Anomaly_Detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Arrivat, Marie - Rajput, Rehan Razzaque</h3>\n",
    "<h1>Anomaly Detection (AD)</h1>\n",
    "<hr style=\"height:2px;border:none;color:#333;background-color:#333;\"/>\n",
    "<i>3<sup>th</sup> May 2019</i>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Anomaly detection (AD) refers to the process of detecting data points that do not conform with the rest of observations. Applications of anomaly detection include fraud and fault detection, surveillance, diagnosis, data cleanup, predictive maintenance.\n",
    "\n",
    "When we talk about AD, we usually look at it as an unsupervised (or semi-supervised) task, where the concept of anomaly is often not well defined or, in the best case, just few samples are labeled as anomalous. In this challenge, you will look at AD from a different perspective!\n",
    "\n",
    "The dataset you are going to work on consists of monitoring data generated by IT systems; such data is then processed by a monitoring system that executes some checks and detects a series of anomalies. This is a multi-label classification problem, where each check is a binary label corresponding to a specific type of anomaly. Your goal is to develop a machine learning model (or multiple ones) to accurately detect such anomalies.\n",
    "\n",
    "This will also involve a mixture of data exploration, pre-processing, model selection, and performance evaluation. You will also be asked to try one or more rule learning models, and compare them with other ML models both in terms of predictive performances and interpretability. Interpreatibility is indeed a strong requirement especially in applications like AD where understanding the output of a model is as important as the output itself.\n",
    "\n",
    "Please, bear in mind that the purpose of this challenge is not simply to find the best-performing model. You should rather make sure to understand the difficulties that come with this AD task."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"\">\n",
    "    <h2>Dataset Description</h2>\n",
    "    <hr style=\"height:1px;border:none;color:#333;background-color:#333;\"/>    \n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### * Location of the Dataset on zoe\n",
    "The data for this challenge is located at: `/mnt/datasets/anomaly`\n",
    "\n",
    "#### * Files\n",
    "\n",
    "You have a unique csv file with 36 features and 8 labels.\n",
    "Each record contains aggregate features computed over a given amount of time.\n",
    "\n",
    "#### * Attributes\n",
    "\n",
    "A brief outline of the available attributes is given below.\n",
    "\n",
    "1. <i>SessionNumber</i> (INTEGER): it identifies the session on which data is collected;\n",
    "* <i>SystemID</i> (INTEGER): it identifies the system generating the data;\n",
    "* <i>Date</i> (DATE): collection date;\n",
    "* <i>HighPriorityAlerts</i> (INTEGER [0, N]): number of high priority alerts in the session;\n",
    "* <i>Dumps</i> (INTEGER [0, N]): number of memory dumps;\n",
    "* <i>CleanupOOMDumps</i> (INTEGER) [0, N]): number of cleanup OOM dumps;\n",
    "* <i>CompositeOOMDums</i> (INTEGER [0, N]): number of composite OOM dumps;\n",
    "* <i>IndexServerRestarts</i> (INTEGER [0, N]): number of restarts of the index server;\n",
    "* <i>NameServerRestarts</i> (INTEGER [0, N]): number of restarts of the name server;\n",
    "* <i>XSEngineRestarts</i> (INTEGER [0, N]): number of restarts of the XSEngine;\n",
    "* <i>PreprocessorRestarts</i> (INTEGER [0, N]): number of restarts of the preprocessor;\n",
    "* <i>DaemonRestarts</i> (INTEGER [0, N]): number of restarts of the daemon process;\n",
    "* <i>StatisticsServerRestarts</i> (INTEGER [0, N]): number of restarts of the statistics server;\n",
    "* <i>CPU</i> (FLOAT [0, 100]): cpu usage;\n",
    "* <i>PhysMEM</i> (FLOAT [0, 100]): physical memory;\n",
    "* <i>InstanceMEM</i> (FLOAT [0, 100]): memory usage of one instance of the system;\n",
    "* <i>TablesAllocation</i> (FLOAT [0, 100]): memory allocated for tables;\n",
    "* <i>IndexServerAllocationLimit</i> (FLOAT [0, 100]): level of memory used by index server;\n",
    "* <i>ColumnUnloads</i> (INTEGER [0, N]): number of columns unloaded from the tables;\n",
    "* <i>DeltaSize</i> (INTEGER [0, N]): size of the delta store;\n",
    "* <i>MergeErrors</i> BOOLEAN [0, 1]: 1 if there are merge errors;\n",
    "* <i>BlockingPhaseSec</i> (INTEGER [0, N]): blocking phase duration in seconds;\n",
    "* <i>Disk</i> (FLOAT [0, 100]): disk usage;\n",
    "* <i>LargestTableSize</i> (INTEGER [0, N]): size of the largest table;\n",
    "* <i>LargestPartitionSize</i> (INTEGER [0, N]): size of the largest partition of a table;\n",
    "* <i>DiagnosisFiles</i> (INTEGER [0, N]): number of diagnosis files;\n",
    "* <i>DiagnosisFilesSize</i> (INTEGER [0, N]): size of diagnosis files;\n",
    "* <i>DaysWithSuccessfulDataBackups</i> (INTEGER [0, N]): number of days with successful data backups;\n",
    "* <i>DaysWithSuccessfulLogBackups</i> (INTEGER [0, N]): number of days with successful log backups;\n",
    "* <i>DaysWithFailedDataBackups</i> (INTEGER [0, N]): number of days with failed data backups;\n",
    "* <i>DaysWithFailedfulLogBackups</i> (INTEGER [0, N]): number of days with failed log backups;\n",
    "* <i>MinDailyNumberOfSuccessfulDataBackups</i> (INTEGER [0, N]): minimum number of successful data backups per day;\n",
    "* <i>MinDailyNumberOfSuccessfulLogBackups</i> (INTEGER [0, N]): minimum number of successful log backups per day;\n",
    "* <i>MaxDailyNumberOfFailedDataBackups</i> (INTEGER [0, N]): maximum number of failed data backups per day;\n",
    "* <i>MaxDailyNumberOfFailedLogBackups</i> (INTEGER [0, N]): maximum number of failed log backups per day;\n",
    "* <i>LogSegmentChange</i> (INTEGER [0, N]): changes in the number of log segments.\n",
    "\n",
    "#### * Labels\n",
    "\n",
    "Labels are binary. Each label refers to a different anomaly.\n",
    "\n",
    "* <i>Check1</i>;\n",
    "* <i>Check2</i>;\n",
    "* <i>Check3</i>;\n",
    "* <i>Check4</i>;\n",
    "* <i>Check5</i>;\n",
    "* <i>Check6</i>;\n",
    "* <i>Check7</i>;\n",
    "* <i>Check8</i>;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: fim in /mnt/workspace/.local/lib/python3.5/site-packages (6.27)\n",
      "\u001b[33mYou are using pip version 18.0, however version 20.0.2 is available.\n",
      "You should consider upgrading via the 'pip install --upgrade pip' command.\u001b[0m\n",
      "Requirement already satisfied: ruleset in /mnt/workspace/.local/lib/python3.5/site-packages (1.0.1)\n",
      "Requirement already satisfied: sklearn in /mnt/workspace/.local/lib/python3.5/site-packages (from ruleset) (0.0)\n",
      "Requirement already satisfied: numpy in /usr/local/lib/python3.5/dist-packages (from ruleset) (1.14.5)\n",
      "Requirement already satisfied: pandas in /usr/local/lib/python3.5/dist-packages (from ruleset) (0.23.4)\n",
      "Requirement already satisfied: scipy in /usr/local/lib/python3.5/dist-packages (from ruleset) (1.1.0)\n",
      "Requirement already satisfied: scikit-learn in /mnt/workspace/.local/lib/python3.5/site-packages (from sklearn->ruleset) (0.20.3)\n",
      "Requirement already satisfied: pytz>=2011k in /usr/local/lib/python3.5/dist-packages (from pandas->ruleset) (2018.5)\n",
      "Requirement already satisfied: python-dateutil>=2.5.0 in /usr/local/lib/python3.5/dist-packages (from pandas->ruleset) (2.7.3)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.5/dist-packages (from python-dateutil>=2.5.0->pandas->ruleset) (1.11.0)\n",
      "\u001b[33mYou are using pip version 18.0, however version 20.0.2 is available.\n",
      "You should consider upgrading via the 'pip install --upgrade pip' command.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "!pip install fim\n",
    "!pip install ruleset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sn\n",
    "import ruleset as rs\n",
    "import sklearn.tree as trees\n",
    "from IPython import display\n",
    "import sklearn.ensemble as ensem\n",
    "import sklearn.metrics as met\n",
    "import sklearn.model_selection as model_selection\n",
    "import sklearn.preprocessing as preprocess\n",
    "import sklearn.decomposition as decom\n",
    "import ruleset as rs\n",
    "from graphviz import Source"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Defining expected types </b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Some values are expected to have a certain type (float between 0 and 1, integer, positive integer...). Here, we split the types in order to check if the types are the good ones later on. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "float_0_100 = [\"CPU\", \"PhysMEM\", \"InstanceMEM\", \"TablesAllocation\", \"IndexServerAllocationLimit\", \"Disk\"]\n",
    "positive_integers = [\"HighPriorityAlerts\", \"Dumps\", \"CleanupOOMDumps\", \"CompositeOOMDums\", \"IndexServerRestarts\", \"NameServerRestarts\", \"XSEngineRestarts\", \"PreprocessorRestarts\", \"DaemonRestarts\", \"StatisticsServerRestarts\",\"ColumnUnloads\", \"DeltaSize \", \"BlockingPhaseSec\", \"LargestTableSize\", \"LargestPartitionSize\", \"DiagnosisFiles\",\"DiagnosisFilesSize\", \"DaysWithSuccessfulDataBackups\",\"DaysWithSuccessfulLogBackups\", \"DaysWithFailedDataBackups\",\"DaysWithFailedfulLogBackups\", \"MinDailyNumberOfSuccessfulDataBackups\",\"MinDailyNumberOfSuccessfulLogBackups\",\"MaxDailyNumberOfFailedDataBackups\", \"MaxDailyNumberOfFailedLogBackups\",\"LogSegmentChange\"]\n",
    "integers = [\"SessionNumber\", \"SystemID\"]\n",
    "all_integers = integers + positive_integers\n",
    "booleans = [\"MergeErrors\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Checking the data </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "head_tags = ['SessionNumber','SystemID','Date','HighPriorityAlerts','Dumps','CleanupOOMDumps','CompositeOOMDums','IndexServerRestarts','NameServerRestarts',\\\n",
    "             'XSEngineRestarts','PreprocessorRestarts','DaemonRestarts','StatisticsServerRestarts','CPU','PhysMEM','InstanceMEM','TablesAllocation','IndexServerAllocationLimit',\\\n",
    "             'ColumnUnloads','DeltaSize','MergeErrors','BlockingPhaseSec','Disk','LargestTableSize','LargestPartitionSize','DiagnosisFiles','DiagnosisFilesSize',\\\n",
    "            'DaysWithSuccessfulDataBackups','DaysWithSuccessfulLogBackups','DaysWithFailedDataBackups','DaysWithFailedfulLogBackups','MinDailyNumberOfSuccessfulDataBackups',\\\n",
    "             'MinDailyNumberOfSuccessfulLogBackups','MaxDailyNumberOfFailedDataBackups','MaxDailyNumberOfFailedLogBackups','LogSegmentChange','Check1','Check2','Check3','Check4',\\\n",
    "             'Check5','Check6','Check7','Check8']\n",
    "label_tags = head_tags[-8:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data= pd.read_csv('/mnt/datasets/anomaly/data.csv',sep=';',names=head_tags)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(['Date'],inplace=True,axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> We considered that the 'Date' had nothing to offer to us, so we dropped it. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "for tag in label_tags:\n",
    "    data = data[data[tag].notna()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Checking the Nans </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dumps\n",
      "CleanupOOMDumps\n",
      "CompositeOOMDums\n",
      "CPU\n",
      "PhysMEM\n",
      "InstanceMEM\n",
      "IndexServerAllocationLimit\n",
      "DeltaSize\n",
      "BlockingPhaseSec\n",
      "Disk\n",
      "LargestTableSize\n",
      "LargestPartitionSize\n",
      "DiagnosisFiles\n",
      "DiagnosisFilesSize\n",
      "LogSegmentChange\n"
     ]
    }
   ],
   "source": [
    "for column in data.columns:\n",
    "    if(data[column].isna().sum() >= 1):\n",
    "        print(column)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> All the columns above have some Nan values. We can notice that all of the column types are or integers or floats, it means that we can fill them back with a number. </p>\n",
    "<p> In order to fill the Nans back, we calculate the mean of all the values of the same column, so the new value won't be an outlier neither have an impact on the mean of the values. </p>\n",
    "<p> Carefull: Since the columns have specific types, we shouldn't put a float mean if the type is an int. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "for column in data.columns:\n",
    "    if(data[column].isna().sum() >= 1):\n",
    "        m = data[column].mean()\n",
    "        if(column in all_integers):\n",
    "            data[column] = data[column].fillna(int(m))\n",
    "        else:\n",
    "            data[column] = data[column].fillna(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "for column in data.columns:\n",
    "    if(data[column].isna().sum() >= 1):\n",
    "        print(column)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The above cell doesn't print anything, it means that all the Nans have been removed/replaced. </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Merging columns </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(236611, 43)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "feat_tags = list(set(data.columns).difference(set(label_tags)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['SessionNumber', 'SystemID', 'HighPriorityAlerts', 'Dumps',\n",
       "       'CleanupOOMDumps', 'CompositeOOMDums', 'IndexServerRestarts',\n",
       "       'NameServerRestarts', 'XSEngineRestarts', 'PreprocessorRestarts',\n",
       "       'DaemonRestarts', 'StatisticsServerRestarts', 'CPU', 'PhysMEM',\n",
       "       'InstanceMEM', 'TablesAllocation', 'IndexServerAllocationLimit',\n",
       "       'ColumnUnloads', 'DeltaSize', 'MergeErrors', 'BlockingPhaseSec', 'Disk',\n",
       "       'LargestTableSize', 'LargestPartitionSize', 'DiagnosisFiles',\n",
       "       'DiagnosisFilesSize', 'DaysWithSuccessfulDataBackups',\n",
       "       'DaysWithSuccessfulLogBackups', 'DaysWithFailedDataBackups',\n",
       "       'DaysWithFailedfulLogBackups', 'MinDailyNumberOfSuccessfulDataBackups',\n",
       "       'MinDailyNumberOfSuccessfulLogBackups',\n",
       "       'MaxDailyNumberOfFailedDataBackups', 'MaxDailyNumberOfFailedLogBackups',\n",
       "       'LogSegmentChange', 'Check1', 'Check2', 'Check3', 'Check4', 'Check5',\n",
       "       'Check6', 'Check7', 'Check8'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Changing the type of the data </b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> In order to work better on the data, we set a coomon type that is float. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.apply(lambda x:np.float32(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Visualizing the data </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.5/dist-packages/statsmodels/nonparametric/kde.py:488: RuntimeWarning: invalid value encountered in true_divide\n",
      "  binned = fast_linbin(X, a, b, gridsize) / (delta * nobs)\n",
      "/usr/local/lib/python3.5/dist-packages/statsmodels/nonparametric/kdetools.py:34: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  FAC1 = 2*(np.pi*bw/RANGE)**2\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x2880 with 44 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = len(data.columns) + 1\n",
    "r = 4\n",
    "n1 = n//r\n",
    "fig, axs = plt.subplots(n1, r,figsize=(20,40))\n",
    "fig.subplots_adjust(wspace = 0.5, hspace = 0.5)\n",
    "k = 2\n",
    "for i in range(n1):\n",
    "    for j in range(r):\n",
    "        if(k>= n-1):\n",
    "            break\n",
    "        sn.distplot(data[data.columns[k]].values,ax=axs[i][j])\n",
    "        axs[i][j].set_title(data.columns[k])\n",
    "        k = k+1\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> We realize that some of the features keep a near constant value for all the rows, so we remove them </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(['CleanupOOMDumps','PreprocessorRestarts','DaemonRestarts'],inplace=True,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['SessionNumber', 'SystemID', 'HighPriorityAlerts', 'Dumps',\n",
       "       'CompositeOOMDums', 'IndexServerRestarts', 'NameServerRestarts',\n",
       "       'XSEngineRestarts', 'StatisticsServerRestarts', 'CPU', 'PhysMEM',\n",
       "       'InstanceMEM', 'TablesAllocation', 'IndexServerAllocationLimit',\n",
       "       'ColumnUnloads', 'DeltaSize', 'MergeErrors', 'BlockingPhaseSec', 'Disk',\n",
       "       'LargestTableSize', 'LargestPartitionSize', 'DiagnosisFiles',\n",
       "       'DiagnosisFilesSize', 'DaysWithSuccessfulDataBackups',\n",
       "       'DaysWithSuccessfulLogBackups', 'DaysWithFailedDataBackups',\n",
       "       'DaysWithFailedfulLogBackups', 'MinDailyNumberOfSuccessfulDataBackups',\n",
       "       'MinDailyNumberOfSuccessfulLogBackups',\n",
       "       'MaxDailyNumberOfFailedDataBackups', 'MaxDailyNumberOfFailedLogBackups',\n",
       "       'LogSegmentChange', 'Check1', 'Check2', 'Check3', 'Check4', 'Check5',\n",
       "       'Check6', 'Check7', 'Check8'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indexes = data.columns\n",
