COVID_19_Model_Comparison.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "COVID-19 Model Comparison",
      "provenance": [],
      "collapsed_sections": [
        "wuEnp-EiF_X9",
        "Jmgcf3iLeK0z",
        "mO_GHtYI0rMw",
        "nRZPxsBo1F_c",
        "drOJRhkW2voo",
        "aejoz6yW3R_1",
        "MC2MzVvI6iMm",
        "WKx7g9wQ513H"
      ]
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "MyiRwr5uJHEf",
        "colab_type": "code",
        "outputId": "da4a0546-bf5e-4f0a-f808-621c114c6690",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/gdrive')\n",
        "%cd /gdrive/My Drive/cvd"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /gdrive\n",
            "/gdrive/My Drive/cvd\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7PLKtB7CJe6l",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import datetime as dt\n",
        "import sklearn\n",
        "from scipy import stats\n",
        "from sklearn import preprocessing\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import recall_score as rs\n",
        "from sklearn.metrics import precision_score as ps\n",
        "from sklearn.metrics import f1_score as fs\n",
        "from sklearn.metrics import log_loss\n",
        "\n",
        "encoder = preprocessing.LabelEncoder()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wuEnp-EiF_X9",
        "colab_type": "text"
      },
      "source": [
        "###Data Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AJG6HG94N-HB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data = pd.read_csv('data.csv')\n",
        "data = data.drop('id',axis=1)\n",
        "data = data.fillna(np.nan,axis=0)\n",
        "data['location'] = encoder.fit_transform(data['location'].astype(str))\n",
        "data['country'] = encoder.fit_transform(data['country'].astype(str))\n",
        "data['gender'] = encoder.fit_transform(data['gender'].astype(str))\n",
        "data[['symptom1']] = encoder.fit_transform(data['symptom1'].astype(str))\n",
        "data[['symptom2']] = encoder.fit_transform(data['symptom2'].astype(str))\n",
        "data[['symptom3']] = encoder.fit_transform(data['symptom3'].astype(str))\n",
        "data[['symptom4']] = encoder.fit_transform(data['symptom4'].astype(str))\n",
        "data[['symptom5']] = encoder.fit_transform(data['symptom5'].astype(str))\n",
        "data[['symptom6']] = encoder.fit_transform(data['symptom6'].astype(str))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PSV2sW29OEff",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data['sym_on'] = pd.to_datetime(data['sym_on'])\n",
        "data['hosp_vis'] = pd.to_datetime(data['hosp_vis'])\n",
        "data['sym_on']= data['sym_on'].map(dt.datetime.toordinal)\n",
        "data['hosp_vis']= data['hosp_vis'].map(dt.datetime.toordinal)\n",
        "data['diff_sym_hos']= data['hosp_vis'] - data['sym_on']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rgeHinTzOJg0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data['diff_symp_hos'] = data['hosp_vis']-data['sym_on']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zmjM6wk7Q6jD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data = data.drop(['sym_on','hosp_vis'],axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7yWKs3_ZQxUH",
        "colab_type": "code",
        "outputId": "7414216e-1914-4eca-f268-d8bdba91cb03",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        }
      },
      "source": [
        "print(data.dtypes)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "location           int64\n",
            "country            int64\n",
            "gender             int64\n",
            "age              float64\n",
            "vis_wuhan          int64\n",
            "from_wuhan       float64\n",
            "death              int64\n",
            "recov              int64\n",
            "symptom1           int64\n",
            "symptom2           int64\n",
            "symptom3           int64\n",
            "symptom4           int64\n",
            "symptom5           int64\n",
            "symptom6           int64\n",
            "diff_sym_hos       int64\n",
            "diff_symp_hos      int64\n",
            "dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jmgcf3iLeK0z",
        "colab_type": "text"
      },
      "source": [
        "###Training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WlP3cXzdd3OR",
        "colab_type": "code",
        "outputId": "428f16da-6eb9-4477-b38e-082058f28395",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "tdata = pd.read_csv('train.csv')\n",
        "print(tdata.head())"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   id      location country gender  ...  symptom3 symptom4 symptom5  symptom6\n",
            "0  49  Wuhan, Hubei   China   male  ...       NaN      NaN      NaN       NaN\n",
            "1  50  Wuhan, Hubei   China   male  ...       NaN      NaN      NaN       NaN\n",
            "2  51  Wuhan, Hubei   China   male  ...       NaN      NaN      NaN       NaN\n",
            "3  52  Wuhan, Hubei   China   male  ...       NaN      NaN      NaN       NaN\n",
            "4  53  Wuhan, Hubei   China   male  ...       NaN      NaN      NaN       NaN\n",
            "\n",
            "[5 rows x 16 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "XDpFK6U1eXcD",
        "colab": {}
      },
      "source": [
        "tdata = pd.read_csv('train.csv')\n",
        "tdata = tdata.drop('id',axis=1)\n",
        "tdata = tdata.fillna(np.nan,axis=0)\n",
        "tdata['age'] = tdata['age'].fillna(value=tdata['age'].mean())\n",
        "tdata['location'] = encoder.fit_transform(tdata['location'].astype(str))\n",
        "tdata['country'] = encoder.fit_transform(tdata['country'].astype(str))\n",
        "tdata['gender'] = encoder.fit_transform(tdata['gender'].astype(str))\n",
        "tdata[['symptom1']] = encoder.fit_transform(tdata['symptom1'].astype(str))\n",
