Predictive Maintenance using Machine Learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Sources\n",
    "\n",
    "Common data sources for predictive maintenance problems are :\n",
    "\n",
    "* **Failure history:** The failure history of a machine or component within the machine.\n",
    "* **Maintenance history:** The repair history of a machine, e.g. error codes, previous maintenance activities or component replacements.\n",
    "* **Machine conditions and usage:** The operating conditions of a machine e.g. data collected from sensors.\n",
    "* **Machine features:** The features of a machine, e.g. engine size, make and model, location.\n",
    "* **Operator features:** The features of the operator, e.g. gender, past experience\n",
    "The data for this example comes from 4 different sources which are real-time telemetry data collected from machines, error messages, historical maintenance records that include failures and machine information such as type and age."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "telemetry = pd.read_csv('PdM_telemetry.csv')\n",
    "errors = pd.read_csv('PdM_errors.csv')\n",
    "maint = pd.read_csv('PdM_maint.csv')\n",
    "failures = pd.read_csv('PdM_failures.csv')\n",
    "machines = pd.read_csv('PdM_machines.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of telemetry records: 876100\n",
      "             datetime  machineID        volt      rotate    pressure  \\\n",
      "0 2015-01-01 06:00:00          1  176.217853  418.504078  113.077935   \n",
      "1 2015-01-01 07:00:00          1  162.879223  402.747490   95.460525   \n",
      "2 2015-01-01 08:00:00          1  170.989902  527.349825   75.237905   \n",
      "3 2015-01-01 09:00:00          1  162.462833  346.149335  109.248561   \n",
      "4 2015-01-01 10:00:00          1  157.610021  435.376873  111.886648   \n",
      "\n",
      "   vibration  \n",
      "0  45.087686  \n",
      "1  43.413973  \n",
      "2  34.178847  \n",
      "3  41.122144  \n",
      "4  25.990511  \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>volt</th>\n",
       "      <th>rotate</th>\n",
       "      <th>pressure</th>\n",
       "      <th>vibration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>170.777736</td>\n",
       "      <td>446.605119</td>\n",
       "      <td>100.858668</td>\n",
       "      <td>40.385007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.866087</td>\n",
       "      <td>15.509114</td>\n",
       "      <td>52.673886</td>\n",
       "      <td>11.048679</td>\n",
       "      <td>5.370361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>97.333604</td>\n",
       "      <td>138.432075</td>\n",
       "      <td>51.237106</td>\n",
       "      <td>14.877054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.750000</td>\n",
       "      <td>160.304927</td>\n",
       "      <td>412.305714</td>\n",
       "      <td>93.498181</td>\n",
       "      <td>36.777299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>170.607338</td>\n",
       "      <td>447.558150</td>\n",
       "      <td>100.425559</td>\n",
       "      <td>40.237247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.250000</td>\n",
       "      <td>181.004493</td>\n",
       "      <td>482.176600</td>\n",
       "      <td>107.555231</td>\n",
       "      <td>43.784938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>255.124717</td>\n",
       "      <td>695.020984</td>\n",
       "      <td>185.951998</td>\n",
       "      <td>76.791072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           machineID           volt         rotate       pressure  \\\n",
       "count  876100.000000  876100.000000  876100.000000  876100.000000   \n",
       "mean       50.500000     170.777736     446.605119     100.858668   \n",
       "std        28.866087      15.509114      52.673886      11.048679   \n",
       "min         1.000000      97.333604     138.432075      51.237106   \n",
       "25%        25.750000     160.304927     412.305714      93.498181   \n",
       "50%        50.500000     170.607338     447.558150     100.425559   \n",
       "75%        75.250000     181.004493     482.176600     107.555231   \n",
       "max       100.000000     255.124717     695.020984     185.951998   \n",
       "\n",
       "           vibration  \n",
       "count  876100.000000  \n",
       "mean       40.385007  \n",
       "std         5.370361  \n",
       "min        14.877054  \n",
       "25%        36.777299  \n",
       "50%        40.237247  \n",
       "75%        43.784938  \n",
       "max        76.791072  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# format datetime field which comes in as string\n",
    "telemetry['datetime'] = pd.to_datetime(telemetry['datetime'], format=\"%Y-%m-%d %H:%M:%S\")\n",
    "\n",
    "print(\"Total number of telemetry records: %d\" % len(telemetry.index))\n",
    "print(telemetry.head())\n",
    "telemetry.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Telemetry**\n",
    "The first data source is the telemetry time-series data which consists of **voltage, rotation, pressure, and vibration** measurements collected from 100 machines in **real time averaged over every hour collected during the year 2015**. Below, we display the first 10 records in the dataset. A summary of the whole dataset is also provided."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x2564dac0780>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJAAAAHfCAYAAAAP5iI3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nPwS/4r/0OcKfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564d9b00b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "plot_df = telemetry.loc[(telemetry['machineID'] == 1) & \n",
    "                        (telemetry['datetime'] > pd.to_datetime('2015-01-01')) & \n",
    "                        (telemetry['datetime'] <pd.to_datetime('2015-02-01')),\n",
    "                        ['datetime','volt']]\n",
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(20, 8))\n",
    "plt.plot(plot_df['datetime'], plot_df['volt'])\n",
    "plt.ylabel('voltage')\n",
    "\n",
    "# make x-axis ticks legible\n",
    "adf = plt.gca().get_xaxis().get_major_formatter()\n",
    "adf.scaled[1.0] = '%m-%d-%Y'\n",
    "plt.xlabel('Date')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Errors**\n",
    "The second major data source is the error logs. These are **non-breaking errors thrown while the machine is still operational and do not constitute as failures.** The **error date and times** are rounded to the closest hour since the telemetry data is collected at an hourly rate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Number of error records: 3919\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>errorID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-03 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-03 20:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-04 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-10 15:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-22 10:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime  machineID errorID\n",
       "0 2015-01-03 07:00:00          1  error1\n",
       "1 2015-01-03 20:00:00          1  error3\n",
       "2 2015-01-04 06:00:00          1  error5\n",
       "3 2015-01-10 15:00:00          1  error4\n",
       "4 2015-01-22 10:00:00          1  error4"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# format of datetime field which comes in as string\n",
    "errors['datetime'] = pd.to_datetime(errors['datetime'],format = '%Y-%m-%d %H:%M:%S')\n",
    "errors['errorID'] = errors['errorID'].astype('category')\n",
    "print(\"Total Number of error records: %d\" %len(errors.index))\n",
    "errors.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "error1    1010\n",
       "error2     988\n",
       "error3     838\n",
       "error4     727\n",
       "error5     356\n",
       "Name: errorID, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJUAAAHjCAYAAACady6lAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKgEAAABQTFQCAAAAoNj/A1BuUtv7wchiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564dcb02e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(20, 8))\n",
    "errors['errorID'].value_counts().plot(kind='bar')\n",
    "plt.ylabel('Count')\n",
    "errors['errorID'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Maintenance**\n",
