.\n",
    "\n",
    "In conclusion, this does seem to be an effective model, it just has a different purpose (and thus different error) than the others. It would do well in predicting the order of magnitude of comments, rather than the actual amounts."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}