        "tdata[['symptom2']] = encoder.fit_transform(tdata['symptom2'].astype(str))\n",
        "tdata[['symptom3']] = encoder.fit_transform(tdata['symptom3'].astype(str))\n",
        "tdata[['symptom4']] = encoder.fit_transform(tdata['symptom4'].astype(str))\n",
        "tdata[['symptom5']] = encoder.fit_transform(tdata['symptom5'].astype(str))\n",
        "tdata[['symptom6']] = encoder.fit_transform(tdata['symptom6'].astype(str))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "cxxY76mHeXcV",
        "colab": {}
      },
      "source": [
        "tdata['sym_on'] = pd.to_datetime(tdata['sym_on'])\n",
        "tdata['hosp_vis'] = pd.to_datetime(tdata['hosp_vis'])\n",
        "tdata['sym_on']= tdata['sym_on'].map(dt.datetime.toordinal)\n",
        "tdata['hosp_vis']= tdata['hosp_vis'].map(dt.datetime.toordinal)\n",
        "tdata['diff_sym_hos']= tdata['hosp_vis'] - tdata['sym_on']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "wrUP1Kx3eXcp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        },
        "outputId": "7f720c49-7151-47ab-c6c1-722542bf4d4d"
      },
      "source": [
        "tdata = tdata.drop(['sym_on','hosp_vis'],axis=1)\n",
        "print(tdata)\n"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "     location  country  gender  ...  symptom5  symptom6  diff_sym_hos\n",
            "0          38        2       1  ...         0         0             7\n",
            "1          38        2       1  ...         0         0             4\n",
            "2          38        2       1  ...         0         0             0\n",
            "3          38        2       1  ...         0         0             0\n",
            "4          38        2       1  ...         0         0             6\n",
            "..        ...      ...     ...  ...       ...       ...           ...\n",
            "217         2        3       2  ...         0         0             0\n",
            "218        22        6       2  ...         0         0             0\n",
            "219        22        6       2  ...         0         0             0\n",
            "220        22        6       2  ...         0         0             0\n",
            "221        22        6       2  ...         0         0             0\n",
            "\n",
            "[222 rows x 14 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "URBOdKUteXcu",
        "outputId": "1f56bbb5-819c-40b1-e93b-e3b45f7f1216",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print(tdata.isna().sum())"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "location        0\n",
            "country         0\n",
            "gender          0\n",
            "age             0\n",
            "vis_wuhan       0\n",
            "from_wuhan      0\n",
            "death           0\n",
            "symptom1        0\n",
            "symptom2        0\n",
            "symptom3        0\n",
            "symptom4        0\n",
            "symptom5        0\n",
            "symptom6        0\n",
            "diff_sym_hos    0\n",
            "dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HEPS4gxb2nmI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import recall_score as rs\n",
        "from sklearn.metrics import precision_score as ps\n",
        "from sklearn.metrics import f1_score as fs\n",
        "from sklearn.metrics import balanced_accuracy_score as bas\n",
        "from sklearn.metrics import confusion_matrix as cm"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mO_GHtYI0rMw",
        "colab_type": "text"
      },
      "source": [
        "###Logisitic Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZzhuAHDGgdzJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.linear_model import LogisticRegression as lr\n",
        "\n",
        "classifier = lr()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I2Eqf4CIfqGb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = tdata[['location','country','gender','age','vis_wuhan','from_wuhan','symptom1','symptom2','symptom3','symptom4','symptom5','symptom6','diff_sym_hos']]\n",
        "Y = tdata['death']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yWroRhrfZQy7",
        "colab_type": "code",
        "outputId": "c3678c1f-782a-4619-ab68-f6ace1410a66",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292
        }
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(X,Y,test_size=0.2,random_state=10)\n",
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/linear_model/_logistic.py:940: 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",
            "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WYVlHGmQGXh8",
        "colab_type": "text"
      },
      "source": [
        "#####Evaluation of model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eTFY9atzcmO-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "pred = np.array(classifier.predict(X_test))\n",
        "\n",
        "recall_lr = rs(Y_test,pred)\n",
        "precision_lr = ps(Y_test,pred)\n",
        "f1_lr = fs(Y_test,pred)\n",
        "ma_lr = classifier.score(X_test,Y_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4dtmxNcbcoBF",
        "colab_type": "code",
        "outputId": "63adc286-3b65-4d4f-e92e-7a2b1162aa53",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print('*** Evaluation metrics for test dataset ***\\n')\n",
        "print('Recall Score: ',recall_lr)\n",
        "print('Precision Score: ',precision_lr)\n",
        "print('F1 Score: ',f1_lr)\n",
        "print('Accuracy: ',ma_lr)\n",
        "a = pd.DataFrame(Y_test)\n",
        "a['pred']= classifier.predict(X_test)\n",
        "print('\\n\\tTable 3\\n')\n",
        "print(a.head())"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "*** Evaluation metrics for test dataset ***\n",
            "\n",
            "Recall Score:  0.7272727272727273\n",
            "Precision Score:  0.6666666666666666\n",
            "F1 Score:  0.6956521739130435\n",