    "indexes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Checking columns anomalies </b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The idea now is to check if anomalies are detected when a value doesn't fit its expected type. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "Check = ['Check1',\"Check2\",\"Check3\",\"Check4\",\"Check5\",\"Check6\",\"Check7\",\"Check8\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> We create a new column that gives 1 if an anomaly has been detected by any of the Checks, 0 otherwise. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['Check'] = data['Check1'] + data['Check2'] + data['Check3'] + data['Check4'] + data['Check5'] + data['Check6'] + data['Check7']+ data['Check8']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['Check'] = np.where(data['Check'] > 0, np.ones(data.shape[0]),np.zeros(data.shape[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['CPU',\n",
       " 'PhysMEM',\n",
       " 'InstanceMEM',\n",
       " 'TablesAllocation',\n",
       " 'IndexServerAllocationLimit',\n",
       " 'Disk']"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## We work only on the columns that have not been removed\n",
    "\n",
    "final_float_0_100 = []\n",
    "for column_name in float_0_100:\n",
    "    if(column_name in indexes):\n",
    "        final_float_0_100.append(column_name)\n",
    "float_0_100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For CPU, 6.0 anomalies are detected out of 6 values greater than 100.\n",
      "For CPU, 0 anomalies are detected out of 0 negative values.\n",
      "For PhysMEM, 3.0 anomalies are detected out of 3 values greater than 100.\n",
      "For PhysMEM, 0 anomalies are detected out of 0 negative values.\n",
      "For InstanceMEM, 0 anomalies are detected out of 0 values greater than 100.\n",
      "For InstanceMEM, 0 anomalies are detected out of 0 negative values.\n",
      "For TablesAllocation, 0 anomalies are detected out of 0 values greater than 100.\n",
      "For TablesAllocation, 0 anomalies are detected out of 0 negative values.\n",
      "For IndexServerAllocationLimit, 0 anomalies are detected out of 0 values greater than 100.\n",
      "For IndexServerAllocationLimit, 0 anomalies are detected out of 0 negative values.\n",
      "For Disk, 1395.0 anomalies are detected out of 3209 values greater than 100.\n",
      "For Disk, 0 anomalies are detected out of 0 negative values.\n"
     ]
    }
   ],
   "source": [
    "for column_name in final_float_0_100:\n",
    "    \n",
    "    above_out = data[data[column_name] > 100]\n",
    "    above_tot_out_of_range = len(above_out[\"Check\"])\n",
    "    above_tot_anomalies = 0\n",
    "    for i in range(above_tot_out_of_range):\n",
    "        above_tot_anomalies += above_out[\"Check\"].iloc[i]\n",
    "            \n",
    "    under_out = data[data[column_name]< 0]\n",
    "    under_tot_out_of_range = len(under_out[\"Check\"])\n",
    "    under_tot_anomalies = 0\n",
    "    for i in range(under_tot_out_of_range):\n",
    "        under_tot_anomalies += under_out[\"Check\"].iloc[i]\n",
    "            \n",
    "    print('For ' + column_name + \", \" + str(above_tot_anomalies) + \" anomalies are detected out of \" + str(above_tot_out_of_range) + \" values greater than 100.\")\n",
    "    print('For ' + column_name + \", \" + str(under_tot_anomalies) + \" anomalies are detected out of \" + str(under_tot_out_of_range) + \" negative values.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> We notice that <i>Disk</i> has a lot of values that are greater than 100. </p>\n",
    "<p> Not all those anomalies are detected. It means that being in the expected range is not a criterion for the anomaly detection for <i>Disk</i>. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['HighPriorityAlerts',\n",
       " 'Dumps',\n",
       " 'CompositeOOMDums',\n",
       " 'IndexServerRestarts',\n",
       " 'NameServerRestarts',\n",
       " 'XSEngineRestarts',\n",
       " 'StatisticsServerRestarts',\n",
       " 'ColumnUnloads',\n",
       " 'BlockingPhaseSec',\n",
       " 'LargestTableSize',\n",
       " 'LargestPartitionSize',\n",
       " 'DiagnosisFiles',\n",
       " 'DiagnosisFilesSize',\n",
       " 'DaysWithSuccessfulDataBackups',\n",
       " 'DaysWithSuccessfulLogBackups',\n",
       " 'DaysWithFailedDataBackups',\n",
       " 'DaysWithFailedfulLogBackups',\n",
       " 'MinDailyNumberOfSuccessfulDataBackups',\n",
       " 'MinDailyNumberOfSuccessfulLogBackups',\n",
       " 'MaxDailyNumberOfFailedDataBackups',\n",
       " 'MaxDailyNumberOfFailedLogBackups',\n",
       " 'LogSegmentChange']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## We work only on the columns that have not been removed\n",
    "\n",
    "final_positive_integers = []\n",
    "for column_name in positive_integers:\n",
    "    if(column_name in indexes):\n",
    "        final_positive_integers.append(column_name)\n",
    "final_positive_integers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data[data['LogSegmentChange'].apply(float.is_integer) == False]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For HighPriorityAlerts, 0 anomalies are detected out of 0 not integers values.\n",
      "For HighPriorityAlerts, 0 anomalies are detected out of 0 negative values.\n",
      "For Dumps, 0 anomalies are detected out of 0 not integers values.\n",
      "For Dumps, 0 anomalies are detected out of 0 negative values.\n",
      "For CompositeOOMDums, 0 anomalies are detected out of 0 not integers values.\n",
      "For CompositeOOMDums, 0 anomalies are detected out of 0 negative values.\n",
      "For IndexServerRestarts, 0 anomalies are detected out of 0 not integers values.\n",
      "For IndexServerRestarts, 0 anomalies are detected out of 0 negative values.\n",
      "For NameServerRestarts, 0 anomalies are detected out of 0 not integers values.\n",
      "For NameServerRestarts, 0 anomalies are detected out of 0 negative values.\n",
      "For XSEngineRestarts, 0 anomalies are detected out of 0 not integers values.\n",
      "For XSEngineRestarts, 0 anomalies are detected out of 0 negative values.\n",
      "For StatisticsServerRestarts, 0 anomalies are detected out of 0 not integers values.\n",
      "For StatisticsServerRestarts, 0 anomalies are detected out of 0 negative values.\n",
      "For ColumnUnloads, 0 anomalies are detected out of 0 not integers values.\n",
      "For ColumnUnloads, 0 anomalies are detected out of 0 negative values.\n",
      "For BlockingPhaseSec, 0 anomalies are detected out of 0 not integers values.\n",
      "For BlockingPhaseSec, 0 anomalies are detected out of 0 negative values.\n",
      "For LargestTableSize, 0 anomalies are detected out of 0 not integers values.\n",
      "For LargestTableSize, 0 anomalies are detected out of 0 negative values.\n",
      "For LargestPartitionSize, 0 anomalies are detected out of 0 not integers values.\n",
      "For LargestPartitionSize, 0 anomalies are detected out of 0 negative values.\n",
      "For DiagnosisFiles, 0 anomalies are detected out of 0 not integers values.\n",
      "For DiagnosisFiles, 0 anomalies are detected out of 0 negative values.\n",
      "For DiagnosisFilesSize, 0 anomalies are detected out of 0 not integers values.\n",
      "For DiagnosisFilesSize, 0 anomalies are detected out of 0 negative values.\n",
      "For DaysWithSuccessfulDataBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For DaysWithSuccessfulDataBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For DaysWithSuccessfulLogBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For DaysWithSuccessfulLogBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For DaysWithFailedDataBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For DaysWithFailedDataBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For DaysWithFailedfulLogBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For DaysWithFailedfulLogBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For MinDailyNumberOfSuccessfulDataBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For MinDailyNumberOfSuccessfulDataBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For MinDailyNumberOfSuccessfulLogBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For MinDailyNumberOfSuccessfulLogBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For MaxDailyNumberOfFailedDataBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For MaxDailyNumberOfFailedDataBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For MaxDailyNumberOfFailedLogBackups, 0 anomalies are detected out of 0 not integers values.\n",
      "For MaxDailyNumberOfFailedLogBackups, 0 anomalies are detected out of 0 negative values.\n",
      "For LogSegmentChange, 0 anomalies are detected out of 0 not integers values.\n",
      "For LogSegmentChange, 2819.0 anomalies are detected out of 5277 negative values.\n"
     ]
    }
   ],
   "source": [
    "for column_name in final_positive_integers:\n",
    "    not_integers = data[data[column_name].apply(float.is_integer) == False]\n",
    "    not_integers_tot = len(not_integers[\"Check\"])\n",
    "    not_integers_anomalies = 0\n",
    "    for i in range(not_integers_tot):\n",
    "        not_integers_anomalies += not_integers[\"Check\"].iloc[i]\n",
    "\n",
    "    \n",
    "    under_out = data[data[column_name]< 0]\n",
    "    under_tot_out_of_range = len(under_out[\"Check\"])\n",
    "    under_tot_anomalies = 0\n",
    "    for i in range(under_tot_out_of_range):\n",
    "        under_tot_anomalies += under_out[\"Check\"].iloc[i]\n",
    "\n",
    "    print('For ' + column_name + \", \" + str(not_integers_anomalies) + \" anomalies are detected out of \" + str(not_integers_tot) + \" not integers values.\")\n",
    "    print('For ' + column_name + \", \" + str(under_tot_anomalies) + \" anomalies are detected out of \" + str(under_tot_out_of_range) + \" negative values.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> We notice that <i>LogSegmentChange</i> has a lot of values that are negatives. </p>\n",
    "<p> Not all those anomalies are detected. It means that being in the expected range is not a criterion for the anomaly detection for <i>LogSegmentChange</i>. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['SessionNumber', 'SystemID']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## We work only on the columns that have not been removed\n",
    "\n",
    "final_integers = []\n",
    "for column_name in integers:\n",
    "    if(column_name in indexes):\n",
    "        final_integers.append(column_name)\n",
    "final_integers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For SessionNumber, 0 anomalies are detected out of 0 not integers values.\n",
      "For SystemID, 0 anomalies are detected out of 0 not integers values.\n"
     ]
    }
   ],
   "source": [
    "for column_name in final_integers:\n",
    "    not_integers = data[data[column_name].apply(float.is_integer) == False]\n",
    "    not_integers_tot = len(not_integers[\"Check\"])\n",
    "    not_integers_anomalies = 0\n",
    "    for i in range(not_integers_tot):\n",
    "        not_integers_anomalies += not_integers[\"Check\"].iloc[i]\n",
    "\n",
    "    print('For ' + column_name + \", \" + str(not_integers_anomalies) + \" anomalies are detected out of \" + str(not_integers_tot) + \" not integers values.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> No integer anomalies exists in the dataset. </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SessionNumber</th>\n",
       "      <th>SystemID</th>\n",
       "      <th>HighPriorityAlerts</th>\n",
       "      <th>Dumps</th>\n",
       "      <th>CompositeOOMDums</th>\n",
       "      <th>IndexServerRestarts</th>\n",
       "      <th>NameServerRestarts</th>\n",
       "      <th>XSEngineRestarts</th>\n",
       "      <th>StatisticsServerRestarts</th>\n",
       "      <th>CPU</th>\n",
       "      <th>...</th>\n",
       "      <th>LogSegmentChange</th>\n",
       "      <th>Check1</th>\n",
       "      <th>Check2</th>\n",
       "      <th>Check3</th>\n",
       "      <th>Check4</th>\n",
       "      <th>Check5</th>\n",
       "      <th>Check6</th>\n",
       "      <th>Check7</th>\n",
       "      <th>Check8</th>\n",
       "      <th>Check</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>0 rows × 41 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [SessionNumber, SystemID, HighPriorityAlerts, Dumps, CompositeOOMDums, IndexServerRestarts, NameServerRestarts, XSEngineRestarts, StatisticsServerRestarts, CPU, PhysMEM, InstanceMEM, TablesAllocation, IndexServerAllocationLimit, ColumnUnloads, DeltaSize, MergeErrors, BlockingPhaseSec, Disk, LargestTableSize, LargestPartitionSize, DiagnosisFiles, DiagnosisFilesSize, DaysWithSuccessfulDataBackups, DaysWithSuccessfulLogBackups, DaysWithFailedDataBackups, DaysWithFailedfulLogBackups, MinDailyNumberOfSuccessfulDataBackups, MinDailyNumberOfSuccessfulLogBackups, MaxDailyNumberOfFailedDataBackups, MaxDailyNumberOfFailedLogBackups, LogSegmentChange, Check1, Check2, Check3, Check4, Check5, Check6, Check7, Check8, Check]\n",
       "Index: []\n",
       "\n",
       "[0 rows x 41 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[~data.MergeErrors.isin([0,1])]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> No boolean anomalies exist in the dataset. </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Correlation Matrixes </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f787d9570b8>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2880x1440 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_corr = data.corr()\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(40,20))\n",
    "sn.heatmap(data_corr,xticklabels=data_corr.columns,yticklabels=data_corr.columns,ax=ax,cmap=sn.cm.rocket_r)\n",
    "# sn.pairplot(data_corr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col0 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col1 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col2 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col3 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col4 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col5 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col6 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col7 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col0 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col1 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col2 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col4 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col5 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col6 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col7 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col0 {\n",
       "            background-color:  #5b7ae5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col1 {\n",
       "            background-color:  #85a8fc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col2 {\n",
       "            background-color:  #6e90f2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col3 {\n",
       "            background-color:  #bfd3f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col4 {\n",
       "            background-color:  #799cf8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col5 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col6 {\n",
       "            background-color:  #94b6ff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col7 {\n",
       "            background-color:  #7b9ff9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col0 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col1 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col2 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col3 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col4 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col5 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col6 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col7 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col0 {\n",
       "            background-color:  #3f53c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col1 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col2 {\n",
       "            background-color:  #4961d2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col3 {\n",
       "            background-color:  #5673e0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col4 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col5 {\n",
       "            background-color:  #5e7de7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col6 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col7 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col1 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col2 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col3 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col4 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col5 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col6 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col7 {\n",
       "            background-color:  #b3cdfb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col1 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col2 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col3 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col4 {\n",
       "            background-color:  #4961d2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col5 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col6 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col7 {\n",
