Artificial_Neural_network.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyOYEyHhzm4kAHXkg8Tq+t5M",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/OfekYa/DataMining/blob/main/Artificial_Neural_network.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install graphviz\n"
      ],
      "metadata": {
        "id": "qs3D7mPWqEus",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fed63ee7-d1e0-4fe7-af80-620b254dcbc5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: graphviz in /usr/local/lib/python3.10/dist-packages (0.20.1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install keras_sequential_ascii"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sQAKLHfc2FrS",
        "outputId": "e61c3a83-8cf2-453d-e216-5b22db19c98e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: keras_sequential_ascii in /usr/local/lib/python3.10/dist-packages (0.1.1)\n",
            "Requirement already satisfied: keras in /usr/local/lib/python3.10/dist-packages (from keras_sequential_ascii) (2.12.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install ann_visualizer\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rky8sVFMpgL3",
        "outputId": "d1185e29-6f89-4620-d59e-ede9c219b881"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting ann_visualizer\n",
            "  Downloading ann_visualizer-2.5.tar.gz (4.7 kB)\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Building wheels for collected packages: ann_visualizer\n",
            "  Building wheel for ann_visualizer (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for ann_visualizer: filename=ann_visualizer-2.5-py3-none-any.whl size=4167 sha256=c8d364c0dd22ce44575983997d7b19344de88dc5300acd70812b0b3559954155\n",
            "  Stored in directory: /root/.cache/pip/wheels/6e/0f/ae/f5dba91db71b1b32bf03d0ad18c32e86126093aba5ec6b6488\n",
            "Successfully built ann_visualizer\n",
            "Installing collected packages: ann_visualizer\n",
            "Successfully installed ann_visualizer-2.5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EmsBleA4IkQ5",
        "outputId": "77b3560b-1ca5-46c3-c413-9439db83ce31"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn import metrics\n",
        "from sklearn.metrics import accuracy_score, confusion_matrix, precision_score, recall_score, classification_report\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "file_path_bd = '/content/drive/MyDrive/Colab Notebooks//Data Mining/full_data.csv' \n",
        "\n",
        "df_bd = pd.read_csv(file_path_bd, index_col=0)\n",
        "df_bd[df_bd.columns[-1]].replace({'ckd':0,'notckd':1},inplace=True)\n",
        "df_bd = pd.get_dummies(df_bd, drop_first=True)\n",
        "\n",
        "X_bd, y_bd = df_bd[df_bd.columns[:-1]], df_bd[df_bd.columns[-1]]\n",
        "X_train_bd, X_test_bd, y_train_bd, y_test_bd = train_test_split(X_bd, y_bd, test_size=0.33)"
      ],
      "metadata": {
        "id": "fCYDl3G3JDPj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "file_path_ad = '/content/drive/MyDrive/Colab Notebooks//Data Mining/full_data_after_Discretization.csv'\n",
        "\n",
        "df_ad = pd.read_csv(file_path_ad, index_col=0)\n",
        "#print(df_ad[df_ad.columns[-1]].head())\n",
        "df_ad[df_ad.columns[-1]].replace({'ckd':0,'notckd':1},inplace=True)\n",
        "#print(df_ad[df_ad.columns[-1]].head())\n",
        "df_ad = pd.get_dummies(df_ad, drop_first=True)\n",
        "\n",
        "X_ad, y_ad = df_ad[df_ad.columns[:-1]], df_ad[df_ad.columns[-1]]\n",
        "\n",
        "X_train_ad, X_test_ad, y_train_ad, y_test_ad = train_test_split(X_ad, y_ad, test_size=0.33, shuffle=True)\n",
        "#print(df_ad[df_ad.columns[-1]].head())\n"
      ],
      "metadata": {
        "id": "vXIlZkdAJKsc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.layers import Dense,Activation\n",
        "from tensorflow.keras.models import Sequential \n",
        "from tensorflow.keras.callbacks import EarlyStopping  \n",
        "from sklearn.preprocessing import StandardScaler\n"
      ],
      "metadata": {
        "id": "O3RyuHOAf8is"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import keras.backend as K\n"
      ],
      "metadata": {
        "id": "P5r9qxlfmLUi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "callbacks = tf.keras.callbacks.EarlyStopping(\n",
        "    monitor=\"val_loss\",\n",
        "    min_delta=0,\n",
        "    patience=0,\n",
        "    verbose=0,\n",
        "    mode=\"auto\",\n",
        "    baseline=None,\n",
        "    restore_best_weights=False,\n",
        ")\n",
        "\n",
        "def train_ANN(X_train, Y_train,X_test, Y_test, epochs=100):\n",
        "    BATCH_SIZE=32\n",
        "    EPOCHS = epochs\n",
        "    IN_FEATURES = len(X_train.columns)\n",
        "    \n",
        "    print(\"shape of training data: \",X_train.shape)\n",
        "\n",
        "    sc = StandardScaler()\n",
        "    X_train = sc.fit_transform(X_train)\n",
        "\n",
        "    model = Sequential()\n",
        "  \n",
        "    #Adding First Hidden Layer\n",
        "    model.add(tf.keras.layers.Dense(units=IN_FEATURES ,input_dim=IN_FEATURES ,activation=\"relu\" ,use_bias=True))\n",
        "\n",
        "    #Adding Second Hidden Layer\n",
        "    model.add(tf.keras.layers.Dense(units=IN_FEATURES ,activation=\"relu\"))\n",
        "\n",
        "    #Adding Output Layer\n",
        "    model.add(tf.keras.layers.Dense(units=1,activation=\"sigmoid\"))\n",
        "\n",
        "    #Compiling ANN\n",
        "    model.compile(optimizer=\"adam\",loss=\"binary_crossentropy\",metrics=['accuracy'])\n",
        "    print(K.eval(model.optimizer.lr))\n",
        "\n",
        "    #Fitting ANN\n",
        "    model.fit(X_train, Y_train, batch_size=BATCH_SIZE, validation_split=0.2, epochs=EPOCHS, verbose=0,callbacks=[callbacks])\n",
        "   \n",
        "    plt.style.use('ggplot')\n",
        "\n",
        "    plt.plot(model.history.history['accuracy'], label='Accuracy')\n",
        "    plt.plot(model.history.history['loss'] ,label='Loss')     \n",
        "    plt.plot(model.history.history['val_accuracy'], label='val_Accuracy')\n",
        "    plt.plot(model.history.history['val_loss'] ,label='val_Loss')\n",
        "    plt.title('Model Loss and Accuracy')\n",
        "    plt.ylabel('Accuracy and Loss')\n",
        "    plt.xlabel('Epochs')   \n",
        "    plt.legend();\n",
        "    plt.show()\n",
        "\n",
        "    hist_df = pd.DataFrame(model.history.history)\n",
        "    hist_df['epoch'] = hist_df.index + 1\n",
        "    cols = list(hist_df.columns)\n",
        "    cols = [cols[-1]] + cols[:-1]\n",
        "    hist_df = hist_df[cols]\n",
        "    #hist_df.to_csv(checkpoint_no + '/' + 'history_df_' + model_name + '.csv')\n",
        "    print(hist_df.head(100))\n",
        "    values_of_best_model = hist_df[hist_df.val_loss == hist_df.val_loss.min()]\n",
        "    print('\\n   values_of_best_model: \\n',values_of_best_model)\n",
        "\n",
        "    return model, sc"
      ],
      "metadata": {
        "id": "Ws4xhiDP3MXt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ann_ad, sc_ad = train_ANN(X_train_ad, y_train_ad, X_test_ad, y_test_ad, epochs=100)"
      ],
      "metadata": {
        "id": "lkIciQLX36op",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "47be0e29-e27b-4f7c-dbc0-82075c359460"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "shape of training data:  (268, 29)\n",
            "0.001\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "    epoch      loss  accuracy  val_loss  val_accuracy\n",
            "0       1  1.106617  0.182243  0.957930      0.222222\n",
            "1       2  0.905346  0.238318  0.804904      0.351852\n",
            "2       3  0.764220  0.471963  0.688044      0.629630\n",
            "3       4  0.657675  0.658879  0.602039      0.759259\n",
            "4       5  0.577423  0.794393  0.537506      0.851852\n",
            "5       6  0.513894  0.859813  0.485086      0.833333\n",
            "6       7  0.464365  0.873832  0.440963      0.907407\n",
            "7       8  0.422018  0.869159  0.404377      0.907407\n",
            "8       9  0.386232  0.869159  0.375010      0.907407\n",
            "9      10  0.358104  0.873832  0.349558      0.907407\n",
            "10     11  0.333951  0.873832  0.330365      0.907407\n",
            "11     12  0.314584  0.883178  0.313367      0.907407\n",
            "12     13  0.297748  0.897196  0.298866      0.907407\n",
            "13     14  0.284044  0.897196  0.286585      0.907407\n",
            "14     15  0.271866  0.901869  0.277252      0.907407\n",
            "15     16  0.261627  0.901869  0.269587      0.888889\n",
            "16     17  0.252107  0.906542  0.262432      0.888889\n",
            "17     18  0.243539  0.911215  0.256780      0.870370\n",
            "18     19  0.235929  0.911215  0.251000      0.870370\n",
            "19     20  0.228320  0.920561  0.246534      0.870370\n",
            "20     21  0.221112  0.920561  0.243370      0.870370\n",
            "21     22  0.214134  0.920561  0.238815      0.870370\n",
            "22     23  0.207651  0.920561  0.234367      0.870370\n",
            "23     24  0.200721  0.920561  0.231375      0.870370\n",
            "24     25  0.194786  0.915888  0.229383      0.870370\n",
            "25     26  0.188775  0.915888  0.227082      0.888889\n",
            "26     27  0.183407  0.925234  0.224652      0.888889\n",
            "27     28  0.178103  0.925234  0.223120      0.888889\n",
            "28     29  0.173405  0.925234  0.222007      0.888889\n",
            "29     30  0.168699  0.925234  0.220013      0.888889\n",
            "30     31  0.163967  0.934579  0.219112      0.888889\n",
            "31     32  0.159921  0.934579  0.219038      0.888889\n",
            "32     33  0.155602  0.939252  0.217001      0.907407\n",
            "33     34  0.151217  0.939252  0.217265      0.907407\n",
            "\n",
            "   values_of_best_model: \n",
            "     epoch      loss  accuracy  val_loss  val_accuracy\n",