    "These are the **scheduled and unscheduled** maintenance records which correspond to both **regular inspection of components as well as failures.** A **record is generated if a component is replaced during the scheduled inspection or replaced due to a breakdown.** The **records that are created due to breakdowns will be called failures** which is explained in the later sections. Maintenance data has both 2014 and 2015 records."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Number of maintenance Records: 3286\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>comp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014-06-01 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2014-07-16 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2014-07-31 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2014-12-13 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-05 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime  machineID   comp\n",
       "0 2014-06-01 06:00:00          1  comp2\n",
       "1 2014-07-16 06:00:00          1  comp4\n",
       "2 2014-07-31 06:00:00          1  comp3\n",
       "3 2014-12-13 06:00:00          1  comp1\n",
       "4 2015-01-05 06:00:00          1  comp4"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "maint['datetime'] = pd.to_datetime(maint['datetime'], format='%Y-%m-%d %H:%M:%S')\n",
    "maint['comp'] = maint['comp'].astype('category')\n",
    "print(\"Total Number of maintenance Records: %d\" %len(maint.index))\n",
    "maint.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "comp2    863\n",
       "comp4    811\n",
       "comp3    808\n",
       "comp1    804\n",
       "Name: comp, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmIAAAEMCAYAAABnQNpRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsFYROM5ATAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564c18b0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(10, 4))\n",
    "maint['comp'].value_counts().plot(kind='bar')\n",
    "plt.ylabel('Count')\n",
    "maint['comp'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Machines**\n",
    "This data set includes some information about the machines: model type and age (years in service)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of machines: 100\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>model</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>model4</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>model3</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>model3</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>model3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   machineID   model  age\n",
       "0          1  model3   18\n",
       "1          2  model4    7\n",
       "2          3  model3    8\n",
       "3          4  model3    7\n",
       "4          5  model3    2"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "machines['model'] = machines['model'].astype('category')\n",
    "\n",
    "print(\"Total number of machines: %d\" % len(machines.index))\n",
    "machines.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2564dda9898>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3MAAAFyCAYAAABfioP2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564dd2ee10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(15, 6))\n",
    "_, bins, _ = plt.hist([machines.loc[machines['model'] == 'model1', 'age'],\n",
    "                       machines.loc[machines['model'] == 'model2', 'age'],\n",
    "                       machines.loc[machines['model'] == 'model3', 'age'],\n",
    "                       machines.loc[machines['model'] == 'model4', 'age']],\n",
    "                       20, stacked=True, label=['model1', 'model2', 'model3', 'model4'])\n",
    "plt.xlabel('Age (yrs)')\n",
    "plt.ylabel('Count')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Failures**\n",
    "These are the records of component replacements **due to failures.** Each record has a **date and time, machine ID, and failed component type.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of failures: 761\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>failure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-05 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-03-06 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-04-20 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-06-19 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-09-02 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime  machineID failure\n",
       "0 2015-01-05 06:00:00          1   comp4\n",
       "1 2015-03-06 06:00:00          1   comp1\n",
       "2 2015-04-20 06:00:00          1   comp2\n",
       "3 2015-06-19 06:00:00          1   comp4\n",
       "4 2015-09-02 06:00:00          1   comp4"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# format datetime field which comes in as string\n",
    "failures['datetime'] = pd.to_datetime(failures['datetime'], format=\"%Y-%m-%d %H:%M:%S\")\n",
    "failures['failure'] = failures['failure'].astype('category')\n",
    "\n",
    "print(\"Total number of failures: %d\" % len(failures.index))\n",
    "failures.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "comp2    259\n",
       "comp1    192\n",
       "comp4    179\n",
       "comp3    131\n",
       "Name: failure, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3kAAAEMCAYAAAB5gH+OAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564c147400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(15, 4))\n",
    "failures['failure'].value_counts().plot(kind='bar')\n",
    "plt.ylabel('Count')\n",
    "failures['failure'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering\n",
    "The first step in predictive maintenance applications is feature engineering which requires bringing the different data sources together to create features that best describe a machines's health condition at a given point in time. In the next sections, several feature engineering methods are used to create features based on the properties of each data source."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Lag Features from Telemetry\n",
    "Telemetry data almost always comes with time-stamps which makes it suitable for calculating lagging features. A common method is to pick a window size for the lag features to be created and compute rolling aggregate measures such as mean, standard deviation, minimum, maximum, etc. to represent the short term history of the telemetry over the lag window. In the following, rolling mean and standard deviation of the telemetry data over the last 3 hour lag window is calculated for every 3 hours."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:8: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).mean()\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "# Calculate mean values for telemetry features\n",
    "temp = []\n",
    "fields = ['volt', 'rotate', 'pressure', 'vibration']\n",
    "for col in fields:\n",
    "    temp.append(pd.pivot_table(telemetry,\n",
    "                               index='datetime',\n",
    "                               columns='machineID',\n",
    "                               values=col).resample('3H', closed='left', label='right', how='mean').unstack())\n",
    "telemetry_mean_3h = pd.concat(temp, axis=1)\n",
    "telemetry_mean_3h.columns = [i + 'mean_3h' for i in fields]\n",
    "telemetry_mean_3h.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:8: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).mean()\n",
      "  \n",
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:19: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).std()\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 09:00:00</td>\n",
       "      <td>170.028993</td>\n",
       "      <td>449.533798</td>\n",
       "      <td>94.592122</td>\n",
       "      <td>40.893502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 12:00:00</td>\n",
       "      <td>164.192565</td>\n",
       "      <td>403.949857</td>\n",
       "      <td>105.687417</td>\n",
       "      <td>34.255891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 15:00:00</td>\n",
       "      <td>168.134445</td>\n",
       "      <td>435.781707</td>\n",
       "      <td>107.793709</td>\n",
       "      <td>41.239405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 18:00:00</td>\n",
       "      <td>165.514453</td>\n",
       "      <td>430.472823</td>\n",
       "      <td>101.703289</td>\n",