       "            background-color:  #cad8ef;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col1 {\n",
       "            background-color:  #5f7fe8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col2 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col3 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col4 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col5 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col6 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col7 {\n",
       "            background-color:  #bfd3f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col1 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col2 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col3 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col4 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col5 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col6 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col7 {\n",
       "            background-color:  #6b8df0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col0 {\n",
       "            background-color:  #b2ccfb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col1 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col2 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col3 {\n",
       "            background-color:  #7093f3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col4 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col5 {\n",
       "            background-color:  #6282ea;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col6 {\n",
       "            background-color:  #5f7fe8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col7 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col0 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col1 {\n",
       "            background-color:  #7093f3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col2 {\n",
       "            background-color:  #8badfd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col3 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col4 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col5 {\n",
       "            background-color:  #7699f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col6 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col7 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col0 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col1 {\n",
       "            background-color:  #b5cdfa;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col2 {\n",
       "            background-color:  #688aef;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col3 {\n",
       "            background-color:  #d3dbe7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col4 {\n",
       "            background-color:  #7b9ff9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col5 {\n",
       "            background-color:  #81a4fb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col6 {\n",
       "            background-color:  #5875e1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col7 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col0 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col1 {\n",
       "            background-color:  #b3cdfb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col2 {\n",
       "            background-color:  #6485ec;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col3 {\n",
       "            background-color:  #d5dbe5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col4 {\n",
       "            background-color:  #a6c4fe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col5 {\n",
       "            background-color:  #8caffe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col6 {\n",
       "            background-color:  #5e7de7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col7 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col0 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col1 {\n",
       "            background-color:  #a3c2fe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col2 {\n",
       "            background-color:  #6e90f2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col3 {\n",
       "            background-color:  #e5d8d1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col4 {\n",
       "            background-color:  #7699f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col5 {\n",
       "            background-color:  #96b7ff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col6 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col7 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col0 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col1 {\n",
       "            background-color:  #5f7fe8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col2 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col3 {\n",
       "            background-color:  #88abfd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col4 {\n",
       "            background-color:  #5875e1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col5 {\n",
       "            background-color:  #5673e0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col6 {\n",
       "            background-color:  #7295f4;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col7 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col0 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col1 {\n",
       "            background-color:  #6180e9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col2 {\n",
       "            background-color:  #6180e9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col3 {\n",
       "            background-color:  #688aef;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col4 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col5 {\n",
       "            background-color:  #5d7ce6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col6 {\n",
       "            background-color:  #5b7ae5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col7 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col1 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col2 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col3 {\n",
       "            background-color:  #5f7fe8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col4 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col5 {\n",
       "            background-color:  #5d7ce6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col6 {\n",
       "            background-color:  #6384eb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col7 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col0 {\n",
       "            background-color:  #3f53c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col1 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col2 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col3 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col4 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col5 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col6 {\n",
       "            background-color:  #5d7ce6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col7 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col1 {\n",
       "            background-color:  #3f53c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col2 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col3 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col4 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col5 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col6 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col7 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col0 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col1 {\n",
       "            background-color:  #7295f4;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col2 {\n",
       "            background-color:  #7597f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col3 {\n",
       "            background-color:  #9abbff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col4 {\n",
       "            background-color:  #6485ec;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col5 {\n",
       "            background-color:  #7b9ff9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col6 {\n",
       "            background-color:  #5f7fe8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col7 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col0 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col1 {\n",
       "            background-color:  #7699f6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col2 {\n",
       "            background-color:  #779af7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col3 {\n",
       "            background-color:  #9fbfff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col4 {\n",
       "            background-color:  #6b8df0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col5 {\n",
       "            background-color:  #7ea1fa;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col6 {\n",
       "            background-color:  #6282ea;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col7 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col1 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col2 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col3 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col4 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col5 {\n",
       "            background-color:  #6180e9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col6 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col7 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col1 {\n",
       "            background-color:  #4961d2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col2 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col3 {\n",
       "            background-color:  #4961d2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col4 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col5 {\n",
       "            background-color:  #6384eb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col6 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col7 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col0 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col1 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col2 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col3 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col4 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col5 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col6 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col7 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col0 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col1 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col2 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col3 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col4 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col5 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col6 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col7 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col1 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col2 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col3 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col4 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col5 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col6 {\n",
       "            background-color:  #6f92f3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col7 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col1 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col2 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col3 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col4 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col5 {\n",
       "            background-color:  #5875e1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col6 {\n",
       "            background-color:  #7da0f9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col7 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col1 {\n",
       "            background-color:  #4358cb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col2 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col3 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col4 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col5 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col6 {\n",
       "            background-color:  #4c66d6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col7 {\n",
       "            background-color:  #3b4cc0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col0 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col1 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col2 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col3 {\n",
       "            background-color:  #5673e0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col4 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col5 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col6 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col7 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col0 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col1 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col2 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col3 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col4 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col5 {\n",
       "            background-color:  #4b64d5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col6 {\n",
       "            background-color:  #6e90f2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col7 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col1 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col2 {\n",
       "            background-color:  #3f53c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col3 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col4 {\n",
       "            background-color:  #3c4ec2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col5 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col6 {\n",
       "            background-color:  #7093f3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col7 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col0 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col1 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col2 {\n",
       "            background-color:  #3d50c3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col3 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col4 {\n",
       "            background-color:  #3f53c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col5 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col6 {\n",
       "            background-color:  #6f92f3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col7 {\n",
       "            background-color:  #3e51c5;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col0 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col1 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col2 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col3 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col4 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col5 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col6 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col7 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col0 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col1 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col2 {\n",
       "            background-color:  #516ddb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col3 {\n",
       "            background-color:  #ebd3c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col4 {\n",
       "            background-color:  #a7c5fe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col5 {\n",
       "            background-color:  #6180e9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col6 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col7 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col0 {\n",
       "            background-color:  #4055c8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col1 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col2 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col3 {\n",
       "            background-color:  #6c8ff1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col4 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col5 {\n",
       "            background-color:  #536edd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col6 {\n",
       "            background-color:  #5572df;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col7 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col0 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col1 {\n",
       "            background-color:  #ebd3c6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col2 {\n",
       "            background-color:  #6b8df0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col3 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col4 {\n",
       "            background-color:  #a1c0ff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col5 {\n",
       "            background-color:  #7da0f9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col6 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col7 {\n",
