predict_180days_Diet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jxsbMqBx7wy9"
   },
   "source": [
    "#### <span style=\"color:#AD07FF\"> This notebook creates prediction about is user on 180 days program or not. It's important becouse we can detect people who is going to use app at least 180 days and it's really great information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "XEiabK-yVL_C",
    "outputId": "2ddd6665-055a-4faa-ec05-5d7294c808e9"
   },
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from plotly.subplots import make_subplots\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split, KFold\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.metrics import f1_score,precision_score,recall_score,accuracy_score,mean_squared_error, r2_score\n",
    "from sklearn.metrics import classification_report,confusion_matrix\n",
    "from sklearn.tree import DecisionTreeClassifier,DecisionTreeRegressor\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "K0BA_3DWVL_D"
   },
   "outputs": [],
   "source": [
    "# The data is large and this command will help us to see the whole result \n",
    "pd.set_option('display.max_rows', 500)\n",
    "pd.set_option('display.max_columns', 500)\n",
    "pd.set_option('display.width', 1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "xh5fPED8VL_D"
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('../data/data_for_predict_180days_diet.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vAoLuhwIF-oY",
    "outputId": "a03a6d04-f282-479b-ccbc-68876c654b2f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    2219\n",
       "1.0    1524\n",
       "Name: is_180_diet, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[\"is_180_diet\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OLAnbrwZEJ8r"
   },
   "source": [
    "#### <span style=\"color:#00CC00\"> As you can see, the people who were on 180 day program are twice less than the people who were not."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aMbOOXWbEew8"
   },
   "source": [
    "#### <span style=\"color:#FF7B07\"> Let's drop unused columns becouse if we won't do this, there will be approximately 200 features ! :) <br>  First of all, our new features containing \"diff\" consists both \"goal\" and \"total \" becouse it's their difference and I think in my algorithm differences between total and goal are more important then these columns itself. So I'm gonna drop them ."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "KnhDrbZbVL_E"
   },
   "outputs": [],
   "source": [
    "col1 = list(filter(lambda x:   (\"goal\" in x) or (\"total\" in x)  or (\"sugar\" in x), data.columns))\n",
    "X = data.drop(columns=['is_180_diet','logged_frequency','user_id','days_missed'])\n",
    "X = X.drop(columns=col1)\n",
    "y = data['is_180_diet']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cDwPzhVaEm-k"
   },
   "source": [
    "#### <span style=\"color:#00CC00\"> Look's Like our data needs to be balanced . <br>So let's do this with smote"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "h2RAAHJaU1uB"
   },
   "outputs": [],
   "source": [
    "def prepare_data(X,y,columnsTobeDropped):\n",
    "    if(columnsTobeDropped):\n",
    "        X = X.drop(columns=columnsTobeDropped)\n",
    "    X_train, X_test,y_train,y_test = train_test_split(X, y, test_size=0.2, random_state = 42)\n",
    "    sm = SMOTE(random_state=42)\n",
    "    X_train_s,y_train_s =sm.fit_sample(X_train,y_train)\n",
    "    X_test_s,y_test_s =sm.fit_sample(X_test,y_test)\n",
    "    return X_train_s, y_train_s, X_test_s,y_test_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "w6An5dUkVL_H"
   },
   "outputs": [],
   "source": [
    "def classification_metrics(yTrueTrain,yPredictTrain,yTrueTest,yPredictTest,heatmap=False)->None:\n",
    "    \"\"\"\n",
    "    this function prints accuracy, precision, recall, F1 scre \n",
    "    and create confusion matrix  heatmap for both train and test sets\n",
    "    \n",
    "    Arguments:\n",
    "    yTrueTrain -- the true value of prediction from the train set\n",
    "    yPredictTrain -- predicted value from the train set\n",
    "    yTrueTest -- he true value of prediction from the test set\n",
    "    yPredictTest -- predicted value from the test set\n",
    "    heatmap -- boolean which tells us if plot heatmap or not\n",
    "    \n",
    "    Returns:\n",
    "    None\n",
    "    \"\"\"\n",
    "    print(\"train data:\\t  \"+ \"\\t\"*6+ \"test data:\\t\\n\")\n",
    "    print((\"accuracy:\\t {0} \"+ \"\\t\"*6+ \"accuracy:\\t {1}\").format(accuracy_score(yTrueTrain, yPredictTrain).round(2), \n",
    "                                                                 accuracy_score(yTrueTest, yPredictTest).round(2)))\n",
    "          \n",
    "    print((\"precision:\\t {0} \"+ \"\\t\"*6+ \"precision:\\t {1}\").format(precision_score(yTrueTrain, yPredictTrain).round(2), \n",
    "                                                                   precision_score(yTrueTest, yPredictTest).round(2)))\n",
    "    \n",
    "    print((\"recall:\\t\\t {0} \"+ \"\\t\"*6+ \"recall:\\t\\t {1}\").format(recall_score(yTrueTrain, yPredictTrain).round(2), \n",
    "                                                                 recall_score(yTrueTest, yPredictTest).round(2)))\n",
    "    \n",
    "    print((\"F1:\\t\\t {0} \"+ \"\\t\"*6+ \"F1:\\t\\t {1}\").format(f1_score(yTrueTrain, yPredictTrain).round(2), \n",
    "                                                         f1_score(yTrueTest, yPredictTest).round(2)))\n",
    "    if(heatmap):\n",
    "        f, ax  = plt.subplots(1,2,figsize = (18,6))\n",