            "32     33  0.155602  0.939252  0.217001      0.907407\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ann_ad.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FVsNk4peEDkD",
        "outputId": "9cc25e97-3b4c-47a8-eea6-7239df1e7a85"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_84\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_252 (Dense)           (None, 29)                870       \n",
            "                                                                 \n",
            " dense_253 (Dense)           (None, 29)                870       \n",
            "                                                                 \n",
            " dense_254 (Dense)           (None, 1)                 30        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1,770\n",
            "Trainable params: 1,770\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from keras_sequential_ascii import keras2ascii\n",
        "keras2ascii(ann_ad)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kJ6tsG6U2DIY",
        "outputId": "79839256-ed55-4420-e54b-49caf2ac3ff8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "           OPERATION           DATA DIMENSIONS   WEIGHTS(N)   WEIGHTS(%)\n",
            "\n",
            "               Input   #####          29\n",
            "               Dense   XXXXX -------------------       870    49.2%\n",
            "                relu   #####          29\n",
            "               Dense   XXXXX -------------------       870    49.2%\n",
            "                relu   #####          29\n",
            "               Dense   XXXXX -------------------        30     1.7%\n",
            "             sigmoid   #####           1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import graphviz"
      ],
      "metadata": {
        "id": "0D3Mb4zIsbHX"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from ann_visualizer.visualize import ann_viz;\n",
        "\n",
        "ann_viz(ann_ad, title=\"My neural network\", filename='network.gv',view=True, );\n",
        "graph_file = graphviz.Source.from_file('network.gv')\n",
        "graph_file"
      ],
      "metadata": {
        "id": "a3TsgXdarZ_b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d365906b-3f5e-4500-a1ff-489dd04e6a10"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/svg+xml": "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n<!-- Generated by graphviz version 2.43.0 (0)\n -->\n<!-- Title: g Pages: 1 -->\n<svg width=\"1077pt\" height=\"737pt\"\n viewBox=\"0.00 0.00 1077.00 736.99\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 732.99)\">\n<title>g</title>\n<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-732.99 1073,-732.99 1073,4 -4,4\"/>\n<g id=\"clust1\" class=\"cluster\">\n<title>cluster_input</title>\n<polygon fill=\"none\" stroke=\"white\" points=\"22,-579.99 22,-720.99 1048,-720.99 1048,-579.99 22,-579.99\"/>\n<text text-anchor=\"middle\" x=\"535\" y=\"-705.79\" font-family=\"Times,serif\" font-size=\"14.00\">My neural network</text>\n<text text-anchor=\"middle\" x=\"535\" y=\"-642.79\" font-family=\"Times,serif\" font-size=\"14.00\">Input Layer (+19)</text>\n</g>\n<g id=\"clust2\" class=\"cluster\">\n<title>cluster_1</title>\n<polygon fill=\"none\" stroke=\"white\" points=\"8,-373.99 8,-451.99 1061,-451.99 1061,-373.99 8,-373.99\"/>\n<text text-anchor=\"middle\" x=\"1035.5\" y=\"-381.79\" font-family=\"Times,serif\" font-size=\"14.00\"> (+19)</text>\n</g>\n<g id=\"clust3\" class=\"cluster\">\n<title>cluster_2</title>\n<polygon fill=\"none\" stroke=\"white\" points=\"8,-191 8,-268.99 1061,-268.99 1061,-191 8,-191\"/>\n<text text-anchor=\"middle\" x=\"1035.5\" y=\"-198.8\" font-family=\"Times,serif\" font-size=\"14.00\"> (+19)</text>\n</g>\n<g id=\"clust4\" class=\"cluster\">\n<title>cluster_output</title>\n<polygon fill=\"none\" stroke=\"white\" points=\"490,-8 490,-86 579,-86 579,-8 490,-8\"/>\n<text text-anchor=\"middle\" x=\"534.5\" y=\"-15.8\" font-family=\"Times,serif\" font-size=\"14.00\">Output Layer</text>\n</g>\n<!-- 1 -->\n<g id=\"node1\" class=\"node\">\n<title>1</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"588\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"588\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">1</text>\n</g>\n<!-- 11 -->\n<g id=\"node11\" class=\"node\">\n<title>11</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"146\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"146\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">11</text>\n</g>\n<!-- 1&#45;&gt;11 -->\n<g id=\"edge1\" class=\"edge\">\n<title>1&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M571.48,-599.72C502.21,-571.36 235.45,-462.12 163.8,-432.78\"/>\n</g>\n<!-- 12 -->\n<g id=\"node12\" class=\"node\">\n<title>12</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"35\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"35\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">12</text>\n</g>\n<!-- 1&#45;&gt;12 -->\n<g id=\"edge11\" class=\"edge\">\n<title>1&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M571.15,-600.97C489.78,-574.34 138.04,-459.22 53.38,-431.51\"/>\n</g>\n<!-- 13 -->\n<g id=\"node13\" class=\"node\">\n<title>13</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"1034\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"1034\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">13</text>\n</g>\n<!-- 1&#45;&gt;13 -->\n<g id=\"edge21\" class=\"edge\">\n<title>1&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M604.4,-599.83C673.87,-571.64 943.56,-462.2 1016,-432.8\"/>\n</g>\n<!-- 14 -->\n<g id=\"node14\" class=\"node\">\n<title>14</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"923\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"923\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">14</text>\n</g>\n<!-- 1&#45;&gt;14 -->\n<g id=\"edge31\" class=\"edge\">\n<title>1&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M603.71,-598C659.59,-567.81 848.48,-465.75 906.14,-434.6\"/>\n</g>\n<!-- 15 -->\n<g id=\"node15\" class=\"node\">\n<title>15</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"812\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"812\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">15</text>\n</g>\n<!-- 1&#45;&gt;15 -->\n<g id=\"edge41\" class=\"edge\">\n<title>1&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M601.68,-595.44C641.28,-563.43 755.89,-470.83 797.06,-437.57\"/>\n</g>\n<!-- 16 -->\n<g id=\"node16\" class=\"node\">\n<title>16</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"701\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"701\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">16</text>\n</g>\n<!-- 1&#45;&gt;16 -->\n<g id=\"edge51\" class=\"edge\">\n<title>1&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M597.25,-591.68C618.23,-558.07 669.28,-476.29 690.98,-441.54\"/>\n</g>\n<!-- 17 -->\n<g id=\"node17\" class=\"node\">\n<title>17</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"590\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"590\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">17</text>\n</g>\n<!-- 1&#45;&gt;17 -->\n<g id=\"edge61\" class=\"edge\">\n<title>1&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M588.19,-589.46C588.56,-555.49 589.4,-479.55 589.79,-444.22\"/>\n</g>\n<!-- 18 -->\n<g id=\"node18\" class=\"node\">\n<title>18</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"479\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"479\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">18</text>\n</g>\n<!-- 1&#45;&gt;18 -->\n<g id=\"edge71\" class=\"edge\">\n<title>1&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M579.08,-591.68C558.84,-558.07 509.59,-476.29 488.67,-441.54\"/>\n</g>\n<!-- 19 -->\n<g id=\"node19\" class=\"node\">\n<title>19</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"368\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"368\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">19</text>\n</g>\n<!-- 1&#45;&gt;19 -->\n<g id=\"edge81\" class=\"edge\">\n<title>1&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M574.57,-595.44C535.67,-563.43 423.11,-470.83 382.68,-437.57\"/>\n</g>\n<!-- 20 -->\n<g id=\"node20\" class=\"node\">\n<title>20</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"257\" cy=\"-424.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"257\" y=\"-420.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">20</text>\n</g>\n<!-- 1&#45;&gt;20 -->\n<g id=\"edge91\" class=\"edge\">\n<title>1&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M572.48,-598C517.39,-567.88 331.49,-466.23 274.05,-434.82\"/>\n</g>\n<!-- 2 -->\n<g id=\"node2\" class=\"node\">\n<title>2</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"480\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"480\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">2</text>\n</g>\n<!-- 2&#45;&gt;11 -->\n<g id=\"edge2\" class=\"edge\">\n<title>2&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M464.34,-598C408.62,-567.81 220.3,-465.75 162.81,-434.6\"/>\n</g>\n<!-- 2&#45;&gt;12 -->\n<g id=\"edge12\" class=\"edge\">\n<title>2&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M463.63,-599.83C394.33,-571.64 125.24,-462.2 52.96,-432.8\"/>\n</g>\n<!-- 2&#45;&gt;13 -->\n<g id=\"edge22\" class=\"edge\">\n<title>2&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M496.88,-600.97C578.39,-574.34 930.77,-459.22 1015.59,-431.51\"/>\n</g>\n<!-- 2&#45;&gt;14 -->\n<g id=\"edge32\" class=\"edge\">\n<title>2&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M496.29,-599.83C565.29,-571.64 833.16,-462.2 905.12,-432.8\"/>\n</g>\n<!-- 2&#45;&gt;15 -->\n<g id=\"edge42\" class=\"edge\">\n<title>2&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M495.57,-598C550.95,-567.81 738.15,-465.75 795.29,-434.6\"/>\n</g>\n<!-- 2&#45;&gt;16 -->\n<g id=\"edge52\" class=\"edge\">\n<title>2&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M493.49,-595.44C532.57,-563.43 645.64,-470.83 686.26,-437.57\"/>\n</g>\n<!-- 2&#45;&gt;17 -->\n<g id=\"edge62\" class=\"edge\">\n<title>2&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M489,-591.68C509.42,-558.07 559.13,-476.29 580.25,-441.54\"/>\n</g>\n<!-- 2&#45;&gt;18 -->\n<g id=\"edge72\" class=\"edge\">\n<title>2&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M479.91,-589.46C479.72,-555.49 479.3,-479.55 479.1,-444.22\"/>\n</g>\n<!-- 2&#45;&gt;19 -->\n<g id=\"edge82\" class=\"edge\">\n<title>2&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M470.83,-591.68C450.04,-558.07 399.44,-476.29 377.93,-441.54\"/>\n</g>\n<!-- 2&#45;&gt;20 -->\n<g id=\"edge92\" class=\"edge\">\n<title>2&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M466.39,-595.44C426.95,-563.43 312.86,-470.83 271.88,-437.57\"/>\n</g>\n<!-- 3 -->\n<g id=\"node3\" class=\"node\">\n<title>3</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"372\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"372\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">3</text>\n</g>\n<!-- 3&#45;&gt;11 -->\n<g id=\"edge3\" class=\"edge\">\n<title>3&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M358.2,-595.44C318.14,-563.36 202.06,-470.39 160.78,-437.33\"/>\n</g>\n<!-- 3&#45;&gt;12 -->\n<g id=\"edge13\" class=\"edge\">\n<title>3&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M356.2,-598C299.98,-567.81 109.97,-465.75 51.96,-434.6\"/>\n</g>\n<!-- 3&#45;&gt;13 -->\n<g id=\"edge23\" class=\"edge\">\n<title>3&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M389.37,-601.74C482.19,-576.36 919.13,-456.9 1015.26,-430.61\"/>\n</g>\n<!-- 3&#45;&gt;14 -->\n<g id=\"edge33\" class=\"edge\">\n<title>3&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M388.79,-600.97C469.86,-574.34 820.33,-459.22 904.69,-431.51\"/>\n</g>\n<!-- 3&#45;&gt;15 -->\n<g id=\"edge43\" class=\"edge\">\n<title>3&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M388.45,-599.72C457.4,-571.36 722.96,-462.12 794.28,-432.78\"/>\n</g>\n<!-- 3&#45;&gt;16 -->\n<g id=\"edge53\" class=\"edge\">\n<title>3&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M387.43,-598C442.18,-567.88 626.96,-466.23 684.05,-434.82\"/>\n</g>\n<!-- 3&#45;&gt;17 -->\n<g id=\"edge63\" class=\"edge\">\n<title>3&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M385.64,-595.17C424.41,-562.97 535.09,-471.08 575.23,-437.75\"/>\n</g>\n<!-- 3&#45;&gt;18 -->\n<g id=\"edge73\" class=\"edge\">\n<title>3&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M380.76,-591.68C400.62,-558.07 448.97,-476.29 469.51,-441.54\"/>\n</g>\n<!-- 3&#45;&gt;19 -->\n<g id=\"edge83\" class=\"edge\">\n<title>3&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M371.62,-589.46C370.87,-555.49 369.19,-479.55 368.41,-444.22\"/>\n</g>\n<!-- 3&#45;&gt;20 -->\n<g id=\"edge93\" class=\"edge\">\n<title>3&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M362.78,-591.98C341.52,-558.51 289.15,-476.1 267.08,-441.35\"/>\n</g>\n<!-- 4 -->\n<g id=\"node4\" class=\"node\">\n<title>4</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"264\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"264\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">4</text>\n</g>\n<!-- 4&#45;&gt;11 -->\n<g id=\"edge4\" class=\"edge\">\n<title>4&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M254.54,-591.98C232.72,-558.51 178.99,-476.1 156.34,-441.35\"/>\n</g>\n<!-- 4&#45;&gt;12 -->\n<g id=\"edge14\" class=\"edge\">\n<title>4&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M250.36,-595.71C210.11,-563.9 92.04,-470.57 50.04,-437.38\"/>\n</g>\n<!-- 4&#45;&gt;13 -->\n<g id=\"edge24\" class=\"edge\">\n<title>4&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M281.53,-602.37C384.39,-578.19 908.69,-454.95 1015.13,-429.93\"/>\n</g>\n<!-- 4&#45;&gt;14 -->\n<g id=\"edge34\" class=\"edge\">\n<title>4&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M281.29,-601.74C373.69,-576.36 808.65,-456.9 904.35,-430.61\"/>\n</g>\n<!-- 4&#45;&gt;15 -->\n<g id=\"edge44\" class=\"edge\">\n<title>4&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M281,-600.87C362.12,-574.08 709.33,-459.4 793.54,-431.59\"/>\n</g>\n<!-- 4&#45;&gt;16 -->\n<g id=\"edge54\" class=\"edge\">\n<title>4&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M280.33,-599.72C348.67,-571.42 611.37,-462.61 682.92,-432.98\"/>\n</g>\n<!-- 4&#45;&gt;17 -->\n<g id=\"edge64\" class=\"edge\">\n<title>4&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M279.72,-597.76C334.54,-567.32 516.94,-466.06 573.28,-434.78\"/>\n</g>\n<!-- 4&#45;&gt;18 -->\n<g id=\"edge74\" class=\"edge\">\n<title>4&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M277.45,-595.17C315.69,-562.97 424.85,-471.08 464.43,-437.75\"/>\n</g>\n<!-- 4&#45;&gt;19 -->\n<g id=\"edge84\" class=\"edge\">\n<title>4&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M272.69,-591.37C292.07,-557.64 338.7,-476.49 358.67,-441.73\"/>\n</g>\n<!-- 4&#45;&gt;20 -->\n<g id=\"edge94\" class=\"edge\">\n<title>4&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M263.34,-589.46C262.03,-555.49 259.09,-479.55 257.72,-444.22\"/>\n</g>\n<!-- 5 -->\n<g id=\"node5\" class=\"node\">\n<title>5</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"156\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"156\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">5</text>\n</g>\n<!-- 5&#45;&gt;11 -->\n<g id=\"edge5\" class=\"edge\">\n<title>5&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M155.06,-589.46C153.18,-555.49 148.99,-479.55 147.03,-444.22\"/>\n</g>\n<!-- 5&#45;&gt;12 -->\n<g id=\"edge15\" class=\"edge\">\n<title>5&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M146.3,-591.98C123.87,-558.43 68.55,-475.68 45.43,-441.09\"/>\n</g>\n<!-- 5&#45;&gt;13 -->\n<g id=\"edge25\" class=\"edge\">\n<title>5&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M173.54,-602.87C285.44,-579.8 898.6,-453.4 1014.94,-429.42\"/>\n</g>\n<!-- 5&#45;&gt;14 -->\n<g id=\"edge35\" class=\"edge\">\n<title>5&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M173.46,-602.37C275.92,-578.19 798.18,-454.95 904.21,-429.93\"/>\n</g>\n<!-- 5&#45;&gt;15 -->\n<g id=\"edge45\" class=\"edge\">\n<title>5&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M173.21,-601.74C265.19,-576.36 698.17,-456.9 793.43,-430.61\"/>\n</g>\n<!-- 5&#45;&gt;16 -->\n<g id=\"edge55\" class=\"edge\">\n<title>5&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M172.91,-600.87C253.58,-574.08 598.89,-459.4 682.65,-431.59\"/>\n</g>\n<!-- 5&#45;&gt;17 -->\n<g id=\"edge65\" class=\"edge\">\n<title>5&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M172.48,-599.61C240.77,-571.14 501.17,-462.54 572.08,-432.96\"/>\n</g>\n<!-- 5&#45;&gt;18 -->\n<g id=\"edge75\" class=\"edge\">\n<title>5&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M171.58,-597.76C225.89,-567.32 406.61,-466.06 462.43,-434.78\"/>\n</g>\n<!-- 5&#45;&gt;19 -->\n<g id=\"edge85\" class=\"edge\">\n<title>5&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M169.26,-595.17C206.97,-562.97 314.6,-471.08 353.64,-437.75\"/>\n</g>\n<!-- 5&#45;&gt;20 -->\n<g id=\"edge95\" class=\"edge\">\n<title>5&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M164.44,-591.37C183.21,-557.72 228.32,-476.89 247.79,-442\"/>\n</g>\n<!-- 6 -->\n<g id=\"node6\" class=\"node\">\n<title>6</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"48\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"48\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">6</text>\n</g>\n<!-- 6&#45;&gt;11 -->\n<g id=\"edge6\" class=\"edge\">\n<title>6&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M56.19,-591.37C74.4,-557.72 118.17,-476.89 137.06,-442\"/>\n</g>\n<!-- 6&#45;&gt;12 -->\n<g id=\"edge16\" class=\"edge\">\n<title>6&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M46.78,-589.46C44.33,-555.4 38.85,-479.15 36.33,-443.95\"/>\n</g>\n<!-- 6&#45;&gt;13 -->\n<g id=\"edge26\" class=\"edge\">\n<title>6&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M65.53,-603.27C185.93,-581.17 889.43,-452.03 1014.87,-429\"/>\n</g>\n<!-- 6&#45;&gt;14 -->\n<g id=\"edge36\" class=\"edge\">\n<title>6&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M65.48,-602.87C177,-579.8 788.06,-453.4 904.01,-429.42\"/>\n</g>\n<!-- 6&#45;&gt;15 -->\n<g id=\"edge46\" class=\"edge\">\n<title>6&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M65.39,-602.37C167.34,-578.21 686.54,-455.21 792.93,-430.01\"/>\n</g>\n<!-- 6&#45;&gt;16 -->\n<g id=\"edge56\" class=\"edge\">\n<title>6&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M65.13,-601.74C156.59,-576.39 586.74,-457.16 682.2,-430.7\"/>\n</g>\n<!-- 6&#45;&gt;17 -->\n<g id=\"edge66\" class=\"edge\">\n<title>6&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M64.82,-600.87C145.04,-574.08 488.45,-459.4 571.75,-431.59\"/>\n</g>\n<!-- 6&#45;&gt;18 -->\n<g id=\"edge76\" class=\"edge\">\n<title>6&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M64.37,-599.61C132.18,-571.14 390.78,-462.54 461.2,-432.96\"/>\n</g>\n<!-- 6&#45;&gt;19 -->\n<g id=\"edge86\" class=\"edge\">\n<title>6&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M63.43,-597.76C117.12,-567.39 295.45,-466.53 351.2,-435\"/>\n</g>\n<!-- 6&#45;&gt;20 -->\n<g id=\"edge96\" class=\"edge\">\n<title>6&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M61.39,-594.89C98.78,-562.51 204.07,-471.33 242.62,-437.94\"/>\n</g>\n<!-- 7 -->\n<g id=\"node7\" class=\"node\">\n<title>7</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"1022\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"1022\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">7</text>\n</g>\n<!-- 7&#45;&gt;11 -->\n<g id=\"edge7\" class=\"edge\">\n<title>7&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1004.5,-602.87C892.85,-579.8 281.09,-453.4 165.02,-429.42\"/>\n</g>\n<!-- 7&#45;&gt;12 -->\n<g id=\"edge17\" class=\"edge\">\n<title>7&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1004.45,-603.27C883.93,-581.17 179.72,-452.03 54.15,-429\"/>\n</g>\n<!-- 7&#45;&gt;13 -->\n<g id=\"edge27\" class=\"edge\">\n<title>7&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1023.13,-589.46C1025.38,-555.49 1030.42,-479.55 1032.76,-444.22\"/>\n</g>\n<!-- 7&#45;&gt;14 -->\n<g id=\"edge37\" class=\"edge\">\n<title>7&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1013.73,-591.37C995.33,-557.72 951.12,-476.89 932.03,-442\"/>\n</g>\n<!