       "            background-color:  #5d7ce6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col0 {\n",
       "            background-color:  #4a63d3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col1 {\n",
       "            background-color:  #aac7fd;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col2 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col3 {\n",
       "            background-color:  #a1c0ff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col4 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col5 {\n",
       "            background-color:  #5977e3;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col6 {\n",
       "            background-color:  #5875e1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col7 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col0 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col1 {\n",
       "            background-color:  #5e7de7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col2 {\n",
       "            background-color:  #4e68d8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col3 {\n",
       "            background-color:  #799cf8;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col4 {\n",
       "            background-color:  #5470de;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col5 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col6 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col7 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col0 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col1 {\n",
       "            background-color:  #465ecf;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col2 {\n",
       "            background-color:  #455cce;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col3 {\n",
       "            background-color:  #4961d2;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col4 {\n",
       "            background-color:  #485fd1;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col5 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col6 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col7 {\n",
       "            background-color:  #445acc;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col0 {\n",
       "            background-color:  #4257c9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col1 {\n",
       "            background-color:  #5d7ce6;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col2 {\n",
       "            background-color:  #4f69d9;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col3 {\n",
       "            background-color:  #5e7de7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col4 {\n",
       "            background-color:  #506bda;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col5 {\n",
       "            background-color:  #5673e0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col6 {\n",
       "            background-color:  #5673e0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col7 {\n",
       "            background-color:  #b40426;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col0 {\n",
       "            background-color:  #5e7de7;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col1 {\n",
       "            background-color:  #92b4fe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col2 {\n",
       "            background-color:  #8caffe;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col3 {\n",
       "            background-color:  #cedaeb;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col4 {\n",
       "            background-color:  #6b8df0;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col5 {\n",
       "            background-color:  #e9785d;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col6 {\n",
       "            background-color:  #98b9ff;\n",
       "        }    #T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col7 {\n",
       "            background-color:  #6687ed;\n",
       "        }</style>  \n",
       "<table id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02\" > \n",
       "<thead>    <tr> \n",
       "        <th class=\"blank level0\" ></th> \n",
       "        <th class=\"col_heading level0 col0\" >Check1</th> \n",
       "        <th class=\"col_heading level0 col1\" >Check2</th> \n",
       "        <th class=\"col_heading level0 col2\" >Check3</th> \n",
       "        <th class=\"col_heading level0 col3\" >Check4</th> \n",
       "        <th class=\"col_heading level0 col4\" >Check5</th> \n",
       "        <th class=\"col_heading level0 col5\" >Check6</th> \n",
       "        <th class=\"col_heading level0 col6\" >Check7</th> \n",
       "        <th class=\"col_heading level0 col7\" >Check8</th> \n",
       "    </tr></thead> \n",
       "<tbody>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row0\" class=\"row_heading level0 row0\" >SessionNumber</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col0\" class=\"data row0 col0\" >-0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col1\" class=\"data row0 col1\" >-0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col2\" class=\"data row0 col2\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col3\" class=\"data row0 col3\" >0.009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col4\" class=\"data row0 col4\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col5\" class=\"data row0 col5\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col6\" class=\"data row0 col6\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row0_col7\" class=\"data row0 col7\" >-0.001</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row1\" class=\"row_heading level0 row1\" >SystemID</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col0\" class=\"data row1 col0\" >-0.004</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col1\" class=\"data row1 col1\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col2\" class=\"data row1 col2\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col3\" class=\"data row1 col3\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col4\" class=\"data row1 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col5\" class=\"data row1 col5\" >-0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col6\" class=\"data row1 col6\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row1_col7\" class=\"data row1 col7\" >-0.004</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row2\" class=\"row_heading level0 row2\" >HighPriorityAlerts</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col0\" class=\"data row2 col0\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col1\" class=\"data row2 col1\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col2\" class=\"data row2 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col3\" class=\"data row2 col3\" >0.4</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col4\" class=\"data row2 col4\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col5\" class=\"data row2 col5\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col6\" class=\"data row2 col6\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row2_col7\" class=\"data row2 col7\" >0.2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row3\" class=\"row_heading level0 row3\" >Dumps</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col0\" class=\"data row3 col0\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col1\" class=\"data row3 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col2\" class=\"data row3 col2\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col3\" class=\"data row3 col3\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col4\" class=\"data row3 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col5\" class=\"data row3 col5\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col6\" class=\"data row3 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row3_col7\" class=\"data row3 col7\" >0.04</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row4\" class=\"row_heading level0 row4\" >CompositeOOMDums</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col0\" class=\"data row4 col0\" >0.009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col1\" class=\"data row4 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col2\" class=\"data row4 col2\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col3\" class=\"data row4 col3\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col4\" class=\"data row4 col4\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col5\" class=\"data row4 col5\" >0.09</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col6\" class=\"data row4 col6\" >0.009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row4_col7\" class=\"data row4 col7\" >0.04</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row5\" class=\"row_heading level0 row5\" >IndexServerRestarts</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col0\" class=\"data row5 col0\" >0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col1\" class=\"data row5 col1\" >0.009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col2\" class=\"data row5 col2\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col3\" class=\"data row5 col3\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col4\" class=\"data row5 col4\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col5\" class=\"data row5 col5\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col6\" class=\"data row5 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row5_col7\" class=\"data row5 col7\" >0.4</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row6\" class=\"row_heading level0 row6\" >NameServerRestarts</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col0\" class=\"data row6 col0\" >0.004</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col1\" class=\"data row6 col1\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col2\" class=\"data row6 col2\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col3\" class=\"data row6 col3\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col4\" class=\"data row6 col4\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col5\" class=\"data row6 col5\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col6\" class=\"data row6 col6\" >0.008</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row6_col7\" class=\"data row6 col7\" >0.4</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row7\" class=\"row_heading level0 row7\" >XSEngineRestarts</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col0\" class=\"data row7 col0\" >0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col1\" class=\"data row7 col1\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col2\" class=\"data row7 col2\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col3\" class=\"data row7 col3\" >0.09</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col4\" class=\"data row7 col4\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col5\" class=\"data row7 col5\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col6\" class=\"data row7 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row7_col7\" class=\"data row7 col7\" >0.4</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row8\" class=\"row_heading level0 row8\" >StatisticsServerRestarts</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col0\" class=\"data row8 col0\" >-0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col1\" class=\"data row8 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col2\" class=\"data row8 col2\" >0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col3\" class=\"data row8 col3\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col4\" class=\"data row8 col4\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col5\" class=\"data row8 col5\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col6\" class=\"data row8 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row8_col7\" class=\"data row8 col7\" >0.1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row9\" class=\"row_heading level0 row9\" >CPU</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col0\" class=\"data row9 col0\" >0.4</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col1\" class=\"data row9 col1\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col2\" class=\"data row9 col2\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col3\" class=\"data row9 col3\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col4\" class=\"data row9 col4\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col5\" class=\"data row9 col5\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col6\" class=\"data row9 col6\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row9_col7\" class=\"data row9 col7\" >0.06</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row10\" class=\"row_heading level0 row10\" >PhysMEM</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col0\" class=\"data row10 col0\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col1\" class=\"data row10 col1\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col2\" class=\"data row10 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col3\" class=\"data row10 col3\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col4\" class=\"data row10 col4\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col5\" class=\"data row10 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col6\" class=\"data row10 col6\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row10_col7\" class=\"data row10 col7\" >0.06</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row11\" class=\"row_heading level0 row11\" >InstanceMEM</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col0\" class=\"data row11 col0\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col1\" class=\"data row11 col1\" >0.4</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col2\" class=\"data row11 col2\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col3\" class=\"data row11 col3\" >0.5</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col4\" class=\"data row11 col4\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col5\" class=\"data row11 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col6\" class=\"data row11 col6\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row11_col7\" class=\"data row11 col7\" >0.08</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row12\" class=\"row_heading level0 row12\" >TablesAllocation</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col0\" class=\"data row12 col0\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col1\" class=\"data row12 col1\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col2\" class=\"data row12 col2\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col3\" class=\"data row12 col3\" >0.5</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col4\" class=\"data row12 col4\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col5\" class=\"data row12 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col6\" class=\"data row12 col6\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row12_col7\" class=\"data row12 col7\" >0.08</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row13\" class=\"row_heading level0 row13\" >IndexServerAllocationLimit</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col0\" class=\"data row13 col0\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col1\" class=\"data row13 col1\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col2\" class=\"data row13 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col3\" class=\"data row13 col3\" >0.5</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col4\" class=\"data row13 col4\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col5\" class=\"data row13 col5\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col6\" class=\"data row13 col6\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row13_col7\" class=\"data row13 col7\" >0.08</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row14\" class=\"row_heading level0 row14\" >ColumnUnloads</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col0\" class=\"data row14 col0\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col1\" class=\"data row14 col1\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col2\" class=\"data row14 col2\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col3\" class=\"data row14 col3\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col4\" class=\"data row14 col4\" >0.09</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col5\" class=\"data row14 col5\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col6\" class=\"data row14 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row14_col7\" class=\"data row14 col7\" >0.05</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row15\" class=\"row_heading level0 row15\" >DeltaSize</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col0\" class=\"data row15 col0\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col1\" class=\"data row15 col1\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col2\" class=\"data row15 col2\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col3\" class=\"data row15 col3\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col4\" class=\"data row15 col4\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col5\" class=\"data row15 col5\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col6\" class=\"data row15 col6\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row15_col7\" class=\"data row15 