       "      <td>40.373739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 21:00:00</td>\n",
       "      <td>168.809347</td>\n",
       "      <td>437.111120</td>\n",
       "      <td>90.911060</td>\n",
       "      <td>41.738542</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   machineID            datetime  voltmean_3h  rotatemean_3h  pressuremean_3h  \\\n",
       "0          1 2015-01-01 09:00:00   170.028993     449.533798        94.592122   \n",
       "1          1 2015-01-01 12:00:00   164.192565     403.949857       105.687417   \n",
       "2          1 2015-01-01 15:00:00   168.134445     435.781707       107.793709   \n",
       "3          1 2015-01-01 18:00:00   165.514453     430.472823       101.703289   \n",
       "4          1 2015-01-01 21:00:00   168.809347     437.111120        90.911060   \n",
       "\n",
       "   vibrationmean_3h  \n",
       "0         40.893502  \n",
       "1         34.255891  \n",
       "2         41.239405  \n",
       "3         40.373739  \n",
       "4         41.738542  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate mean values for telemetry features\n",
    "temp = []\n",
    "fields = ['volt', 'rotate', 'pressure', 'vibration']\n",
    "for col in fields:\n",
    "    temp.append(pd.pivot_table(telemetry,\n",
    "                               index='datetime',\n",
    "                               columns='machineID',\n",
    "                               values=col).resample('3H', closed='left', label='right', how='mean').unstack())\n",
    "telemetry_mean_3h = pd.concat(temp, axis=1)\n",
    "telemetry_mean_3h.columns = [i + 'mean_3h' for i in fields]\n",
    "telemetry_mean_3h.reset_index(inplace=True)\n",
    "\n",
    "# repeat for standard deviation\n",
    "temp = []\n",
    "for col in fields:\n",
    "    temp.append(pd.pivot_table(telemetry,\n",
    "                               index='datetime',\n",
    "                               columns='machineID',\n",
    "                               values=col).resample('3H', closed='left', label='right', how='std').unstack())\n",
    "telemetry_sd_3h = pd.concat(temp, axis=1)\n",
    "telemetry_sd_3h.columns = [i + 'sd_3h' for i in fields]\n",
    "telemetry_sd_3h.reset_index(inplace=True)\n",
    "\n",
    "telemetry_mean_3h.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "For capturing a longer term effect, 24 hour lag features are also calculated as below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:7: FutureWarning: pd.rolling_mean is deprecated for DataFrame and will be removed in a future version, replace with \n",
      "\tDataFrame.rolling(window=24,center=False).mean()\n",
      "  import sys\n",
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:10: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).first()\n",
      "  # Remove the CWD from sys.path while we load stuff.\n",
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:23: FutureWarning: pd.rolling_std is deprecated for DataFrame and will be removed in a future version, replace with \n",
      "\tDataFrame.rolling(window=24,center=False).std()\n",
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:26: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).first()\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_24h</th>\n",
       "      <th>rotatemean_24h</th>\n",
       "      <th>pressuremean_24h</th>\n",
       "      <th>vibrationmean_24h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 06:00:00</td>\n",
       "      <td>169.733809</td>\n",
       "      <td>445.179865</td>\n",
       "      <td>96.797113</td>\n",
       "      <td>40.385160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 09:00:00</td>\n",
       "      <td>170.614862</td>\n",
       "      <td>446.364859</td>\n",
       "      <td>96.849785</td>\n",
       "      <td>39.736826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 12:00:00</td>\n",
       "      <td>169.893965</td>\n",
       "      <td>447.009407</td>\n",
       "      <td>97.715600</td>\n",
       "      <td>39.498374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 15:00:00</td>\n",
       "      <td>171.243444</td>\n",
       "      <td>444.233563</td>\n",
       "      <td>96.666060</td>\n",
       "      <td>40.229370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 18:00:00</td>\n",
       "      <td>170.792486</td>\n",
       "      <td>448.440437</td>\n",
       "      <td>95.766838</td>\n",
       "      <td>40.055214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 21:00:00</td>\n",
       "      <td>170.556674</td>\n",
       "      <td>452.267095</td>\n",
       "      <td>98.065860</td>\n",
       "      <td>40.033247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 00:00:00</td>\n",
       "      <td>168.460525</td>\n",
       "      <td>451.031783</td>\n",
       "      <td>99.273286</td>\n",
       "      <td>38.903462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 03:00:00</td>\n",
       "      <td>169.772951</td>\n",
       "      <td>447.502464</td>\n",
       "      <td>99.005946</td>\n",
       "      <td>39.389725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 06:00:00</td>\n",
       "      <td>170.900562</td>\n",
       "      <td>453.864597</td>\n",
       "      <td>100.877342</td>\n",
       "      <td>38.696225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 09:00:00</td>\n",
       "      <td>169.533156</td>\n",
       "      <td>454.785072</td>\n",
       "      <td>100.050567</td>\n",
       "      <td>39.449734</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    machineID            datetime  voltmean_24h  rotatemean_24h  \\\n",
       "7           1 2015-01-02 06:00:00    169.733809      445.179865   \n",
       "8           1 2015-01-02 09:00:00    170.614862      446.364859   \n",
       "9           1 2015-01-02 12:00:00    169.893965      447.009407   \n",
       "10          1 2015-01-02 15:00:00    171.243444      444.233563   \n",
       "11          1 2015-01-02 18:00:00    170.792486      448.440437   \n",
       "12          1 2015-01-02 21:00:00    170.556674      452.267095   \n",
       "13          1 2015-01-03 00:00:00    168.460525      451.031783   \n",
       "14          1 2015-01-03 03:00:00    169.772951      447.502464   \n",
       "15          1 2015-01-03 06:00:00    170.900562      453.864597   \n",
       "16          1 2015-01-03 09:00:00    169.533156      454.785072   \n",
       "\n",
       "    pressuremean_24h  vibrationmean_24h  \n",
       "7          96.797113          40.385160  \n",
       "8          96.849785          39.736826  \n",
       "9          97.715600          39.498374  \n",
       "10         96.666060          40.229370  \n",
       "11         95.766838          40.055214  \n",
       "12         98.065860          40.033247  \n",
       "13         99.273286          38.903462  \n",
       "14         99.005946          39.389725  \n",
       "15        100.877342          38.696225  \n",
       "16        100.050567          39.449734  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = []\n",
    "fields = ['volt', 'rotate', 'pressure', 'vibration']\n",
    "for col in fields:\n",
    "    temp.append(pd.rolling_mean(pd.pivot_table(telemetry,\n",
    "                                               index='datetime',\n",
    "                                               columns='machineID',\n",
    "                                               values=col), window=24).resample('3H',\n",
    "                                                                                closed='left',\n",
    "                                                                                label='right',\n",
    "                                                                                how='first').unstack())\n",
    "telemetry_mean_24h = pd.concat(temp, axis=1)\n",
    "telemetry_mean_24h.columns = [i + 'mean_24h' for i in fields]\n",
    "telemetry_mean_24h.reset_index(inplace=True)\n",
    "telemetry_mean_24h = telemetry_mean_24h.loc[-telemetry_mean_24h['voltmean_24h'].isnull()]\n",
    "\n",
    "# repeat for standard deviation\n",
    "temp = []\n",
    "fields = ['volt', 'rotate', 'pressure', 'vibration']\n",
    "for col in fields:\n",
    "    temp.append(pd.rolling_std(pd.pivot_table(telemetry,\n",
    "                                               index='datetime',\n",