    "        sns.heatmap(confusion_matrix(yTrueTrain,yPredictTrain),cmap='coolwarm',annot=True,ax=ax[0])\n",
    "        sns.heatmap(confusion_matrix(yTrueTest,yPredictTest),cmap='coolwarm',annot=True,ax=ax[1])\n",
    "        plt.tight_layout(rect=[0, 0.03, 1, 0.95])\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "OnIIBQ1Kq0di"
   },
   "outputs": [],
   "source": [
    "def depthChecker(columnsTobeDropped,X,y):\n",
    "    X_train, y_train, X_test,y_test =prepare_data(X,y,columnsTobeDropped);\n",
    "    cv = KFold(n_splits=10,shuffle = True,random_state=42)\n",
    "    accuracies = list()\n",
    "    depth_range = range(1, 10)\n",
    "    # Testing max_depths from 1 to 10\n",
    "    for depth in depth_range:\n",
    "        fold_accuracy = []\n",
    "        tree_model = DecisionTreeClassifier(max_depth = depth,random_state = 42)\n",
    "        for train_fold, valid_fold in cv.split(X_train):\n",
    "            f_train_x = X_train[train_fold] \n",
    "            f_train_y = y_train[train_fold]\n",
    "          \n",
    "            f_valid_x = X_train[valid_fold]\n",
    "            f_valid_y = y_train[valid_fold]\n",
    "\n",
    "            model = tree_model.fit(X = f_train_x,y = f_train_y) # We fit the model with the fold train data\n",
    "            valid_acc = model.score(X = f_valid_x,y = f_valid_y) # We calculate accuracy with the fold validation data\n",
    "            fold_accuracy.append(valid_acc)\n",
    "        avg = sum(fold_accuracy)/len(fold_accuracy)\n",
    "        accuracies.append(avg)\n",
    "      \n",
    "    # Just to show results conveniently\n",
    "    df = pd.DataFrame({\"Max Depth\": depth_range, \"Average Accuracy\": accuracies})\n",
    "    df = df[[\"Max Depth\", \"Average Accuracy\"]]\n",
    "    print(df.to_string(index=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 601
    },
    "id": "vZE-io92qvWx",
    "outputId": "0c981443-3a38-4a66-e083-06654b37597e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data:\t  \t\t\t\t\t\ttest data:\t\n",
      "\n",
      "accuracy:\t 0.68 \t\t\t\t\t\taccuracy:\t 0.67\n",
      "precision:\t 0.67 \t\t\t\t\t\tprecision:\t 0.66\n",
      "recall:\t\t 0.73 \t\t\t\t\t\trecall:\t\t 0.71\n",
      "F1:\t\t 0.7 \t\t\t\t\t\tF1:\t\t 0.68\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train, y_train, X_test,y_test =prepare_data(X,y,{})\n",
    "decision_tree = DecisionTreeClassifier(max_depth=3,random_state=42)\n",
    "decision_tree.fit(X_train, y_train)\n",
    "\n",
    "# Predicting results for test dataset\n",
    "y_pred_train = decision_tree.predict(X_train)\n",
    "y_pred_test = decision_tree.predict(X_test)\n",
    "classification_metrics(y_train,y_pred_train,y_test,y_pred_test,True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zdu6qSYuBKa5"
   },
   "source": [
    "#### <span style=\"color:#FF7B07\"> In the previous algorithm (predict reach goal) we saw with pca that the perfect quantity of columns is approximateli 10-15.<br> So I'm gonna follow this idea and I'm gonna drop some days and let's see if we'll improve our acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 601
    },
    "id": "KBeBxnBlCA4b",
    "outputId": "25923899-876b-4e8f-9ee9-8e63c805c961"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data:\t  \t\t\t\t\t\ttest data:\t\n",
      "\n",
      "accuracy:\t 0.67 \t\t\t\t\t\taccuracy:\t 0.7\n",
      "precision:\t 0.68 \t\t\t\t\t\tprecision:\t 0.7\n",
      "recall:\t\t 0.67 \t\t\t\t\t\trecall:\t\t 0.72\n",
      "F1:\t\t 0.67 \t\t\t\t\t\tF1:\t\t 0.71\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "col2= list(filter(lambda x:   (\"_2\" in x ) or (\"_3\" in x) or (\"_4\" in x) or (\"_5\" in x) or (\"_6\" in x) or (\"_7\" in x) or (\"_8\" in x) ,X.columns))\n",
    "X_train, y_train, X_test,y_test =prepare_data(X,y,col2)\n",
    "decision_tree = DecisionTreeClassifier(max_depth=3,random_state=42)\n",
    "decision_tree.fit(X_train, y_train)\n",
    "\n",
    "# Predicting results for test dataset\n",
    "y_pred_train = decision_tree.predict(X_train)\n",
    "y_pred_test = decision_tree.predict(X_test)\n",
    "classification_metrics(y_train,y_pred_train,y_test,y_pred_test,True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ReaiZfvCpCSg",
    "outputId": "c1abd56b-63fe-4501-be2f-6cd5acc366b9"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>foods_len_9</td>\n",
       "      <td>0.126784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>healtyDistrib</td>\n",
       "      <td>0.069655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>carbs_diff</td>\n",
       "      <td>0.068995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>calories_diff</td>\n",
       "      <td>0.065760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>sodium_diff</td>\n",
       "      <td>0.061724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>fat_diff</td>\n",
       "      <td>0.061650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>protein_diff</td>\n",
       "      <td>0.051339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foods_len</td>\n",
       "      <td>0.047744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>sodium_diff_9</td>\n",
       "      <td>0.045123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>foods_len_1</td>\n",
       "      <td>0.042863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>fat_diff_1</td>\n",
       "      <td>0.039409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>carbs_diff_9</td>\n",
       "      <td>0.038629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>healtyDistrib_1</td>\n",