-- 7&#45;&gt;15 -->\n<g id=\"edge47\" class=\"edge\">\n<title>7&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1008.86,-595.17C971.6,-563.05 865.41,-471.52 826.51,-438\"/>\n</g>\n<!-- 7&#45;&gt;16 -->\n<g id=\"edge57\" class=\"edge\">\n<title>7&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1006.52,-597.76C952.67,-567.39 773.78,-466.53 717.86,-435\"/>\n</g>\n<!-- 7&#45;&gt;17 -->\n<g id=\"edge67\" class=\"edge\">\n<title>7&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1005.59,-599.61C937.63,-571.14 678.42,-462.54 607.84,-432.96\"/>\n</g>\n<!-- 7&#45;&gt;18 -->\n<g id=\"edge77\" class=\"edge\">\n<title>7&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1005.15,-600.87C924.78,-574.08 580.74,-459.4 497.29,-431.59\"/>\n</g>\n<!-- 7&#45;&gt;19 -->\n<g id=\"edge87\" class=\"edge\">\n<title>7&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1004.84,-601.74C913.14,-576.36 481.49,-456.9 386.51,-430.61\"/>\n</g>\n<!-- 7&#45;&gt;20 -->\n<g id=\"edge97\" class=\"edge\">\n<title>7&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1004.58,-602.37C902.39,-578.19 381.5,-454.95 275.75,-429.93\"/>\n</g>\n<!-- 8 -->\n<g id=\"node8\" class=\"node\">\n<title>8</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"914\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"914\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">8</text>\n</g>\n<!-- 8&#45;&gt;11 -->\n<g id=\"edge8\" class=\"edge\">\n<title>8&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M896.52,-602.37C793.92,-578.19 270.98,-454.95 164.82,-429.93\"/>\n</g>\n<!-- 8&#45;&gt;12 -->\n<g id=\"edge18\" class=\"edge\">\n<title>8&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M896.44,-602.87C784.41,-579.8 170.55,-453.4 54.08,-429.42\"/>\n</g>\n<!-- 8&#45;&gt;13 -->\n<g id=\"edge28\" class=\"edge\">\n<title>8&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M923.62,-591.98C945.86,-558.43 1000.72,-475.68 1023.66,-441.09\"/>\n</g>\n<!-- 8&#45;&gt;14 -->\n<g id=\"edge38\" class=\"edge\">\n<title>8&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M914.85,-589.46C916.54,-555.49 920.31,-479.55 922.07,-444.22\"/>\n</g>\n<!-- 8&#45;&gt;15 -->\n<g id=\"edge48\" class=\"edge\">\n<title>8&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M905.48,-591.37C886.47,-557.64 840.74,-476.49 821.15,-441.73\"/>\n</g>\n<!-- 8&#45;&gt;16 -->\n<g id=\"edge58\" class=\"edge\">\n<title>8&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M900.68,-595.17C862.79,-562.97 754.65,-471.08 715.43,-437.75\"/>\n</g>\n<!-- 8&#45;&gt;17 -->\n<g id=\"edge68\" class=\"edge\">\n<title>8&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M898.37,-597.76C843.89,-567.32 662.61,-466.06 606.62,-434.78\"/>\n</g>\n<!-- 8&#45;&gt;18 -->\n<g id=\"edge78\" class=\"edge\">\n<title>8&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M897.74,-599.72C829.72,-571.42 568.22,-462.61 497,-432.98\"/>\n</g>\n<!-- 8&#45;&gt;19 -->\n<g id=\"edge88\" class=\"edge\">\n<title>8&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M897.06,-600.87C816.24,-574.08 470.3,-459.4 386.39,-431.59\"/>\n</g>\n<!-- 8&#45;&gt;20 -->\n<g id=\"edge98\" class=\"edge\">\n<title>8&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M896.76,-601.74C804.64,-576.36 371.01,-456.9 275.6,-430.61\"/>\n</g>\n<!-- 9 -->\n<g id=\"node9\" class=\"node\">\n<title>9</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"806\" cy=\"-607.49\" rx=\"18\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"806\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">9</text>\n</g>\n<!-- 9&#45;&gt;11 -->\n<g id=\"edge9\" class=\"edge\">\n<title>9&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M788.69,-601.74C696.14,-576.36 260.53,-456.9 164.68,-430.61\"/>\n</g>\n<!-- 9&#45;&gt;12 -->\n<g id=\"edge19\" class=\"edge\">\n<title>9&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M788.45,-602.37C685.45,-578.19 160.47,-454.95 53.89,-429.93\"/>\n</g>\n<!-- 9&#45;&gt;13 -->\n<g id=\"edge29\" class=\"edge\">\n<title>9&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M819.92,-595.44C860.33,-563.36 977.45,-470.39 1019.09,-437.33\"/>\n</g>\n<!-- 9&#45;&gt;14 -->\n<g id=\"edge39\" class=\"edge\">\n<title>9&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M815.38,-591.98C837.01,-558.51 890.29,-476.1 912.75,-441.35\"/>\n</g>\n<!-- 9&#45;&gt;15 -->\n<g id=\"edge49\" class=\"edge\">\n<title>9&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M806.56,-589.46C807.69,-555.49 810.21,-479.55 811.38,-444.22\"/>\n</g>\n<!-- 9&#45;&gt;16 -->\n<g id=\"edge59\" class=\"edge\">\n<title>9&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M797.23,-591.37C777.66,-557.64 730.58,-476.49 710.42,-441.73\"/>\n</g>\n<!-- 9&#45;&gt;17 -->\n<g id=\"edge69\" class=\"edge\">\n<title>9&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M792.49,-595.17C754.07,-562.97 644.41,-471.08 604.63,-437.75\"/>\n</g>\n<!-- 9&#45;&gt;18 -->\n<g id=\"edge79\" class=\"edge\">\n<title>9&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M790.67,-598C736.25,-567.88 552.59,-466.23 495.85,-434.82\"/>\n</g>\n<!-- 9&#45;&gt;19 -->\n<g id=\"edge89\" class=\"edge\">\n<title>9&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M789.63,-599.72C721.14,-571.42 457.83,-462.61 386.12,-432.98\"/>\n</g>\n<!-- 9&#45;&gt;20 -->\n<g id=\"edge99\" class=\"edge\">\n<title>9&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M788.97,-600.87C707.61,-574.05 359.09,-459.15 275.21,-431.49\"/>\n</g>\n<!-- 10 -->\n<g id=\"node10\" class=\"node\">\n<title>10</title>\n<ellipse fill=\"#2ecc71\" stroke=\"#2ecc71\" cx=\"697\" cy=\"-607.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"697\" y=\"-603.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#2ecc71\">10</text>\n</g>\n<!-- 10&#45;&gt;11 -->\n<g id=\"edge10\" class=\"edge\">\n<title>10&#45;&gt;11</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M678.67,-600.47C594.82,-572.92 248.39,-459.12 164.4,-431.54\"/>\n</g>\n<!-- 10&#45;&gt;12 -->\n<g id=\"edge20\" class=\"edge\">\n<title>10&#45;&gt;12</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M678.25,-601.36C582.52,-575.19 148.94,-456.64 53.58,-430.57\"/>\n</g>\n<!-- 10&#45;&gt;13 -->\n<g id=\"edge30\" class=\"edge\">\n<title>10&#45;&gt;13</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M714.18,-597.26C771.98,-566.21 959.1,-465.72 1016.86,-434.7\"/>\n</g>\n<!-- 10&#45;&gt;14 -->\n<g id=\"edge40\" class=\"edge\">\n<title>10&#45;&gt;14</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M711.83,-594.61C752.78,-561.82 867.1,-470.26 908.12,-437.41\"/>\n</g>\n<!-- 10&#45;&gt;15 -->\n<g id=\"edge50\" class=\"edge\">\n<title>10&#45;&gt;15</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M707,-590.74C728.75,-556.51 780.13,-475.64 801.94,-441.33\"/>\n</g>\n<!-- 10&#45;&gt;16 -->\n<g id=\"edge60\" class=\"edge\">\n<title>10&#45;&gt;16</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M697.41,-587.8C698.18,-553.03 699.82,-478.77 700.59,-444.08\"/>\n</g>\n<!-- 10&#45;&gt;17 -->\n<g id=\"edge70\" class=\"edge\">\n<title>10&#45;&gt;17</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M687.5,-590.43C667.25,-556.16 620,-476.24 599.63,-441.78\"/>\n</g>\n<!-- 10&#45;&gt;18 -->\n<g id=\"edge80\" class=\"edge\">\n<title>10&#45;&gt;18</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M682.36,-594.33C642.64,-561.36 533.22,-470.51 493.58,-437.6\"/>\n</g>\n<!-- 10&#45;&gt;19 -->\n<g id=\"edge90\" class=\"edge\">\n<title>10&#45;&gt;19</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M680.23,-597.26C623.8,-566.21 441.12,-465.72 384.73,-434.7\"/>\n</g>\n<!-- 10&#45;&gt;20 -->\n<g id=\"edge100\" class=\"edge\">\n<title>10&#45;&gt;20</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M679.21,-599.17C608.3,-570 346.31,-462.23 275.01,-432.9\"/>\n</g>\n<!-- 21 -->\n<g id=\"node21\" class=\"node\">\n<title>21</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"146\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"146\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">21</text>\n</g>\n<!-- 11&#45;&gt;21 -->\n<g id=\"edge101\" class=\"edge\">\n<title>11&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M146,-404.8C146,-370.03 146,-295.78 146,-261.08\"/>\n</g>\n<!-- 22 -->\n<g id=\"node22\" class=\"node\">\n<title>22</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"35\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"35\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">22</text>\n</g>\n<!-- 11&#45;&gt;22 -->\n<g id=\"edge111\" class=\"edge\">\n<title>11&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M136.15,-407.43C115.09,-373.08 65.87,-292.84 44.83,-258.52\"/>\n</g>\n<!-- 23 -->\n<g id=\"node23\" class=\"node\">\n<title>23</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"1034\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"1034\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">23</text>\n</g>\n<!-- 11&#45;&gt;23 -->\n<g id=\"edge121\" class=\"edge\">\n<title>11&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M164.96,-419.63C281.34,-395.91 897.89,-270.24 1014.84,-246.4\"/>\n</g>\n<!-- 24 -->\n<g id=\"node24\" class=\"node\">\n<title>24</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"923\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"923\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">24</text>\n</g>\n<!-- 11&#45;&gt;24 -->\n<g id=\"edge131\" class=\"edge\">\n<title>11&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M164.82,-419.11C271.32,-394.3 797.33,-271.77 904.08,-246.9\"/>\n</g>\n<!-- 25 -->\n<g id=\"node25\" class=\"node\">\n<title>25</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"812\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"812\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">25</text>\n</g>\n<!-- 11&#45;&gt;25 -->\n<g id=\"edge141\" class=\"edge\">\n<title>11&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M164.86,-418.37C261.18,-392.19 697.38,-273.65 793.31,-247.58\"/>\n</g>\n<!-- 26 -->\n<g id=\"node26\" class=\"node\">\n<title>26</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"701\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"701\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">26</text>\n</g>\n<!-- 11&#45;&gt;26 -->\n<g id=\"edge151\" class=\"edge\">\n<title>11&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M164.47,-417.47C249.01,-389.9 598.65,-275.87 682.75,-248.45\"/>\n</g>\n<!