    "                                               columns='machineID',\n",
    "                                               values=col), window=24).resample('3H',\n",
    "                                                                                closed='left',\n",
    "                                                                                label='right',\n",
    "                                                                                how='first').unstack())\n",
    "telemetry_sd_24h = pd.concat(temp, axis=1)\n",
    "telemetry_sd_24h.columns = [i + 'sd_24h' for i in fields]\n",
    "telemetry_sd_24h = telemetry_sd_24h.loc[-telemetry_sd_24h['voltsd_24h'].isnull()]\n",
    "telemetry_sd_24h.reset_index(inplace=True)\n",
    "\n",
    "# Notice that a 24h rolling average is not available at the earliest timepoints\n",
    "telemetry_mean_24h.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, the columns of the feature datasets created earlier are merged to create the final feature set from telemetry."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>voltmean_24h</th>\n",
       "      <th>rotatemean_24h</th>\n",
       "      <th>pressuremean_24h</th>\n",
       "      <th>vibrationmean_24h</th>\n",
       "      <th>voltsd_24h</th>\n",
       "      <th>rotatesd_24h</th>\n",
       "      <th>pressuresd_24h</th>\n",
       "      <th>vibrationsd_24h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.380935</td>\n",
       "      <td>170.774427</td>\n",
       "      <td>446.609386</td>\n",
       "      <td>100.858340</td>\n",
       "      <td>40.383609</td>\n",
       "      <td>13.300173</td>\n",
       "      <td>44.453951</td>\n",
       "      <td>8.885780</td>\n",
       "      <td>4.440575</td>\n",
       "      <td>170.775661</td>\n",
       "      <td>446.609874</td>\n",
       "      <td>100.857574</td>\n",
       "      <td>40.383881</td>\n",
       "      <td>14.919452</td>\n",
       "      <td>49.950788</td>\n",
       "      <td>10.046380</td>\n",
       "      <td>5.002089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.798424</td>\n",
       "      <td>9.498824</td>\n",
       "      <td>33.119738</td>\n",
       "      <td>7.411701</td>\n",
       "      <td>3.475512</td>\n",
       "      <td>6.966389</td>\n",
       "      <td>23.214291</td>\n",
       "      <td>4.656364</td>\n",
       "      <td>2.319989</td>\n",
       "      <td>4.720237</td>\n",
       "      <td>18.070458</td>\n",
       "      <td>4.737293</td>\n",
       "      <td>2.058059</td>\n",
       "      <td>2.261097</td>\n",
       "      <td>7.684305</td>\n",
       "      <td>1.713206</td>\n",
       "      <td>0.799599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>125.532506</td>\n",
       "      <td>211.811184</td>\n",
       "      <td>72.118639</td>\n",
       "      <td>26.569635</td>\n",
       "      <td>0.025509</td>\n",
       "      <td>0.078991</td>\n",
       "      <td>0.027417</td>\n",
       "      <td>0.015278</td>\n",
       "      <td>155.812721</td>\n",
       "      <td>266.010419</td>\n",
       "      <td>91.057429</td>\n",
       "      <td>35.060087</td>\n",
       "      <td>6.380619</td>\n",
       "      <td>18.385248</td>\n",
       "      <td>4.145308</td>\n",
       "      <td>2.144863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.000000</td>\n",
       "      <td>164.447794</td>\n",
       "      <td>427.564793</td>\n",
       "      <td>96.239534</td>\n",
       "      <td>38.147458</td>\n",
       "      <td>8.028675</td>\n",
       "      <td>26.906319</td>\n",
       "      <td>5.369959</td>\n",
       "      <td>2.684556</td>\n",
       "      <td>168.072275</td>\n",
       "      <td>441.542561</td>\n",
       "      <td>98.669734</td>\n",
       "      <td>39.354077</td>\n",
       "      <td>13.359069</td>\n",
       "      <td>44.669022</td>\n",
       "      <td>8.924165</td>\n",
       "      <td>4.460675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.000000</td>\n",
       "      <td>170.432407</td>\n",
       "      <td>448.380260</td>\n",
       "      <td>100.235357</td>\n",
       "      <td>40.145874</td>\n",
       "      <td>12.495542</td>\n",
       "      <td>41.793798</td>\n",
       "      <td>8.345801</td>\n",
       "      <td>4.173704</td>\n",
       "      <td>170.212704</td>\n",
       "      <td>449.206885</td>\n",
       "      <td>100.099533</td>\n",
       "      <td>40.072618</td>\n",
       "      <td>14.854186</td>\n",
       "      <td>49.617459</td>\n",
       "      <td>9.921332</td>\n",
       "      <td>4.958793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.000000</td>\n",
       "      <td>176.610017</td>\n",
       "      <td>468.443933</td>\n",
       "      <td>104.406534</td>\n",
       "      <td>42.226898</td>\n",
       "      <td>17.688520</td>\n",
       "      <td>59.092354</td>\n",
       "      <td>11.789358</td>\n",
       "      <td>5.898512</td>\n",
       "      <td>172.462228</td>\n",
       "      <td>456.366349</td>\n",
       "      <td>101.613047</td>\n",
       "      <td>40.833112</td>\n",
       "      <td>16.395372</td>\n",
       "      <td>54.826993</td>\n",
       "      <td>10.980250</td>\n",
       "      <td>5.484430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>241.420717</td>\n",
       "      <td>586.682904</td>\n",
       "      <td>162.309656</td>\n",
       "      <td>69.311324</td>\n",
       "      <td>58.444332</td>\n",
       "      <td>179.903039</td>\n",
       "      <td>35.659369</td>\n",
       "      <td>18.305595</td>\n",
       "      <td>220.782618</td>\n",
       "      <td>499.096975</td>\n",
       "      <td>152.310351</td>\n",
       "      <td>61.932124</td>\n",
       "      <td>27.664538</td>\n",
       "      <td>103.819404</td>\n",
       "      <td>28.654103</td>\n",
       "      <td>12.325783</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           machineID    voltmean_3h  rotatemean_3h  pressuremean_3h  \\\n",
       "count  290601.000000  290601.000000  290601.000000    290601.000000   \n",
       "mean       50.380935     170.774427     446.609386       100.858340   \n",
       "std        28.798424       9.498824      33.119738         7.411701   \n",
       "min         1.000000     125.532506     211.811184        72.118639   \n",
       "25%        25.000000     164.447794     427.564793        96.239534   \n",
       "50%        50.000000     170.432407     448.380260       100.235357   \n",
       "75%        75.000000     176.610017     468.443933       104.406534   \n",
       "max       100.000000     241.420717     586.682904       162.309656   \n",
       "\n",
       "       vibrationmean_3h      voltsd_3h    rotatesd_3h  pressuresd_3h  \\\n",
       "count     290601.000000  290601.000000  290601.000000  290601.000000   \n",
       "mean          40.383609      13.300173      44.453951       8.885780   \n",
       "std            3.475512       6.966389      23.214291       4.656364   \n",
       "min           26.569635       0.025509       0.078991       0.027417   \n",
       "25%           38.147458       8.028675      26.906319       5.369959   \n",
       "50%           40.145874      12.495542      41.793798       8.345801   \n",
       "75%           42.226898      17.688520      59.092354      11.789358   \n",
       "max           69.311324      58.444332     179.903039      35.659369   \n",
       "\n",
       "       vibrationsd_3h   voltmean_24h  rotatemean_24h  pressuremean_24h  \\\n",
       "count   290601.000000  290601.000000   290601.000000     290601.000000   \n",
       "mean         4.440575     170.775661      446.609874        100.857574   \n",
       "std          2.319989       4.720237       18.070458          4.737293   \n",
       "min          0.015278     155.812721      266.010419         91.057429   \n",
       "25%          2.684556     168.072275      441.542561         98.669734   \n",
       "50%          4.173704     170.212704      449.206885        100.099533   \n",
       "75%          5.898512     172.462228      456.366349        101.613047   \n",
       "max         18.305595     220.782618      499.096975        152.310351   \n",
       "\n",
       "       vibrationmean_24h     voltsd_24h   rotatesd_24h  pressuresd_24h  \\\n",
       "count      290601.000000  290601.000000  290601.000000   290601.000000   \n",
       "mean           40.383881      14.919452      49.950788       10.046380   \n",