col7\" >0.05</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row16\" class=\"row_heading level0 row16\" >MergeErrors</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col0\" class=\"data row16 col0\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col1\" class=\"data row16 col1\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col2\" class=\"data row16 col2\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col3\" class=\"data row16 col3\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col4\" class=\"data row16 col4\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col5\" class=\"data row16 col5\" >0.09</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col6\" class=\"data row16 col6\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row16_col7\" class=\"data row16 col7\" >0.02</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row17\" class=\"row_heading level0 row17\" >BlockingPhaseSec</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col0\" class=\"data row17 col0\" >0.007</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col1\" class=\"data row17 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col2\" class=\"data row17 col2\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col3\" class=\"data row17 col3\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col4\" class=\"data row17 col4\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col5\" class=\"data row17 col5\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col6\" class=\"data row17 col6\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row17_col7\" class=\"data row17 col7\" >0.02</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row18\" class=\"row_heading level0 row18\" >Disk</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col0\" class=\"data row18 col0\" >-0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col1\" class=\"data row18 col1\" >-0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col2\" class=\"data row18 col2\" >-0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col3\" class=\"data row18 col3\" >-0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col4\" class=\"data row18 col4\" >-0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col5\" class=\"data row18 col5\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col6\" class=\"data row18 col6\" >0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row18_col7\" class=\"data row18 col7\" >-0.002</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row19\" class=\"row_heading level0 row19\" >LargestTableSize</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col0\" class=\"data row19 col0\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col1\" class=\"data row19 col1\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col2\" class=\"data row19 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col3\" class=\"data row19 col3\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col4\" class=\"data row19 col4\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col5\" class=\"data row19 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col6\" class=\"data row19 col6\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row19_col7\" class=\"data row19 col7\" >0.07</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row20\" class=\"row_heading level0 row20\" >LargestPartitionSize</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col0\" class=\"data row20 col0\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col1\" class=\"data row20 col1\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col2\" class=\"data row20 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col3\" class=\"data row20 col3\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col4\" class=\"data row20 col4\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col5\" class=\"data row20 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col6\" class=\"data row20 col6\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row20_col7\" class=\"data row20 col7\" >0.08</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row21\" class=\"row_heading level0 row21\" >DiagnosisFiles</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col0\" class=\"data row21 col0\" >0.004</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col1\" class=\"data row21 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col2\" class=\"data row21 col2\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col3\" class=\"data row21 col3\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col4\" class=\"data row21 col4\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col5\" class=\"data row21 col5\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col6\" class=\"data row21 col6\" >-0.004</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row21_col7\" class=\"data row21 col7\" >0.01</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row22\" class=\"row_heading level0 row22\" >DiagnosisFilesSize</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col0\" class=\"data row22 col0\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col1\" class=\"data row22 col1\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col2\" class=\"data row22 col2\" >0.0009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col3\" class=\"data row22 col3\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col4\" class=\"data row22 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col5\" class=\"data row22 col5\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col6\" class=\"data row22 col6\" >0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row22_col7\" class=\"data row22 col7\" >0.03</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row23\" class=\"row_heading level0 row23\" >DaysWithSuccessfulDataBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col0\" class=\"data row23 col0\" >-0.007</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col1\" class=\"data row23 col1\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col2\" class=\"data row23 col2\" >-0.007</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col3\" class=\"data row23 col3\" >-0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col4\" class=\"data row23 col4\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col5\" class=\"data row23 col5\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col6\" class=\"data row23 col6\" >-0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row23_col7\" class=\"data row23 col7\" >-0.005</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row24\" class=\"row_heading level0 row24\" >DaysWithSuccessfulLogBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col0\" class=\"data row24 col0\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col1\" class=\"data row24 col1\" >0.006</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col2\" class=\"data row24 col2\" >-0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col3\" class=\"data row24 col3\" >0.007</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col4\" class=\"data row24 col4\" >-0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col5\" class=\"data row24 col5\" >-0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col6\" class=\"data row24 col6\" >-0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row24_col7\" class=\"data row24 col7\" >0.003</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row25\" class=\"row_heading level0 row25\" >DaysWithFailedDataBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col0\" class=\"data row25 col0\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col1\" class=\"data row25 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col2\" class=\"data row25 col2\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col3\" class=\"data row25 col3\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col4\" class=\"data row25 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col5\" class=\"data row25 col5\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col6\" class=\"data row25 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row25_col7\" class=\"data row25 col7\" >0.03</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row26\" class=\"row_heading level0 row26\" >DaysWithFailedfulLogBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col0\" class=\"data row26 col0\" >0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col1\" class=\"data row26 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col2\" class=\"data row26 col2\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col3\" class=\"data row26 col3\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col4\" class=\"data row26 col4\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col5\" class=\"data row26 col5\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col6\" class=\"data row26 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row26_col7\" class=\"data row26 col7\" >0.02</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row27\" class=\"row_heading level0 row27\" >MinDailyNumberOfSuccessfulDataBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col0\" class=\"data row27 col0\" >0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col1\" class=\"data row27 col1\" >0.008</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col2\" class=\"data row27 col2\" >-0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col3\" class=\"data row27 col3\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col4\" class=\"data row27 col4\" >0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col5\" class=\"data row27 col5\" >-0.0009</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col6\" class=\"data row27 col6\" >-0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row27_col7\" class=\"data row27 col7\" >-0.008</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row28\" class=\"row_heading level0 row28\" >MinDailyNumberOfSuccessfulLogBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col0\" class=\"data row28 col0\" >-0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col1\" class=\"data row28 col1\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col2\" class=\"data row28 col2\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col3\" class=\"data row28 col3\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col4\" class=\"data row28 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col5\" class=\"data row28 col5\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col6\" class=\"data row28 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row28_col7\" class=\"data row28 col7\" >0.03</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row29\" class=\"row_heading level0 row29\" >MaxDailyNumberOfFailedDataBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col0\" class=\"data row29 col0\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col1\" class=\"data row29 col1\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col2\" class=\"data row29 col2\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col3\" class=\"data row29 col3\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col4\" class=\"data row29 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col5\" class=\"data row29 col5\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col6\" class=\"data row29 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row29_col7\" class=\"data row29 col7\" >0.03</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row30\" class=\"row_heading level0 row30\" >MaxDailyNumberOfFailedLogBackups</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col0\" class=\"data row30 col0\" >-0.001</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col1\" class=\"data row30 col1\" >-0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col2\" class=\"data row30 col2\" >0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col3\" class=\"data row30 col3\" >-0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col4\" class=\"data row30 col4\" >-0.004</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col5\" class=\"data row30 col5\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col6\" class=\"data row30 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row30_col7\" class=\"data row30 col7\" >0.007</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row31\" class=\"row_heading level0 row31\" >LogSegmentChange</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col0\" class=\"data row31 col0\" >-0.0005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col1\" class=\"data row31 col1\" >0.003</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col2\" class=\"data row31 col2\" >-0.0008</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col3\" class=\"data row31 col3\" >-0.002</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col4\" class=\"data row31 col4\" >0.007</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col5\" class=\"data row31 col5\" >0.008</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col6\" class=\"data row31 col6\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row31_col7\" class=\"data row31 col7\" >0.005</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row32\" class=\"row_heading level0 row32\" >Check1</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col0\" class=\"data row32 col0\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col1\" class=\"data row32 col1\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col2\" class=\"data row32 col2\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col3\" class=\"data row32 col3\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col4\" class=\"data row32 col4\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col5\" class=\"data row32 col5\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col6\" class=\"data row32 col6\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row32_col7\" class=\"data row32 col7\" >0.02</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row33\" class=\"row_heading level0 row33\" >Check2</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col0\" class=\"data row33 col0\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col1\" class=\"data row33 col1\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col2\" class=\"data row33 col2\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col3\" class=\"data row33 col3\" >0.6</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col4\" class=\"data row33 col4\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col5\" class=\"data row33 col5\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col6\" class=\"data row33 col6\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row33_col7\" class=\"data row33 col7\" >0.09</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row34\" class=\"row_heading level0 row34\" >Check3</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col0\" class=\"data row34 col0\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col1\" class=\"data row34 col1\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col2\" class=\"data row34 col2\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col3\" class=\"data row34 col3\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col4\" class=\"data row34 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col5\" class=\"data row34 col5\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col6\" class=\"data row34 col6\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row34_col7\" class=\"data row34 col7\" >0.06</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row35\" class=\"row_heading level0 row35\" >Check4</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col0\" class=\"data row35 col0\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col1\" class=\"data row35 col1\" >0.6</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col2\" class=\"data row35 col2\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col3\" class=\"data row35 col3\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col4\" class=\"data row35 col4\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col5\" class=\"data row35 col5\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col6\" class=\"data row35 col6\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row35_col7\" class=\"data row35 col7\" >0.1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row36\" class=\"row_heading level0 row36\" >Check5</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col0\" class=\"data row36 col0\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col1\" class=\"data row36 col1\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col2\" class=\"data row36 col2\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col3\" class=\"data row36 col3\" >0.3</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col4\" class=\"data row36 col4\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col5\" class=\"data row36 col5\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col6\" class=\"data row36 col6\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row36_col7\" class=\"data row36 col7\" >0.06</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row37\" class=\"row_heading level0 row37\" >Check6</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col0\" class=\"data row37 col0\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col1\" class=\"data row37 col1\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col2\" class=\"data row37 col2\" >0.05</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col3\" class=\"data