            "Accuracy:  0.8444444444444444\n",
            "\n",
            "\tTable 3\n",
            "\n",
            "     death  pred\n",
            "184      0     0\n",
            "170      1     0\n",
            "142      0     0\n",
            "182      0     0\n",
            "49       1     1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LbfWZ1Hyrua6",
        "colab_type": "code",
        "outputId": "4789e656-9b96-41a9-dfde-286a0f4b1709",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.bar(['Accuracy','F1 Score','Recall Score','Precision Score'],[ma_lr,f1_lr,recall_lr,precision_lr],color=['red','green','purple','orange'])\n",
        "plt.plot([ma_lr,f1_lr,recall_lr,precision_lr],color='black')\n",
        "plt.title('Evaluation Metrics for Logistic Regression')"
      ],
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Evaluation Metrics for Logistic Regression')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "69vkWD86tDoU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#print(pd.DataFrame({'Val':Y_test,'Pred':classifier.predict(X_test)}))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nRZPxsBo1F_c",
        "colab_type": "text"
      },
      "source": [
        "###Decision Tree Classifier"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "X6Wboj3t1P_p",
        "colab": {}
      },
      "source": [
        "from sklearn.tree import DecisionTreeClassifier as dtc\n",
        "classifier = dtc(max_depth=2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "HEMkNq_b1P_4",
        "colab": {}
      },
      "source": [
        "X = tdata[['location','country','gender','age','vis_wuhan','from_wuhan','symptom1','symptom2','symptom3','symptom4','symptom5','symptom6','diff_sym_hos']]\n",
        "Y = tdata['death']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "94054250-6202-48a2-ca3b-30a8cb14f005",
        "id": "TaJSM50y1P_-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(X,Y,test_size=0.2,random_state=10)\n",
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
              "                       max_depth=2, max_features=None, max_leaf_nodes=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                       random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "AqRIHx9S1QAF"
      },
      "source": [
        "######Evaluation of the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "MCUMzAhi1QAG",
        "colab": {}
      },
      "source": [
        "pred = np.array(classifier.predict(X_test))\n",
        "\n",
        "recall_dtc = rs(Y_test,pred)\n",
        "precision_dtc = ps(Y_test,pred)\n",
        "f1_dtc = fs(Y_test,pred)\n",
        "ma_dtc = classifier.score(X_test,Y_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "bf1b63d3-d584-437f-98e5-6579f5749e25",
        "id": "18K9mjoX1QAM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print('*** Evaluation metrics for test dataset ***\\n')\n",
        "print('Recall Score: ',recall_dtc)\n",
        "print('Precision Score: ',precision_dtc)\n",
        "print('F1 Score: ',f1_dtc)\n",
        "print('Accuracy: ',ma_dtc)\n",
        "a = pd.DataFrame(Y_test)\n",
        "a['pred']= classifier.predict(X_test)\n",
        "print('\\n\\tTable 3\\n')\n",
        "print(a.head())"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "*** Evaluation metrics for test dataset ***\n",
            "\n",
            "Recall Score:  0.5454545454545454\n",
            "Precision Score:  0.8571428571428571\n",
            "F1 Score:  0.6666666666666665\n",
            "Accuracy:  0.8666666666666667\n",
            "\n",
            "\tTable 3\n",
            "\n",
            "     death  pred\n",
            "184      0     0\n",
            "170      1     0\n",
            "142      0     0\n",
            "182      0     0\n",
            "49       1     0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "QbBSo0GH1QAR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "outputId": "c2543ee5-4179-4363-ddfa-a679402a2965"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.bar(['Accuracy','F1 Score','Recall Score','Precision Score'],[ma_dtc,f1_dtc,recall_dtc,precision_dtc],color=['red','green','purple','orange'])\n",
        "plt.plot([ma_dtc,f1_dtc,recall_dtc,precision_dtc],color='black')\n",
        "plt.title('Evaluation Metrics for Decision Tree')"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Evaluation Metrics for Decision Tree')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 56
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GUdAzXF3offb",
        "colab_type": "text"
      },
      "source": [
        "######Visualizing the decision trees from random forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uAdfJJPMkNL-",
        "colab_type": "code",
        "outputId": "1be08991-245b-4771-b827-e5305280c06b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        }
      },
      "source": [
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n",
        "from sklearn.externals.six import StringIO  \n",
        "from IPython.display import Image  \n",
        "from sklearn.tree import export_graphviz\n",
        "import pydotplus\n",
        "\n",
        "estimator = classifier\n",
        "dot_data = StringIO()\n",
        "export_graphviz(estimator, out_file=dot_data,  \n",
        "                filled=True, rounded=True,\n",
        "                special_characters=True)\n",
        "graph = pydotplus.graph_from_dot_data(dot_data.getvalue())  \n",
        "Image(graph.create_png(),width=250,height=200)\n"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/externals/six.py:31: FutureWarning: The module is deprecated in version 0.21 and will be removed in version 0.23 since we've dropped support for Python 2.7. Please rely on the official version of six (https://pypi.org/project/six/).\n",
            "  \"(https://pypi.org/project/six/).\", FutureWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": [],