       "std             2.058059       2.261097       7.684305        1.713206   \n",
       "min            35.060087       6.380619      18.385248        4.145308   \n",
       "25%            39.354077      13.359069      44.669022        8.924165   \n",
       "50%            40.072618      14.854186      49.617459        9.921332   \n",
       "75%            40.833112      16.395372      54.826993       10.980250   \n",
       "max            61.932124      27.664538     103.819404       28.654103   \n",
       "\n",
       "       vibrationsd_24h  \n",
       "count    290601.000000  \n",
       "mean          5.002089  \n",
       "std           0.799599  \n",
       "min           2.144863  \n",
       "25%           4.460675  \n",
       "50%           4.958793  \n",
       "75%           5.484430  \n",
       "max          12.325783  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# merge columns of feature sets created earlier\n",
    "telemetry_feat = pd.concat([telemetry_mean_3h,\n",
    "                            telemetry_sd_3h.ix[:, 2:6],\n",
    "                            telemetry_mean_24h.ix[:, 2:6],\n",
    "                            telemetry_sd_24h.ix[:, 2:6]], axis=1).dropna()\n",
    "telemetry_feat.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>voltmean_24h</th>\n",
       "      <th>rotatemean_24h</th>\n",
       "      <th>pressuremean_24h</th>\n",
       "      <th>vibrationmean_24h</th>\n",
       "      <th>voltsd_24h</th>\n",
       "      <th>rotatesd_24h</th>\n",
       "      <th>pressuresd_24h</th>\n",
       "      <th>vibrationsd_24h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 06:00:00</td>\n",
       "      <td>180.133784</td>\n",
       "      <td>440.608320</td>\n",
       "      <td>94.137969</td>\n",
       "      <td>41.551544</td>\n",
       "      <td>21.322735</td>\n",
       "      <td>48.770512</td>\n",
       "      <td>2.135684</td>\n",
       "      <td>10.037208</td>\n",
       "      <td>169.733809</td>\n",
       "      <td>445.179865</td>\n",
       "      <td>96.797113</td>\n",
       "      <td>40.385160</td>\n",
       "      <td>15.726970</td>\n",
       "      <td>39.648116</td>\n",
       "      <td>11.904700</td>\n",
       "      <td>5.601191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 09:00:00</td>\n",
       "      <td>176.364293</td>\n",
       "      <td>439.349655</td>\n",
       "      <td>101.553209</td>\n",
       "      <td>36.105580</td>\n",
       "      <td>18.952210</td>\n",
       "      <td>51.329636</td>\n",
       "      <td>13.789279</td>\n",
       "      <td>6.737739</td>\n",
       "      <td>170.614862</td>\n",
       "      <td>446.364859</td>\n",
       "      <td>96.849785</td>\n",
       "      <td>39.736826</td>\n",
       "      <td>15.635083</td>\n",
       "      <td>41.828592</td>\n",
       "      <td>11.326412</td>\n",
       "      <td>5.583521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 12:00:00</td>\n",
       "      <td>160.384568</td>\n",
       "      <td>424.385316</td>\n",
       "      <td>99.598722</td>\n",
       "      <td>36.094637</td>\n",
       "      <td>13.047080</td>\n",
       "      <td>13.702496</td>\n",
       "      <td>9.988609</td>\n",
       "      <td>1.639962</td>\n",
       "      <td>169.893965</td>\n",
       "      <td>447.009407</td>\n",
       "      <td>97.715600</td>\n",
       "      <td>39.498374</td>\n",
       "      <td>13.995465</td>\n",
       "      <td>40.843882</td>\n",
       "      <td>11.036546</td>\n",
       "      <td>5.561553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 15:00:00</td>\n",
       "      <td>170.472461</td>\n",
       "      <td>442.933997</td>\n",
       "      <td>102.380586</td>\n",
       "      <td>40.483002</td>\n",
       "      <td>16.642354</td>\n",
       "      <td>56.290447</td>\n",
       "      <td>3.305739</td>\n",
       "      <td>8.854145</td>\n",
       "      <td>171.243444</td>\n",
       "      <td>444.233563</td>\n",
       "      <td>96.666060</td>\n",
       "      <td>40.229370</td>\n",
       "      <td>13.100364</td>\n",
       "      <td>43.409841</td>\n",
       "      <td>10.972862</td>\n",
       "      <td>6.068674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 18:00:00</td>\n",
       "      <td>163.263806</td>\n",
       "      <td>468.937558</td>\n",
       "      <td>102.726648</td>\n",
       "      <td>40.921802</td>\n",
       "      <td>17.424688</td>\n",
       "      <td>38.680380</td>\n",
       "      <td>9.105775</td>\n",
       "      <td>3.060781</td>\n",
       "      <td>170.792486</td>\n",
       "      <td>448.440437</td>\n",
       "      <td>95.766838</td>\n",
       "      <td>40.055214</td>\n",
       "      <td>13.808489</td>\n",
       "      <td>43.742304</td>\n",
       "      <td>10.988704</td>\n",
       "      <td>7.286129</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    machineID            datetime  voltmean_3h  rotatemean_3h  \\\n",
       "7           1 2015-01-02 06:00:00   180.133784     440.608320   \n",
       "8           1 2015-01-02 09:00:00   176.364293     439.349655   \n",
       "9           1 2015-01-02 12:00:00   160.384568     424.385316   \n",
       "10          1 2015-01-02 15:00:00   170.472461     442.933997   \n",
       "11          1 2015-01-02 18:00:00   163.263806     468.937558   \n",
       "\n",
       "    pressuremean_3h  vibrationmean_3h  voltsd_3h  rotatesd_3h  pressuresd_3h  \\\n",
       "7         94.137969         41.551544  21.322735    48.770512       2.135684   \n",
       "8        101.553209         36.105580  18.952210    51.329636      13.789279   \n",
       "9         99.598722         36.094637  13.047080    13.702496       9.988609   \n",
       "10       102.380586         40.483002  16.642354    56.290447       3.305739   \n",
       "11       102.726648         40.921802  17.424688    38.680380       9.105775   \n",
       "\n",
       "    vibrationsd_3h  voltmean_24h  rotatemean_24h  pressuremean_24h  \\\n",
       "7        10.037208    169.733809      445.179865         96.797113   \n",
       "8         6.737739    170.614862      446.364859         96.849785   \n",
       "9         1.639962    169.893965      447.009407         97.715600   \n",
       "10        8.854145    171.243444      444.233563         96.666060   \n",
       "11        3.060781    170.792486      448.440437         95.766838   \n",
       "\n",
       "    vibrationmean_24h  voltsd_24h  rotatesd_24h  pressuresd_24h  \\\n",
       "7           40.385160   15.726970     39.648116       11.904700   \n",
       "8           39.736826   15.635083     41.828592       11.326412   \n",
       "9           39.498374   13.995465     40.843882       11.036546   \n",
       "10          40.229370   13.100364     43.409841       10.972862   \n",
       "11          40.055214   13.808489     43.742304       10.988704   \n",
       "\n",
       "    vibrationsd_24h  \n",
       "7          5.601191  \n",
       "8          5.583521  \n",
       "9          5.561553  \n",
       "10         6.068674  \n",
       "11         7.286129  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "telemetry_feat.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Lag Features from Errors\n",