row37 col3\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col4\" class=\"data row37 col4\" >0.08</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col5\" class=\"data row37 col5\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col6\" class=\"data row37 col6\" >0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row37_col7\" class=\"data row37 col7\" >0.07</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row38\" class=\"row_heading level0 row38\" >Check7</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col0\" class=\"data row38 col0\" >0.01</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col1\" class=\"data row38 col1\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col2\" class=\"data row38 col2\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col3\" class=\"data row38 col3\" >0.04</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col4\" class=\"data row38 col4\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col5\" class=\"data row38 col5\" >0.005</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col6\" class=\"data row38 col6\" >1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row38_col7\" class=\"data row38 col7\" >0.03</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row39\" class=\"row_heading level0 row39\" >Check8</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col0\" class=\"data row39 col0\" >0.02</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col1\" class=\"data row39 col1\" >0.09</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col2\" class=\"data row39 col2\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col3\" class=\"data row39 col3\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col4\" class=\"data row39 col4\" >0.06</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col5\" class=\"data row39 col5\" >0.07</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col6\" class=\"data row39 col6\" >0.03</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row39_col7\" class=\"data row39 col7\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02level0_row40\" class=\"row_heading level0 row40\" >Check</th> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col0\" class=\"data row40 col0\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col1\" class=\"data row40 col1\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col2\" class=\"data row40 col2\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col3\" class=\"data row40 col3\" >0.4</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col4\" class=\"data row40 col4\" >0.1</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col5\" class=\"data row40 col5\" >0.8</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col6\" class=\"data row40 col6\" >0.2</td> \n",
       "        <td id=\"T_822323f8_6c7c_11ea_b1f2_02420a051a02row40_col7\" class=\"data row40 col7\" >0.1</td> \n",
       "    </tr></tbody> \n",
       "</table> "
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f7877d20b00>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_corr[label_tags].style.background_gradient(cmap='coolwarm').set_precision(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Checking the balance of anomalies </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The fraction of anomalies among the data is equal to 0.3605580467518416.\n"
     ]
    }
   ],
   "source": [
    "nb_glob_anomalies = len(data[data[\"Check\"] == 1])\n",
    "nb_non_anomalies = len(data[data[\"Check\"] == 0])\n",
    "nb_data = len(data)\n",
    "\n",
    "fraction_anomalies = nb_glob_anomalies/nb_data\n",
    "print(\"The fraction of anomalies among the data is equal to \" + str(fraction_anomalies) + \".\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The fraction of anomalies from Check1 among the data is equal to 0.00649589410466969.\n",
      "The fraction of anomalies from Check2 among the data is equal to 0.033975597077058974.\n",
      "The fraction of anomalies from Check3 among the data is equal to 0.03136794147355786.\n",
      "The fraction of anomalies from Check4 among the data is equal to 0.0979413467674791.\n",
      "The fraction of anomalies from Check5 among the data is equal to 0.0119943705068657.\n",
      "The fraction of anomalies from Check6 among the data is equal to 0.27511400568866196.\n",
      "The fraction of anomalies from Check7 among the data is equal to 0.03009581126828423.\n",
      "The fraction of anomalies from Check8 among the data is equal to 0.009919234524176814.\n",
      "The fraction of anomalies from Check among the data is equal to 0.3605580467518416.\n"
     ]
    }
   ],
   "source": [
    "nb_1_anomalies = len(data[data[\"Check1\"] == 1])\n",
    "nb_2_anomalies = len(data[data[\"Check2\"] == 1])\n",
    "nb_3_anomalies = len(data[data[\"Check3\"] == 1])\n",
    "nb_4_anomalies = len(data[data[\"Check4\"] == 1])\n",
    "nb_5_anomalies = len(data[data[\"Check5\"] == 1])\n",
    "nb_6_anomalies = len(data[data[\"Check6\"] == 1])\n",
    "nb_7_anomalies = len(data[data[\"Check7\"] == 1])\n",
    "nb_8_anomalies = len(data[data[\"Check8\"] == 1])\n",
    "nb_all_anomalies = len(data[data[\"Check\"] == 1])\n",
    "\n",
    "fraction_anomalies_1 = nb_1_anomalies/nb_data\n",
    "fraction_anomalies_2 = nb_2_anomalies/nb_data\n",
    "fraction_anomalies_3 = nb_3_anomalies/nb_data\n",
    "fraction_anomalies_4 = nb_4_anomalies/nb_data\n",
    "fraction_anomalies_5 = nb_5_anomalies/nb_data\n",
    "fraction_anomalies_6 = nb_6_anomalies/nb_data\n",
    "fraction_anomalies_7 = nb_7_anomalies/nb_data\n",
    "fraction_anomalies_8 = nb_8_anomalies/nb_data\n",
    "fraction_anomalies_all = nb_all_anomalies/nb_data\n",
    "\n",
    "\n",
    "\n",
    "print(\"The fraction of anomalies from Check1 among the data is equal to \" + str(fraction_anomalies_1) + \".\")\n",
    "print(\"The fraction of anomalies from Check2 among the data is equal to \" + str(fraction_anomalies_2) + \".\")\n",
    "print(\"The fraction of anomalies from Check3 among the data is equal to \" + str(fraction_anomalies_3) + \".\")\n",
    "print(\"The fraction of anomalies from Check4 among the data is equal to \" + str(fraction_anomalies_4) + \".\")\n",
    "print(\"The fraction of anomalies from Check5 among the data is equal to \" + str(fraction_anomalies_5) + \".\")\n",
    "print(\"The fraction of anomalies from Check6 among the data is equal to \" + str(fraction_anomalies_6) + \".\")\n",
    "print(\"The fraction of anomalies from Check7 among the data is equal to \" + str(fraction_anomalies_7) + \".\")\n",
    "print(\"The fraction of anomalies from Check8 among the data is equal to \" + str(fraction_anomalies_8) + \".\")\n",
    "print(\"The fraction of anomalies from Check among the data is equal to \" + str(fraction_anomalies_all) + \".\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_fraction = [fraction_anomalies_1*100, fraction_anomalies_2*100, fraction_anomalies_3*100, fraction_anomalies_4*100, fraction_anomalies_5*100, fraction_anomalies_6*100, fraction_anomalies_7*100, fraction_anomalies_8*100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAHktJREFUeJzt3XuUXFWZ/vHvAyQQciE4hBjDJYAYDSoBAiKoE0AUCDfnJzqKDiIOsOSqjAqKA3gZYX4M4CCKERAcIywUUBDkIkI0omCCgRACg2IghJCEa8I9Ce/8cXZB0XZ173TnnNPd9XzWqtXnUufst6qSemvvc/beigjMzKx9rVV3AGZmVi8nAjOzNudEYGbW5pwIzMzanBOBmVmbcyIwM2tzTgTWb0k6WNKNXex/r6T7q4wplTte0mxJyyUdW3X5a4KkcZJC0jpp/VeSDqk7LiuH3I/AWpE0HxgNrAKeBa4Hjo6IZ2uIZRzwN2BQRKxs8ZwAto6Iv1QYWmdxXAgsi4jP1RlHb+S83zZwuEZg3dkvIoYBE4HtgJOqDqDxq7Qf2RyYW3cQZrmcCCxLRDwG3ECREACQtK6kMyU9LGmxpPMlDUn7Jkt6RNKXJT0uab6kg5uOnSLpz5KWSVog6dSmfY1micMkPQz8Bvht2v20pGclvVvSpyTNSMc09t+V9n+0EUPTed8m6VZJT0uaK2n/pn0XSzpP0rWpSed2SVu1ej8k7Z/O8XQ659vS9t8AuwHfSXG8pZNjD5U0L5XzoKQjmvY13rcTJC2RtEjSoU37N5D0I0lLJT0k6WRJa6V9n5L0e0lnp7gelLRL2r4gne+QpnO1/Aw6iflWSZ9pWv90eg1PSbpB0uZpu1L5S9J550h6e6vzWh8REX740ekDmA+8Py1vAswBvt20/2zgauANwHDgGuBbad9kYCVwFrAu8I/Ac8D4pv3voPgx8k5gMXBg2jcOCOBHwFBgSNO2dZrK/xQwo2k9gDc3rU8GHknLg4C/AF8GBgO7A8ub4rkYeALYCVgHmAZc1uJ9eUt6LXum834xnXtw2n8r8Jku3tcpwFaA0vvyPLB9h/fta+nc+6T9G6b9PwJ+kd7vccD/Aoc1vR8rgUOBtYFvAA8D56XP4APpNQ9bjc9gnY6vCTggvd63pffqZOC2tO+DwCxgZHp9bwPG1P1v2Y9u/q/XHYAfffdBkQieTV8eAdwMjEz7lL4Mt2p6/ruBv6Xlxhfa0Kb9lwNfbVHWOcDZabnxJbRl0/7XfTGlbZ8iPxG8F3gMWKtp/6XAqWn5YuCCpn37APe1iPWrwOVN62sBC4HJaf3VL83M9/nnwHFNMb/Q4XUuAXZOX+4vAxOa9h0B3Nr0fjzQtO8d6T0Z3bTtCWDianwGnSWCX5GST9Prf56iSWx3iuS0c/N77UfffrhpyLpzYEQMp/iCeiuwUdo+ClgfmJWaIZ6muJg8qunYpyLiuab1h4A3AUh6l6RbUhPHM8CRTeduWLAGX8ebgAUR8UqHeMY2rT/WtPw8MKyLcz3UWEnnXNDhXC1J2lvSHyU9md63fXj9a38iXn+BthHLRhS1hIea9nV8DYubll9I8XXcNizFkfMZdGZz4NtNn/uTFD8MxkbEb4DvUNRClkiaKmlExjmtRk4EliUiplP8aj4zbXqc4ktlm4gYmR4bRHFhuWFDSUOb1jcDHk3LP6FoVto0IjYAzqf4MnldsS2We+JRYNNGe3pTPAt7eK7NGyuSBGyacy5J6wJXULyPoyNiJHAdf//aO/M4sKK5bHr+GiDvM+jMAuCIps99ZEQMiYjbACLivyNiB2ACRTPaF3oYn1XEicBWxznAnpK2Tb+CfwCcLWljAEljJX2wwzGnSRos6b3AvsBP0/bhwJMR8aKknYCPd1P2UuAVYMsunrO4i/23U/yy/qKkQZImA/sBl3VTbmcuB6ZI2kPSIOAE4CXgtoxjB1O01y8FVkram6LtvlsRsSqV/U1Jw9MF2s8DP+7Ba4DV/wwazgdOkrQNvHoB+6C0vGOqaQyiaDp8keJzsz7MicCyRcRSiouV/542fYniouEfJS0Dfg2MbzrkMeApil/Q04AjI+K+tO+zwNckLU/nu7ybsp8Hvgn8PjVJ7NzJ004FLkn7P9Lh+Jcpvvj3pvhl/V3gX5riyRYR9wOfAM5N59qP4jbblzOOXQ4cS/F6n6L48r16NYo/huIL9kFgBsWv+otWJ/4mq/UZNETEVcAZwGXpc7+H4n0FGEHxA+EpimarJ4D/38P4rCLuUGalSL+4fxwRm9Qdi5l1zTUCM7M250RgZtbm3DRkZtbmXCMwM2tz/WIwr4022ijGjRtXdxhmZv3KrFmzHo+IUd09r18kgnHjxjFz5sy6wzAz61ckPdT9s9w0ZGbW9pwIzMzanBOBmVmbcyIwM2tzTgRmZm3OicDMrM05EZiZtTknAjOzNudEYGbW5vpFz2Iz671xJ15bW9nzT59SW9nWPdcIzMzanBOBmVmbcyIwM2tz3SYCSUMlrZWW3yJpf0mDyg/NzMyqkFMj+C2wnqSxwI3AJ4GLywzKzMyqk5MIFBHPA/8EfDciDgK2KTcsMzOrSlYikPRu4GCgcf/Z2uWFZGZmVcpJBMcDJwFXRcRcSVsCt5QblpmZVaXbDmURMR2YLmn9tP4gcGzZgZmZWTVy7hp6t6R7gfvS+raSvlt6ZGZmVomcpqFzgA8CTwBExF3A+8oMyszMqpPVoSwiFnTYtKqEWMzMrAY5g84tkLQLEKkj2XHAvHLDMjOzquTUCI4EjgLGAguBiWndzMwGgG4TQUQ8HhEHR8ToiNg4Ij4REU90d5ykTSXdIuleSXMlHZe2nyppoaTZ6bHPmnghZmbWMy2bhiR9MSL+U9K5QHTcHxHd3UK6EjghIu6UNByYJemmtO/siDizx1Gbmdka09U1gsZ1gJk9OXFELAIWpeXlkuZRNC+ZmVkf0jIRRMQ16e8lvS1E0jhgO+B2YFfgaEn/QpFkToiIp3pbhpmZ9UxXTUPX0EmTUENE7J9TgKRhwBXA8RGxTNL3gK+nc38d+C/g050cdzhwOMBmm22WU5SZmfVAV01DvW7DT7ebXgFMi4grASJicdP+HwC/7OzYiJgKTAWYNGlSy4RkZma901XT0PTenFiSgAuBeRFxVtP2Men6AcCHgHt6U46ZmfVOtx3KJG0NfAuYAKzX2B4RW3Zz6K4Uk9jMkTQ7bfsy8DFJEymahuYDR6x+2GZmtqbk9Cz+IXAKcDawG3Aoef0PZgDqZNd1qxOgmZmVK6dn8ZCIuJliprKHIuJUYEq5YZmZWVVyagQvpcnrH5B0NMUwE8PKDcvMzKqSUyM4DlifYjKaHSja/Q8pMygzM6tOzgxlf0qLz1JcHzAzswEk566hScBXgM2bnx8R7ywxLjMzq0jONYJpwBeAOcAr5YZjZmZVy0kESyPi6tIjMTOzWuQkglMkXQDcDLzU2NgYMsLMzPq3nERwKPBWYBCvNQ0F4ERgZjYA5CSCHSNifOmRmJlZLXL6EdwmaULpkZiZWS1yagQ7A7Ml/Y3iGoGA8O2jZmYDQ04i2Kv0KMzMrDY5o4g+BIwE9kuPkWmbmZkNAN0mAknHUXQq2zg9fizpmLIDMzOzauQ0DR0GvCsingOQdAbwB+DcMgMzM7Nq5Nw1JGBV0/oqOp9wxszM+qHcGcpul3RVWj+QYi5iMzMbAHKGoT5L0nSKOYgBDo2IP5cblpmZVSWnRgAwG1jUeL6kzSLi4dKiMjOzyuTMR3AMxeT1i3nt+kAA7lBmZjYA5NQIjgPGR8QTZQdjZmbVy7lraAHwTNmBmJlZPXJqBA8Ct0q6ltfPR3BWaVGZmVllchLBw+kxOD3MzGwAybl99LQqAjEzs3rk3DU0CvgisA2wXmN7ROxeYlxmZlaRnIvF04D7gC2A04D5wJ9KjMnMzCqUkwj+ISIuBFZExPSI+DTg2oCZ2QCRkwhWpL+LJE2RtB3whu4OkrSppFsk3StpbhrOGklvkHSTpAfS3w17Eb+ZmfVSTiL4hqQNgBOAfwMuAD6XcdxK4ISImEAx3eVRae7jE4GbI2Jr4Oa0bmZmNcm5a+iXafEZYLfcE0fEIorxiYiI5ZLmAWOBA4DJ6WmXALcCX8qO2MzM1qicGkGvSRoHbAfcDoxOSQLgMWB0i2MOlzRT0sylS5dWEaaZWVsqPRFIGgZcARwfEcua90VEUAxg93ciYmpETIqISaNGjSo7TDOztlVqIpA0iCIJTIuIK9PmxZLGpP1jgCVlxmBmZl3L6VD2+U42PwPMiojZXRwnipnM5nUYl+hq4BDg9PT3F6sVsZmZrVE5Yw1NSo9r0vq+wN3AkZJ+GhH/2eK4XYFPAnMkNRLGlykSwOWSDgMeAj7S0+DNzKz3chLBJsD2EfEsgKRTgGuB9wGzgE4TQUTMoPUk93usfqhmZlaGnGsEG9M0/DRFB7PREfFCh+1mZtYP5dQIpgG3S2q05e8H/ETSUODe0iIzM7NK5HQo+7qk64Fd0qYjI2JmWj64tMjMzKwSOTUCgDuBhY3nS9osIh4uLSozM6tMzu2jxwCnAIuBVRQXgAN4Z7mhmZlZFXJqBMcB4yPiibKDMTOz6uXcNbSAogOZmZkNQDk1ggeBWyVdS9Ptoh16C5uZWT+VkwgeTo/B6WFmZgNIzu2jp1URiJmZ1aNlIpB0TkQcL+kaOhkqOiL2LzUyMzOrRFc1gv9Jf8+sIhAzM6tHy0QQEbPS3+mShgCbRcT9lUVmZmaV6Pb2UUn7AbOB69P6RElXlx2YmZlVI6cfwanATsDTAGkymi1KjMnMzCqUkwhWRETHDmWdzjNsZmb9T04/grmSPg6sLWlr4FjgtnLDMjOzquTUCI4BtqHoVXwpsAw4vsygzMysOjkdyp4HviLpjGI1lpcflpmZVSXnrqEdJc2hmLB+jqS7JO1QfmhmZlaFnGsEFwKfjYjfAUh6D/BDPB+BmdmAkHONYFUjCQBExAxgZXkhmZlZlboaa2j7tDhd0vcpLhQH8FHg1vJDMzOzKnTVNPRfHdZPaVp2PwIzswGiq7GGdqsyEDMzq0fONQIzMxvAnAjMzNqcE4GZWZvL6VB2kKThaflkSVc23VFkZmb9XE6N4KsRsTx1JHs/RQez73V3kKSLJC2RdE/TtlMlLZQ0Oz326XnoZma2JmR1KEt/pwBTI+JaYHDGcRcDe3Wy/eyImJge1+WFaWZmZclJBAtTh7KPAtdJWjfnuIj4LfBkL+MzM7OS5SSCjwA3AB+MiKeBNwBf6EWZR0u6OzUdbdjqSZIOlzRT0sylS5f2ojgzM+tKzi/754ElwHvSppXAAz0s73vAVsBEYBF/33u5udypETEpIiaNGjWqh8WZmVl3cu4aOgX4EnBS2jQI+HFPCouIxRGxKiJeAX5AMReymZnVKKdp6EPA/sBzABHxKDC8J4VJGtPhvPe0eq6ZmVUjZz6ClyMiJAWApKE5J5Z0KTAZ2EjSIxSD1k2WNJFi0Lr5wBE9CdrMzNacnERwebpraKSkfwU+TdGs06WI+Fgnmy9czfjMzKxkOXMWnylpT4pJ68cD/x4RN5UemZmZVSKnRkD64veXv5nZANRtIpC0nNcmohlMcdfQcxExoszAzMysGjlNQ6/eISRJwAHAzmUGZWZm1VmtYaij8HPggyXFY2ZmFctpGvqnptW1gEnAi6VFZGZmlcq5WLxf0/JKivv/DyglGjMzq1zONYJDqwjEzMzq0TIRSDqX1+4W+jsRcWwpEZmZWaW6qhHMrCwKMzOrTctEEBGXVBmImZnVI+euoVEUw1BPANZrbI+I3UuMy8zMKpLTj2AaMA/YAjiN4q6hP5UYk5mZVSgnEfxDRFwIrIiI6RHxacC1ATOzASKnH8GK9HeRpCnAoxTzFpuZ2QCQkwi+IWkD4ATgXGAE8LlSozIzs8rkJILbI+IZ4Blgt5LjMTOziuVcI/i9pBslHSZpw9IjMjOzSnWbCCLiLcDJwDbALEm/lPSJ0iMzM7NKZA1DHRF3RMTngZ2AJwF3NjMzGyC6TQSSRkg6RNKvgNuARRQJwczMBoCci8V3AT8HvhYRfyg5HjMzq1hOItgyIlqOQmpmZv1bzsViJwEzswFsteYsNjOzgadlIpB0Rvp7UHXhmJlZ1bqqEewjScBJVQVjZmbV6+pi8fXAU8AwScsAUUxdKYpLByMqiM/MzErWskYQEV+IiJHAtRExIiKGN//t7sSSLpK0RNI9TdveIOkmSQ+kvx6ywsysZjl3DR0gabSkfdNjVOa5Lwb26rDtRODmiNgauDmtm5lZjXJ6Fh8E3AEcBHwEuEPSh7s7LiJ+SzEcRbMDeG14ikuAA1crWjMzW+NyOpSdDOwYEUvg1TmMfw38rAfljY6IRWn5MWB0D85hZmZrUE4/grUaSSB5IvO4LqWOai07q0k6XNJMSTOXLl3a2+LMzKyFnBrB9ZJuAC5N6x8FrutheYsljYmIRZLGAEtaPTEipgJTASZNmuTezfaqcSdeW1vZ80+fUlvZZmXJuVj8BeD7wDvTY2pEfKmH5V0NHJKWDwF+0cPzmJnZGpJTIyAirgSuXJ0TS7oUmAxsJOkR4BTgdOBySYcBD1FcfDYzsxplJYKeiIiPtdi1R1llmpnZ6vOgc2ZmbS4rEUgaIml82cGYmVn1cjqU7QfMphh7CEkTJV1ddmBmZlaNnBrBqRRzFD8NEBGzgS1KjMnMzCqUkwhWRMQzHbb5vn4zswEi566huZI+DqwtaWvgWOC2csMyM7Oq5NQIjgG2AV6i6F28DDi+zKDMzKw63dYIIuJ54CvpYWZmA0y3iUDSNfz9NYFngJnA9yPixTICMzOzauQ0DT0IPAv8ID2WAcuBt6R1MzPrx3IuFu8SETs2rV8j6U8RsaOkuWUFZmZm1cipEQyTtFljJS0PS6svlxKVmZlVJqdGcAIwQ9JfAVF0JvuspKG8Nu2kmZn1Uzl3DV2X+g+8NW26v+kC8TmlRWZmZpXIHYZ6a2A8sB6wrSQi4kflhWVmZlXJuX30FIoJZiZQTFG5NzADcCIwMxsAci4Wf5hiMpnHIuJQYFtgg1KjMjOzyuQkghci4hVgpaQRFBPOb1puWGZmVpWcawQzJY2k6Dw2i6Jz2R9KjcrMzCqTc9fQZ9Pi+ZKuB0ZExN3lhmVmZlXJmaHs5sZyRMyPiLubt5mZWf/WskYgaT1gfWAjSRtSdCYDGAGMrSA2MzOrQFdNQ0dQzDvwJoprA41EsAz4TslxmZlZRVomgoj4NvBtScdExLkVxmRmZhXKuVh8rqRdgHHNz3fPYjOzgSGnZ/H/AFsBs4FVaXPgnsVmZgNCTj+CScCEiOg4S5mZmQ0AOT2L7wHeWHYgZmZWj5wawUbAvZLuAF5qbIyI/UuLyszMKpOTCE5d04VKmk8x7/EqYGVETFrTZZiZWZ6cu4amS9oc2Doifi1pfWDtNVD2bhHx+Bo4j5mZ9ULOEBP/CvwM+H7aNBb4eZlBmZlZdXIuFh8F7ErRo5iIeADYuJflBnCjpFmSDu/sCZIOlzRT0sylS5f2sjgzM2slJxG8FBEvN1YkrUPxRd4b74mI7SlmOztK0vs6PiEipkbEpIiYNGrUqF4WZ2ZmreQkgumSvgwMkbQn8FPgmt4UGhEL098lwFXATr05n5mZ9VxOIjgRWArMoRiI7jrg5J4WKGmopOGNZeADFH0VzMysBjm3jw4BLoqIHwBIWjtte76HZY4GrpLUKP8nEXF9D89lZma9lJMIbgbeTzFFJRRJ4EZgl54UGBEPAtv25FgzM1vzcpqG1ouIRhIgLa9fXkhmZlalnETwnKTtGyuSdgBeKC8kMzOrUk7T0HHATyU9SjFL2RuBj5YalZmZVabLRCBpLWAw8FZgfNp8f0SsKDswMzOrRpeJICJekXReRGyHb/E0MxuQcq4R3Czp/ynd72lmZgNLzjWCI4DPA6skvUBxnSAiYkSpkbWBcSdeW1vZ80+fUlvZZta35AxDPbyKQMzMrB45k9cLOBjYIiK+LmlTYExE3FF6dFYb11bM2kdO09B3gVeA3YGvU/QwPg/YscS4zKyN1PnDA/zjIycRvCsitpf0Z4CIeErS4JLjMjOziuTcNbQiDTQXAJJGUdQQzMxsAMhJBP9NMWfAxpK+CcwA/qPUqMzMrDI5dw1NkzQL2IPi1tEDI2Je6ZGZmVklWiYCSesBRwJvppiU5vsRsbKqwMzMrBpdNQ1dAkyiSAJ7A2dWEpGZmVWqq6ahCRHxDgBJFwLuN2BmNgB1VSN4dYRRNwmZmQ1cXdUItpW0LC0LGJLWPdaQmdkA0jIRRMTaVQZiZmb1yOlHYGZmA5gTgZlZm3MiMDNrczmDzplZJg/fbf2RE4GZWRfaYYhsNw2ZmbU5JwIzszY34JuG3GZrZtY11wjMzNpcLYlA0l6S7pf0F0kn1hGDmZkVKk8EadrL8yiGtp4AfEzShKrjMDOzQh01gp2Av0TEgxHxMnAZcEANcZiZGaCIqLZA6cPAXhHxmbT+SeBdEXF0h+cdDhyeVscD91ca6Gs2Ah6vqezuOLaecWw949h6ps7YNo+IUd09qc/eNRQRU4GpdcchaWZETKo7js44tp5xbD3j2HqmL8fWUEfT0EJg06b1TdI2MzOrQR2J4E/A1pK2kDQY+Gfg6hriMDMzamgaioiVko4GbgDWBi6KiLlVx7Eaam+e6oJj6xnH1jOOrWf6cmxADReLzcysb3HPYjOzNudEYGbW5pwIWpB0kaQlku6pO5ZmkjaVdIukeyXNlXRc3TE1SFpP0h2S7kqxnVZ3TB1JWlvSnyX9su5YmkmaL2mOpNmSZtYdTzNJIyX9TNJ9kuZJenfdMTVIGp/es8ZjmaTj644LQNLn0v+DeyRdKmm9umNqxdcIWpD0PuBZ4EcR8fa642mQNAYYExF3ShoOzAIOjIh7aw4NSQKGRsSzkgYBM4DjIuKPNYf2KkmfByYBIyJi37rjaZA0H5gUEX2uU5SkS4DfRcQF6U6/9SPi6brj6igNX7OQooPqQzXHMpbi3/+EiHhB0uXAdRFxcZ1xteIaQQsR8Vvgybrj6CgiFkXEnWl5OTAPGFtvVIUoPJtWB6VHn/mlIWkTYApwQd2x9BeSNgDeB1wIEBEv98UkkOwB/LXuJNBkHWCIpHWA9YFHa46nJSeCfkzSOGA74PZ6I3lNanqZDSwBboqIPhMbcA7wReCVugPpRAA3SpqVhlfpK7YAlgI/TE1qF0gaWndQLfwzcGndQQBExELgTOBhYBHwTETcWG9UrTkR9FOShgFXAMdHxLK642mIiFURMZGix/hOkvpEs5qkfYElETGr7lhaeE9EbE8xKu9RqWmyL1gH2B74XkRsBzwH9Lmh41OT1f7AT+uOBUDShhSDaW4BvAkYKukT9UbVmhNBP5Ta368ApkXElXXH05nUfHALsFfdsSS7AvuntvjLgN0l/bjekF6TfkESEUuAqyhG6e0LHgEeaarZ/YwiMfQ1ewN3RsTiugNJ3g/8LSKWRsQK4Epgl5pjasmJoJ9JF2QvBOZFxFl1x9NM0ihJI9PyEGBP4L56oypExEkRsUlEjKNoQvhNRPSJX2iShqYL/6Rmlw8AfeJutYh4DFggaXzatAdQ+40JnfgYfaRZKHkY2FnS+un/7B4U1/P6JCeCFiRdCvwBGC/pEUmH1R1TsivwSYpftI1b5vapO6hkDHCLpLspxpS6KSL61G2afdRoYIaku4A7gGsj4vqaY2p2DDAtfa4Tgf+oOZ7XSclzT4pf3X1CqkH9DLgTmEPxXdtnh5rw7aNmZm3ONQIzszbnRGBm1uacCMzM2pwTgZlZm3MiMDNrc04E1tYkvVHSZZL+moZ3uE7S4WtidFJJk/vaKKdmnXEisLaVOvpcBdwaEVtFxA7ASRT39Zu1DScCa2e7ASsi4vzGhoi4C/gdMKxpDP5pKWkgaQdJ01Pt4YY0LDiS3izp12kuhjslbdVckKQd06BtW0n6x6bOgH9u9Co2q4sTgbWzt1PM59CZ7YDjgQnAlsCuaYync4EPp9rDRcA30/OnAedFxLYUY8osapxI0i7A+cABEfFX4N+Ao9LgfO8FXljTL8xsdaxTdwBmfdQdEfEIQBpWexzwNEXyuClVENYGFqVf9GMj4iqAiHgxHQfwNoqhBT4QEY3x6H8PnCVpGnBloxyzurhGYO1sLrBDi30vNS2vovjRJGBuRExMj3dExAe6KWMR8CJFDQOAiDgd+AwwBPi9pLf29AWYrQlOBNbOfgOs2zwRjKR3UjTXdOZ+YFRjzl5JgyRtk2aKe0TSgWn7upLWT8c8TTEr2rckTU77t4qIORFxBsXgfE4EVisnAmtbUYy4+CHg/en20bnAt4DHWjz/ZeDDwBlppNDZvDbG/CeBY9MInbcBb2w6bjGwL3CepHcBx6cJze8GVgC/KuUFmmXy6KNmZm3ONQIzszbnRGBm1uacCMzM2pwTgZlZm3MiMDNrc04EZmZtzonAzKzN/R/Fo5ymIuakDQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar( np.arange(1,9),anom_fraction, capsize=1, align='center', orientation='vertical')\n",