            "image/png": {
              "width": 250,
              "height": 200
            }
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "drOJRhkW2voo",
        "colab_type": "text"
      },
      "source": [
        "###SVM"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "WbRACbyI2xCS",
        "colab": {}
      },
      "source": [
        "from sklearn import svm\n",
        "classifier = svm.SVC()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "4AFSOOqw2xCg",
        "colab": {}
      },
      "source": [
        "X = tdata[['location','country','gender','age','vis_wuhan','from_wuhan','symptom1','symptom2','symptom3','symptom4','symptom5','symptom6','diff_sym_hos']]\n",
        "Y = tdata['death']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "b0de150c-b2f6-4f76-a7dc-69dc4615439e",
        "id": "4y3yODJQ2xCu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        }
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(X,Y,test_size=0.2,random_state=10)\n",
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "3zKgizqW2xC3"
      },
      "source": [
        "######Evaluation of the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "YaRJTyH82xC4",
        "colab": {}
      },
      "source": [
        "pred = np.array(classifier.predict(X_test))\n",
        "\n",
        "recall_svm = rs(Y_test,pred)\n",
        "precision_svm = ps(Y_test,pred)\n",
        "f1_svm = fs(Y_test,pred)\n",
        "ma_svm = classifier.score(X_test,Y_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "f43674c6-2280-4e24-9f3c-2344bca3c6af",
        "id": "EY_BbL8a2xC-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print('*** Evaluation metrics for test dataset ***\\n')\n",
        "print('Recall Score: ',recall_svm)\n",
        "print('Precision Score: ',precision_svm)\n",
        "print('F1 Score: ',f1_svm)\n",
        "print('Accuracy: ',ma_svm)\n",
        "a = pd.DataFrame(Y_test)\n",
        "a['pred']= classifier.predict(X_test)\n",
        "print('\\n\\tTable 3\\n')\n",
        "print(a.head())"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "*** Evaluation metrics for test dataset ***\n",
            "\n",
            "Recall Score:  0.6363636363636364\n",
            "Precision Score:  0.7\n",
            "F1 Score:  0.6666666666666666\n",
            "Accuracy:  0.8444444444444444\n",
            "\n",
            "\tTable 3\n",
            "\n",
            "     death  pred\n",
            "184      0     0\n",
            "170      1     0\n",
            "142      0     0\n",
            "182      0     0\n",
            "49       1     1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "tjJp35F12xDD",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "outputId": "282d21b8-46c3-4c07-f068-f7d6df2316ff"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.bar(['Accuracy','F1 Score','Recall Score','Precision Score'],[ma_svm,f1_svm,recall_svm,precision_svm],color=['red','green','purple','orange'])\n",
        "plt.plot([ma_svm,f1_svm,recall_svm,precision_svm],color='black')\n",
        "plt.title('Evaluation Metrics for Support Vector Machine')"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Evaluation Metrics for Support Vector Machine')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aejoz6yW3R_1",
        "colab_type": "text"
      },
      "source": [
        "###Gaussian Naive Bayes"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "FgV2wh503ikt",
        "colab": {}
      },
      "source": [
        "from sklearn.naive_bayes import GaussianNB\n",
        "classifier = GaussianNB()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "6D8PjlV73ik7",
        "colab": {}
      },
      "source": [
        "X = tdata[['location','country','gender','age','vis_wuhan','from_wuhan','symptom1','symptom2','symptom3','symptom4','symptom5','symptom6','diff_sym_hos']]\n",
        "Y = tdata['death']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "0423d6c0-9bf5-4579-abf5-bfcdff9d7094",
        "id": "2NtkH2CO3ilE",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(X,Y,test_size=0.2,random_state=10)\n",
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GaussianNB(priors=None, var_smoothing=1e-09)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "9O6RJysN3ilP"
      },
      "source": [
        "######Evaluation of the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "uEu8rlTw3ilR",
        "colab": {}
      },
      "source": [
        "pred = np.array(classifier.predict(X_test))\n",
        "\n",
        "recall_gnb = rs(Y_test,pred)\n",
        "precision_gnb = ps(Y_test,pred)\n",
        "f1_gnb = fs(Y_test,pred)\n",
        "ma_gnb = classifier.score(X_test,Y_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "3d8bd104-415a-400b-caed-ed5b1cde71fb",
        "id": "eovrIeK73ilb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print('*** Evaluation metrics for test dataset ***\\n')\n",
        "print('Recall Score: ',recall_gnb)\n",
        "print('Precision Score: ',precision_gnb)\n",
        "print('F1 Score: ',f1_gnb)\n",
        "print('Accuracy: ',ma_gnb)\n",
        "a = pd.DataFrame(Y_test)\n",
        "a['pred']= classifier.predict(X_test)\n",
        "print('\\n\\tTable 3\\n')\n",
        "print(a.head())"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "*** Evaluation metrics for test dataset ***\n",
            "\n",
            "Recall Score:  0.7272727272727273\n",
            "Precision Score:  0.5333333333333333\n",
            "F1 Score:  0.6153846153846153\n",
            "Accuracy:  0.7777777777777778\n",
            "\n",
            "\tTable 3\n",
            "\n",
            "     death  pred\n",
            "184      0     0\n",
            "170      1     0\n",
            "142      0     0\n",
            "182      0     0\n",
            "49       1     1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "847b0b44-dbe4-4866-d3f5-a1c9b323035b",