       "      <td>0.036967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>protein_diff_1</td>\n",
       "      <td>0.035688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>carbs_diff_1</td>\n",
       "      <td>0.031636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>sodium_diff_1</td>\n",
       "      <td>0.030746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>calories_diff_9</td>\n",
       "      <td>0.026692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>healtyDistrib_9</td>\n",
       "      <td>0.026580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>start_date</td>\n",
       "      <td>0.024649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>calories_diff_1</td>\n",
       "      <td>0.022632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>fat_diff_9</td>\n",
       "      <td>0.022383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>protein_diff_9</td>\n",
       "      <td>0.012583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>FoodGroup</td>\n",
       "      <td>0.009769</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            feature  importance\n",
       "14      foods_len_9    0.126784\n",
       "1     healtyDistrib    0.069655\n",
       "3        carbs_diff    0.068995\n",
       "2     calories_diff    0.065760\n",
       "6       sodium_diff    0.061724\n",
       "4          fat_diff    0.061650\n",
       "5      protein_diff    0.051339\n",
       "0         foods_len    0.047744\n",
       "20    sodium_diff_9    0.045123\n",
       "7       foods_len_1    0.042863\n",
       "11       fat_diff_1    0.039409\n",
       "17     carbs_diff_9    0.038629\n",
       "8   healtyDistrib_1    0.036967\n",
       "12   protein_diff_1    0.035688\n",
       "10     carbs_diff_1    0.031636\n",
       "13    sodium_diff_1    0.030746\n",
       "16  calories_diff_9    0.026692\n",
       "15  healtyDistrib_9    0.026580\n",
       "21       start_date    0.024649\n",
       "9   calories_diff_1    0.022632\n",
       "18       fat_diff_9    0.022383\n",
       "19   protein_diff_9    0.012583\n",
       "22        FoodGroup    0.009769"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = DecisionTreeRegressor()\n",
    "col2= list(filter(lambda x:   (\"_2\" in x ) or (\"_3\" in x) or (\"_4\" in x) or (\"_5\" in x) or (\"_6\" in x) or (\"_7\" in x) or (\"_8\" in x) ,X.columns))\n",
    "X_train, y_train, X_test,y_test =prepare_data(X,y,col2)\n",
    "model.fit(X_train, y_train)\n",
    "importance = model.feature_importances_\n",
    "importance_df = pd.DataFrame(columns=['feature','importance'])\n",
    "importance_df['feature'] = X.drop(columns=col2).columns\n",
    "importance_df['importance'] = importance\n",
    "importance_df.sort_values(by=['importance'],ascending = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 634
    },
    "id": "4_tAw3cHkkkA",
    "outputId": "af035c01-bd4f-4c31-e1fc-1f8e90e65ba2"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Raffael Kalandadze\\anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:762: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data:\t  \t\t\t\t\t\ttest data:\t\n",
      "\n",
      "accuracy:\t 0.66 \t\t\t\t\t\taccuracy:\t 0.69\n",
      "precision:\t 0.66 \t\t\t\t\t\tprecision:\t 0.68\n",
      "recall:\t\t 0.68 \t\t\t\t\t\trecall:\t\t 0.72\n",
      "F1:\t\t 0.67 \t\t\t\t\t\tF1:\t\t 0.7\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "clf = LogisticRegression(random_state=0).fit(X_train, y_train)\n",
    "classification_metrics(y_train,clf.predict(X_train),y_test,clf.predict(X_test),True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5J2HvF0pCcSb"
   },
   "source": [
    "#### <span style=\"color:#00CC00\"> As we see without unused days, we improved our accuracy !<br>Let's check how another algorithms behave with our data :) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Rs7l9Hywn_Vb",
    "outputId": "17608648-51be-44e4-a1ca-648d1b66a6c6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC()"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "svclassifier = SVC(kernel='rbf')\n",
    "svclassifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 466
    },
    "id": "hootvWJ0mOgC",
    "outputId": "e5dbc7b7-0a32-474b-9c78-504cfaed6641"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data:\t  \t\t\t\t\t\ttest data:\t\n",
      "\n",
      "accuracy:\t 0.65 \t\t\t\t\t\taccuracy:\t 0.7\n",
      "precision:\t 0.65 \t\t\t\t\t\tprecision:\t 0.69\n",
      "recall:\t\t 0.67 \t\t\t\t\t\trecall:\t\t 0.73\n",
      "F1:\t\t 0.66 \t\t\t\t\t\tF1:\t\t 0.71\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABNcAAAGFCAYAAADelRV/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAA+2klEQVR4nO3dd5xddZ3/8ddnSnoljUASCRC6tCCLIsqCCuuKYMENlkUEsygqKK40/aEoyK4LWHGNAoIFpArrKi6igJSAgPQQEooQCQnpvczc7++Pc4GbMjNhMrlz8s3r+Xicx9z7veeekgcT3vmcb4mUEpIkSZIkSZJev4buvgBJkiRJkiRpc2VxTZIkSZIkSeoki2uSJEmSJElSJ1lckyRJkiRJkjrJ4pokSZIkSZLUSRbXJEmSJEmSpE5q2tQneOTdB6dNfQ5J3evrO1/W3ZcgaRO75qKx0V3n/t/mnbskS/zz6qnddg85eu6EI814Usa+MfaS7r4ESXXwk7OGZp/xIqIXcAfQk6IOdm1K6eyI+BZwBLAKeBo4LqW0ICK2A6YAU6uHmJxSOrG9c9hzTZIkSZIkSblaCRySUtoL2Bs4PCIOAG4B9kgp7Qk8BZxR852nU0p7V7d2C2tQh55rkiRJGyOa7XAmSZKUm3plvJRSApZU3zZXt5RS+r+a3SYDH+zsOSyuSZKkUmtosrgmSZKUm67KeBExEZhY0zQppTRprX0agQeAHYEfpJTuXeswnwB+VfN+bET8FVgEfDml9Of2rsHimiRJKrVodhYLSZKk3HRVxqsW0iZ1sE8rsHdEDAJuiIg9UkqPAUTEWUAL8Ivq7jOBMSmluRExHvh1ROyeUlrU1vFNq5IkSZIkScpeSmkBcBtwOEBEHAu8B/hIdfgoKaWVKaW51dcPUCx2sFN7x7XnmiRJKjWHhUqSJOWnXhkvIoYBq6srgfYG3gH8R0QcDpwGvD2ltGyt/eellFojYntgHPBMe+ewuCZJkkrNBQ0kSZLyU8eMNxK4vDrvWgNwdUrpNxExHegJ3BIRAJOrK4O+DTgnIlqAVuDElNK89k5gcU2SJEmSJElZSik9AuyznvYd29j/OuC613MOi2uSJKnUHBYqSZKUn5wynsU1SZJUag