-- 27 -->\n<g id=\"node27\" class=\"node\">\n<title>27</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"590\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"590\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">27</text>\n</g>\n<!-- 11&#45;&gt;27 -->\n<g id=\"edge161\" class=\"edge\">\n<title>11&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M163.96,-416.17C235.66,-386.94 501.07,-278.75 572.31,-249.71\"/>\n</g>\n<!-- 28 -->\n<g id=\"node28\" class=\"node\">\n<title>28</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"479\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"479\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">28</text>\n</g>\n<!-- 11&#45;&gt;28 -->\n<g id=\"edge171\" class=\"edge\">\n<title>11&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M162.97,-414.27C220.09,-383.22 404.99,-282.72 462.06,-251.7\"/>\n</g>\n<!-- 29 -->\n<g id=\"node29\" class=\"node\">\n<title>29</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"368\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"368\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">29</text>\n</g>\n<!-- 11&#45;&gt;29 -->\n<g id=\"edge181\" class=\"edge\">\n<title>11&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M160.91,-411.33C201.35,-378.36 312.79,-287.51 353.15,-254.6\"/>\n</g>\n<!-- 30 -->\n<g id=\"node30\" class=\"node\">\n<title>30</title>\n<ellipse fill=\"#3498db\" stroke=\"#3498db\" cx=\"257\" cy=\"-241.49\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"257\" y=\"-237.79\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#3498db\">30</text>\n</g>\n<!-- 11&#45;&gt;30 -->\n<g id=\"edge191\" class=\"edge\">\n<title>11&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M155.85,-407.43C176.91,-373.08 226.13,-292.84 247.17,-258.52\"/>\n</g>\n<!-- 12&#45;&gt;21 -->\n<g id=\"edge102\" class=\"edge\">\n<title>12&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M44.85,-407.43C65.91,-373.08 115.13,-292.84 136.17,-258.52\"/>\n</g>\n<!-- 12&#45;&gt;22 -->\n<g id=\"edge112\" class=\"edge\">\n<title>12&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M35,-404.8C35,-370.03 35,-295.78 35,-261.08\"/>\n</g>\n<!-- 12&#45;&gt;23 -->\n<g id=\"edge122\" class=\"edge\">\n<title>12&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M54.06,-420.04C179.77,-397.26 888.42,-268.87 1014.74,-245.99\"/>\n</g>\n<!-- 12&#45;&gt;24 -->\n<g id=\"edge132\" class=\"edge\">\n<title>12&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M53.96,-419.63C170.34,-395.91 786.89,-270.24 903.84,-246.4\"/>\n</g>\n<!-- 12&#45;&gt;25 -->\n<g id=\"edge142\" class=\"edge\">\n<title>12&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M53.82,-419.11C160.32,-394.3 686.33,-271.77 793.08,-246.9\"/>\n</g>\n<!-- 12&#45;&gt;26 -->\n<g id=\"edge152\" class=\"edge\">\n<title>12&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M53.86,-418.37C150.18,-392.19 586.38,-273.65 682.31,-247.58\"/>\n</g>\n<!-- 12&#45;&gt;27 -->\n<g id=\"edge162\" class=\"edge\">\n<title>12&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M53.47,-417.47C138.01,-389.9 487.65,-275.87 571.75,-248.45\"/>\n</g>\n<!-- 12&#45;&gt;28 -->\n<g id=\"edge172\" class=\"edge\">\n<title>12&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M52.96,-416.17C124.66,-386.94 390.07,-278.75 461.31,-249.71\"/>\n</g>\n<!-- 12&#45;&gt;29 -->\n<g id=\"edge182\" class=\"edge\">\n<title>12&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M51.97,-414.27C109.09,-383.22 293.99,-282.72 351.06,-251.7\"/>\n</g>\n<!-- 12&#45;&gt;30 -->\n<g id=\"edge192\" class=\"edge\">\n<title>12&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M49.91,-411.33C90.35,-378.36 201.79,-287.51 242.15,-254.6\"/>\n</g>\n<!-- 13&#45;&gt;21 -->\n<g id=\"edge103\" class=\"edge\">\n<title>13&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1015.04,-419.63C898.66,-395.91 282.11,-270.24 165.16,-246.4\"/>\n</g>\n<!-- 13&#45;&gt;22 -->\n<g id=\"edge113\" class=\"edge\">\n<title>13&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1014.94,-420.04C889.23,-397.26 180.58,-268.87 54.26,-245.99\"/>\n</g>\n<!-- 13&#45;&gt;23 -->\n<g id=\"edge123\" class=\"edge\">\n<title>13&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1034,-404.8C1034,-370.03 1034,-295.78 1034,-261.08\"/>\n</g>\n<!-- 13&#45;&gt;24 -->\n<g id=\"edge133\" class=\"edge\">\n<title>13&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1024.15,-407.43C1003.09,-373.08 953.87,-292.84 932.83,-258.52\"/>\n</g>\n<!-- 13&#45;&gt;25 -->\n<g id=\"edge143\" class=\"edge\">\n<title>13&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1019.09,-411.33C978.65,-378.36 867.21,-287.51 826.85,-254.6\"/>\n</g>\n<!-- 13&#45;&gt;26 -->\n<g id=\"edge153\" class=\"edge\">\n<title>13&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1017.03,-414.27C959.91,-383.22 775.01,-282.72 717.94,-251.7\"/>\n</g>\n<!-- 13&#45;&gt;27 -->\n<g id=\"edge163\" class=\"edge\">\n<title>13&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1016.04,-416.17C944.34,-386.94 678.93,-278.75 607.69,-249.71\"/>\n</g>\n<!-- 13&#45;&gt;28 -->\n<g id=\"edge173\" class=\"edge\">\n<title>13&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1015.53,-417.47C930.99,-389.9 581.35,-275.87 497.25,-248.45\"/>\n</g>\n<!-- 13&#45;&gt;29 -->\n<g id=\"edge183\" class=\"edge\">\n<title>13&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1015.14,-418.37C918.82,-392.19 482.62,-273.65 386.69,-247.58\"/>\n</g>\n<!-- 13&#45;&gt;30 -->\n<g id=\"edge193\" class=\"edge\">\n<title>13&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1015.18,-419.11C908.68,-394.3 382.67,-271.77 275.92,-246.9\"/>\n</g>\n<!-- 14&#45;&gt;21 -->\n<g id=\"edge104\" class=\"edge\">\n<title>14&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M904.18,-419.11C797.68,-394.3 271.67,-271.77 164.92,-246.9\"/>\n</g>\n<!-- 14&#45;&gt;22 -->\n<g id=\"edge114\" class=\"edge\">\n<title>14&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M904.04,-419.63C787.66,-395.91 171.11,-270.24 54.16,-246.4\"/>\n</g>\n<!-- 14&#45;&gt;23 -->\n<g id=\"edge124\" class=\"edge\">\n<title>14&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M932.85,-407.43C953.91,-373.08 1003.13,-292.84 1024.17,-258.52\"/>\n</g>\n<!-- 14&#45;&gt;24 -->\n<g id=\"edge134\" class=\"edge\">\n<title>14&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M923,-404.8C923,-370.03 923,-295.78 923,-261.08\"/>\n</g>\n<!-- 14&#45;&gt;25 -->\n<g id=\"edge144\" class=\"edge\">\n<title>14&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M913.15,-407.43C892.09,-373.08 842.87,-292.84 821.83,-258.52\"/>\n</g>\n<!-- 14&#45;&gt;26 -->\n<g id=\"edge154\" class=\"edge\">\n<title>14&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M908.09,-411.33C867.65,-378.36 756.21,-287.51 715.85,-254.6\"/>\n</g>\n<!-- 14&#45;&gt;27 -->\n<g id=\"edge164\" class=\"edge\">\n<title>14&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M906.03,-414.27C848.91,-383.22 664.01,-282.72 606.94,-251.7\"/>\n</g>\n<!-- 14&#45;&gt;28 -->\n<g id=\"edge174\" class=\"edge\">\n<title>14&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M905.04,-416.17C833.34,-386.94 567.93,-278.75 496.69,-249.71\"/>\n</g>\n<!-- 14&#45;&gt;29 -->\n<g id=\"edge184\" class=\"edge\">\n<title>14&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M904.53,-417.47C819.99,-389.9 470.35,-275.87 386.25,-248.45\"/>\n</g>\n<!-- 14&#45;&gt;30 -->\n<g id=\"edge194\" class=\"edge\">\n<title>14&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M904.14,-418.37C807.82,-392.19 371.62,-273.65 275.69,-247.58\"/>\n</g>\n<!-- 15&#45;&gt;21 -->\n<g id=\"edge105\" class=\"edge\">\n<title>15&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M793.14,-418.37C696.82,-392.19 260.62,-273.65 164.69,-247.58\"/>\n</g>\n<!-- 15&#45;&gt;22 -->\n<g id=\"edge115\" class=\"edge\">\n<title>15&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M793.18,-419.11C686.68,-394.3 160.67,-271.77 53.92,-246.9\"/>\n</g>\n<!-- 15&#45;&gt;23 -->\n<g id=\"edge125\" class=\"edge\">\n<title>15&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M826.91,-411.33C867.35,-378.36 978.79,-287.51 1019.15,-254.6\"/>\n</g>\n<!-- 15&#45;&gt;24 -->\n<g id=\"edge135\" class=\"edge\">\n<title>15&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M821.85,-407.43C842.91,-373.08 892.13,-292.84 913.17,-258.52\"/>\n</g>\n<!-- 15&#45;&gt;25 -->\n<g id=\"edge145\" class=\"edge\">\n<title>15&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M812,-404.8C812,-370.03 812,-295.78 812,-261.08\"/>\n</g>\n<!-- 15&#45;&gt;26 -->\n<g id=\"edge155\" class=\"edge\">\n<title>15&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M802.15,-407.43C781.09,-373.08 731.87,-292.84 710.83,-258.52\"/>\n</g>\n<!-- 15&#45;&gt;27 -->\n<g id=\"edge165\" class=\"edge\">\n<title>15&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M797.09,-411.33C756.65,-378.36 645.21,-287.51 604.85,-254.6\"/>\n</g>\n<!-- 15&#45;&gt;28 -->\n<g id=\"edge175\" class=\"edge\">\n<title>15&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M795.03,-414.27C737.91,-383.22 553.01,-282.72 495.94,-251.7\"/>\n</g>\n<!-- 15&#45;&gt;29 -->\n<g id=\"edge185\" class=\"edge\">\n<title>15&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M794.04,-416.17C722.34,-386.94 456.93,-278.75 385.69,-249.71\"/>\n</g>\n<!-- 15&#45;&gt;30 -->\n<g id=\"edge195\" class=\"edge\">\n<title>15&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M793.53,-417.47C708.99,-389.9 359.35,-275.87 275.25,-248.45\"/>\n</g>\n<!-- 16&#45;&gt;21 -->\n<g id=\"edge106\" class=\"edge\">\n<title>16&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M682.53,-417.47C597.99,-389.9 248.35,-275.87 164.25,-248.45\"/>\n</g>\n<!-- 16&#45;&gt;22 -->\n<g id=\"edge116\" class=\"edge\">\n<title>16&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M682.14,-418.37C585.82,-392.19 149.62,-273.65 53.69,-247.58\"/>\n</g>\n<!-- 16&#45;&gt;23 -->\n<g id=\"edge126\" class=\"edge\">\n<title>16&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M717.97,-414.27C775.09,-383.22 959.99,-282.72 1017.06,-251.7\"/>\n</g>\n<!-- 16&#45;&gt;24 -->\n<g id=\"edge136\" class=\"edge\">\n<title>16&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M715.91,-411.33C756.35,-378.36 867.79,-287.51 908.15,-254.6\"/>\n</g>\n<!-- 16&#45;&gt;25 -->\n<g id=\"edge146\" class=\"edge\">\n<title>16&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M710.85,-407.43C731.91,-373.08 781.13,-292.84 802.17,-258.52\"/>\n</g>\n<!-- 16&#45;&gt;26 -->\n<g id=\"edge156\" class=\"edge\">\n<title>16&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M701,-404.8C701,-370.03 701,-295.78 701,-261.08\"/>\n</g>\n<!-- 16&#45;&gt;27 -->\n<g id=\"edge166\" class=\"edge\">\n<title>16&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M691.15,-407.43C670.09,-373.08 620.87,-292.84 599.83,-258.52\"/>\n</g>\n<!-- 16&#45;&gt;28 -->\n<g id=\"edge176\" class=\"edge\">\n<title>16&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M686.09,-411.33C645.65,-378.36 534.21,-287.51 493.85,-254.6\"/>\n</g>\n<!-- 16&#45;&gt;29 -->\n<g id=\"edge186\" class=\"edge\">\n<title>16&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M684.03,-414.27C626.91,-383.22 442.01,-282.72 384.94,-251.7\"/>\n</g>\n<!-- 16&#45;&gt;30 -->\n<g id=\"edge196\" class=\"edge\">\n<title>16&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M683.04,-416.17C611.34,-386.94 345.93,-278.75 274.69,-249.71\"/>\n</g>\n<!-- 17&#45;&gt;21 -->\n<g id=\"edge107\" class=\"edge\">\n<title>17&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M572.04,-416.17C500.34,-386.94 234.93,-278.75 163.69,-249.71\"/>\n</g>\n<!-- 17&#45;&gt;22 -->\n<g id=\"edge117\" class=\"edge\">\n<title>17&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M571.53,-417.47C486.99,-389.9 137.35,-275.87 53.25,-248.45\"/>\n</g>\n<!-- 17&#45;&gt;23 -->\n<g id=\"edge127\" class=\"edge\">\n<title>17&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M607.96,-416.17C679.66,-386.94 945.07,-278.75 1016.31,-249.71\"/>\n</g>\n<!-- 17&#45;&gt;24 -->\n<g id=\"edge137\" class=\"edge\">\n<title>17&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M606.97,-414.27C664.09,-383.22 848.99,-282.72 906.06,-251.7\"/>\n</g>\n<!-- 17&#45;&gt;25 -->\n<g id=\"edge147\" class=\"edge\">\n<title>17&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M604.91,-411.33C645.35,-378.36 756.79,-287.51 797.15,-254.6\"/>\n</g>\n<!-- 17&#45;&gt;26 -->\n<g id=\"edge157\" class=\"edge\">\n<title>17&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M599.85,-407.43C620.91,-373.08 670.13,-292.84 691.17,-258.52\"/>\n</g>\n<!-- 17&#45;&gt;27 -->\n<g id=\"edge167\" class=\"edge\">\n<title>17&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M590,-404.8C590,-370.03 590,-295.78 590,-261.08\"/>\n</g>\n<!-- 17&#45;&gt;28 -->\n<g id=\"edge177\" class=\"edge\">\n<title>17&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M580.15,-407.43C559.09,-373.08 509.87,-292.84 488.83,-258.52\"/>\n</g>\n<!-- 17&#45;&gt;29 -->\n<g id=\"edge187\" class=\"edge\">\n<title>17&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M575.09,-411.33C534.65,-378.36 423.21,-287.51 382.85,-254.6\"/>\n</g>\n<!-- 17&#45;&gt;30 -->\n<g id=\"edge197\" class=\"edge\">\n<title>17&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M573.03,-414.27C515.91,-383.22 331.01,-282.72 273.94,-251.7\"/>\n</g>\n<!-- 18&#45;&gt;21 -->\n<g id=\"edge108\" class=\"edge\">\n<title>18&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M462.03,-414.27C404.91,-383.22 220.01,-282.72 162.94,-251.7\"/>\n</g>\n<!-- 18&#45;&gt;22 -->\n<g id=\"edge118\" class=\"edge\">\n<title>18&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M461.04,-416.17C389.34,-386.94 123.93,-278.75 52.69,-249.71\"/>\n</g>\n<!-- 18&#45;&gt;23 -->\n<g id=\"edge128\" class=\"edge\">\n<title>18&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M497.47,-417.47C582.01,-389.9 931.65,-275.87 1015.75,-248.45\"/>\n</g>\n<!-- 18&#45;&gt;24 -->\n<g id=\"edge138\" class=\"edge\">\n<title>18&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M496.96,-416.17C568.66,-386.94 834.07,-278.75 905.31,-249.71\"/>\n</g>\n<!-- 18&#45;&gt;25 -->\n<g id=\"edge148\" class=\"edge\">\n<title>18&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M495.97,-414.27C553.09,-383.22 737.99,-282.72 795.06,-251.7\"/>\n</g>\n<!-- 18&#45;&gt;26 -->\n<g id=\"edge158\" class=\"edge\">\n<title>18&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M493.91,-411.33C534.35,-378.36 645.79,-287.51 686.15,-254.6\"/>\n</g>\n<!-- 18&#45;&gt;27 -->\n<g id=\"edge168\" class=\"edge\">\n<title>18&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M488.85,-407.43C509.91,-373.08 559.13,-292.84 580.17,-258.52\"/>\n</g>\n<!-- 18&#45;&gt;28 -->\n<g id=\"edge178\" class=\"edge\">\n<title>18&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M479,-404.8C479,-370.03 479,-295.78 479,-261.08\"/>\n</g>\n<!-- 18&#45;&gt;29 -->\n<g id=\"edge188\" class=\"edge\">\n<title>18&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M469.15,-407.43C448.09,-373.08 398.87,-292.84 377.83,-258.52\"/>\n</g>\n<!-- 18&#45;&gt;30 -->\n<g id=\"edge198\" class=\"edge\">\n<title>18&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M464.09,-411.33C423.65,-378.36 312.21,-287.51 271.85,-254.6\"/>\n</g>\n<!-- 19&#45;&gt;21 -->\n<g id=\"edge109\" class=\"edge\">\n<title>19&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M353.09,-411.33C312.65,-378.36 201.21,-287.51 160.85,-254.6\"/>\n</g>\n<!-- 19&#45;&gt;22 -->\n<g id=\"edge119\" class=\"edge\">\n<title>19&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M351.03,-414.27C293.91,-383.22 109.01,-282.72 51.94,-251.7\"/>\n</g>\n<!-- 19&#45;&gt;23 -->\n<g id=\"edge129\" class=\"edge\">\n<title>19&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M386.86,-418.37C483.18,-392.19 919.38,-273.65 1015.31,-247.58\"/>\n</g>\n<!-- 19&#45;&gt;24 -->\n<g id=\"edge139\" class=\"edge\">\n<title>19&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M386.47,-417.47C471.01,-389.9 820.65,-275.87 904.75,-248.45\"/>\n</g>\n<!-- 19&#45;&gt;25 -->\n<g id=\"edge149\" class=\"edge\">\n<title>19&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M385.96,-416.17C457.66,-386.94 723.07,-278.75 794.31,-249.71\"/>\n</g>\n<!-- 19&#45;&gt;26 -->\n<g id=\"edge159\" class=\"edge\">\n<title>19&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M384.97,-414.27C442.09,-383.22 626.99,-282.72 684.06,-251.7\"/>\n</g>\n<!-- 19&#45;&gt;27 -->\n<g id=\"edge169\" class=\"edge\">\n<title>19&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M382.91,-411.33C423.35,-378.36 534.79,-287.51 575.15,-254.6\"/>\n</g>\n<!-- 19&#45;&gt;28 -->\n<g id=\"edge179\" class=\"edge\">\n<title>19&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M377.85,-407.43C398.91,-373.08 448.13,-292.84 469.17,-258.52\"/>\n</g>\n<!-- 19&#45;&gt;29 -->\n<g id=\"edge189\" class=\"edge\">\n<title>19&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M368,-404.8C368,-370.03 368,-295.78 368,-261.08\"/>\n</g>\n<!-- 19&#45;&gt;30 -->\n<g id=\"edge199\" class=\"edge\">\n<title>19&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M358.15,-407.43C337.09,-373.08 287.87,-292.84 266.83,-258.52\"/>\n</g>\n<!-- 20&#45;&gt;21 -->\n<g id=\"edge110\" class=\"edge\">\n<title>20&#45;&gt;21</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M247.15,-407.43C226.09,-373.08 176.87,-292.84 155.83,-258.52\"/>\n</g>\n<!-- 20&#45;&gt;22 -->\n<g id=\"edge120\" class=\"edge\">\n<title>20&#45;&gt;22</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M242.09,-411.33C201.65,-378.36 90.21,-287.51 49.85,-254.6\"/>\n</g>\n<!-- 20&#45;&gt;23 -->\n<g id=\"edge130\" class=\"edge\">\n<title>20&#45;&gt;23</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M275.82,-419.11C382.32,-394.3 908.33,-271.77 1015.08,-246.9\"/>\n</g>\n<!-- 20&#45;&gt;24 -->\n<g id=\"edge140\" class=\"edge\">\n<title>20&#45;&gt;24</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M275.86,-418.37C372.18,-392.19 808.38,-273.65 904.31,-247.58\"/>\n</g>\n<!-- 20&#45;&gt;25 -->\n<g id=\"edge150\" class=\"edge\">\n<title>20&#45;&gt;25</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M275.47,-417.47C360.01,-389.9 709.65,-275.87 793.75,-248.45\"/>\n</g>\n<!-- 20&#45;&gt;26 -->\n<g id=\"edge160\" class=\"edge\">\n<title>20&#45;&gt;26</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M274.96,-416.17C346.66,-386.94 612.07,-278.75 683.31,-249.71\"/>\n</g>\n<!-- 20&#45;&gt;27 -->\n<g id=\"edge170\" class=\"edge\">\n<title>20&#45;&gt;27</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M273.97,-414.27C331.09,-383.22 515.99,-282.72 573.06,-251.7\"/>\n</g>\n<!-- 20&#45;&gt;28 -->\n<g id=\"edge180\" class=\"edge\">\n<title>20&#45;&gt;28</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M271.91,-411.33C312.35,-378.36 423.79,-287.51 464.15,-254.6\"/>\n</g>\n<!-- 20&#45;&gt;29 -->\n<g id=\"edge190\" class=\"edge\">\n<title>20&#45;&gt;29</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M266.85,-407.43C287.91,-373.08 337.13,-292.84 358.17,-258.52\"/>\n</g>\n<!-- 20&#45;&gt;30 -->\n<g id=\"edge200\" class=\"edge\">\n<title>20&#45;&gt;30</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M257,-404.8C257,-370.03 257,-295.78 257,-261.08\"/>\n</g>\n<!-- 31 -->\n<g id=\"node31\" class=\"node\">\n<title>31</title>\n<ellipse fill=\"#e74c3c\" stroke=\"#e74c3c\" cx=\"534\" cy=\"-58.5\" rx=\"19.5\" ry=\"19.5\"/>\n<text text-anchor=\"middle\" x=\"534\" y=\"-54.8\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"#e74c3c\">31</text>\n</g>\n<!-- 21&#45;&gt;31 -->\n<g id=\"edge201\" class=\"edge\">\n<title>21&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M163.43,-232.37C227.86,-202.31 452.26,-97.63 516.62,-67.6\"/>\n</g>\n<!-- 22&#45;&gt;31 -->\n<g id=\"edge202\" class=\"edge\">\n<title>22&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M53.05,-233.95C130.98,-205.68 438.09,-94.29 515.97,-66.04\"/>\n</g>\n<!-- 23&#45;&gt;31 -->\n<g id=\"edge203\" class=\"edge\">\n<title>23&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M1015.91,-233.95C937.82,-205.68 630.1,-94.29 552.07,-66.04\"/>\n</g>\n<!-- 24&#45;&gt;31 -->\n<g id=\"edge204\" class=\"edge\">\n<title>24&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M905.53,-232.37C840.92,-202.31 615.95,-97.63 551.42,-67.6\"/>\n</g>\n<!-- 25&#45;&gt;31 -->\n<g id=\"edge205\" class=\"edge\">\n<title>25&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M795.85,-229.98C746.72,-197.99 599.39,-102.07 550.19,-70.04\"/>\n</g>\n<!-- 26&#45;&gt;31 -->\n<g id=\"edge206\" class=\"edge\">\n<title>26&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M688.17,-226.59C657.09,-192.9 578,-107.19 546.87,-73.45\"/>\n</g>\n<!-- 27&#45;&gt;31 -->\n<g id=\"edge207\" class=\"edge\">\n<title>27&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M584.43,-222.48C573.69,-187.79 550.27,-112.08 539.55,-77.44\"/>\n</g>\n<!-- 28&#45;&gt;31 -->\n<g id=\"edge208\" class=\"edge\">\n<title>28&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M484.47,-222.48C495.02,-187.79 518.02,-112.08 528.55,-77.44\"/>\n</g>\n<!-- 29&#45;&gt;31 -->\n<g id=\"edge209\" class=\"edge\">\n<title>29&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M380.75,-226.59C411.65,-192.9 490.26,-107.19 521.21,-73.45\"/>\n</g>\n<!-- 30&#45;&gt;31 -->\n<g id=\"edge210\" class=\"edge\">\n<title>30&#45;&gt;31</title>\n<path fill=\"none\" stroke=\"#707070\" d=\"M273.09,-229.98C322.05,-197.99 468.85,-102.07 517.87,-70.04\"/>\n</g>\n</g>\n</svg>\n",