        "id": "E62591Tw3ilh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.bar(['Accuracy','F1 Score','Recall Score','Precision Score'],[ma_gnb,f1_gnb,recall_gnb,precision_gnb],color=['red','green','purple','orange'])\n",
        "plt.plot([ma_gnb,f1_gnb,recall_gnb,precision_gnb],color='black')\n",
        "plt.title('Evaluation Metrics for Gaussian Naive Bayes')"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Evaluation Metrics for Gaussian Naive Bayes')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3dd5wV1fnH8c+XLqKgsoaqkKWEJh0kdhSFKKCxgRU0olFRflgxioCiCcZeUFQiRUVRNAhERBGj2FikN7MSCMWCgiJYEHh+f5xZvaxb7sJd7jI8b177Yu6ZuTPPnDv3uWfONJkZzjnn9nyl0h2Ac8651PCE7pxzMeEJ3TnnYsITunPOxYQndOeciwlP6M45FxOe0IuJpBmS/lRM875J0hPFMe/dSdImSb9N4fwaSpor6VtJV6VqviWJpHMlvZbuOPIi6VFJt6Q7jr3ZXp/QJa2Q9H2UXHL+Hkp3XDkkHStpdWKZmd1hZin/sZDUS5JJujdXefeo/Kkk55PUj5mZVTKz5TsZbl6uB940s/3M7IFUzFBSfUnjJK2TtFHSfyQ9KKlWKuZfVGb2tJmdWBzzjj7jBZJKJZTdnuznbmaXmdltxRBX4nd0g6TJkmqnejlxsNcn9EjXKLnk/F2Z7oDS6BPgLEllEsouBD5O1QJyzTuVDgUW7cwb84pJUj3gA2At0NLM9geOINTRkbsQZ0lWA+iR7iDy0NXMKgHVgc+BB9McT4nkCT0fkspL+lpS04SyjKilcLCkAyRNilpuG6LhPFttkgZJGpvwuk7UGioTve4taUnUVbBc0qVR+b7Av4AaCXsPNfKYXzdJi6J4Z0hqlDBuhaRrJc2X9I2k5yRVKGDVPwMWACdF7z8Q+D0wMdc6HS7p3WiZ8yQdG5UPBY4CHkrc24nW9wpJ/wH+k1BWLxreR9LdklZGcb4TlVWQNFbSV9GyZkn6TR51PB04LmG5DSRVljQ6+oxWSro5p/UZ7Y3MlHSvpK+AQXnUxSBgppn1N7PVAGb2hZndZ2bjovkUuB1E9X9CwuufP7uC1i2Kb3m0TfxX0rkJ5e8kzO9+SauivYfZko7Ktaznozr4NtpG2hTw2QMMAwbn96Mrabykz6LP6N+SmiSMe0rS7dHwEkmnJIwrE9VRq+h1nttPYczsB+AFoHHCvE+WNCeqg1WSBiWMmyypb651mC/ptGj4d5KmSVovaZmksxKm+4OkxVHdrZF0bTIxppMn9HyY2Y/ABKBnQvFZwFtm9gWh7v5BaBUeAnwP7GxXzRfAKcD+QG/gXkmtzGwz0AVYm7D3sDbxjZIaAM8C/YAMYArwiqRyueLuDNQFDgN6FRLPaOCCaLgH8E/gx4Rl1gQmA7cDBwLXAi9KyjCzvwBvA1fmsbdzKtCehC9jgr8DrQk/HgcSuk+2E/YOKgO1gYOAywh1vQMz65hruR8TWnGVgd8Cx0Tr1Dvhbe2B5cBvgKF5xHQC8GIe5Yl2ZTvIc90UfsgfALqY2X6EOpmbzzxmAS0IdfYMMD7XD3Y3YBxQhfCjXFhsE4CN5L+N/AuoDxwMfAQ8nc90z7Ljd+ck4Esz+6ig7aeQ2JBUETgbeD+heDPhs60CnAz8WdKp0bhRwHkJ728O1AQmR/U8jVBvBxO29Uck5WyfTwKXRp9BU2B6YfGlmyf04OWopZDzd0lU/gw77n6eE5VhZl+Z2Ytm9p2ZfUtICMfszMLNbLKZfWLBW8BrhFZuMs4GJpvZNDP7iZAY9yEkgRwPmNlaM1sPvEJIAAV5CThWUmXCF2V0rvHnAVPMbIqZbTezaUAW8IdC5nunma03sx0SctRqvgi42szWmNk2M3s3+lH9iZDs6kXls81sYyHLQVJpwmc3wMy+NbMVwN3A+QmTrTWzB81sa+6YIlUJeyw587wy2j42SXocdnk7KGjdtgNNJe1jZp+aWZ5dSWY2Nophq5ndDZQHGiZM8k70OW0DxgDNC4nJgFuAW3I1CnKWNzKqzx8JezDNo+0kt2eAblEChvDdeTYa3pnt52VJXwPfAJ2AuxJimmFmC6J5zY+Wk/MZTAQaSKofvT4feM7MthAaUSvM7B9R/c0h/ICfGU37E9BY0v5mtsHMPiogvhLBE3pwqplVSfh7PCp/E6goqb2kOoRE+BKEloKkx6Jd+Y3Av4EqUSIpEkldJL0f7fZ9Tdiwqyb59hrAypwXZrYdWEVoheT4LGH4O6BSQTOMkttk4GbgIDObmWuSQ4EzE38ECX3K1QuJdVU+5VWBCoS+6dzGAFOBcZLWShomqWwhy8mZZ1kS6iYaTqyX/OLJ8RUJ62RmD5lZFeC+aN67uh3kuW7RntnZhBb7p1G3we/ymoFCd9qSqAvka0KLP3Hbyf3ZV8ivOyVhPacAq4FLcy2rtKS/SvokWtcV0ahfbatmlg0sAbpGSb0bUWOIndt+To3qvgJwJfCWpGpRXO0lvRl16XxDqLeqURw/AM8B50UNh56Ees+Jo32uOM4FqkXjTyd8F1dKektSh4LqrSTwhF6AqFXzPGEj6AlMilphANcQWkLto4NlR0flymNWm4GKCa9zNhgklSe0Cv4O/CbaaKckzKew22GuJWyYOfMTYRd+TWHrV4jRhHUcm8e4VcCYXD+C+5rZXwuJOb/yL4EfgMxfvcHsJzMbbGaNCXsdp/BLd1BBviS0sA5NKDuEHeulsLp9A/hjIdMUth3k+9kXtG5mNtXMOhGS3FLgcXJR6C+/ntCldkC07XxD3ttgUf0FuClX7OcA3QldUZWBOjmh5DOPnG6X7sDiKMlD4dtPvqI9mQnANn45MP0MoSVe28wqA4/mimkUIVEfD3xnZu8lxPFWrjgqmdmfo2XNMrPuhO6Ylwm5oETzhF64ZwitpXP5pYUBsB+hv/RrhQOHtxYwj7nA0ZIOiXZPBySMK0fYTV4HbJXUBUg8Le1z4KB8dmshbGQnSzo+arleQ+jvfjfZFczHW4Rd27zOJhhLaHmdFLXaKiicXplzMPBzQr91UqK9ipHAPQoHfUtL6qBwYPo4Sc2iFu9GQpLensQ8c36Mh0raT9KhQH/y/oHKzyDgKEn3RP2+SKoKNEqYprDtYC7QQ1JZhQOSZ+SMyG/dJP1G4VTRfQmf5aZ81nk/YCth2ykjaSDhOMwuM7MZwEJCP3/i8n4k7LlUBO4oZDbjCNvyn9nxu1PY9pMvBd2BAwh7ADlxrTezHyS1I/zwJK7Le4T6u5tfWucAkwjdMedHn09ZSW0lNZJUTuGc/8pRV+ZGktju0s0TevCKdjwP/aWcEWb2AaGVVYNwQCjHfYS+6i8JB2hezW/mUR/hc8B8YDZhQ8oZ9y1wFSH5bCBsjBMTxi8ltHSWR7uFNXLNexmhT/LBKJauhFO8thS1EnLN18zsjajfPfe4VYRW102EZLIKuI5ftqf7gTMUzvpI9nzwawln18wC1gN/i+ZXjXBWw0bCF/gtdvxSFqQv4bNbDrxDSCojk3wvFg6stgdqAfMkfQvMJOwV5VxAU9h2cAthz2MDMJgdE1t+61aK8OOzllAXxxCSYm5To+V9TOhO+oHCu5GK4mbCQcsco6PlrAEWs+OByV8xs0+B9wh7H88llBe2/eTlFUmbCHU1FLgw4bjC5cCQ6PMZSN4t6dFAMxJ+0KPv3omEYy1rCd1TfyM0sCD0t6+IupcuIzTqSjSZP+DCORdzki4A+phZXK8fALyF7pyLueig7OXAiHTHUtw8oTvnYkvSSYRunc/ZsbsrlrzLxTnnYsJb6M45FxPFdZOkQlWtWtXq1KmTrsU759weafbs2V+aWZ63SUhbQq9Tpw5ZWVnpWrxzzu2RJK3Mb1xSXS6SOivciSxb0o15jD8kuvR2jsKdzAq7p4dzzrkUKzShR1exPUy4619joKd+uRtZjpuB582sJdEdy1IdqHPOuYIl00JvB2Sb2fLo6sNxhKu8Ehm/XHJcmXDVlXPOud0omYRekx0vJ17Njnesg3DPi/MUHpU2hXDJ9a9I6iMpS1LWunXrdiJc55xz+UnVaYs9gafMrBbhdpNjlPBcwhxmNsLM2phZm4yMQu9l75xzrgiSSehrCLdjzVGLX9+a9WKiG+JEdzarQPL383bOOZcCyST0WUB9SXUVnmDSg1zPlwT+R7jXMArPs6xAuNzWOefcblJoQjezrYQnhEwl3OLzeTNbJGmIpG7RZNcAl0iaR7jVay+L6T0FfvrpJz777LPCJ3TOud0sqT706Nl/Dcws08yGRmUDzWxiNLzYzI4ws+Zm1sLMXivOoNPpySefJDMzk4EDB7JxY6GPtnTOud0mbTfnatOmje30laJKxRO2dk424dlczwMZhKcXXEp47NBuE8+dH+dcEiTNNrM2eY3zm3MVUT3Co1c+BJoSHjX0O0I/U4l/PpVzLtY8oe+ktoQnCL9KuKLqHKANMC2dQTnn9mqe0HeBgJOAjwgPKtxAeEDhiVGZc87tTp7QU6AU4emxS4F7Ccm8NaHVvjyNcTnn9i6e0FOoPNAP+IRw4PRlQv/6VcAXaYzLObd38IReDCoDtxPOiLmIcOvJTOA2YFMa43LOxZsn9GJUA3gUWEjoVx9IOEtmOPBTGuNyzsWTJ/Td4HfAi8B7QAPgcqAJMJ5w32HnnEsFT+i70eHAW8ArhAuRzgLaAzPSGJNzLj48oe9mAk4B5gH/AD4FjiPcc3h+GuNyzu35PKGnSWmgF/AxcBfwPtACuABYkbaonHN7Mk/oabYPcC3hVMfrCPeIaQj0B75KY1yu5Nm+fTtvvPEGgwYNIjs7O93huBLIb85VwqwiPM/vKaAScCNwNVAxcSK/OVdaDdbg3bq8L/mSedG/jYQ7fJahDMdyLB3oQGlK79Z4AG61W3f7Ml1Q0M25yuzuYFzBagNPElroN0V/DxGSfG/8A9tbfM/3LGQh85jHalYjRCaZdKITNanJa7zG67zOIhbRjW5Up3q6Q3YlgOeHEqoJ8E/gbeAGoA9wD3An0N0MxXgvZW+1jW18wifMYx5LWco2tpFBBp3oRDOasT/7/zzt2ZzNYhYzhSmMYARHcATHcAxlKZvGNXDp5gm9hDsKmElI7gOA04AORxzBsGHDOPLII9Mam0uNz/mcecxjPvPZxCb2YR9a05rmNKcGNRC//vEWoglNqEtdXuM13uEdlrCErnSlDnV2/0q4EiGphC6pM3A/4eSMJ8zsr7nG30s4+w5Cd+/BZlYllYHuzQScSjjd8R/ArStWcNRRR9G1a1fuvPNOmjRpkt4AXZFtZjMLWMA85vEpn1KKUtSnPi1oQX3qUybJtlZFKnIqp9KMZrzCKzzFU7ShDSdwAhWoUMxr4UqaQrcaSaWBh4FOwGpglqSJZrY4Zxoz+7+E6fsCLYsh1r1eGeAS4NzsbO6//37++te/cthhh9GrVy8GDx5MrVq10h2iK8BWtvIf/sM85vExH7Od7VSjGp3pTFOaUolKOz3vTDK5nMuZznQ+4AOWsYxTOIWGNEzhGriSLpnTFtsB2Wa23My2AOOA7gVM35PwAB9XTCpWrMiAAQNYvnw5/fr1Y+zYsdSvX58bbriBDRs2pDs8l8Aw1rKWKUzhHu7hOZ5jFatoT3sui/4dzuG7lMxzlKMcnenMxVzMPuzDszzLeMazyW8Jt9dIJqHXJJxNl2N1VPYrkg4F6gLT8xnfR1KWpKx169YVNVaXy0EHHcTdd9/Nxx9/zJlnnsldd93Fb3/7W+666y5++OGHdIe3V/uWb5nJTIYznBGMYDazqUMdzuEc+tOfkziJalQrlmXXohZ96MNxHMdSlvIwDzOXuZjfOSj2Un1hUQ/gBTPbltdIMxthZm3MrE1GRkaKF733OvTQQxk9ejRz5syhQ4cOXH/99TRo0ICnnnqKbdvy/ChcMfiJn1jIQsYylnu4h2lMoyxlOZmTuYZrOIuzaECD3XLeeBnKcAzHcBmXUZWqvMzLjGUsG/A9uDhLJqGvIZwenaNWVJaXHnh3S9o0b96cKVOmMH36dKpVq0bv3r1p3rw5kyZNIl0XkMWdYaxiFa/wCndzNy/wAl/wBUdwBFdwBZdwCW1pS8UdLw3bbTLIoDe96UIXVrGKR3iE93mf7f5I81hK5lD6LKC+pLqERN6D8HS1HUj6HXAA4S6xLo2OO+44PvjgA1544QVuuukmunbtytFHH83f/vY3Dj/88HSHFwvf8M3PV29+xVeUoQyNaEQLWlCXupQqQXfVKEUp2tOehjRkEpN4lVdZwAK6052DOTjd4bkUKnSrM7OtwJXAVGAJ8LyZLZI0RFK3hEl7AOPMm4IlgiTOPPNMFi9ezCOPPMKyZcvo0KEDp59+OsuWLUt3eHukzZs3M2bMGEYxinu5l+lMZ1/2pRvduJZrOZ3TySSzRCXzRFWowrmcyx/5I+tZz6M8ypu8yVa2pjs0lyJ+L5c90U58Zps2beLee+9l2LBhfP/991x88cXceuut1KhRoxgCjI/t27fz9ttvM2rUKMaPH8+mTZuoQhWaR/8O5MB0h7hTNrP555Z6Bhl0oxu1d+hZLZjfyyV9CrqXiyf0PdEufGZffPEFt99+O48++ihlypShf//+XHfddVSuXDmFAe75li9fzujRoxk1ahQrVqygUqVKnHnmmfTq1Ys3jnmjxLbCi+pjPmYSk9jIRtrTno50pDzlC32fJ/T0KSihx2OrdEk7+OCDeeCBB1iyZAndu3dn6NChZGZmct999/Hjjz+mO7y02rhxI08++SRHH300mZmZDBkyhHr16jFmzBg+++wzRo4cydFHHx2bZA7QgAZcwRW0pS0f8AGP8AjZ+K1591Tx2TJdkWRmZvLss8+SlZVFy5Yt+b//+z8aNmzI2LFj2b597zkDYtu2bUybNo1zzz2XatWq8ac//YnPP/+coUOHsnLlSqZNm8Z5553Hvvvum+5Qi015ynMyJ3MRF1GWsoxlLBOYwHd8l+7QXBF5Qt/LtW7dmmnTpvHaa69x4IEHcv7559OqVSumTp0a61Mdly5dyoABAzj00EM58cQTmTJlChdeeCHvvfceS5cu5aabbqJ27eT7lOPgEA7hUi7lKI5iIQt5iIdYyEK/IGkP4gndAdCpUyeysrJ45pln2LhxI507d+aEE05gp49zlEDr169n+PDhHH744TRq1Ihhw4bRvHlznnvuOT799NOfx+3NtyYuS1mO53j60IcqVOEFXuBZnuUbvkl3aC4JntDdz0qVKkXPnj1ZsmQJ999/P/Pnz6dt27acffbZe+wjz7Zu3cqkSZM488wzqV69OpdffjmbN2/m73//O2vWrGHy5MmcddZZVKjgdyZMVI1q/Ik/cSInspzlPMzDzGKWX5BUwnlCd79Svnx5rrrqKj755BNuueUWJk2aRKNGjbjiiiv4/PPP0x1eUubPn88111xDrVq16Nq1KzNmzOCyyy5j9uzZP4+rVq147qUSF6Uoxe/5PZdzOTWpyWQmM4pRfMmX6Q7N5cMTusvX/vvvz5AhQ8jOzuaSSy7hscceIzMzk0GDBvHtt9+mO7xfWbduHffffz8tW7akefPmPPDAA3To0IGXXnqJNWvWcP/999OqVau9uktlZxzIgVzABXSnO5/zOcMZzh133MFPP/2U7tBcLp7QXaGqV6/OI488wuLFi+nSpQuDBw8mMzOThx56iC1btqQ1ti1btjBhwgS6d+9OjRo16NevH6VKleKBBx7g008/5aWXXuLUU0+lXLlyaY1zTydES1pyBVfQkIb85S9/oW3btsyePTvdobkEntBd0ho0aMD48eN5//33ady4MX379qVRo0aMGzdut57qaGZkZWXRt29fqlevzumnn86HH35Iv379WLBgAbNnz6Zv375UrVp1t8W0t9iP/TiLs5gwYQJffPEF7dq14/rrr+e77/wUx5LAE7orsvbt2/Pmm28yefJk9t13X3r27Em7du144403inW5a9euZdiwYTRt2pS2bdvy+OOP06lTJ6ZMmcKqVau46667aNq0abHG4ILTTjuNxYsXc9FFF3HXXXdx2GGH8eabb6Y7rL2eJ3S3UyTxhz/8gTlz5jBq1CjWrVvHCSecwEknncScOXNStpzvv/+ecePG0aVLF2rXrs0NN9xA5cqVefTRR/n0009/HlemjD/vfHerUqUKjz/+ONOnh+fZdOzYkUsuuYSvv/46zZHtvTyhu11SunRpLrjgApYtW8bdd99NVlYWrVq14rzzzuO///3vTs3TzJg5cyZ9+vShevXq9OzZk0WLFnHjjTeydOlS3n33XS699FIOOOCAFK+N2xnHHXcc8+fP57rrrmPkyJE0btyYl156Kd1h7ZU8obuUqFChAv379+eTTz7hxhtv5MUXX6Rhw4b069ePZB83uHLlSm6//XYaNmzIkUceydNPP023bt14/fXXWbFiBUOHDqVhQ3/ocUlUsWJFhg0bxocffsjBBx/MH//4R8444ww+++yzdIe2V/GE7lKqSpUq3HnnnWRnZ3PhhRfy4IMPkpmZye23387mzZt/Nf2mTZsYPXo0HTt2pE6dOtxyyy3UqFGDkSNH8tlnnzF69GiOP/54SpXyTXVP0Lp1a2bNmsUdd9zx8/ULI0eOjPVtJEoS/5a4YlGzZk0ef/xxFi5cSMeOHbnllluoV68ejz32GFu2bOHNN9+kV69eVKtWjQsvvJCVK1cyePBgli9fzowZM+jduzf77bdfulfD7YSyZcsyYMAA5s2bR7Nmzbj44ovp1KkTy5cvT3dosecJ3RWrRo0a8fLLL/POO++QmZnJZZddRuXKlenYsSMTJkygR48e/Pvf/yY7O5uBAwdSt27ddIfsUqRhw4bMmDGD4cOH8+GHH9K0aVPuuecef3B5MfIHXOyJdvEz0+A01Z8By4ClwG+B3wFpuN7Hbt21+huswSmKZM9V1AdcrF69mssvv5xXXnmFtm3b8sQTT3DYYYcVU3TxtssPuJDUWdIySdmSbsxnmrMkLZa0SNIzuxKwiykRkvipwGGkJZm79KhVqxb//Oc/GTduHCtWrKB169bcfPPN/PDDD+kOLVYKTeiSSgMPA12AxkBPSY1zTVMfGAAcYWZNgH7FEKtzbg8mibPPPpslS5ZwzjnnMHToUFq2bMk777yT7tBiI5kWejsg28yWm9kWYBzQPdc0lwAPm9kGADP7IrVhOufi4qCDDmLUqFG8+uqrfP/99xx11FFcccUVbNy4Md2h7fGSSeg1gVUJr1dHZYkaAA0kzZT0vqTOec1IUh9JWZKykj032TkXTyeddBILFy7k6quvZvjw4TRp0oTJkyenO6w9WqrOcikD1AeOBXoCj0uqknsiMxthZm3MrE1GRkaKFu2c21NVqlSJ++67j5kzZ7L//vtzyimncO655yZ9MZrbUTIJfQ2Q+HDFWlFZotXARDP7ycz+C3xMSPDOOVeoDh068NFHH3Hrrbcyfvx4GjVqxNNPP+0XJBVRMgl9FlBfUl1J5YAewMRc07xMaJ0jqSqhC8avInDOJa18+fIMGjSIOXPmUK9ePc477zxOPvlk/ve//6U7tD1GoQndzLYCVwJTgSXA82a2SNIQSd2iyaYCX0laDLwJXGdmXxVX0M65+GrSpAkzZ87kvvvu46233qJJkyY89NBDu/We+3uqpPrQzWyKmTUws0wzGxqVDTSzidGwmVl/M2tsZs3MbFxxBu2ci7fSpUtz9dVXs2jRIn7/+9/Tt29fjjrqKJYsWZLu0Eo0v/TfOVdi1alTh1dffZVRo0axdOlSWrRowW233Zb2Rx+WVJ7QnXMlmiQuuOAClixZwmmnncbAgQNp3bo1H374YbpDK3E8oTvn9ggHH3ww48aNY+LEiWzYsIEOHTrQv3//PG/LvLfyhO6c26N07dqVRYsW0adPH+69916aNWvG66+/nu6wSgRP6M65PU7lypUZPnw4b731FmXLlqVTp0707t2b9evXpzu0tPKE7pzbYx199NHMmzePAQMGMGbMGBo3bsz48eP32guSPKE75/ZoFSpU4I477iArK4uaNWty1llncdppp7F27dp0h7bbeUJ3zsVCixYt+OCDDxg2bBhTp06lUaNGjBgxYq+6IMkTunMuNsqUKcN1113HggULaNWqFZdeeinHH3882dnZ6Q5tt/CE7pyLnXr16jF9+nRGjBjBnDlzaNasGcOGDWPr1q3pDq1YeUJ3zsWSJC655BIWL15M586dueGGG2jfvj1z585Nd2jFxhO6cy7WatSowYQJE3jhhRdYs2YNbdq0YezYsekOq1iUSXcAzrm90DParYsTcDrQ8Ta46Tk4ZsP58Mz5uzWGHZxTPKdVekJ3zu01DtgXhl+U7iiKj3e5OOdcTHhCd865mPCE7pxzMeEJ3TnnYiKphC6ps6RlkrIl3ZjH+F6S1kmaG/39KfWhOuecK0ihZ7lIKg08DHQCVgOzJE00s8W5Jn3OzK4shhidc84lIZkWejsg28yWm9kWYBzQvXjDcs45V1TJJPSawKqE16ujstxOlzRf0guSauc1I0l9JGVJylq3bt1OhOuccy4/qToo+gpQx8wOA6YBo/KayMxGmFkbM2uTkZGRokU755yD5BL6GiCxxV0rKvuZmX1lZj9GL58AWqcmPOecc8lKJqHPAupLqiupHNADmJg4gaTqCS+7AUtSF6JzzrlkFHqWi5ltlXQlMBUoDYw0s0WShgBZZjYRuEpSN2ArsB7oVYwxO+ecy0NSN+cysynAlFxlAxOGBwADUhuac865ovArRZ1zLiY8oTvnXEx4QnfOuZjwhO6cczHhCd0552LCE7pzzsWEJ3TnnIsJT+jOORcTntCdcy4mPKE751xMeEJ3zrmY8ITunHMx4QndOediwhO6c87FhCd055yLCU/ozjkXE57QnXMuJpJK6JI6S1omKVvSjQVMd7okk9QmdSE655xLRqEJXVJp4GGgC9AY6CmpcR7T7QdcDXyQ6iCdc84VLpkWejsg28yWm9kWYBzQPY/pbgP+BvyQwvicc84lKZmEXhNYlfB6dVT2M0mtgNpmNrmgGUnqIylLUta6deuKHKxzzrn87fJBUUmlgHuAawqb1sxGmFkbM2uTkZGxq4t2zjmXIJmEvgaonfC6VlSWYz+gKTBD0r7QGzgAAA9ESURBVArgcGCiHxh1zrndK5mEPguoL6mupHJAD2Bizkgz+8bMqppZHTOrA7wPdDOzrGKJ2DnnXJ4KTehmthW4EpgKLAGeN7NFkoZI6lbcATrnnEtOmWQmMrMpwJRcZQPzmfbYXQ/LOedcUfmVos45FxOe0J1zLiY8oTvnXEx4QnfOuZjwhO6cczHhCd0552LCE7pzzsWEJ3TnnIsJT+jOORcTntCdcy4mPKE751xMeEJ3zrmY8ITunHMx4QndOediwhO6c87FhCd055yLCU/ozjkXE57QnXMuJpJK6JI6S1omKVvSjXmMv0zSAklzJb0jqXHqQ3XOOVeQQhO6pNLAw0AXoDHQM4+E/YyZNTOzFsAw4J6UR+qcc65AybTQ2wHZZrbczLYA44DuiROY2caEl/sClroQnXPOJaNMEtPUBFYlvF4NtM89kaQrgP5AOaBjXjOS1AfoA3DIIYcUNVbnnHMFSNlBUTN72MwygRuAm/OZZoSZtTGzNhkZGalatHPOOZJL6GuA2gmva0Vl+RkHnLorQTnnnCu6ZBL6LKC+pLqSygE9gImJE0iqn/DyZOA/qQvROedcMgrtQzezrZKuBKYCpYGRZrZI0hAgy8wmAldKOgH4CdgAXFicQTvnnPu1ZA6KYmZTgCm5ygYmDF+d4ricc84VkV8p6pxzMeEJ3TnnYsITunPOxYQndOeciwlP6M45FxOe0J1zLiY8oTvnXEx4QnfOuZjwhO6cczHhCd0552LCE7pzzsWEJ3TnnIsJT+jOORcTntCdcy4mPKE751xMeEJ3zrmY8ITunHMxkVRCl9RZ0jJJ2ZJuzGN8f0mLJc2X9IakQ1MfqnPOuYIUmtAllQYeBroAjYGekhrnmmwO0MbMDgNeAIalOlDnnHMFS6aF3g7INrPlZrYFGAd0T5zAzN40s++il+8DtVIbpnPOucIkk9BrAqsSXq+OyvJzMfCvXQnKOedc0ZVJ5cwknQe0AY7JZ3wfoA/AIYcckspFO+fcXi+ZFvoaoHbC61pR2Q4knQD8BehmZj/mNSMzG2FmbcysTUZGxs7E65xzLh/JJPRZQH1JdSWVA3oAExMnkNQSeIyQzL9IfZjOOecKU2hCN7OtwJXAVGAJ8LyZLZI0RFK3aLK7gErAeElzJU3MZ3bOOeeKSVJ96GY2BZiSq2xgwvAJKY7LOedcEfmVos45FxOe0J1zLiY8oTvnXEx4QnfOuZjwhO6cczHhCd0552LCE7pzzsWEJ3TnnIsJT+jOORcTntCdcy4mPKE751xMeEJ3zrmY8ITunHMx4QndOediwhO6c87FhCd055yLCU/ozjkXE57QnXMuJpJK6JI6S1omKVvSjXmMP1rSR5K2Sjoj9WE655wrTKEJXVJp4GGgC9AY6Cmpca7J/gf0Ap5JdYDOOeeSk8xDotsB2Wa2HEDSOKA7sDhnAjNbEY3bXgwxOuecS0IyXS41gVUJr1dHZUUmqY+kLElZ69at25lZOOecy8duPShqZiPMrI2ZtcnIyNidi3bOudhLJqGvAWonvK4VlTnnnCtBkknos4D6kupKKgf0ACYWb1jOOeeKqtCEbmZbgSuBqcAS4HkzWyRpiKRuAJLaSloNnAk8JmlRcQbtnHPu15I5ywUzmwJMyVU2MGF4FqErxjnnXJr4laLOORcTntCdcy4mPKE751xMeEJ3zrmY8ITunHMx4QndOediwhO6c87FhCd055yLCU/ozjkXE57QnXMuJjyhO+dcTHhCd865mPCE7pxzMeEJ3TnnYsITunPOxYQndOeciwlP6M45FxOe0J1zLiaSSuiSOktaJilb0o15jC8v6blo/AeS6qQ6UOeccwUrNKFLKg08DHQBGgM9JTXONdnFwAYzqwfcC/wt1YE655wrWDIt9HZAtpktN7MtwDige65pugOjouEXgOMlKXVhOuecK0yZJKapCaxKeL0aaJ/fNGa2VdI3wEHAl4kTSeoD9IlebpK0bGeCLgGqkmvddqs9/7cyrfWnQXt8/UGa63CQBqVr0amS3u/wubu0DR6a34hkEnrKmNkIYMTuXGZxkJRlZm3SHceeyutv13kd7pq41l8yXS5rgNoJr2tFZXlOI6kMUBn4KhUBOuecS04yCX0WUF9SXUnlgB7AxFzTTAQujIbPAKabmaUuTOecc4UptMsl6hO/EpgKlAZGmtkiSUOALDObCDwJjJGUDawnJP042+O7jdLM62/XeR3umljWn7wh7Zxz8eBXijrnXEx4QnfOuZjYKxK6pFMlmaTfpTuWPZGkbZLmJvzVkXSQpDclbZL0UAHvPUXSHEnzJC2WdOnujL24JdTNQkmvSKqS4vmvkFQ1Gt6UzzR/kbRI0vwoltzXiaRVrjoaL6liCuY5RNIJBYy/TNIFKVhOKUkPRLEvkDRLUt1dnW9x2Sv60CU9B9QgnH1zazEto7SZbSuOeaebpE1mVilX2b5AS6Ap0NTMrszjfWWBlUA7M1stqTxQx8x2+oKy6Apkmdn2nZ1HKiXWjaRRwMdmNjSF818BtDGzL/P5HDoA9wDHmtmPUfIvZ2Zrd2GZZcxs6y4FvuP8EuvoaWC2md1TXMtLJUk9gdOBs8xsu6RawGYz27AL8yy29Y19C11SJeBIwv1mekRlpSX9PfrVnS+pb1TeVtK7UWvyQ0n7SeqV2AKVNEnSsdHwJkl3S5oHdJA0MPoFXyhpRM7tDyTVk/R6NN+PJGVKGi3p1IT5Pi0p9y0VSiwz22xm7wA/FDDZfoQzqb6K3vNjTjKX9BtJL0V1Mk/S76Py/lH9LZTULyqro3BzuNHAQqC2pOuiup4vaXBxrmsRvEe4aproM35V0mxJb+fsHRaw3i9H0y5SuKI6WdWBL83sRwAz+zInmeezPVeQ9I+otTlH0nHRtL0kTZQ0HXhD0r6SRkbvm5PCbfNtoJ6kY6N6mQgsjr6TdyV8pj/vyUm6IYp3nqS/RmVPSTojGv6rwt7ffEl/j8oGSbo2Gm4h6f1o/EuSDojKZ0j6W7SOH0s6Kp/6/TSnAWFmq3OSucJNCz+K4nojKjsw+iznR8s8LCGeMZJmEs4IzJD0YrS+syQdkZLaNbNY/wHnAk9Gw+8CrYE/E+45UyYqPxAoBywH2kZl+xOSUS/goYT5TSK0hgCM8MudM+7AhOExQNdo+APgtGi4AlAROAZ4OSqrDPw3J56S9gdsA+ZGfy/lGrdD/eTx3ieAL4Bno8+iVFT+HNAvGi4d1UFrYAGwL1AJWETYC6gDbAcOj6Y/kXDamQiNkknA0Wmqm00J6zAe6By9fgOoHw23J+wd5rneidsOsA/hR+ug6PUKoGrisnItv1L0uXwMPAIcE5Xntz1fQzj1GOB3wP+ibbIX4bYeOXHcAZwXDVeJ5r/vLtZRGeCfhO/fscBmoG40rg9wczRcHsgC6hJuCvguUDFXPT1FuOblIGAZv/Q2VIn+HwRcGw3PT6iXIcB90fAM4O5o+A/A63nEXiv6DOYCdwMto/IMwu1O6uaK60Hg1mi4IzA3IZ7ZwD7R62eAI6PhQ4Alqdged+ul/2nSE7g/Gh4Xva4LPGrRbo+ZrZfUjPBLPCsq2wiggu+bsg14MeH1cZKuJyTsA4FFkmYANc3spWi+OS3atyQ9IimDsEv3opXQ3U7gezNrsTNvNLM/RXV7AnAt0ImQPDoCF0TTbAO+kXQk4QdjM4CkCcBRhAvXVprZ+9FsT4z+5kSvKwH1gX/vTIy7aB9Jcwkt8yXANIW9wt8D4xO2n/LR/79a76j8KkmnRcO1CetT6NXWZrZJUmtCPR0HPKdwi+vZ5L09H0lIOpjZUkkrgQbR7KaZ2fpo+ESgW04rl5D0D4nWsahy6ghCC/1JQv18aGb/TVjeYTmtbsIPfH3CdvMPM/suink9O/qGsJf4pKRJhB/3n0mqTEjyb0VFowg/vDkmRP/PJjQcdmChq7Ah4XPrSNh7OZPwHf93TvwJcR1J+D5jZtMVjjXtH42baGbfR8MnAI0Tto/9JVUyszyPkyQr1gld0oGED6GZJCO0iIxw9WuytrJj11SFhOEfoi8lkioQWkhtzGyVpEG5ps3LaOA8QldQ7yLEtEcxswXAAkljCHsivXZiNpsThgXcaWaPpSC8XfW9mbVQONA3FbiC0Hr8OtkfQYUuvBOADmb2XdQIKGzb+Vm0Dc4AZkhaQLhqe3YR1iFH7jo+3XbheEeCXzUIokSWe3l9zWxqrulOKmjGFi58bAccT2ixX0n4zifrx+j/beSTDy10Z/0L+Jekz4FTgdeKsIwcietbirDHWVCXZZHFvQ/9DGCMmR1qZnXMrDYhocwDLlW470xO4l8GVJfUNirbLxq/AmihcLS7NuF2wnnJ+QJ+GbXQzgAws2+B1Yr6yxUeBpJzlP8poF803eIUrneJIKlSlKxytCAcJIXQJfHnaLrSUUvqbeBUSRUVDrqeFpXlNhW4KKpnJNWUdHAxrUZSohbkVYQuje+A/0YtORQ0jybNa70rE54n8J1CX/vhyS5XUkNJ9ROKcuo4v+35bULXF5IaEFrdeSXtqUBf6efjQC2TjWknTQX+rHAgHUkNom1gGtA75zsTfVd/Fm0Dlc1sCvB/QPPE8Wb2DbAhoX/8fOAtkiSplaQa0XAp4DBC/b4PHK3ojJeEuBLr91jC8Y2Necz6NaBvwnJ2ag84t1i30AndK7kftvEi0IjQdzhf0k/A42b2kKSzgQcl7QN8T2g1zST8CCwm7G5+lNeCzOxrSY8T+j8/Y8e9gPOBxxRul/ATcCaw3Mw+l7QEeDkla7ubKZyBsT9QLvrBOjHXD5OA6yU9RqjPzfzSOr8aGCHpYkLr6M9m9p6kp4APo2meMLM5yvUELDN7TVIj4L0o32wi7Ol8kep1LIoo1vmE7e5cYLikm4GyhO6+eeSx3sCrwGXRtrCMkCySVYmwzVYh7E1mA33MbEs+2/MjUVwLoul7WTg7Jvd8bwPuI3xHShG+A6cUrUaK5AlCl8dH0Y/IOuBUM3s1SnZZkrYAU4CbEt63H/DPaA9ZQP885n0h8Gj0o7Ccou0NHww8rnCGFoRt8yEz+0Hh4PWEqH6+IHQnDgJGRtvBd/xyj6vcrgIejqYrQ+guvKwIceVprzhtsaSKNrAFQKuoJeGcczst7l0uJZbCRRFLgAc9mTvnUsFb6M45FxPeQnfOuZjwhO6cczHhCd0552LCE7pzzsWEJ3TnnIuJ/wfr80/RHWUJsQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MC2MzVvI6iMm",
        "colab_type": "text"
      },
      "source": [
        "###Boosted Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "afhmdljp6gfO",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import recall_score as rs\n",
        "from sklearn.metrics import precision_score as ps\n",
        "from sklearn.metrics import f1_score as fs\n",
        "from sklearn.metrics import balanced_accuracy_score as bas\n",
        "from sklearn.metrics import confusion_matrix as cm"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "CaQx21TE6gfV",
        "colab": {}
      },
      "source": [
        "rf = RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
        "                       criterion='gini', max_depth=2, max_features='auto',\n",
        "                       max_leaf_nodes=None, max_samples=None,\n",
        "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
        "                       min_samples_leaf=2, min_samples_split=2,\n",
        "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
        "                       n_jobs=None, oob_score=False, random_state=None,\n",
        "                       verbose=0, warm_start=False)\n",