4LlSRJyk9OGc/imiRJKrWcnmpKkiSpkFPGa+juC5AkSZIkSZI2V/ZckyRJpRaN+TzVlCRJUiGnjGdxTZIklVpDRsFLkiRJhZwynsNCJUmSJEmSpE6y55okSSq1aMjnqaYkSZIKOWU8i2uSJKnUotGO9pIkSbnJKeNZXJMkSaWW03wckiRJKuSU8fIpE0qSJEmSJEl1Zs81SZJUajnNxyFJkqRCThnP4pokSSq1nIYMSJIkqZBTxnNYqCRJkiRJktRJ9lyTJEmlFhk91ZQkSVIhp4xncU2SJJVaNNjRXpIkKTc5ZTyLa5IkqdRymuxWkiRJhZwyXj5lQkmSJEmSJKnO7LkmSZJKLaeVpCRJklTIKeNZXJMkSaWW05ABSZIkFXLKeA4LlSRJkiRJkjrJnmuSJKnUclpJSpIkSYWcMp7FNUmSVGo5DRmQJElSIaeMZ3FNkiSVWk6T3UqSJKmQU8bLpw+eJEmSJEmSVGf2XJMkSaWW05ABSZIkFXLKeBbXJElSqeU02a0kSZIKOWW8fO5EkiRJkiRJqjN7rkmSpFLLaciAJEmSCjllPItrkiSp1HIKXpIkSSrklPEsrkmSpFLLKXhJkiSpkFPGc841SZIkSZIkqZPsuSZJkkotp5WkJEmSVMgp41lckyRJpdbQmM+QAUmSJBVyynj5lAklSZIkSZKkOrO4JkmSSi0aoks2SZIklUe9Ml5E9IqI+yLi4Yh4PCK+Vm3fKiJuiYhp1Z+Da75zRkRMj4ipEXFYR+dwWKgkSSq1nObjkCRJUqGOGW8lcEhKaUlENAN3RsTvgPcDt6aUzo+I04HTgdMiYjdgArA7sA3wh4jYKaXU2tYJTKuSJKnU6vhU89KImB0Rj9W0HV19wlmJiP3W2n+9TzQjYnxEPFr97LsRYbc5SZKktdQr46XCkurb5uqWgCOBy6vtlwNHVV8fCVyVUlqZUnoWmA7s3945LK5JkiQVfgocvlbbYxRPNe+obVzriebhwMUR0Vj9+IfARGBcdVv7mJIkSeoiETExIu6v2SauZ5/GiHgImA3cklK6FxiRUpoJUP05vLr7tsALNV+fUW1rk8NCJUlSqdVrvrSU0h0Rsd1abVMA1tP57NUnmsCzETEd2D8ingMGpJTuqX7vCoqnoL/bpBcvSZK0memqjJdSmgRM6mCfVmDviBgE3BARe7R3aes7RHvHt7gmSZJKravm46g+xax9kjmpGsY6Y1tgcs37V55orq6+XrtdkiRJNbpjXt2U0oKIuI1iZMGsiBiZUpoZESMperVBkd9G13xtFPBie8d1WKgkSdoipJQmpZT2q9k6W1iDtp9ovu4nnZIkSdp0ImJYtccaEdEbeAfwJHATcGx1t2OBG6uvbwImRETPiBhLMc3Hfe2dw55rkiSp1Oo1LPR1auuJ5ozq67XbJUmSVKOOGW8kcHl1ftwG4OqU0m8i4h7g6og4HngeOBogpfR4RFwNPAG0ACe1t1IoWFyTJEkl1x1DBjbATcAvI+JCiiXaxwH3pZRaI2JxRBwA3Av8K/C9brxOSZKkUqpXxkspPQLss572ucChbXznXODcDT2HxTVJklRu6y4msIlOE1cCBwNDI2IGcDYwj6I4Ngz434h4KKV0WAdPND9FsfJob4qFDFzMQJIkaW11ynj1YHFNkiQJSCkd08ZHN7Sx/3qfaKaU7gfaW4FKkiRJGbG4JkmSSq2kc65JkiRpI+SU8SyuSZKkUivpnGuSJEnaCDllvHzuRJIkSZIkSaoze65JkqRSy2nIgCRJkgo5ZTyLa5IkqdRyGjIgSZKkQk4Zz+KaJEkqtZyeakqSJKmQU8bLp0woSZIkSZIk1Zk91yRJUqnl9FRTkiRJhZwynsU1SZJUbhnNxyFJkqSqjDJePnciSZIkSZIk1Zk91yRJUqlF5DNkQJIkSYWcMp7FNUmSVGo5LdMuSZKkQk4Zz+KaJEkqtZwmu5UkSVIhp4yXT5lQkiRJkiRJqjN7rkmSpHLLaMiAJEmSqjLKeBbXJElSqeU0ZECSJEmFnDJePmVCSZIkSZIkqc7suSZJkkotwmeBkiRJuckp41lckyRJ5ZbRkAFJkiRVZZTxLK5JkqRSi4wmu5UkSVIhp4yXz51IkiRJkiRJdWbPNUmSVGo5rSQlSZKkQk4Zz+KaJEkqt4wmu5UkSVJVRhkvnzuRJEmSJEmS6syea5IkqdRyGjIgSZKkQk4Zz+KaJEkqt4xWkpIkSVJVRhnP4pokSSq1iHyeakqSJKmQU8bLp0woSZIkSZIk1Zk91yRJUrllNGRAkiRJVRllPItrkiSp1HKa7FaSJEmFnDKexTUx6pQvMWD/N9OyYAFPffq4dT7vOWoMoz5/Gr13HMdLl1/CnOt/tdHnjKZmRn/xDHrvuDOtixfyt2+ew+rZL9E8fARvOOscoqGRaGpkzv/cwLzf3rTR55O2dH16NfCpCUMZvXUzCfjhlXN46m8r19lvh9E9OO+UbbjoitlMfnjZRp2zqRE++5FhbD+qJ4uXtXLR5S/z8vwWttumB588egi9ezVQqcD1tyzg7oeWbtS5JEkdi6Zmtj7tPKKpGRoaWfbA3Sy46cqNOmbft/wjg/75QwAs+N+rWXr3nwAYesIX6LndjqTWFlY+O425P7sYWls3+h4kvebj7+nHnjv2YPHSCmf/eME6n+88ppmTju7PnIUVAB58ciW/uXP5Rp2zqRGOf29/3rB1E0uWV/jRDYuZu7DC6BGNfPTwfvTqGaQK/O9dy/jLlFUbdS5pc2JxTcz/w83M/Z8bGH3qmev9vGXxIl787+8y4M1vfd3Hbh6+NaO/cDrPnH7KGu1bHfZuWpcsYeoJH2Hg2w5h5Ccm8vz559Ayby5Pn/oZUstqGnr1ZqcfXsaiyXfRMm9uZ25NUtVx79+Kv05ZxgU/XUJTI/RoXrcLdkPAR4/YioeefH2ha9jgJk768FC++oOX1mg/5ID+LFle4bPnzeAt+/Tlo0cM5qIrXmbl6grf+8XLvDSnhcEDGvmPU7fhoSeXs2xFZaPuURmLfIYMSN0ptazmpf/6CmnlCmhsZORp57P8sQdY+cxTHX5363//BnMu/S4tc2e/2tbQtx+DjpjAzG+cCikx8isXsvyh+6gsW8rSe29nzk8uBGDoJ0+l/0HvZPFtN2+ye5O2RHc9vII/3r+c44/o3+Y+015o4XtXL3rdxx4ysIFPHNGfb/184Rrtb927F0tXVDjzh/N50249+OAhffnRDYtZtTpxyU2LmT2/wsB+DXzl+EE89sx8lq9Mr/vc2oJklPEsromljz1C8/Ct2/y8deECli9cwID937zOZ4P+8Z0Mfe/7iaZmlk19gr9f/G2odPwP5AEHHMisX/wUgIV33s62nzoZgNTS8uo+0dwMGa0eInWX3j2D3bbvxQ9+OQeAllZoaV339/TwgwYw+eGl7Dim5xrtB43vy7sPGkhTE0z720p+cu1cKhuQk960Rx+uuXkBAJMfXsrx7x8CwMyXX/s9n7+olYWLWxnQr8HimtqW0ZABqbullSsAiMZGaGwkJWgatjVDPvJvNPQfQFq5krlX/IDVL/29w2P13n0fVjzxEJWlSwBY8cRD9N5jX5be92eWP/rAq/utem4ajYOHbpobkrZg015oYcjAzhUnDtijJ4e+qReNjcGzf1/Nz29eStqAfLf3uB7c9OdidMMDU1bx4cP6ATBr3ms5buGSCouXVujfJyyuqX0ZZbwOi2sRsQtwJLAtkIAXgZtSSlM28bWp5HqOHsOgt/0j07/4GWhtZZtPn8Kgg9/Bgj/+X4ffbR4yjNUvv1y8qbTSumwJjQMG0rpoIc1Dh7Hd186n58htmXnpf9trTdpII4Y0s2hJhZOOGcobtunBMzNWcdkNc1m56rWws9XARv7hjX342sUvrVFc23Z4M2/Zpx9f/u6LtFbghA8M4a3j+3HH/Us6PO9WA5uYs6AopFUqsGxFhf59G1i89LXwteOYHjQ1BbPmtrR1GInI6KlmmZjxtlDRwDZfuYCm4SNZ/KffsurZpxhx6jnM/dkPaZk9kx5jd2Krj5zIrAu+0uGhGgcNoWXenFfft8yfS+OgIWvt1EjfAw5m3lU/6eo7kbQBdti2ibNPGMSCJRWu+cNSXpzTysghjbxpt56cf/lCWivwkcP7csAePbnn0XWnDFnb4P4NzF9UZLlKguUrE/16B0uWv5Yrx27TRFMjvDzfB6dqX70yXkSMBq4AtgYqwKSU0nci4lfAztXdBgELUkp7R8R2wBRgavWzySmlE9s7R7vFtYg4DTgGuAq4r9o8CrgyIq5KKZ3fxvcmAhMBvrL7OD44Zpv2TqPNVL+9xtN7x50Y9+0fAdDQswetCxcA8IYvf50eI0YSzU00DxvBuO8VgWrOTdcy/5abYX0F6uqjktVzXmbaScfTtNUQtvvKN1h45+20LJhfj1uSstTQCGNH9eCS6+cy/fmVHPe+rTjq0IH86ncLXt3n40cN4ee/mb9Oj7Q37tSL7Uf14PwvFH+P92gOFi4p5sz59+OGM3xIE02NwdDBTXzri8U+/3vHIm67b8l6O57WPhEdNKCRz35kGN//5ZwNelIqqet0RcY798A9+fAu29XhatWlUoUXz/k8Db37MuykM2jeZgw9d9iF4Sd+6bV9mpsB6HfgoQw49D0ANA0fyfCT/x+0rGb1nNm8fPE32xhhsOZf6EM+ciIrn3qcldOe2FR3JKkNf3uphdO+P4+Vq+GNOzRz0tEDOOuH89llbDNv2LqRsz4xCIAeTbB4afG7++kP9mfooEaaGoqHr//vhGKfW+9bzl2PrFzvv+Nqf+sH9guOf28/Lr1pCcY7lUgLcGpK6cGI6A88EBG3pJT+5ZUdIuICoHYc9NMppb039AQd9Vw7Htg9pbS6tjEiLgQeB9YbvFJKk4BJAI+8+2B/p3IVMP/W3/PST3+8zkd/+0bxtLOtOddWz3mZ5mHDWD33ZWhopLFPP1oXrzkXQMu8uax4/jn67r4nC++6fZPdhpS7eQtambuwhenPF08j73l4Ke87dNAa++wwugen/OswAAb0bWSfXfvQ2joHIrj9L0v45f+uW+D+1mXFvDttzbk2d0ELQwc1MW9hKw0NxaIKS5YVTzB79wzO+OQIrvztfKatZ2EFaQ0ZDRkokY3OeM+dcKQZbzNWWb6UFVMfpc++b6aybCkvnvP5dfZZctetLLnrVmD9c661zp9Dr533ePV90+AhrJj62KvvBx7xLzT2H8Dsn128Ce9EUltW1IxSePTp1XykAfr1DgK4+5GVXH/buotXXXztYqDtOdfmL6oweEAD8xdXaIgi0y2t9lrr1SP43L8M5IbblvHMi45K0AaoU8ZLKc0EZlZfL46IKRQ9958AiIgAPgQc0tlzdNQHrwKsr9vZyOpn2oIteehBBh74dhoHDgKgsV9/moeP2KDvLrr3bga/43AABr717Sx55EGgGC4aPXpUj9ePvrvtwcq/P9/1Fy9tQRYsbmXugla2GVb0RHjjuN7MeGnN1ZtO+sYMTvp6sU1+eCk/uW4Of3lsGY89tZwD9urLgH7F/y769Wlg6OANm67z/seW8fb9i3k4DtirL49NLxZKaGqEf//ECG7/y5KNXpFUW4ZoaOiSTWsw422BGvoNoKF3XwCiuQe9d92LVc8/Tcuc2fQZ/5ZX92setd0GHW/543+l12770NCnLw19+tJrt31Y/vhfAeh30Dvpvfu+vDzpAuyeLHWPAX1fK1yM3aaJCFiyPDHludWM37Un/fsUn/ftFWw1YMP+P/nwtFW8Zc9iCpHxu/bgyeeKZzSNDXDSB/tzzyMreOBJVwnVhumqjBcREyPi/pptYpvnLIZ87gPcW9N8EDArpTStpm1sRPw1Im6PiIM6upeO/oV0CnBrREwDXqi2jQF2BD7T0cG1eRjzpa/Qd8+9aRowkF2uuIZZP7+MaCr+05j325toGrwVO37nRzT26QOVxNCjPshT/3YsK1/4Gy/97BK2/8Z/QUOQWlp48eLvsHr2rA7POe/3v2X0F89k55/8gtbFi3j+P84BoOeYMWx3wqeLEBbBy9f9ihXPPbtJ71/aElx63Vw+97FhNDUGs+au5uIr5/DOtxQrS91y9+I2vzdj1mqu+u18vnLi1kQEra2Jn1w3lznzO34a+cd7l/DZjwzje2eOYsmyChf9rOjt8Oa9+7LrDr3o37eBf6wW337wyzk896JBTKqjUzDjbXEaBw1m6CdOKYrNESz9y10sf+R+Vr/4PEM++ikGvedD0NjE0vv+zMIZz3V4vMrSJSz8za8Y+eULAFj4m1+9urjBkI9+ipa5sxl5xn8AsPTBySz8za822b1JW6JPHtWfnd/QTL/ewX9+djA33bGMxsaiYHb7gysYv2tPDt63F5UKrGpJTLqhyHwz57Ty69uW8vkPD6QBaK3AL25ewrxFHT9b+fNDKzjhyP6c96nBLF1R4UfVY75pt56MG9NM3z4NvGWvXgBc9j+LeWFW66a5ealGbc/69kREP+A64JSUUu3QuWOAK2vezwTGpJTmRsR44NcRsfta31nz2KmDJ0lRzDC3P0WXuQBmAH9JKW3Qb4nDQqX8fX3ny7r7EiRtYtdcNLbbxmYuu/TsLskSfT7xNceX1tjYjOewUClv3xh7SXdfgqQ6+MlZQ7eIjBcRzcBvgN+nlC6saW8C/g6MTynNaOO7twFfTCnd39bxOxzbk1KqAJM72k+SJGmTcEjnJmHGkyRJ3apOGa86p9olwJTawlrVO4AnawtrETEMmJdSao2I7YFxwDPtnWPDJs6RJEnqLutdkVCSJEmbtfplvAOBjwGPRsRD1bYzU0q/BSaw5pBQgLcB50REC9AKnJhSmtfeCSyuSZIkSZIkKUsppTsppsBY32cfX0/bdRRzs20wi2uSJKnUXOlTkiQpPzllPItrkiSp3CKf4CVJkqSqjDJePnciSZIkSZIk1Zk91yRJUrk1uKCBJElSdjLKeBbXJElSqUVGQwYkSZJUyCnjWVyTJEnlltFTTUmSJFVllPHyKRNKkiRJkiRJdWbPNUmSVG4ZDRmQJElSVUYZz+KaJEkqt8hnyIAkSZKqMsp4+ZQJJUmSNkJEXBoRsyPisZq2rSLiloiYVv05uOazMyJiekRMjYjDatrHR8Sj1c++G5FRcpQkSdI6LK5JkqRya2jomq1jPwUOX6vtdODWlNI44NbqeyJiN2ACsHv1OxdHRGP1Oz8EJgLjqtvax5QkSVL9Mt4mV46rkCRJaks0dM3WgZTSHcC8tZqPBC6vvr4cOKqm/aqU0sqU0rPAdGD/iBgJDEgp3ZNSSsAVNd+RJEnSK+qU8erBOdckSVK5ddEy7RExkaJH2SsmpZQmdfC1ESmlmQAppZkRMbzavi0wuWa/GdW21dXXa7dLkiSpVhdlvDKwuCZJkrYI1UJaR8W0DbW+NJjaaZckSVKmLK5JkqRy697u/rMiYmS119pIYHa1fQYwuma/UcCL1fZR62mXJElSrZIM6ewK+dyJJEnKU0TXbJ1zE3Bs9fWxwI017RMiomdEjKVYuOC+6hDSxRFxQHWV0H+t+Y4kSZJe0b0Zr0vZc02SJAmIiCuBg4GhETEDOBs4H7g6Io4HngeOBkgpPR4RVwNPAC3ASSml1uqhPkWx8mhv4HfVTZIkSZmyuCZJksqtTkusp5SOaeOjQ9vY/1zg3PW03w/s0YWXJkmSlJ86Zbx6sLgmSZLKrSTd/SVJktSFMsp4FtckSVK5ZTTZrSRJkqoyynj53IkkSZIkSZJUZ/ZckyRJ5ZbRfBySJEmqyijjWVyTJEnlltF8HJIkSarKKOPlUyaUJEmSJEmS6syea5IkqdwymuxWkiRJVRllPItrkiSp3DIaMiBJkqSqjDKexTVJklRuGU12K0mSpKqMMl4+dyJJkiRJkiTVmT3XJElSqaWMhgxIkiSpkFPGs7gmSZLKLaPJbiVJklSVUcbL504kSZIkSZKkOrPnmiRJKreMnmpKkiSpKqOMZ3FNkiSVWk7zcUiSJKmQU8azuCZJksoto6eakiRJqsoo4+VzJ5IkSZIkSVKdWVyTJEnlFtE1myRJksqjThkvIkZHxJ8iYkpEPB4RJ1fbvxoRf4+Ih6rbu2u+c0ZETI+IqRFxWEfncFioJEkqtwafBUqSJGWnfhmvBTg1pfRgRPQHHoiIW6qfXZRS+q/anSNiN2ACsDuwDfCHiNgppdTa1glMq5IkSZIkScpSSmlmSunB6uvFwBRg23a+ciRwVUppZUrpWWA6sH9757C4JkmSSi1FdMkmSZKk8uiqjBcREyPi/pptYlvnjIjtgH2Ae6tNn4mIRyLi0ogYXG3bFnih5mszaL8Y57BQSZJUchmtJCVJkqSqLsp4KaVJwKQOTxfRD7gOOCWltCgifgh8HUjVnxcAnwDW91Q2tXdsi2uSJKnUksU1SZKk7NQz40VEM0Vh7RcppesBUkqzaj7/MfCb6tsZwOiar48CXmzv+KZVSZIkSZIkZSkiArgEmJJSurCmfWTNbu8DHqu+vgmYEBE9I2IsMA64r71z2HNNkiSVm/OlSZIk5ad+Ge9A4GPAoxHxULXtTOCYiNibYsjnc8C/AaSUHo+Iq4EnKFYaPam9lULB4pokSSo5h4VKkiTlp14ZL6V0J+ufR+237XznXODcDT2HaVWSJEmSJEnqJHuuSZKkcnNYqCRJUn4yyngW1yRJUrk5LFSSJCk/GWU8i2uSJKnUUkZPNSVJklTIKePlUyaUJEmSJEmS6syea5IkqdwyGjIgSZKkqowynsU1SZJUamm9K6dLkiRpc5ZTxsunTChJkiRJkiTVmT3XJElSqaWMhgxIkiSpkFPGs7gmSZLKLaPgJUmSpKqMMp7FNUmSVGo5LdMuSZKkQk4ZL58yoSRJkiRJklRn9lyTJEmlltN8HJIkSSrklPEsrkmSpHLLaMiAJEmSqjLKePmUCSVJkiRJkqQ6s+eaJEkqtZyGDEiSJKmQU8azuCZJkkotkc+QAUmSJBVyyngW1yRJUqnl9FRTkiRJhZwyXj53IkmSJEmSJNWZPdckSVK5ZbSSlCRJkqoyyngW1yRJUqklO9pLkiRlJ6eMZ3FNkiSVWsroqaYkSZIKOWW8fMqEkiRJkiRJUp3Zc02SJJVaTitJSZIkqZBTxrO4JkmSSi2Rz5ABSZIkFXLKePmUCSVJkjZSRJwcEY9FxOMRcUq1bauIuCUiplV/Dq7Z/4yImB4RUyPisG67cEmSJHUbi2uSJKnUUjR0ydaRiNgD+CSwP7AX8J6IGAecDtyaUhoH3Fp9T0TsBkwAdgcOBy6OiMZN8ocgSZKUmXplvHoox1VIkiS1IUV0ybYBdgUmp5SWpZRagNuB9wFHApdX97kcOKr6+kjgqpTSypTSs8B0isKcJEmSOlDHjLfJWVyTJEmllogu2SJiYkTcX7NNXOtUjwFvi4ghEdEHeDcwGhiRUpoJUP05vLr/tsALNd+fUW2TJElSB7oq45WBCxpIkqQtQkppEjCpnc+nRMR/ALcAS4CHgZZ2Drm+NJc26iIlSZK02bG4JkmSSq2ec2mklC4BLgGIiPMoeqPNioiRKaWZETESmF3dfQZFz7ZXjAJerNvFSpIkbcbKMl9aV8jnTiRJUpbqOWQgIoZXf44B3g9cCdwEHFvd5Vjgxurrm4AJEdEzIsYC44D7uvDWJUmSsuWwUEmSpDxdFxFDgNXASSml+RFxPnB1RBwPPA8cDZBSejwirgaeoBg+elJKqbW7LlySJEndw+KaJEkqtToPCz1oPW1zgUPb2P9c4NxNfV2SJEm5yWlYqMU1SZJUamXp7i9JkqSuk1PGs7gmSZJKLaenmpIkSSrUK+NFxGjgCmBroAJMSil9JyK+BRwBrAKeBo5LKS2IiO2AKcDU6iEmp5RObO8cplVJkiRJkiTlqgU4NaW0K3AAcFJE7AbcAuyRUtoTeAo4o+Y7T6eU9q5u7RbWoA491z7d+LVNfQpJ3eyM7x/e3ZcgaVO7aGrH+2wiOQ0ZyMlHZ32huy9B0iZ0xuUHdvclSKqHs/LPeCmlmcDM6uvFETEF2Dal9H81u00GPtjZc9hzTZIklVqK6JJNkiRJ5dFVGS8iJkbE/TXbxLbOWR3yuQ9w71offQL4Xc37sRHx14i4PSLWWfBqbc65JkmSJEmSpM1SSmkSMKmj/SKiH3AdcEpKaVFN+1kUQ0d/UW2aCYxJKc2NiPHAryNi99rvrM3imiRJKrWU7HUmSZKUm3pmvIhopiis/SKldH1N+7HAe4BDU0qpuK60ElhZff1ARDwN7ATc39bxLa5JkqRSS85iIUmSlJ16ZbyICOASYEpK6cKa9sOB04C3p5SW1bQPA+allFojYntgHPBMe+ewuCZJkkrNBQ0kSZLyU8eMdyDwMeDRiHio2nYm8F2gJ3BLUX9jcnVl0LcB50REC9AKnJhSmtfeCSyuSZIkSZIkKUsppTthvZW837ax/3UUQ0g3mMU1SZJUavZckyRJyk9OGc/imiRJKrWcgpckSZIKOWU8ZwiWJEmSJEmSOsmea5IkqdRyeqopSZKkQk4Zz+KaJEkqtZTyCV6SJEkq5JTxLK5JkqRSy+mppiRJkgo5ZTznXJMkSZIkSZI6yZ5rkiSp1HJ6qilJkqRCThnP4pokSSq1nIKXJEmSCjllPIeFSpIkSZIkSZ1kzzVJklRqOa0kJUmSpEJOGc/imiRJKrVKRkMGJEmSVMgp41lckyRJpZbTfBySJEkq5JTxnHNNkiRJkiRJ6iR7rkmSpFLLaT4OSZIkFXLKeBbXJElSqeU0ZECSJEmFnDKew0IlSZIkSZKkTrLnmiRJKrWchgxIkiSpkFPGs7gmSZJKLachA5IkSSrklPEsrkmSpFLL6ammJEmSCjllPOdckyRJkiRJkjrJnmuSJKnUKt19AZIkSepyOWU8i2uSJKnUchoyIEmSpEJOGc9hoZIkSZIkSVIn2XNNkiSVWk4rSUmSJKmQU8azuCZJkkotpyEDkiRJKuSU8SyuSZKkUsvpqaYkSZIKOWU851yTJEmSJEmSOsmea5IkqdQqqbuvQJIkSV0tp4xncU2SJJVaTkMGJEmSVMgp4zksVJIkSZIkSeoke65JkqRSy2klKUmSJBVyyngW1yRJUqmljObjkCRJUiGnjGdxTZIklVolo/k4JEmSVMgp4znnmiRJkiRJktRJFtckSVKppRRdskmSJKk86pXxImJ0RPwpIqZExOMRcXK1fauIuCUiplV/Dq75zhkRMT0ipkbEYR2dw+KaJEkqtZS6ZpMkSVJ51DHjtQCnppR2BQ4AToqI3YDTgVtTSuOAW6vvqX42AdgdOBy4OCIa2zuBxTVJkiRJkiRlKaU0M6X0YPX1YmAKsC1wJHB5dbfLgaOqr48ErkoprUwpPQtMB/Zv7xwW1yRJUqkloks2SZIklUdXZbyImBgR99dsE9s6Z0RsB+wD3AuMSCnNhKIABwyv7rYt8ELN12ZU29rkaqGSJKnUKg7plCRJyk5XZbyU0iRgUkf7RUQ/4DrglJTSoog2H76u74N2r9bimiRJKjUXI5AkScpPPTNeRDRTFNZ+kVK6vto8KyJGppRmRsRIYHa1fQYwuubro4AX2zu+w0IlSZIkSZKUpSi6qF0CTEkpXVjz0U3AsdXXxwI31rRPiIieETEWGAfc1945LK5JkqRSq+dqoRHx+eoS7Y9FxJUR0asrl2mXJElSoY4Z70DgY8AhEfFQdXs3cD7wzoiYBryz+p6U0uPA1cATwM3ASSml1vZO4LBQSZJUapU6LUYQEdsCnwN2Syktj4irKZZh341imfbzI+J0imXaT1trmfZtgD9ExE4dhS9JkiTVL+OllO5k/fOoARzaxnfOBc7d0HPYc02SJOk1TUDviGgC+lDMr9Fly7RLkiQpPxbXJElSqXXVkIGOlmlPKf0d+C/geWAmsDCl9H904TLtkiRJKtRz6o9NzWGhkiSp1LpqJamOlmmvzqV2JDAWWABcExEfbeeQr3uZdkmSJBVyWhHe4pokSSq1Sv3KVe8Ank0pvQwQEdcDb6ELl2mXJElSoY4Zb5NzWKgkSVLheeCAiOhTXbL9UGAKXbhMuyRJkvJjzzVJklRq9ZpLI6V0b0RcCzwItAB/pRhG2g+4OiKOpyjAHV3d//HqiqJPVPfvcJl2SZIkFcoyX1pXsLgmSZJKLdVpmXaAlNLZwNlrNa+ki5ZplyRJUqGeGW9Tc1ioJEmSJEmS1En2XJMkSaWW02S3kiRJKuSU8SyuSZKkUstpPg5JkiQVcsp4FtckSVKp5RS8JEmSVMgp4znnmiRJkiRJktRJ9lyTJEmlVkn5rCQlSZKkQk4Zz+KaJEkqtZyGDEiSJKmQU8ZzWKgkSZIkSZLUSfZckyRJpZbTU01JkiQVcsp4FtckSVKpVTIKXpIkSSrklPEsrkmSpFJLGU12K0mSpEJOGc851yRJkiRJkqROsueaJEkqtZzm45AkSVIhp4xncU2SJJVaTvNxSJIkqZBTxnNYqCRJkiRJktRJ9lyTJEmlltOQAUmSJBVyyngW1yRJUqnlFLwkSZJUyCnjWVyTJEmlltN8HJIkSSrklPGcc02SJEmSJEnqJHuuSZKkUstpyIAkSZIKOWU8i2uSJKnUKpXuvgJJkiR1tZwynsNCJUmSJEmSpE6y55okSSq1nIYMSJIkqZBTxrO4JkmSSi2n4CVJkqRCThnP4pokSSq1nJZplyRJUiGnjOeca5IkSZIkSVIn2XNNkiSVWuqyMQPRRceRJEnSxsop41lckyRJpZbTfBySJEkq1CvjRcSlwHuA2SmlPaptvwJ2ru4yCFiQUto7IrYDpgBTq59NTimd2NE5LK6Ja37yDyxb3kKlAq2tiRO+8OAan/fv28QZJ+/MNlv3YtXqCt/8zlSefX7ZRp2zuSn48hd2Yecd+rNo8Wr+338+wUuzV7Lj2L588dM70bdPI62tiSuufp4/3vnyRp1L2tLt+ePzGP7ug1k1ey537HPEOp9vc8wR7PDvnwSgdclSHv3MV1n8yNR19ns9Gno0s9dl/8nAfXdn1bwF/PXDn2f53/5O7zHbMP6a70FjIw1NTTx38c95ftJVG3UuSdK6zvjcTrzlTUOYv3A1//qZ+9f5/J1vH85HPjAagOUrWrng4mlMf27pRp3TfCfVV0PPHrz5T7+goWcPorGRmdf/nmnnfG+NfUYccSg7fe1kUqVCamnliVPPY/5dD2zcedvIeQP22oU9vv9Vmvr3I1UqTP/mD5l5ze826lxSF/kp8H3gilcaUkr/8srriLgAWFiz/9Mppb1fzwmcc00AfO6shznu5AfWKawBfOxDY5j2zBI+/rkH+MZFT3LyxB03+LhbD+/J987ba53297xrJIuXtDDh3+7jVzfO4FMf3x6AlSsrfOPCJ/nYSfdz6lcf5XOf3IF+fRs7f2OSmHH59dz3nhPa/Hz5czO455CP8ud938u0c3/IG3/49Q0+du83bMsBf7hinfbRnzia1QsWcduu7+LZ7/yUXc77IgArZr7M3QdN4M79juKuAz/EDv/+SXqOHP76b0pblEqlazZpS/LbW2dx6lcfbfPzmbNW8NkzHubjn3uAy3/1PF/6zE4bfGzznVQOlZWrmPzOY/nz+CP5835HMeywgxj0D2v+bs754z38ed/3cud+R/HIJ89kz//+xgYf//XmvNZlK3jouNO4Y+/3cN8/n8BuF5xJ08D+G3eTylq9Ml5K6Q5g3vo+i4gAPgRcuTH3YnFNHdpudB8eeGQ+AM/PWM7I4b0YPKgZgHcdPJxJF+zDZd8Zz7+fNI6GDfwv6q3/MITf3ToLgNvuepnxew0G4IUXlzNj5nIA5s5bxYKFqxk0oEcX35G0ZZl35/2snrewzc/n3/NXWhYsKl7f+xC9t9361c+2/fB7OfDua3jr/b9mj4u/xob+ko844hBm/OwGAF667vcMPeTNAKTVq6msWg0UT1tjQ//S0BYtpa7ZpC3Jw48vZNHi1W1+/tiTi1i8tAWAx59cxLChPV/9zHwnbT5alxYjiqK5iYbmpnX+h/fK5wCNfXuv8XlX57yl055j2fS/AbBy5mxWvTyPHsO26vzNKXslyXgHAbNSStNq2sZGxF8j4vaIOGhDDuK/akQiceE5e3LJRfvy3sNGrvP59GeX8rY3DwNg13H9GTG8F8OH9OQNo/pw6EHD+dSXHuK4kx+gUkm86+0jNuicw4b0ZPacFQC0VmDp0hYGDlhzlPKu4/rT1BT8/aXlG3mHkjbUmOM+yOzf3wFAv122Z+TR/8TdbzuGO/c7ClorbPvhdYeVrk+vbUaw4oWZAKTWVlYvXEzzkOIfWb1Gbc1BD97Eoc/extP/9WNWzpy9Se5F+aikrtkkrd973rU1kx8oHuib76TNTEMDb73/17zzxbuZ84e7WXDfI+vsMuLId/D2R3/Hm278EQ9PPBPYdDnvFQPf9EYamptZ9vTzG3d/ylpXZbyImBgR99dsE1/HZRzDmr3WZgJjUkr7AF8AfhkRAzo6SKfnXIuI41JKl7Xx2URgIsAObzyVrd+wYb+k6h6f+tJDzJ23ikEDm/n21/fkbzOW8fDjr/Vy+fm1z3PyxB257DvjefpvS5n2zGJaWxPj9xrEzjv04ycX7gtAzx4NzF9QPCE978zdGTmiF01NwYhhvbjsO+MBuOamGfz21lnEehbzqK04Dxncg698YRfO/faT9jaQ6mTI2/+B0cd9kLsP/nDx/pA3M3DfPThw8rUANPbqxcrZcwEYf8336T12FA3NzfQeM5K33v9rAJ773hXMuPx62vslXzHjJf6873vpOXI4+133A2Ze93tWVY8rqfuZ8bYs+7xxEP/8zq359GkPAZjvpM1NpcKd+x1F08D+7HftD+i3+ziWPD5tjV1m3fgHZt34B7Z6637s/NWTuffw4zZZzgPoufUw9r7sWzx8/Gl2HVddpJQmAZNe7/ciogl4PzC+5lgrgZXV1w9ExNPATsC6E5jW2JgFDb4GrDd41d7YW4+43d+mkps7bxUACxau5o575rDbTv3XKK4tW97KN7/z2uTm1/zkH3hx1gr22mMgv/vjLH50xbPrHPPM8x4Hijk5zjplFz575sNrfD57zkqGD+3Fy3NX0dgAffs2sWhxMTShT+9G/vPsPfjxz5/l8amLu/x+Ja2r/xt35o0/+gZ/OeKTrJ63AICIYMbPbmDqly9cZ/8Hjv4MUMzFsdcl32TyO/51jc9X/P0leo0eyYq/zyIaG2ke2P/V475i5czZLH5iGlu9dT9euv73m+S+lAdzed2Z8bYQO2zXl9M/uxNf/Oqjr+awCMx30maoZeFi5t5+L8PfddA6xbVXzLvzfvpsP4bmIYM3Wc5r6t+XN930I6ae/W0W3Pvw2oeW1lCCjPcO4MmU0oxXGiJiGDAvpdQaEdsD44BnOjpQu8NCI+KRNrZHgQ3rH65S69Wzgd69G199/aZ9BvPM39ZcKapf30aamoqnE0e8a2sefnwBy5a38sDDCzj4wKEMGljMv9a/XxMjhvVkQ9x171z+6dDiP6GDDxzGg9U53ZqagvPO2p2b/ziLP901p0vuUVL7eo0eyfirv8fDx32JpdOee7V9zh/vYeT7D3t1rozmwQPpPWabDTrmrN/8kVEfex8AW3/gMOb8aXJxrm1H0NCr+HuiadAABr95X5Y+te4/4KRaqZK6ZNNrzHgaMawn556xO1+/8EleePG1IZrmO2nz0WPo4FcXDGjo1ZOhh76FJVPXrAH02WHMq68H7LMbDT2aWT13/ibJedHczPhrf8CMn9/IS9fdvNH3p/zVK+NFxJXAPcDOETEjIo6vfjSBdRcyeBvwSEQ8DFwLnJhSWu9iCLU66rk2AjgMmL/2tQF3d3Rwld9Wg3pw3lm7A9DYGNxy+2zufXA+Rx5ezL12480zecOovnz5CztTqcBzzy/l/O8+BcBzLyzjxz97jovO2ZMIaG1NXPjf05j18soOz/ubW2bylS/sylU/2p9FS1bz1f+cAsAhbx3G3rsPZGD/Zt59aDGp+rnffpLpz27c0vDSlmzvn13AkLfvT4+hgznk2duZds73iObir//nJ13FuC+fRI8hg9j9e2cDkFpaueuAD7BkytNMPfvb7P+7S4mGBtLq1Tz2uXNY/vyLHZ7zhUuvZe+ffouDp/wfq+cv5MGPfB6AfrvswK7fOr14TBXBMxddyuLHntp0Ny+pLWa8zH31i7uy9xsHMmhAM9dfdgCX/PI5mhqLh6U33jyTj094AwMHNHHqp8YBRY474QsPmu+kzUjPkcPZ69LzicZGIoIXr72Z2b+9jTETJwBFztv6fYcx6qNHUmlpobJ8xauZbFPkvG2O/ie2Omg/mocMYtS/FsW3R44/nUUPP7mJ/gSkDZNSOqaN9o+vp+064LrXe45I7fTDi4hLgMtSSneu57NfppQ+3NEJHDIg5e+Mm1/PfJGSNkf/vHrqeiZYqY//vK5rup196QMN3XYPZWPGk9QR8520ZTDjdY12e66llI5v57MOQ5ckSdLGKsF8HNkx40mSpO6WU8bbmAUNJEmSNrmK86VJkiRlJ6eM1+6CBpIkSZIkSZLaZs81SZJUajkNGZAkSVIhp4xncU2SJJVaTsFLkiRJhZwynsNCJUmSJEmSpE6y55okSSq1Sk6PNSVJkgTklfEsrkmSpFJLle6+AkmSJHW1nDKexTVJklRqKaOnmpIkSSrklPGcc02SJEmSJEnqJHuuSZKkUqtkNGRAkiRJhZwynsU1SZJUajkNGZAkSVIhp4znsFBJkiRJkiSpk+y5JkmSSq2Sz0NNSZIkVeWU8ey5JkmSSi1VUpdsHYmInSPioZptUUScEhFbRcQtETGt+nNwzXfOiIjpETE1Ig7bpH8QkiRJGalXxqsHi2uSJKnUUuqarePzpKkppb1TSnsD44FlwA3A6cCtKaVxwK3V90TEbsAEYHfgcODiiGjcFH8GkiRJualXxqsHi2uSJEnrOhR4OqX