    "Like telemetry data, errors come with timestamps. An important difference is that the **error IDs are categorical values** and **should not be averaged over time intervals like the telemetry measurements.** Instead, we count the number of errors of each type in a **lagging window. We begin by reformatting the error data** to have one entry per machine per time at which at least one error occurred:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>errorID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-03 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-03 20:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-04 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-10 15:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-22 10:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2015-01-25 15:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2015-01-27 04:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2015-03-03 22:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2015-03-05 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2015-03-20 18:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2015-03-26 01:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2015-03-31 23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2015-04-19 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2015-04-19 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2015-04-29 19:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2015-05-04 23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>2015-05-12 09:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2015-05-21 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>2015-05-24 02:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>2015-05-25 05:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2015-06-09 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2015-06-18 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2015-06-23 10:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2015-08-23 19:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2015-08-30 01:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>2015-09-01 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2015-09-13 17:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>2015-09-15 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>2015-10-01 23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>2015-10-15 05:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3889</th>\n",
       "      <td>2015-01-16 00:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3890</th>\n",
       "      <td>2015-02-01 10:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3891</th>\n",
       "      <td>2015-02-11 06:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3892</th>\n",
       "      <td>2015-02-12 21:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3893</th>\n",
       "      <td>2015-03-08 15:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3894</th>\n",
       "      <td>2015-04-27 04:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3895</th>\n",
       "      <td>2015-04-27 22:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3896</th>\n",
       "      <td>2015-05-16 23:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3897</th>\n",
       "      <td>2015-05-17 13:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3898</th>\n",
       "      <td>2015-05-22 02:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3899</th>\n",
       "      <td>2015-07-05 16:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3900</th>\n",
       "      <td>2015-07-19 01:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3901</th>\n",
       "      <td>2015-08-14 16:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3902</th>\n",
       "      <td>2015-08-30 15:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3903</th>\n",
       "      <td>2015-09-09 06:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3904</th>\n",
       "      <td>2015-09-14 23:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3905</th>\n",
       "      <td>2015-10-03 05:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3906</th>\n",
       "      <td>2015-10-09 07:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3907</th>\n",
       "      <td>2015-10-17 02:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3908</th>\n",
       "      <td>2015-10-17 12:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3909</th>\n",
       "      <td>2015-10-24 23:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3910</th>\n",
       "      <td>2015-10-27 21:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3911</th>\n",
       "      <td>2015-11-05 02:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3912</th>\n",
       "      <td>2015-11-07 17:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3913</th>\n",
       "      <td>2015-11-12 01:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3914</th>\n",
       "      <td>2015-11-21 08:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3915</th>\n",
       "      <td>2015-12-04 02:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3916</th>\n",
       "      <td>2015-12-08 06:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3917</th>\n",
       "      <td>2015-12-08 06:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3918</th>\n",
       "      <td>2015-12-22 03:00:00</td>\n",
       "      <td>100</td>\n",
       "      <td>error3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3919 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                datetime  machineID errorID\n",
       "0    2015-01-03 07:00:00          1  error1\n",
       "1    2015-01-03 20:00:00          1  error3\n",
       "2    2015-01-04 06:00:00          1  error5\n",
       "3    2015-01-10 15:00:00          1  error4\n",
       "4    2015-01-22 10:00:00          1  error4\n",
       "5    2015-01-25 15:00:00          1  error4\n",
       "6    2015-01-27 04:00:00          1  error1\n",
       "7    2015-03-03 22:00:00          1  error2\n",
       "8    2015-03-05 06:00:00          1  error1\n",
       "9    2015-03-20 18:00:00          1  error1\n",
       "10   2015-03-26 01:00:00          1  error2\n",
       "11   2015-03-31 23:00:00          1  error1\n",
       "12   2015-04-19 06:00:00          1  error2\n",
       "13   2015-04-19 06:00:00          1  error3\n",
       "14   2015-04-29 19:00:00          1  error4\n",
       "15   2015-05-04 23:00:00          1  error2\n",
       "16   2015-05-12 09:00:00          1  error1\n",
       "17   2015-05-21 07:00:00          1  error4\n",
       "18   2015-05-24 02:00:00          1  error3\n",
       "19   2015-05-25 05:00:00          1  error1\n",
       "20   2015-06-09 06:00:00          1  error3\n",
       "21   2015-06-18 06:00:00          1  error5\n",
       "22   2015-06-23 10:00:00          1  error3\n",
       "23   2015-08-23 19:00:00          1  error1\n",
       "24   2015-08-30 01:00:00          1  error3\n",
       "25   2015-09-01 06:00:00          1  error5\n",
       "26   2015-09-13 17:00:00          1  error2\n",
       "27   2015-09-15 06:00:00          1  error1\n",
       "28   2015-10-01 23:00:00          1  error1\n",
       "29   2015-10-15 05:00:00          1  error1\n",
       "...                  ...        ...     ...\n",
       "3889 2015-01-16 00:00:00        100  error4\n",
       "3890 2015-02-01 10:00:00        100  error1\n",
       "3891 2015-02-11 06:00:00        100  error1\n",
       "3892 2015-02-12 21:00:00        100  error1\n",
       "3893 2015-03-08 15:00:00        100  error1\n",
       "3894 2015-04-27 04:00:00        100  error4\n",
       "3895 2015-04-27 22:00:00        100  error5\n",
       "3896 2015-05-16 23:00:00        100  error2\n",
       "3897 2015-05-17 13:00:00        100  error2\n",
       "3898 2015-05-22 02:00:00        100  error3\n",
       "3899 2015-07-05 16:00:00        100  error3\n",
       "3900 2015-07-19 01:00:00        100  error2\n",
       "3901 2015-08-14 16:00:00        100  error4\n",
       "3902 2015-08-30 15:00:00        100  error4\n",
       "3903 2015-09-09 06:00:00        100  error1\n",
       "3904 2015-09-14 23:00:00        100  error3\n",
       "3905 2015-10-03 05:00:00        100  error3\n",
       "3906 2015-10-09 07:00:00        100  error1\n",
       "3907 2015-10-17 02:00:00        100  error3\n",
       "3908 2015-10-17 12:00:00        100  error1\n",
       "3909 2015-10-24 23:00:00        100  error1\n",
       "3910 2015-10-27 21:00:00        100  error2\n",
       "3911 2015-11-05 02:00:00        100  error3\n",
       "3912 2015-11-07 17:00:00        100  error1\n",
       "3913 2015-11-12 01:00:00        100  error1\n",
       "3914 2015-11-21 08:00:00        100  error2\n",
       "3915 2015-12-04 02:00:00        100  error1\n",
       "3916 2015-12-08 06:00:00        100  error2\n",
       "3917 2015-12-08 06:00:00        100  error3\n",
       "3918 2015-12-22 03:00:00        100  error3\n",
       "\n",
       "[3919 rows x 3 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "errors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>error1</th>\n",
       "      <th>error2</th>\n",
       "      <th>error3</th>\n",
       "      <th>error4</th>\n",