    "plt.ylabel(\"Percentage of values being anomalies\")\n",
    "plt.title(\"Repartition of anomalies\")\n",
    "plt.xticks(np.arange(1,9))\n",
    "plt.xlabel(\"Checks\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> \n",
    "</p>\n",
    "<p>\n",
    "</p>\n",
    "<p>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"\">\n",
    "    <h1>Model Selection</h1>\n",
    "    <hr style=\"height:1px;border:none;color:#333;background-color:#333;\"/>    \n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> first of all, divide the dataset into features and labels.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = data[label_tags]\n",
    "X = data.drop(label_tags,axis=1)\n",
    "X.drop(['SessionNumber','SystemID','Check'],axis=1,inplace=True) ## dropping them since they're not useful features\n",
    "X_ind = X.index\n",
    "X_cols = X.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Then we normalize the features </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/mnt/workspace/.local/lib/python3.5/site-packages/sklearn/preprocessing/data.py:334: DataConversionWarning: Data with input dtype float32 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n"
     ]
    }
   ],
   "source": [
    "scaler = preprocess.MinMaxScaler()\n",
    "X = scaler.fit_transform(X)\n",
    "X = pd.DataFrame(X,columns=X_cols,index=X_ind)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b><h5>Note</h5></b>\n",
    "<p> Since, this is a multi-class multi-label problem. We figured that, it would be better to separate the labels are train independent model for each set of label (i.e. When using a particular algorithm, we train 1 model for <i>Check1</i>, 1 model for <i>Check2</i> and so on.... </p><p> This way, we will calculate independent f1 scores and confusion matrices for each label </p> <p> This reasoning is followed for all the algorithms applied to this problem </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Interpretable Model </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>We tried using the models recommended by the problem. However, they were not well documented and they seemed to not work. Therefore, we turned towards using other types of interpretable models.</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The two interpretable models that we're going to use will be <b> Decision Tree </b> and <b> Random Forest</b>.<p> Random Forest is a combination of multiple Decision trees. </p><p>As an example we're going to show how to interpret the results of a very simple Random forest (max depth = 5) on our dataset</p> </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Applying a very simple Random tree on our dataset and trying to interpret the results. We use 'Check1' as our target label and test size equal to 20% of the dataset.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test = model_selection.train_test_split(X,y['Check1'],test_size=0.2,stratify =y['Check1'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.99930919, 0.88245931])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## using a small number of estimators and a shallow depth because this is only an example to explain\n",
    "## the interpretability of the model. The real version with better results is applied below\n",
    "m = ensem.RandomForestClassifier(n_estimators=10,n_jobs=-1,max_depth=5)\n",
    "m.fit(X_train, y_train)\n",
    "pred = m.predict(X_test)\n",
    "met.f1_score(y_test,pred,average=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Generating dot files for each estimator inside the Forest\n",
    "# estimators = m.estimators_\n",
    "# for i,est in enumerate(estimators):\n",
    "#     graph = Source(trees.export_graphviz(decision_tree=est,out_file='tree%d.dot'%(i),feature_names=X_cols))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The above code generated <i>.dot</i> files for the estimators of random forest. we converted the <i>.dot</i> files for the trees into <i>.png</i> files using online resources </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"first_tree_complete.png\" width = 1000>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The Random Forest creates a Tree representing a set of rules based on the features. At every node there is a rule (for instance <i> \"CPU lower than 0.02\" </i>). On the leaves, we find the last conditions, plus the tuple <i>value</i> containing the number of non anomalies, and number of anomalies detected for the criterias of the branch. </p>\n",
    "<p> This technique is interesting because it is interpretable in the sense that we can understand the choices made by the computer. We can also try to understand why some criterias combinations lead to a certain percentage of anomalies.</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"first_tree_left_zoom.png\" width = 700>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"first_tree_right_zoom.png\" width = 700>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> On the leaves, we can notice some different paterns, and the colored ones are the most interesting for us.</p>\n",
    "<p> In red are the leaves indacting a path that gives only anomalies. Some of them show a very big number of anomalies, like <i>value = [0, 113]</i> or <i>value = [0, 384]</i>. Those paths are thus very interesting. For others, like <i>value = [0, 4]</i>, it is hard to know how relatable they are. Indeed it indicated a 100% anomalies, but out of 4 examples only.</p>\n",
    "<p>In orange are the leaves indicating paths that give mostly anomalies. Here, the ratio should be examined, like <i>75 out of 76</i> anomalies, and <i>200 out of 202</i> anomalies. But again, not only the ratio but also the general number of examples should be considered.</p>\n",
    "\n",
    "<p> At the opposite, in blue are the leaves indacting paths that never lead to anomalies. Again, some values like <i>value = [16, 0]</i> can be discussed, but others like <i>value = [9130, 0]</i> show a very accurate set of rules to discriminate anomalies.</p> \n",
    "<p> Plus, the confrontation of two branches of a same node also gives us information, for instance on the path <i>IndexServerAllocationLimit smaller than 0.911, HighPriorityAlerts greater than 0.271, BlockingPhaseSec smaller than 0, PhysMEM smaller than 0.902</i>, on the node <i>CPU lower than 0.02 ?</i>, if the node condition is true we get only non-anomalies, while if it is false we get <i>200 out of 202</i> anomalies  </p>\n",
    "<p>Finally, in green are the leaves indicating paths that give mostly non-anomalies. Again, the ratio and the amount should be examined. Thus, <i>1430 out of 1486</i> non-anomalies, is more interesting than <i>55 out of 65</i> non-anomalies.</p>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"second_tree_zoom.png\" width = 1000>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> By generating several trees, we can find paths leading to better results. Plus, we can also combine and compare paths. For instance, we realise that <i>CPU smaller than 0.2</i> is a rule leading to a 100% non-anomalies set of values. </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Decision Tree</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "## The parameters are self explanatory.\n",
    "## The function returns the model, f1 score and confusion matrix (both calculated on the test set)\n",
    "def DecTreeClass(X,y,to_check='Check1',test_size=0.2,max_depth=None):\n",
    "    #Splitting the dataset\n",
    "    X_train,X_test,y_train,y_test = model_selection.train_test_split(X,y[to_check],test_size=test_size,stratify =y[to_check])\n",
    "    \n",
    "    ## Using all the default parameters except max_depth which could be changed\n",
    "    m = trees.DecisionTreeClassifier(max_depth=max_depth)\n",
    "    ## Fitting on training data\n",
    "    m.fit(X_train,y_train) \n",
    "    \n",
    "    ## Predicting the results of test set\n",
    "    pred = m.predict(X_test)\n",
    "    \n",
    "    ## Calculating metrics\n",
    "    ## Note that the average is set to None, which calculates two f1 scores/label\n",
    "    ## one f1 score for the normal data and one f1 score for the anomalous data\n",
    "    f1 = met.f1_score(y_test,pred,average=None)\n",
    "    conf = met.confusion_matrix(y_pred=pred,y_true=y_test)\n",
    "    \n",
    "    return m,f1,conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "## A dataframe keeping the f1 scores for all the checks\n",
    "DT_f1 = pd.DataFrame(index=y.columns,columns=['Normal','Anomalous']) ### Dataframe which holds the f1_scores for each check using Decision Tree Classifier\n",
    "## A dictionary keeping all the confusion matrices for all the checks\n",
    "DT_conf = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "for check in y.columns: ## iterating on all checks\n",
    "    m,f1,conf = DecTreeClass(X,y,to_check=check,max_depth=5)\n",
    "    DT_f1.loc[check,:] = f1\n",
    "    DT_conf[check] = conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Normal</th>\n",
       "      <th>Anomalous</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Check1</th>\n",
       "      <td>0.999915</td>\n",
       "      <td>0.987138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check2</th>\n",
       "      <td>0.999584</td>\n",
       "      <td>0.988286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check3</th>\n",
       "      <td>0.999225</td>\n",
       "      <td>0.976262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check4</th>\n",
       "      <td>0.999074</td>\n",
       "      <td>0.991523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check5</th>\n",
       "      <td>0.99954</td>\n",
       "      <td>0.962048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check6</th>\n",
       "      <td>0.995846</td>\n",
       "      <td>0.989052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check7</th>\n",
       "      <td>0.989089</td>\n",
       "      <td>0.535237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check8</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Normal Anomalous\n",
       "Check1  0.999915  0.987138\n",
       "Check2  0.999584  0.988286\n",
       "Check3  0.999225  0.976262\n",
       "Check4  0.999074  0.991523\n",
       "Check5   0.99954  0.962048\n",
       "Check6  0.995846  0.989052\n",
       "Check7  0.989089  0.535237\n",
       "Check8         1         1"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DT_f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tCheck1\n",
      "[[47008     8]\n",
      " [    0   307]]\n",
      "\tCheck2\n",
      "[[45682    33]\n",
      " [    5  1603]]\n",
      "\tCheck3\n",
      "[[45792    47]\n",
      " [   24  1460]]\n",
      "\tCheck4\n",
      "[[42624    64]\n",
      " [   15  4620]]\n",
      "\tCheck5\n",
      "[[46735    20]\n",
      " [   23   545]]\n",
      "\tCheck6\n",
      "[[34165   139]\n",
      " [  146 12873]]\n",
      "\tCheck7\n",
      "[[45733   166]\n",
      " [  843   581]]\n",
      "\tCheck8\n",
      "[[46854     0]\n",
      " [    0   469]]\n"
     ]
    }
   ],
   "source": [
    "for keys in y.columns:\n",
    "    print(\"\\t\" + keys)\n",
    "    print(DT_conf[keys])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> Random Forest </h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> For Random Forest we applied a different approach. Looking at the correlation table (in the data exploration part) we realize that some checks are dependent on other checks.</p>\n",
    "<p> i.e. Check2 and Check4 have an intercorrelation of 0.6 . Therefore, when we have to find a model to calculate Check2, we could first make a model for Check4 and then calculate Check2 using the results of the previous model</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Parameters X,y,label and test_size are self explanatory\n",
    "## other_labels -> specify if we need to calculate and keep other anomalies as features before calculating 'label'\n",
    "## n_estimators -> number of estimators for RandomForest. Default is 100\n",
    "## n_jobs -> number of parallel estimator computations. if -1, it means to use the maximum number possible\n",
    "## max_depth -> maximum depth of each estimator, None means it should be the maximum possible\n",
    "## verbose -> If the model should print the results\n",
    "def getRandomForestModel(X,y,label='Check1',other_labels=[],n_estimators=100,n_jobs=-1,max_depth=None,test_size=0.2,verbose=False):\n",
    "    \n",
    "    # Splitting the dataset\n",
    "    X_train,X_test,y_train,y_test = model_selection.train_test_split(X,y[label],test_size=test_size,stratify=y[label])\n",
    "    \n",
    "    ## we would need these if we need to calculate 'other_labels'\n",
    "    old_X_train = X_train\n",
    "    old_X_test = X_test\n",
    "    \n",
    "    for other in other_labels:\n",
    "        \n",
    "        ## Recursively calling the function itself to calculate each label present inside 'other_labels'\n",
    "        model = getRandomForestModel(X,y,label=other,n_estimators=n_estimators,n_jobs=n_jobs,max_depth=max_depth)[0]\n",
    "        \n",
    "        ## the number of rows in the original training set\n",
    "        row = old_X_train.shape[0]\n",
    "        ## predicting using the initial training set i.e. old_X_train\n",
    "        ## The reason for this is that at each iteration we change X_train by appending 'other_labels' as features\n",
    "        pred=model.predict(old_X_train)\n",
    "        ## Appending the 'other_labels' in the original training set\n",
    "        X_train = np.append(X_train,pred.reshape((row,1)),axis=1)\n",
    "        \n",
    "        ## doing the same thing for the test set\n",
    "        row = old_X_test.shape[0]\n",
    "        pred=model.predict(old_X_test)\n",
    "        X_test = np.append(X_test,pred.reshape((row,1)),axis=1)\n",
    "        \n",
    "    ## we don't need these anymore, they just occupy space\n",
    "    del old_X_train,old_X_test\n",
    "    \n",
    "    ## making the model\n",
    "    m = ensem.RandomForestClassifier(n_estimators=n_estimators,n_jobs=n_jobs,max_depth=max_depth)\n",
    "    m.fit(X_train, y_train)\n",
    "    pred  = m.predict(X_test)\n",
    "    \n",
    "    ## Calculating metrics\n",
    "    ## Note that the average is set to None, which calculates two f1 scores/label\n",
    "    ## one f1 score for the normal data and one f1 score for the anomalous data\n",
    "    score = met.f1_score(y_test,pred,average=None)\n",
    "    conf = met.confusion_matrix(y_pred=pred,y_true=y_test)\n",
    "    \n",
    "    if(verbose):\n",
    "        \n",
    "        print(\"Label %s: Validation size %.2f Estimators %d Jobs %d Max_Depth %s\\n\"%(label,test_size,n_estimators,n_jobs,str(max_depth)))\n",
    "        print(\"Score:\" + str(score))\n",
    "        print(\"Confusion Matrix \\n\" + str(conf))\n",
    "    \n",
    "    return m,score,conf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> First, we try the models using each check independently (without taking into account the correlations between them </p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The dataframe keeping f1 scores for all labels\n",
    "RF_f1_indep = pd.DataFrame(index=y.columns,columns=['Normal','Anomalous'])\n",
    "## confusion matrices for all labels\n",
    "RF_conf = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "for check in y.columns:\n",
    "    m,f1,conf = getRandomForestModel(X,y,label=check)\n",
    "    RF_f1_indep.loc[check,:] = f1\n",
    "    RF_conf[check] = conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Normal</th>\n",
       "      <th>Anomalous</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Check1</th>\n",
       "      <td>0.999957</td>\n",
       "      <td>0.993528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check2</th>\n",
       "      <td>0.999748</td>\n",
       "      <td>0.992881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check3</th>\n",
       "      <td>0.999629</td>\n",
       "      <td>0.988636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check4</th>\n",
       "      <td>0.999367</td>\n",
       "      <td>0.994204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check5</th>\n",
       "      <td>0.999775</td>\n",
       "      <td>0.981595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check6</th>\n",
       "      <td>0.998658</td>\n",
       "      <td>0.996476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check7</th>\n",
       "      <td>0.993015</td>\n",
       "      <td>0.736928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check8</th>\n",
       "      <td>0.999851</td>\n",
       "      <td>0.984848</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Normal Anomalous\n",
       "Check1  0.999957  0.993528\n",
       "Check2  0.999748  0.992881\n",
       "Check3  0.999629  0.988636\n",
       "Check4  0.999367  0.994204\n",
       "Check5  0.999775  0.981595\n",
       "Check6  0.998658  0.996476\n",
       "Check7  0.993015  0.736928\n",
       "Check8  0.999851  0.984848"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RF_f1_indep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tCheck1\n",
      "[[47012     4]\n",