0N+BI4PJq++XAUdXXRwJXpZRWppSeBaYD+9f7QiVJktS9nHNNkiSVWqWLuvtHxERgYk3TpJTSpDZ2nwBcWX09IqU0EyClNDMihlfbtwUm13xnRrVNkiRJHeiqjFcGFtckSVKpddUy7dVCWlvFtFdFRA/gvcAZHe26vtN04tIkSZK2OF2V8crAYaGSJElr+ifgwZTSrOr7WRExEqD6c3a1fQYwuuZ7o4AX63aVkiRJKgWLa5IkqdRSpWu21+EYXhsSCnATcGz19bHAjTXtEyKiZ0SMBcYB923c3UqSJG0ZuiHjbTIOC5UkSaVWqeOQgYjoA7wT+Lea5vOBqyPieOB54GiAlNLjEXE18ATQApyUUmqt28VKkiRtxuqZ8TY1i2uSJKnU6jkfR0ppGTBkrba5FKuHrm//c4Fz63BpkiRJWXHONUmSJEmSJEn2XJMkSeWW0zLtkiRJKuSU8SyuSZKkUstoxIAkSZKqcsp4DguVJEmSJEmSOsmea5IkqdRSRkMGJEmSVMgp41lckyRJpZbTMu2SJEkq5JTxLK5JkqRSy+mppiRJkgo5ZTznXJMkSZIkSZI6yZ5rkiSp1HJ6qilJkqRCThnP4pokSSq1jHKXJEmSqnLKeA4LlSRJkiRJkjrJ4pokSSq1VEldskmSJKk86pXxIuLSiJgdEY/VtH01Iv4eEQ9Vt3fXfHZGREyPiKkRcdiG3IvDQiVJUqmljJZplyRJUqGOGe+nwPeBK9Zqvyil9F+1DRGxGzAB2B3YBvhDROyUUmpt7wQW1yRJUqlV7HUmSZKUnXplvJTSHRGx3QbufiRwVUppJfBsREwH9gfuae9LDguVJEmSJEnSZikiJkbE/TXbxA386mci4pHqsNHB1bZtgRdq9plRbWuXPdckSVKpOSxUkiQpP12V8VJKk4BJr/NrPwS+DqTqzwuATwCxvlN0dDCLa5IkqdRcjECSJCk/3ZnxUkqzXnkdET8GflN9OwMYXbPrKODFjo7nsFBJkiRJkiRtMSJiZM3b9wGvrCR6EzAhInpGxFhgHHBfR8ez55okSSo1e65JkiTlp14ZLyKuBA4GhkbEDOBs4OCI2JtiyOdzwL8BpJQej4irgSeAFuCkjlYKBYtrkiSp5CrOuSZJkpSdemW8lNIx62m+pJ39zwXOfT3nsLgmSZJKzZ5rkiRJ+ckp4znnmiRJkiRJktRJ9lyTJEml1lXLtEuSJKk8csp4FtckSVKpVTIaMiBJkqRCThnPYaGSJEmSJElSJ9lzTZIklVpOk91KkiSpkFPGs7gmSZJKLaf5OCRJklTIKeNZXJMkSaWWKpXuvgRJkiR1sZwynnOuSZIkSZIkSZ1kzzVJklRqOa0kJUmSpEJOGc/imiRJKrWc5uOQJElSIaeM57BQSZIkSZIkqZPsuSZJkkotp2XaJUmSVMgp41lckyRJpZZT8JIkSVIhp4xncU2SJJVaJeWzTLskSZIKOWU851yTJEmSJEmSOsmea5IkqdRyGjIgSZKkQk4Zz+KaJEkqtZyClyRJkgo5ZTyHhUqSJEmSJEmdZM81SZJUainl81RTkiRJhZwynsU1SZJUapVKPitJSZIkqZBTxrO4JkmSSi2n+TgkSZJUyCnjOeeaJEmSJEmS1En2XJMkSaWWUj5DBiRJklTIKeNZXJMkSaWW05ABSZIkFXLKeA4LlSRJkiRJkjrJnmuSJKnUcnqqKUmSpEJOGc/imiRJKrVKRvNxSJIkqZBTxrO4JkmSSi2np5qSJEkq5JTxnHNNkiRJkiRJ6iR7rkmSpFJLlXyGDEiSJKmQU8az55okSSq1VEldsm2IiBgUEddGxJMRMSUi3hwRW0XELRExrfpzcM3+Z0TE9IiYGhGHbbI/BEmSpMzUM+NtahbXJEmSXvMd4OaU0i7AXsAU4HTg1pTSOODW6nsiYjdgArA7cDhwcUQ0dstVS5Ikqds4LFSSJJVaqtNKUhExAHgb8PHivGkVsCoijgQOru52OXAbcBpwJHBVSmkl8GxETAf2B+6pywVLkiRtxuqV8erB4pokSSq1Shd194+IicDEmqZJKaVJNe+3B14GLouIvYAHgJOBESmlmQAppZkRMby6/7bA5Jrvz6i2SZIkqQNdlfHKwOKaJEkqta6a7LZaSJvUzi5NwL7AZ1NK90bEd6gOAW1DrO80G3GJkiRJW4x6LWgQEZcC7wFmp5T2qLZ9CzgCWAU8DRyXUloQEdtRTAsytfr1ySmlEzs6h3OuSZIkFWYAM1JK91bfX0tRbJsVESMBqj9n1+w/uub7o4AX63StkiRJ2jA/pZgft9YtwB4ppT2Bp4Azaj57OqW0d3XrsLAGFtckSVLJ1WslqZTSS8ALEbFztelQ4AngJuDYatuxwI3V1zcBEyKiZ0SMBcYB93XlvUuSJOWqjhnvDmDeWm3/l1Jqqb6dTPGQtNMcFipJkkqtzpPdfhb4RUT0AJ4BjqN4GHl1RBwPPA8cXVxXejwirqYowLUAJ6WUWut5sZIkSZurrsp4GzCvbkc+Afyq5v3YiPgrsAj4ckrpzx0dwOKaJElSVUrpIWC/9Xx0aBv7nwucuymvSZIkSW3bgHl12xQRZ1E8JP1FtWkmMCalNDcixgO/jojdU0qL2juOxTVJklRqG9LdX5IkSZuX7s54EXEsxUIHh6aUEkBKaSWwsvr6gYh4GtgJuL+9Y1lckyRJpVavlaQkSZJUP92Z8SLicOA04O0ppWU17cOAeSml1ojYnmJO3Wc6PF61OCd1mYiY+DrHN0vazPh7LklbHv/ul/Ln77lyFBFXAgcDQ4FZwNkUq4P2BOZWd5ucUjoxIj4AnEMxVLQVODul9D8dnsPimrpaRNyfUlrffDWSMuHvuSRtefy7X8qfv+dS5zR09wVIkiRJkiRJmyuLa5IkSZIkSVInWVzTpuAYfSl//p5L0pbHv/ul/Pl7LnWCc65JkiRJkiRJnWTPNUmSJEmSJKmTLK5JkiRJkiRJnWRxTV0qIg6PiKkRMT0iTu/u65HUtSLi0oiYHRGPdfe1SJLqw3wn5c+MJ20ci2vqMhHRCPwA+CdgN+CYiNite69KUhf7KXB4d1+EJKk+zHfSFuOnmPGkTrO4pq60PzA9pfRMSmkVcBVwZDdfk6QulFK6A5jX3dchSaob8520BTDjSRvH4pq60rbACzXvZ1TbJEmStHky30mS1AGLa+pKsZ62VPerkCRJUlcx30mS1AGLa+pKM4DRNe9HAS9207VIkiRp45nvJEnqgMU1daW/AOMiYmxE9AAmADd18zVJkiSp88x3kiR1wOKaukxKqQX4DPB7YApwdUrp8e69KkldKSKuBO4Bdo6IGRFxfHdfkyRp0zHfSVsGM560cSIlp0yQJEmSJEmSOsOea5IkSZIkSVInWVyTJEmSJEmSOsnimiRJkiRJktRJFtckSZIkSZKkTrK4JkmSJEmSJHWSxTVJkiRJkiSpkyyuSZIkSZIkSZ30/wFiJyil1BqKCQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1296x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "classification_metrics(y_train,svclassifier.predict(X_train),y_test,svclassifier.predict(X_test),True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "v3UNucbSy4Md"
   },
   "source": [
    "#### <span style=\"color:#FF7B07\"> Let's see which params to use for SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "CxTLioaoolpI"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 6 candidates, totalling 30 fits\n",
      "[CV] C=10, gamma=1, kernel=linear ....................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ........ C=10, gamma=1, kernel=linear, score=0.643, total=   2.6s\n",
      "[CV] C=10, gamma=1, kernel=linear ....................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    2.5s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ........ C=10, gamma=1, kernel=linear, score=0.660, total=   1.4s\n",
      "[CV] C=10, gamma=1, kernel=linear ....................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    3.9s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ........ C=10, gamma=1, kernel=linear, score=0.654, total=   1.5s\n",
      "[CV] C=10, gamma=1, kernel=linear ....................................\n",
      "[CV] ........ C=10, gamma=1, kernel=linear, score=0.670, total=   1.6s\n",
      "[CV] C=10, gamma=1, kernel=linear ....................................\n",
      "[CV] ........ C=10, gamma=1, kernel=linear, score=0.638, total=   1.9s\n",
      "[CV] C=10, gamma=0.1, kernel=linear ..................................\n",
      "[CV] ...... C=10, gamma=0.1, kernel=linear, score=0.643, total=   2.6s\n",
      "[CV] C=10, gamma=0.1, kernel=linear ..................................\n",
      "[CV] ...... C=10, gamma=0.1, kernel=linear, score=0.660, total=   1.5s\n",
      "[CV] C=10, gamma=0.1, kernel=linear ..................................\n",
      "[CV] ...... C=10, gamma=0.1, kernel=linear, score=0.654, total=   1.5s\n",
      "[CV] C=10, gamma=0.1, kernel=linear ..................................\n",
      "[CV] ...... C=10, gamma=0.1, kernel=linear, score=0.670, total=   1.6s\n",