        "classifier = AdaBoostClassifier(rf,50,0.01,'SAMME.R',10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "7FsgDL6D6gfa",
        "colab": {}
      },
      "source": [
        "X = tdata[['location','country','gender','age','vis_wuhan','from_wuhan','symptom1','symptom2','symptom3','symptom4','symptom5','symptom6','diff_sym_hos']]\n",
        "Y = tdata['death']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "3ca7d132-b1fa-41af-f5a5-2580a5fe0ad5",
        "id": "zKX3NNZJ6gfg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        }
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(X,Y,test_size=0.2,random_state=0)\n",
        "classifier.fit(X_train,np.array(Y_train).reshape(Y_train.shape[0],1))\n"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "AdaBoostClassifier(algorithm='SAMME.R',\n",
              "                   base_estimator=RandomForestClassifier(bootstrap=True,\n",
              "                                                         ccp_alpha=0.0,\n",
              "                                                         class_weight=None,\n",
              "                                                         criterion='gini',\n",
              "                                                         max_depth=2,\n",
              "                                                         max_features='auto',\n",
              "                                                         max_leaf_nodes=None,\n",
              "                                                         max_samples=None,\n",
              "                                                         min_impurity_decrease=0.0,\n",
              "                                                         min_impurity_split=None,\n",
              "                                                         min_samples_leaf=2,\n",
              "                                                         min_samples_split=2,\n",
              "                                                         min_weight_fraction_leaf=0.0,\n",
              "                                                         n_estimators=100,\n",
              "                                                         n_jobs=None,\n",
              "                                                         oob_score=False,\n",
              "                                                         random_state=None,\n",
              "                                                         verbose=0,\n",
              "                                                         warm_start=False),\n",
              "                   learning_rate=0.01, n_estimators=50, random_state=10)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9DJgBXNfG2o8",
        "colab_type": "text"
      },
      "source": [
        "######Evaluation of the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8ddZ5qJO609l",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "pred = np.array(classifier.predict(X_test))\n",
        "\n",
        "recall = rs(Y_test,pred)\n",
        "precision = ps(Y_test,pred)\n",
        "f1 = fs(Y_test,pred)\n",
        "ma = classifier.score(X_test,Y_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7dbf0a1c-fd26-47ce-8479-503598322797",
        "id": "lt1X4VZv6gfm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "source": [
        "print('*** Evaluation metrics for test dataset ***\\n')\n",
        "print('Recall Score: ',recall)\n",
        "print('Precision Score: ',precision)\n",
        "print('F1 Score: ',f1)\n",
        "print('Accuracy: ',ma)\n",
        "a = pd.DataFrame(Y_test)\n",
        "a['pred']= classifier.predict(X_test)\n",
        "print('\\n\\tTable 3\\n')\n",
        "print(a.head())"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "*** Evaluation metrics for test dataset ***\n",
            "\n",
            "Recall Score:  0.75\n",
            "Precision Score:  1.0\n",
            "F1 Score:  0.8571428571428571\n",
            "Accuracy:  0.9333333333333333\n",
            "\n",
            "\tTable 3\n",
            "\n",
            "     death  pred\n",
            "130      0     0\n",
            "203      0     0\n",
            "170      1     0\n",
            "66       0     0\n",
            "181      0     0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WKx7g9wQ513H",
        "colab_type": "text"
      },
      "source": [
        "###Comparison of Evaluation Metrics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7xl4e1c76Rry",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 390
        },
        "outputId": "ea8beffd-1d39-4e2a-aa8f-bbcd0414a835"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "fig = plt.figure(figsize=(8,6))\n",
        "#plt.bar(['Logistic Regression','Decision Tree','SVM','Gaussian NB','Boosted Random Forest'],[f1_lr,f1_dtc,f1_svm,f1_gnb,f1],color=['red','green','purple','orange','Blue'])\n",
        "plt.plot(['Logistic Regression','Decision Tree','Boosted Random Forest','SVM','Gaussian NB'],[f1_lr,f1_dtc,f1,f1_svm,f1_gnb],color='purple',marker='D')\n",
        "plt.plot(['Logistic Regression','Decision Tree','Boosted Random Forest','SVM','Gaussian NB'],[ma_lr,ma_dtc,ma,ma_svm,ma_gnb],color='red',marker='^')\n",
        "plt.plot(['Logistic Regression','Decision Tree','Boosted Random Forest','SVM','Gaussian NB'],[precision_lr,precision_dtc,precision,precision_svm,precision_gnb],color='blue',marker='s')\n",
        "plt.plot(['Logistic Regression','Decision Tree','Boosted Random Forest','SVM','Gaussian NB'],[recall_lr,recall_dtc,recall,recall_svm,recall_gnb],color='green',marker='P')\n",
        "plt.legend(('F1 Score','Accuracy','Precision','Recall'))\n",
        "plt.title('Comparison of various models\\' performance')\n",
        "\n",
        "plt.show(fig)"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dG04IRZLCrC6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}