            "text/plain": [
              "<graphviz.sources.Source at 0x7faf8851ffa0>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def test_ANN(model, sc, X_test, Y_test):\n",
        "    print(\"shape of test data: \",X_test.shape)\n",
        "   \n",
        "    X_test = sc.transform(X_test)\n",
        "\n",
        "    #show actual predicted values\n",
        "    probs_pred = model.predict(X_test)\n",
        "    y_pred = [1 if y>=0.5 else 0 for y in probs_pred]\n",
        "\n",
        "    print('Test Accuracy: ', accuracy_score(Y_test, y_pred))\n",
        "    print('\\nConfusion Matrix: \\n', confusion_matrix(Y_test, y_pred))\n",
        "    print('\\nSummary: \\n', classification_report(Y_test, y_pred))\n",
        "    \n",
        "    tn, fp, fn, tp = confusion_matrix(Y_test, y_pred).ravel()\n",
        "    print(f'True Neg: {tn}')\n",
        "    print(f'False Pos: {fp}')\n",
        "    print(f'False Neg: {fn}')\n",
        "    print(f'True Pos: {tp}')\n",
        "\n",
        "    from sklearn.metrics import roc_curve\n",
        "    from sklearn.metrics import roc_auc_score\n",
        "\n",
        "  \n",
        "    title = 'ROC AUC SCORE: {}'.format(roc_auc_score(Y_test,probs_pred))\n",
        "    fpr, tpr, thresholds = roc_curve(Y_test, probs_pred)\n",
        "    plt.plot([0,1],[0,1],'k--')\n",
        "    plt.plot(fpr,tpr, label='ANN')\n",
        "    plt.xlabel('fpr')\n",
        "    plt.ylabel('tpr')\n",
        "    plt.title(title)\n",
        "    plt.show()\n",
        "\n",
        "    return probs_pred, y_pred \n"
      ],
      "metadata": {
        "id": "JmQF7gd6e0Xo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "#probs_pred_ad = torch.zeros((len(X_test_ad), 1),dtype=torch.int64)\n",
        "#y_pred_ad = torch.zeros(len(X_test_ad), dtype=torch.int64)\n",
        "\n",
        "probs_ad, pred_ad = test_ANN(model=ann_ad, sc=sc_ad, X_test=X_test_ad, Y_test=y_test_ad)\n",
        "#print(probs_ad)\n",
        "#print(pred_ad)\n",
        "\n",
        "labels_ad = torch.tensor(y_test_ad.values, dtype=torch.int64)\n",
        "y_probs_ad = torch.tensor(probs_ad)\n",
        "y_pred_ad = torch.tensor(pred_ad)\n",
        "\n",
        "print('PROB: ', y_probs_ad.shape)\n",
        "print('PRED: ', y_pred_ad)\n",
        "print('LABELS: ', labels_ad.shape)\n",
        "\n",
        "#wrongest_10_predictions = torch.topk(y_probs_ad, k=5)\n",
        "\n",
        "#probs_pred_ad[:] = torch.from_numpy(probs_ad).to(probs_pred_ad)\n",
        "#y_pred_ad[:] = torch.FloatTensor(pred_ad)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "Xlz8qNVWgQuK",
        "outputId": "46024544-84cd-45b9-c0df-99d5f021911b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "shape of test data:  (132, 29)\n",
            "5/5 [==============================] - 0s 3ms/step\n",
            "Test Accuracy:  0.8484848484848485\n",
            "\n",
            "Confusion Matrix: \n",
            " [[107   2]\n",
            " [ 18   5]]\n",
            "\n",
            "Summary: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "           0       0.86      0.98      0.91       109\n",
            "           1       0.71      0.22      0.33        23\n",
            "\n",
            "    accuracy                           0.85       132\n",
            "   macro avg       0.79      0.60      0.62       132\n",
            "weighted avg       0.83      0.85      0.81       132\n",
            "\n",
            "True Neg: 107\n",
            "False Pos: 2\n",
            "False Neg: 18\n",
            "True Pos: 5\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "PROB:  torch.Size([132, 1])\n",
            "PRED:  tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
            "        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,\n",
            "        0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
            "        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0,\n",
            "        0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
            "        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])\n",
            "LABELS:  torch.Size([132])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "def wrongest_predicitions(y_probs, y_preds, Y_test):\n",
        "  \n",
        "    wrong_pred = (y_preds != Y_test)  # If predicted class is wrong then value=True, else False.\n",
        "    wrong_pred_probs = y_probs * wrong_pred  # When the prediction is wrong, the value will be the probability value, else 0.000 .\n",
        "    wrongest_10_predictions = torch.topk(wrong_pred_probs, k=10)  # Take the 10 most extreme errors from 'wrong_pred_probs'.\n",
        "    wrongest_pred_probs, pred_indexs = wrongest_10_predictions\n",
        "\n",
        "    for wrong_pred in range(10): \n",
        "        print('----------', wrong_pred, '----------------')\n",
        "        \n",
        "        sample_pred_prob = wrongest_pred_probs[wrong_pred][wrong_pred]  # The probability of the misclassified sample.\n",
        "        print('sample_pred_prob: ',1-sample_pred_prob)\n",
        "        \n",
        "        sample_index = pred_indexs[wrong_pred][wrong_pred]  # The index of the misclassified sample.\n",
        "        print('sample_index: ', int(sample_index))\n",
        "\n",
        "        print(X_test_ad.iloc[int(sample_index)])\n",
        "\n",
        "        wrong_label_pred =  y_preds[sample_index]\n",
        "        print('wrong_label_pred: ', int(wrong_label_pred))\n",
        "        \n",
        "        correct_label = Y_test[sample_index] \n",
        "        print('correct_label: ', int(correct_label))\n",
        "\n",
        "  \n"
      ],
      "metadata": {
        "id": "t7L3fYpPkeVx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "wrongest_predicitions(y_probs_ad, y_pred_ad, labels_ad)\n"
      ],
      "metadata": {
        "id": "vVujsKqw2Bba",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "bde49e15-eb36-4362-9111-5e262fc3eec0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "---------- 0 ----------------\n",
            "sample_pred_prob:  tensor(0.8974)\n",
            "sample_index:  94\n",
            "Age                              34.00\n",
            "Blood Pressure                   70.00\n",
            "Specific Gravity                  1.02\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            148.00\n",
            "Sodium                          130.00\n",
            "Potassium                         3.80\n",
            "Hemoglobin                        6.00\n",
            "White Blood Cell Count         8406.00\n",
            "Red Blood Cell Count              4.70\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 0.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            0.00\n",
            "Packed Cell Volume_normal         1.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             0.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     0.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 122, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 1 ----------------\n",
            "sample_pred_prob:  tensor(0.9727)\n",
            "sample_index:  12\n",
            "Age                              64.000\n",
            "Blood Pressure                   90.000\n",
            "Specific Gravity                  1.015\n",
            "Albumin                           3.000\n",
            "Sugar                             2.000\n",
            "Blood Glucose Random            463.000\n",
            "Sodium                          135.000\n",
            "Potassium                         4.100\n",
            "Hemoglobin                       12.200\n",
            "White Blood Cell Count         9800.000\n",
            "Red Blood Cell Count              4.600\n",
            "Class                             0.000\n",
            "Red Blood_normal                  1.000\n",
            "Pus Cell_normal                   0.000\n",
            "Pus Cell clumps_present           1.000\n",
            "Bacteria_present                  0.000\n",
            "Blood Urea_high                   1.000\n",
            "Blood Urea_low                    0.000\n",
            "Blood Urea_normal                 0.000\n",
            "Serum Creatinine_high             0.000\n",
            "Serum Creatinine_low              0.000\n",
            "Serum Creatinine_normal           0.000\n",
            "Packed Cell Volume_low            0.000\n",
            "Packed Cell Volume_normal         1.000\n",
            "Hypertension_yes                  1.000\n",
            "Diabetes Mellitus_yes             1.000\n",
            "Coronary Artery Disease_yes       0.000\n",
            "Appetite_poor                     0.000\n",
            "Pedal Edema_yes                   0.000\n",
            "Name: 244, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 2 ----------------\n",
            "sample_pred_prob:  tensor(0.9373)\n",
            "sample_index:  47\n",