      "[CV] C=10, gamma=0.1, kernel=linear ..................................\n",
      "[CV] ...... C=10, gamma=0.1, kernel=linear, score=0.638, total=   1.9s\n",
      "[CV] C=10, gamma=0.01, kernel=linear .................................\n",
      "[CV] ..... C=10, gamma=0.01, kernel=linear, score=0.643, total=   2.6s\n",
      "[CV] C=10, gamma=0.01, kernel=linear .................................\n",
      "[CV] ..... C=10, gamma=0.01, kernel=linear, score=0.660, total=   1.4s\n",
      "[CV] C=10, gamma=0.01, kernel=linear .................................\n",
      "[CV] ..... C=10, gamma=0.01, kernel=linear, score=0.654, total=   1.5s\n",
      "[CV] C=10, gamma=0.01, kernel=linear .................................\n",
      "[CV] ..... C=10, gamma=0.01, kernel=linear, score=0.670, total=   1.7s\n",
      "[CV] C=10, gamma=0.01, kernel=linear .................................\n",
      "[CV] ..... C=10, gamma=0.01, kernel=linear, score=0.638, total=   2.1s\n",
      "[CV] C=100, gamma=1, kernel=linear ...................................\n",
      "[CV] ....... C=100, gamma=1, kernel=linear, score=0.643, total=  14.0s\n",
      "[CV] C=100, gamma=1, kernel=linear ...................................\n",
      "[CV] ....... C=100, gamma=1, kernel=linear, score=0.660, total=   8.8s\n",
      "[CV] C=100, gamma=1, kernel=linear ...................................\n",
      "[CV] ....... C=100, gamma=1, kernel=linear, score=0.654, total=  14.8s\n",
      "[CV] C=100, gamma=1, kernel=linear ...................................\n",
      "[CV] ....... C=100, gamma=1, kernel=linear, score=0.670, total=  10.7s\n",
      "[CV] C=100, gamma=1, kernel=linear ...................................\n",
      "[CV] ....... C=100, gamma=1, kernel=linear, score=0.638, total=  13.8s\n",
      "[CV] C=100, gamma=0.1, kernel=linear .................................\n",
      "[CV] ..... C=100, gamma=0.1, kernel=linear, score=0.643, total=  13.3s\n",
      "[CV] C=100, gamma=0.1, kernel=linear .................................\n",
      "[CV] ..... C=100, gamma=0.1, kernel=linear, score=0.660, total=   8.9s\n",
      "[CV] C=100, gamma=0.1, kernel=linear .................................\n",
      "[CV] ..... C=100, gamma=0.1, kernel=linear, score=0.654, total=   9.1s\n",
      "[CV] C=100, gamma=0.1, kernel=linear .................................\n",
      "[CV] ..... C=100, gamma=0.1, kernel=linear, score=0.670, total=  10.2s\n",
      "[CV] C=100, gamma=0.1, kernel=linear .................................\n",
      "[CV] ..... C=100, gamma=0.1, kernel=linear, score=0.638, total=  11.7s\n",
      "[CV] C=100, gamma=0.01, kernel=linear ................................\n",
      "[CV] .... C=100, gamma=0.01, kernel=linear, score=0.643, total=  12.6s\n",
      "[CV] C=100, gamma=0.01, kernel=linear ................................\n",
      "[CV] .... C=100, gamma=0.01, kernel=linear, score=0.660, total=   7.9s\n",
      "[CV] C=100, gamma=0.01, kernel=linear ................................\n",
      "[CV] .... C=100, gamma=0.01, kernel=linear, score=0.654, total=   8.8s\n",
      "[CV] C=100, gamma=0.01, kernel=linear ................................\n",
      "[CV] .... C=100, gamma=0.01, kernel=linear, score=0.670, total=  11.6s\n",
      "[CV] C=100, gamma=0.01, kernel=linear ................................\n",
      "[CV] .... C=100, gamma=0.01, kernel=linear, score=0.638, total=  15.6s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  30 out of  30 | elapsed:  3.3min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(estimator=SVC(),\n",
       "             param_grid={'C': [10, 100], 'gamma': [1, 0.1, 0.01],\n",
       "                         'kernel': ['linear']},\n",
       "             verbose=3)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV \n",
    "# defining parameter range \n",
    "param_grid = {'C': [10, 100],  \n",
    "              'gamma': [1, 0.1, 0.01], \n",
    "              'kernel': ['linear']}  \n",
    "  \n",
    "grid = GridSearchCV(SVC(), param_grid, refit = True, verbose = 3) \n",
    "  \n",
    "# fitting the model for grid search \n",
    "grid.fit(X_train, y_train) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ho2SpfXbs3W_",
    "outputId": "b4037061-4286-4ba2-d116-4976d5c8cb1e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=10, gamma=1, kernel='linear')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "svclassifier = SVC(kernel='linear',gamma=1,C=10)\n",
    "svclassifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 466
    },
    "id": "eoKzpBVLs9xW",
    "outputId": "89d64a05-4a67-470c-9cb1-d1a7944bee52"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data:\t  \t\t\t\t\t\ttest data:\t\n",
      "\n",
      "accuracy:\t 0.65 \t\t\t\t\t\taccuracy:\t 0.68\n",
      "precision:\t 0.63 \t\t\t\t\t\tprecision:\t 0.65\n",
      "recall:\t\t 0.74 \t\t\t\t\t\trecall:\t\t 0.78\n",
      "F1:\t\t 0.68 \t\t\t\t\t\tF1:\t\t 0.71\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "classification_metrics(y_train,svclassifier.predict(X_train),y_test,svclassifier.predict(X_test),True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CeN0zELME2YW"
   },
   "source": [
    "# <span style=\"color:#FF7B07\"><div align=\"center\">**Result** "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### <span style=\"color:#FF7B07\"> The results are amazing, we can guess if the person is going to be on 180 day program with only her/his first and 10nd day results and our new features ! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "predict_180days_Diet.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}