      " [    0   307]]\n",
      "\tCheck2\n",
      "[[45696    19]\n",
      " [    4  1604]]\n",
      "\tCheck3\n",
      "[[45810    29]\n",
      " [    5  1479]]\n",
      "\tCheck4\n",
      "[[42638    50]\n",
      " [    4  4631]]\n",
      "\tCheck5\n",
      "[[46742    13]\n",
      " [    8   560]]\n",
      "\tCheck6\n",
      "[[34224    80]\n",
      " [   12 13007]]\n",
      "\tCheck7\n",
      "[[45777   122]\n",
      " [  522   902]]\n",
      "\tCheck8\n",
      "[[46854     0]\n",
      " [   14   455]]\n"
     ]
    }
   ],
   "source": [
    "for keys in y.columns:\n",
    "    print(\"\\t\" + keys)\n",
    "    print(RF_conf[keys])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Now we try to use the interdependencies of the Checks </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> To establish dependencies, we used 0.1 correlation as a threshold between checks</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "dependencies = {'Check1':[],\n",
    "               'Check2':['Check4','Check5','Check6'],\n",
    "               'Check3':['Check4'],\n",
    "               'Check4':['Check2','Check3','Check5','Check6','Check8'],\n",
    "               'Check5':['Check2','Check4'],\n",
    "               'Check6':['Check2','Check4'],\n",
    "               'Check7':[],\n",
    "               'Check8':['Check4']}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "RF_f1_dep = pd.DataFrame(index=y.columns,columns=['Normal','Anomalous'])\n",
    "RF_conf_dep = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Making new models but this time using other labels as their features\n",
    "for check in y.columns:\n",
    "    m,f1,conf = getRandomForestModel(X,y,label=check,other_labels=dependencies[check])\n",
    "    RF_f1_dep.loc[check,:] = f1\n",
    "    RF_conf_dep[check] = conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Normal</th>\n",
       "      <th>Anomalous</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Check1</th>\n",
       "      <td>0.999957</td>\n",
       "      <td>0.993528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check2</th>\n",
       "      <td>0.999748</td>\n",
       "      <td>0.992881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check3</th>\n",
       "      <td>0.999629</td>\n",
       "      <td>0.988636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check4</th>\n",
       "      <td>0.999367</td>\n",
       "      <td>0.994204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check5</th>\n",
       "      <td>0.999775</td>\n",
       "      <td>0.981595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check6</th>\n",
       "      <td>0.998658</td>\n",
       "      <td>0.996476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check7</th>\n",
       "      <td>0.993015</td>\n",
       "      <td>0.736928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check8</th>\n",
       "      <td>0.999851</td>\n",
       "      <td>0.984848</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Normal Anomalous\n",
       "Check1  0.999957  0.993528\n",
       "Check2  0.999748  0.992881\n",
       "Check3  0.999629  0.988636\n",
       "Check4  0.999367  0.994204\n",
       "Check5  0.999775  0.981595\n",
       "Check6  0.998658  0.996476\n",
       "Check7  0.993015  0.736928\n",
       "Check8  0.999851  0.984848"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RF_f1_indep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tCheck1\n",
      "[[47011     5]\n",
      " [    0   307]]\n",
      "\tCheck2\n",
      "[[45705    10]\n",
      " [    3  1605]]\n",
      "\tCheck3\n",
      "[[45811    28]\n",
      " [   10  1474]]\n",
      "\tCheck4\n",
      "[[42645    43]\n",
      " [    5  4630]]\n",
      "\tCheck5\n",
      "[[46751     4]\n",
      " [   11   557]]\n",
      "\tCheck6\n",
      "[[34249    55]\n",
      " [   12 13007]]\n",
      "\tCheck7\n",
      "[[45778   121]\n",
      " [  509   915]]\n",
      "\tCheck8\n",
      "[[46854     0]\n",
      " [   13   456]]\n"
     ]
    }
   ],
   "source": [
    "for keys in y.columns:\n",
    "    print(\"\\t\" + keys)\n",
    "    print(RF_conf_dep[keys])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> Using dependencies <b>was not</b> very useful since almost all the classes gave very good results (except Check7). Whereas, Check7 doesn't have any dependency on any other check.</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Non interpretable models</h1>\n",
    "<p> After trying interpretable models, we thought we could also try some non-interpretable models to see how they do </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>PCA and MultiVariate Gaussian</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p> The idea here is to first perform the PCA on the input data and reduce its dimensions. After reducing the dimensions, we could make a multivariate gaussian of all the normal points.</p>\n",
    "<p> The gist is that the normal points will have a higher pdf for this gaussian function, whereas the anomalous points will have a lower pdf</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy.stats as stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Params: X -> input dataset\n",
    "## ratio -> The minimum explained variance ratio. \n",
    "##          the sum of variance explained by all the Principal components should be greater or equal to this\n",
    "def getPCAModel(X,ratio=0.95):\n",
    "     ## iterate over the number of components to keep.\n",
    "    ## Go from 1 to the maximum number of components (maximum number of features of the input dataset)\n",
    "    for i in np.arange(1,X.shape[1]):\n",
    "        model = decom.PCA(n_components=i).fit(X) ## performing PCA and keeping 'i' principal components\n",
    "        ## if the explained variance sums is greater than ratio\n",
    "        ## this is the right model\n",
    "        if( np.sum(model.explained_variance_ratio_) >= ratio ):\n",
    "            return model \n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "## params: X,y,label, test_size are self explanatory\n",
    "## ratio -> The same minimum explained variance ratio that was mentioned in getPCAModel()\n",
    "## Threshold -> When we calculate the pdf of the record on the multi-variate gaussian (of normal records)\n",
    "##             if it is greater than Threshold, then it is a normal record (lies near the centre of the gaussian)\n",
    "##             otherwise, it is an anomaly. (lies far away from the center of the gaussian)\n",
    "def getMVG(X,y,label='Check1',test_size=0.2,ratio=0.95,Threshold=0.1):\n",
    "    \n",
    "    X_train,X_test,y_train,y_test = model_selection.train_test_split(X,y[label],test_size=test_size,stratify=y[label])\n",
    "    \n",
    "    ## Get the PCA model\n",
    "    PCA_Model = getPCAModel(X_train,ratio=ratio)\n",
    "    \n",
    "    ## transform the X_train into principal components.\n",
    "    ## We convert it into a pandas Dataframe because it's easy to manipulate a pandas dataframe\n",
    "    PCA_X = PCA_Model.transform(X_train)\n",
    "    PCA_X = pd.DataFrame(PCA_X, index=X_train.index)\n",
    "    \n",
    "    ## trX is the set of only the normal records present in the training set\n",
    "    trX = PCA_X[y_train==0] ## for training we only need normal points\n",
    "    \n",
    "    ## using the normal records present in the training set (i.e. trX)\n",
    "    ## we make a multivariate gaussian\n",
    "    ## the mean is the mean of trX\n",
    "    ## the the covariance is the covariance of trX\n",
    "    distrib = stats.multivariate_normal(mean=trX.mean(axis=0),cov=trX.cov())\n",
    "    \n",
    "    ## The test set is transformed into principal components using PCA model\n",
    "    tsX = PCA_Model.transform(X_test)\n",
    "    \n",
    "    ## Predictions (initially all zeros)\n",
    "    PCA_pred = pd.Series(np.zeros(X_test.shape[0]),index = X_test.index)\n",
    "    \n",
    "    for i in np.arange(X_test.shape[0]): ## iterating through all the test records\n",
    "        if ( distrib.pdf( tsX[i,:] ) < Threshold): ## calculating the pdf and check if it's lower than 'Threshold'\n",
    "            PCA_pred.iloc[i] = 1 ## if lower, then it's an anomaly.\n",
    "    \n",
    "    ## Calculating metrics\n",
    "    ## Note that the average is set to None, which calculates two f1 scores/label\n",
    "    ## one f1 score for the normal data and one f1 score for the anomalous data\n",
    "    score = met.f1_score(y_test,PCA_pred,average=None)\n",
    "    conf = met.confusion_matrix(y_pred=PCA_pred,y_true=y_test)\n",
    "    \n",
    "    return score,conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0.9579039 , 0.02700637]), array([[43451,  3565],\n",
       "        [  254,    53]]))"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "getMVG(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "PCA_f1 = pd.DataFrame(index=y.columns,columns=['Normal','Anomalous'])\n",
    "PCA_conf = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "for check in y.columns:\n",
    "    f1,conf = getMVG(X,y,label=check,ratio=0.99) ## the minimum variance ratio is set to 99%\n",
    "    PCA_f1.loc[check,:] = f1\n",
    "    PCA_conf[check] = conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Normal</th>\n",
       "      <th>Anomalous</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Check1</th>\n",
       "      <td>0.967641</td>\n",
       "      <td>0.0275591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check2</th>\n",
       "      <td>0.957938</td>\n",
       "      <td>0.13688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check3</th>\n",
       "      <td>0.957072</td>\n",
       "      <td>0.0998143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check4</th>\n",
       "      <td>0.936459</td>\n",
       "      <td>0.30517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check5</th>\n",
       "      <td>0.968495</td>\n",
       "      <td>0.12682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check6</th>\n",
       "      <td>0.82917</td>\n",
       "      <td>0.181696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check7</th>\n",
       "      <td>0.961076</td>\n",
       "      <td>0.169851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check8</th>\n",
       "      <td>0.966633</td>\n",
       "      <td>0.0672989</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Normal  Anomalous\n",
       "Check1  0.967641  0.0275591\n",
       "Check2  0.957938    0.13688\n",
       "Check3  0.957072  0.0998143\n",
       "Check4  0.936459    0.30517\n",
       "Check5  0.968495    0.12682\n",
       "Check6   0.82917   0.181696\n",
       "Check7  0.961076   0.169851\n",
       "Check8  0.966633  0.0672989"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "PCA_f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tCheck1\n",
      "[[44317  2699]\n",
      " [  265    42]]\n",
      "\tCheck2\n",
      "[[43226  2489]\n",
      " [ 1307   301]]\n",
      "\tCheck3\n",
      "[[43230  2609]\n",
      " [ 1269   215]]\n",
      "\tCheck4\n",
      "[[40603  2085]\n",
      " [ 3425  1210]]\n",
      "\tCheck5\n",
      "[[44236  2519]\n",
      " [  359   209]]\n",
      "\tCheck6\n",
      "[[32462  1842]\n",
      " [11534  1485]]\n",
      "\tCheck7\n",
      "[[43444  2455]\n",
      " [ 1064   360]]\n",
      "\tCheck8\n",
      "[[44164  2690]\n",
      " [  359   110]]\n"
     ]
    }
   ],
   "source": [
    "for keys in y.columns:\n",
    "    print(\"\\t\" + keys)\n",
    "    print(PCA_conf[keys])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>Auto Encoder Technique</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import keras.optimizers as optims\n",
    "import keras.utils as utils\n",
    "import keras.layers as layers\n",
    "import keras.models as models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "## params: X,y,label,test_size are self-explanatory\n",
    "## epochs -> number of epochs to train\n",
    "## Thresold -> If the input and output have a difference smaller than this number, then it's not an anomaly\n",
    "## verbose -> if true, prints the model summary and the progress\n",
    "## Note: The shape of the neural network is very simple:\n",
    "##       input(40) -> layer1(15) -> layer2(15) -> output(15)\n",
    "def AutoEncoder(X,y,label='Check1',epochs=5,Thresold=0.1,test_size=0.2,verbose=False):\n",
    "    X_train,X_test,y_train,y_test = model_selection.train_test_split(X,y[label],test_size=test_size,stratify=y[label])\n",
    "    #create model\n",
    "    \n",
    "    model = models.Sequential()\n",
    "    ## Input layer and the first layer. Activation is relu since it's an internal layer\n",
    "    model.add(layers.Dense(15,activation='relu',input_shape=(X_train.shape[1],)))\n",
    "    ## The second layer, activation is relu since it's an internal layer\n",
    "    model.add(layers.Dense(15,activation='relu'))\n",
    "    ## the output layer, activation is sigmoid since it's an output layer\n",
    "    model.add(layers.Dense(X.shape[1],activation='sigmoid'))\n",
    "    if(verbose):\n",
    "        print(model.summary())\n",
    "    # using adam optimizer with default parameters\n",
    "    model.compile(optimizer='adadelta', loss='binary_crossentropy')\n",
    "    \n",
    "    ## ind keeps the index of all the normal records in the training set\n",
    "    ind = y_train[y_train==0].index\n",
    "    ## keeping only the normal records for training\n",
    "    X_train = X_train.loc[ind,:]\n",
    "    \n",
    "    ## fitting the model. Notice the input and output is the same since it's an auto-encoder\n",
    "    model.fit(X_train,X_train,epochs=epochs,verbose=(1 if verbose else 0))\n",
    "    \n",
    "    ## predicting the X_test features\n",
    "    predictions = model.predict(X_test)\n",
    "    \n",
    "    ## The AE_pred are the predictions of the labels. initially all normal\n",
    "    AE_pred = pd.Series(np.zeros(X_test.shape[0]),index=X_test.index)\n",
    "    for i in np.arange(X_test.shape[0]): ## to iterate through the results of the AutoEncoder\n",
    "        \n",
    "        ## If the predictions of features and the test set have a difference greater than threshold\n",
    "        if(np.allclose(predictions[i,:],X_test.iloc[i,:],atol=Thresold) == False):\n",
    "            AE_pred.iloc[i] = 1 # it's an anomaly\n",
    "    \n",
    "    conf = met.confusion_matrix(y_true=y_test,y_pred=AE_pred)\n",
    "    f1 = met.f1_score(y_pred=AE_pred,y_true=y_test,average=None)\n",
    "    \n",
    "    return f1,conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "AE_f1 = pd.DataFrame(index=y.columns,columns=['Normal','Anomalous'])\n",
    "AE_conf = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "for check in y.columns:\n",
    "    f1,conf = AutoEncoder(X,y,label=check)\n",
    "    AE_f1.loc[check,:] = f1\n",
    "    AE_conf[check] = conf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Normal</th>\n",
       "      <th>Anomalous</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Check1</th>\n",
       "      <td>0.8793</td>\n",
       "      <td>0.0200405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check2</th>\n",
       "      <td>0.854162</td>\n",
       "      <td>0.113529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check3</th>\n",
       "      <td>0.866429</td>\n",
       "      <td>0.0911939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check4</th>\n",
       "      <td>0.850722</td>\n",
       "      <td>0.328804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check5</th>\n",
       "      <td>0.844329</td>\n",
       "      <td>0.0575389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check6</th>\n",
       "      <td>0.785559</td>\n",
       "      <td>0.336778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check7</th>\n",
       "      <td>0.830011</td>\n",
       "      <td>0.117609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Check8</th>\n",
       "      <td>0.808969</td>\n",
       "      <td>0.0432933</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Normal  Anomalous\n",
       "Check1    0.8793  0.0200405\n",
       "Check2  0.854162   0.113529\n",
       "Check3  0.866429  0.0911939\n",
       "Check4  0.850722   0.328804\n",
       "Check5  0.844329  0.0575389\n",
       "Check6  0.785559   0.336778\n",
       "Check7  0.830011   0.117609\n",
       "Check8  0.808969  0.0432933"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "AE_f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tCheck1\n",
      "[[37048  9968]\n",
      " [  203   104]]\n",
      "\tCheck2\n",
      "[[34711 11004]\n",
      " [  849   759]]\n",
      "\tCheck3\n",
      "[[35748 10091]\n",
      " [  931   553]]\n",
      "\tCheck4\n",
      "[[32934  9754]\n",
      " [ 1804  2831]]\n",
      "\tCheck5\n",
      "[[34292 12463]\n",
      " [  182   386]]\n",
      "\tCheck6\n",
      "[[28092  6212]\n",
      " [ 9125  3894]]\n",
      "\tCheck7\n",
      "[[32934 12965]\n",
      " [  525   899]]\n",
      "\tCheck8\n",
      "[[31911 14943]\n",
      " [  128   341]]\n"
     ]
    }
   ],
   "source": [
    "for keys in y.columns:\n",
    "    print(\"\\t\" + keys)\n",
    "    print(AE_conf[keys])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}