            "Age                              67.00\n",
            "Blood Pressure                   80.00\n",
            "Specific Gravity                  1.02\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            341.00\n",
            "Sodium                          138.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                       12.30\n",
            "White Blood Cell Count         6900.00\n",
            "Red Blood Cell Count              4.90\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 1.00\n",
            "Serum Creatinine_high             1.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            0.00\n",
            "Packed Cell Volume_normal         1.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     0.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 208, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 3 ----------------\n",
            "sample_pred_prob:  tensor(0.9717)\n",
            "sample_index:  105\n",
            "Age                              61.000\n",
            "Blood Pressure                   90.000\n",
            "Specific Gravity                  1.015\n",
            "Albumin                           0.000\n",
            "Sugar                             2.000\n",
            "Blood Glucose Random            148.000\n",
            "Sodium                          138.000\n",
            "Potassium                         4.400\n",
            "Hemoglobin                       12.500\n",
            "White Blood Cell Count         9800.000\n",
            "Red Blood Cell Count              4.700\n",
            "Class                             0.000\n",
            "Red Blood_normal                  1.000\n",
            "Pus Cell_normal                   1.000\n",
            "Pus Cell clumps_present           0.000\n",
            "Bacteria_present                  0.000\n",
            "Blood Urea_high                   1.000\n",
            "Blood Urea_low                    0.000\n",
            "Blood Urea_normal                 0.000\n",
            "Serum Creatinine_high             0.000\n",
            "Serum Creatinine_low              0.000\n",
            "Serum Creatinine_normal           1.000\n",
            "Packed Cell Volume_low            0.000\n",
            "Packed Cell Volume_normal         1.000\n",
            "Hypertension_yes                  0.000\n",
            "Diabetes Mellitus_yes             1.000\n",
            "Coronary Artery Disease_yes       0.000\n",
            "Appetite_poor                     1.000\n",
            "Pedal Edema_yes                   0.000\n",
            "Name: 113, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 4 ----------------\n",
            "sample_pred_prob:  tensor(0.9985)\n",
            "sample_index:  120\n",
            "Age                              47.00\n",
            "Blood Pressure                  100.00\n",
            "Specific Gravity                  1.01\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            204.00\n",
            "Sodium                          139.00\n",
            "Potassium                         4.20\n",
            "Hemoglobin                        9.70\n",
            "White Blood Cell Count         9200.00\n",
            "Red Blood Cell Count              4.50\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 1.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           1.00\n",
            "Packed Cell Volume_low            1.00\n",
            "Packed Cell Volume_normal         0.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             0.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     0.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 42, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 5 ----------------\n",
            "sample_pred_prob:  tensor(0.9938)\n",
            "sample_index:  17\n",
            "Age                              62.00\n",
            "Blood Pressure                   80.00\n",
            "Specific Gravity                  1.01\n",
            "Albumin                           2.00\n",
            "Sugar                             3.00\n",
            "Blood Glucose Random            423.00\n",
            "Sodium                          138.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                        9.60\n",
            "White Blood Cell Count         7500.00\n",
            "Red Blood Cell Count              4.70\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   1.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 0.00\n",
            "Serum Creatinine_high             1.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            1.00\n",
            "Packed Cell Volume_normal         0.00\n",
            "Hypertension_yes                  0.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     1.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 2, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 6 ----------------\n",
            "sample_pred_prob:  tensor(0.6761)\n",
            "sample_index:  55\n",
            "Age                              65.00\n",
            "Blood Pressure                   90.00\n",
            "Specific Gravity                  1.02\n",
            "Albumin                           2.00\n",
            "Sugar                             1.00\n",
            "Blood Glucose Random            270.00\n",
            "Sodium                          138.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                       12.00\n",
            "White Blood Cell Count         9800.00\n",
            "Red Blood Cell Count              4.90\n",
            "Class                             0.00\n",
            "Red Blood_normal                  0.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 1.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            1.00\n",
            "Packed Cell Volume_normal         0.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     1.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 35, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n",
            "---------- 7 ----------------\n",
            "sample_pred_prob:  tensor(0.5652)\n",
            "sample_index:  89\n",
            "Age                              60.00\n",
            "Blood Pressure                   90.00\n",
            "Specific Gravity                  1.02\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            148.00\n",
            "Sodium                           45.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                       10.90\n",
            "White Blood Cell Count         6200.00\n",
            "Red Blood Cell Count              3.60\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 0.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            1.00\n",
            "Packed Cell Volume_normal         0.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       1.00\n",
            "Appetite_poor                     0.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 21, dtype: float64\n",
            "wrong_label_pred:  1\n",
            "correct_label:  0\n",
            "---------- 8 ----------------\n",
            "sample_pred_prob:  tensor(0.6620)\n",
            "sample_index:  90\n",
            "Age                              68.00\n",
            "Blood Pressure                   90.00\n",
            "Specific Gravity                  1.01\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            238.00\n",
            "Sodium                          138.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                        9.80\n",
            "White Blood Cell Count         8000.00\n",
            "Red Blood Cell Count              3.30\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   1.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 0.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            1.00\n",
            "Packed Cell Volume_normal         0.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     1.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 219, dtype: float64\n",
            "wrong_label_pred:  1\n",
            "correct_label:  0\n",
            "---------- 9 ----------------\n",
            "sample_pred_prob:  tensor(0.9458)\n",
            "sample_index:  13\n",
            "Age                              59.00\n",
            "Blood Pressure                  100.00\n",
            "Specific Gravity                  1.02\n",
            "Albumin                           0.00\n",
            "Sugar                             0.00\n",
            "Blood Glucose Random            148.00\n",
            "Sodium                          138.00\n",
            "Potassium                         4.40\n",
            "Hemoglobin                        6.60\n",
            "White Blood Cell Count         8406.00\n",
            "Red Blood Cell Count              4.70\n",
            "Class                             0.00\n",
            "Red Blood_normal                  1.00\n",
            "Pus Cell_normal                   1.00\n",
            "Pus Cell clumps_present           0.00\n",
            "Bacteria_present                  0.00\n",
            "Blood Urea_high                   0.00\n",
            "Blood Urea_low                    0.00\n",
            "Blood Urea_normal                 0.00\n",
            "Serum Creatinine_high             0.00\n",
            "Serum Creatinine_low              0.00\n",
            "Serum Creatinine_normal           0.00\n",
            "Packed Cell Volume_low            0.00\n",
            "Packed Cell Volume_normal         1.00\n",
            "Hypertension_yes                  1.00\n",
            "Diabetes Mellitus_yes             1.00\n",
            "Coronary Artery Disease_yes       0.00\n",
            "Appetite_poor                     0.00\n",
            "Pedal Edema_yes                   0.00\n",
            "Name: 59, dtype: float64\n",
            "wrong_label_pred:  0\n",
            "correct_label:  1\n"
          ]
        }
      ]
    }
  ]
}