       "      <th>error5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-03 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-03 20:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-04 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-10 15:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-22 10:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2015-01-25 15:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2015-01-27 04:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2015-03-03 22:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2015-03-05 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2015-03-20 18:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2015-03-26 01:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2015-03-31 23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2015-04-19 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              datetime  machineID  error1  error2  error3  error4  error5\n",
       "0  2015-01-03 07:00:00          1       1       0       0       0       0\n",
       "1  2015-01-03 20:00:00          1       0       0       1       0       0\n",
       "2  2015-01-04 06:00:00          1       0       0       0       0       1\n",
       "3  2015-01-10 15:00:00          1       0       0       0       1       0\n",
       "4  2015-01-22 10:00:00          1       0       0       0       1       0\n",
       "5  2015-01-25 15:00:00          1       0       0       0       1       0\n",
       "6  2015-01-27 04:00:00          1       1       0       0       0       0\n",
       "7  2015-03-03 22:00:00          1       0       1       0       0       0\n",
       "8  2015-03-05 06:00:00          1       1       0       0       0       0\n",
       "9  2015-03-20 18:00:00          1       1       0       0       0       0\n",
       "10 2015-03-26 01:00:00          1       0       1       0       0       0\n",
       "11 2015-03-31 23:00:00          1       1       0       0       0       0\n",
       "12 2015-04-19 06:00:00          1       0       1       0       0       0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a column for each error type\n",
    "error_count = pd.get_dummies(errors.set_index('datetime')).reset_index()\n",
    "error_count\n",
    "error_count.columns = ['datetime', 'machineID', 'error1', 'error2', 'error3', 'error4', 'error5']\n",
    "error_count.head(13)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>error1</th>\n",
       "      <th>error2</th>\n",
       "      <th>error3</th>\n",
       "      <th>error4</th>\n",
       "      <th>error5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 20:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 06:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-10 15:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-22 10:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-25 15:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-27 04:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-03 22:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-05 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-20 18:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-26 01:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-31 23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-04-19 06:00:00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    machineID            datetime  error1  error2  error3  error4  error5\n",
       "0           1 2015-01-03 07:00:00       1       0       0       0       0\n",
       "1           1 2015-01-03 20:00:00       0       0       1       0       0\n",
       "2           1 2015-01-04 06:00:00       0       0       0       0       1\n",
       "3           1 2015-01-10 15:00:00       0       0       0       1       0\n",
       "4           1 2015-01-22 10:00:00       0       0       0       1       0\n",
       "5           1 2015-01-25 15:00:00       0       0       0       1       0\n",
       "6           1 2015-01-27 04:00:00       1       0       0       0       0\n",
       "7           1 2015-03-03 22:00:00       0       1       0       0       0\n",
       "8           1 2015-03-05 06:00:00       1       0       0       0       0\n",
       "9           1 2015-03-20 18:00:00       1       0       0       0       0\n",
       "10          1 2015-03-26 01:00:00       0       1       0       0       0\n",
       "11          1 2015-03-31 23:00:00       1       0       0       0       0\n",
       "12          1 2015-04-19 06:00:00       0       1       1       0       0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combine errors for a given machine in a given hour\n",
    "error_count = error_count.groupby(['machineID','datetime']).sum().reset_index()\n",
    "error_count.head(13)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>error1</th>\n",
       "      <th>error2</th>\n",
       "      <th>error3</th>\n",
       "      <th>error4</th>\n",
       "      <th>error5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>0.001153</td>\n",
       "      <td>0.001128</td>\n",
       "      <td>0.000957</td>\n",
       "      <td>0.000830</td>\n",
       "      <td>0.000406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.866087</td>\n",
       "      <td>0.033934</td>\n",
       "      <td>0.033563</td>\n",
       "      <td>0.030913</td>\n",
       "      <td>0.028795</td>\n",
       "      <td>0.020154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.750000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.250000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           machineID         error1         error2         error3  \\\n",
       "count  876100.000000  876100.000000  876100.000000  876100.000000   \n",
       "mean       50.500000       0.001153       0.001128       0.000957   \n",
       "std        28.866087       0.033934       0.033563       0.030913   \n",
       "min         1.000000       0.000000       0.000000       0.000000   \n",
       "25%        25.750000       0.000000       0.000000       0.000000   \n",
       "50%        50.500000       0.000000       0.000000       0.000000   \n",
       "75%        75.250000       0.000000       0.000000       0.000000   \n",
       "max       100.000000       1.000000       1.000000       1.000000   \n",
       "\n",
       "              error4         error5  \n",
       "count  876100.000000  876100.000000  \n",
       "mean        0.000830       0.000406  \n",
       "std         0.028795       0.020154  \n",
       "min         0.000000       0.000000  \n",
       "25%         0.000000       0.000000  \n",
       "50%         0.000000       0.000000  \n",
       "75%         0.000000       0.000000  \n",
       "max         1.000000       1.000000  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_count = telemetry[['datetime', 'machineID']].merge(error_count, on=['machineID', 'datetime'], how='left').fillna(0.0)\n",
    "error_count.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we can compute the total number of errors of each type over the last 24 hours, for timepoints taken every three hours:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:7: FutureWarning: pd.rolling_sum is deprecated for DataFrame and will be removed in a future version, replace with \n",
      "\tDataFrame.rolling(window=24,center=False).sum()\n",
      "  import sys\n",
      "C:\\Program Files\\Microsoft\\ML Server\\PYTHON_SERVER\\lib\\site-packages\\ipykernel_launcher.py:10: FutureWarning: how in .resample() is deprecated\n",
      "the new syntax is .resample(...).first()\n",
      "  # Remove the CWD from sys.path while we load stuff.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>error1count</th>\n",
       "      <th>error2count</th>\n",
       "      <th>error3count</th>\n",
       "      <th>error4count</th>\n",
       "      <th>error5count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>291400.00000</td>\n",
       "      <td>291400.000000</td>\n",
       "      <td>291400.000000</td>\n",
       "      <td>291400.000000</td>\n",
       "      <td>291400.000000</td>\n",
       "      <td>291400.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.50000</td>\n",
       "      <td>0.027649</td>\n",
       "      <td>0.027069</td>\n",
       "      <td>0.022907</td>\n",
       "      <td>0.019904</td>\n",
       "      <td>0.009753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.86612</td>\n",
       "      <td>0.166273</td>\n",
       "      <td>0.164429</td>\n",
       "      <td>0.151453</td>\n",
       "      <td>0.140820</td>\n",
       "      <td>0.098797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.75000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.50000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.25000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.00000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          machineID    error1count    error2count    error3count  \\\n",
       "count  291400.00000  291400.000000  291400.000000  291400.000000   \n",
       "mean       50.50000       0.027649       0.027069       0.022907   \n",
       "std        28.86612       0.166273       0.164429       0.151453   \n",
       "min         1.00000       0.000000       0.000000       0.000000   \n",
       "25%        25.75000       0.000000       0.000000       0.000000   \n",
       "50%        50.50000       0.000000       0.000000       0.000000   \n",
       "75%        75.25000       0.000000       0.000000       0.000000   \n",
       "max       100.00000       2.000000       2.000000       2.000000   \n",
       "\n",
       "         error4count    error5count  \n",
       "count  291400.000000  291400.000000  \n",
       "mean        0.019904       0.009753  \n",
       "std         0.140820       0.098797  \n",
       "min         0.000000       0.000000  \n",
       "25%         0.000000       0.000000  \n",
       "50%         0.000000       0.000000  \n",
       "75%         0.000000       0.000000  \n",
       "max         2.000000       2.000000  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = []\n",
    "fields = ['error%d' % i for i in range(1,6)]\n",
    "for col in fields:\n",
    "    temp.append(pd.rolling_sum(pd.pivot_table(error_count,\n",
    "                                               index='datetime',\n",
    "                                               columns='machineID',\n",
    "                                               values=col), window=24).resample('3H',\n",
    "                                                                             closed='left',\n",
    "                                                                             label='right',\n",
    "                                                                             how='first').unstack())\n",
    "error_count = pd.concat(temp, axis=1)\n",
    "error_count.columns = [i + 'count' for i in fields]\n",
    "error_count.reset_index(inplace=True)\n",
    "error_count = error_count.dropna()\n",
    "error_count.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>error1count</th>\n",
       "      <th>error2count</th>\n",
       "      <th>error3count</th>\n",
       "      <th>error4count</th>\n",
       "      <th>error5count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 06:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 09:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 12:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 15:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 18:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    machineID            datetime  error1count  error2count  error3count  \\\n",
       "7           1 2015-01-02 06:00:00          0.0          0.0          0.0   \n",
       "8           1 2015-01-02 09:00:00          0.0          0.0          0.0   \n",
       "9           1 2015-01-02 12:00:00          0.0          0.0          0.0   \n",
       "10          1 2015-01-02 15:00:00          0.0          0.0          0.0   \n",
       "11          1 2015-01-02 18:00:00          0.0          0.0          0.0   \n",
       "\n",
       "    error4count  error5count  \n",
       "7           0.0          0.0  \n",
       "8           0.0          0.0  \n",
       "9           0.0          0.0  \n",
       "10          0.0          0.0  \n",
       "11          0.0          0.0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_count.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Days Since Last Replacement from Maintenance\n",
    "A crucial data set in this example is the maintenance records which contain the information of component replacement records. Possible features from this data set can be, for example, the number of replacements of each component in the last 3 months to incorporate the frequency of replacements. However, more relevent information would be to calculate how long it has been since a component is last replaced as that would be expected to correlate better with component failures since the longer a component is used, the more degradation should be expected.\n",
    "\n",
    "As a side note, creating lagging features from maintenance data is not as straightforward as for telemetry and errors, so the features from this data are generated in a more custom way. This type of ad-hoc feature engineering is very common in predictive maintenance since domain knowledge plays a big role in understanding the predictors of a problem. In the following, the days since last component replacement are calculated for each component type as features from the maintenance data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "      <td>876100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>53.525185</td>\n",
       "      <td>51.540806</td>\n",
       "      <td>52.725962</td>\n",
       "      <td>53.834191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.866087</td>\n",
       "      <td>62.491679</td>\n",
       "      <td>59.269254</td>\n",
       "      <td>58.873114</td>\n",
       "      <td>59.707978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.750000</td>\n",
       "      <td>13.291667</td>\n",
       "      <td>12.125000</td>\n",
       "      <td>13.125000</td>\n",
       "      <td>13.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.500000</td>\n",
       "      <td>32.791667</td>\n",
       "      <td>29.666667</td>\n",
       "      <td>32.291667</td>\n",
       "      <td>32.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.250000</td>\n",
       "      <td>68.708333</td>\n",
       "      <td>66.541667</td>\n",
       "      <td>67.333333</td>\n",
       "      <td>70.458333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>491.958333</td>\n",
       "      <td>348.958333</td>\n",
       "      <td>370.958333</td>\n",
       "      <td>394.958333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           machineID          comp1          comp2          comp3  \\\n",
       "count  876100.000000  876100.000000  876100.000000  876100.000000   \n",
       "mean       50.500000      53.525185      51.540806      52.725962   \n",
       "std        28.866087      62.491679      59.269254      58.873114   \n",
       "min         1.000000       0.000000       0.000000       0.000000   \n",
       "25%        25.750000      13.291667      12.125000      13.125000   \n",
       "50%        50.500000      32.791667      29.666667      32.291667   \n",
       "75%        75.250000      68.708333      66.541667      67.333333   \n",
       "max       100.000000     491.958333     348.958333     370.958333   \n",
       "\n",
       "               comp4  \n",
       "count  876100.000000  \n",
       "mean       53.834191  \n",
       "std        59.707978  \n",
       "min         0.000000  \n",
       "25%        13.000000  \n",
       "50%        32.500000  \n",
       "75%        70.458333  \n",
       "max       394.958333  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "# create a column for each error type\n",
    "comp_rep = pd.get_dummies(maint.set_index('datetime')).reset_index()\n",
    "comp_rep.columns = ['datetime', 'machineID', 'comp1', 'comp2', 'comp3', 'comp4']\n",
    "\n",
    "# combine repairs for a given machine in a given hour\n",
    "comp_rep = comp_rep.groupby(['machineID', 'datetime']).sum().reset_index()\n",
    "\n",
    "# add timepoints where no components were replaced\n",
    "comp_rep = telemetry[['datetime', 'machineID']].merge(comp_rep,\n",
    "                                                      on=['datetime', 'machineID'],\n",
    "                                                      how='outer').fillna(0).sort_values(by=['machineID', 'datetime'])\n",
    "\n",
    "components = ['comp1', 'comp2', 'comp3', 'comp4']\n",
    "for comp in components:\n",
    "    # convert indicator to most recent date of component change\n",
    "    comp_rep.loc[comp_rep[comp] < 1, comp] = None\n",
    "    comp_rep.loc[-comp_rep[comp].isnull(), comp] = comp_rep.loc[-comp_rep[comp].isnull(), 'datetime']\n",
    "    \n",
    "    # forward-fill the most-recent date of component change\n",
    "    comp_rep[comp] = comp_rep[comp].fillna(method='ffill')\n",
    "\n",
    "# remove dates in 2014 (may have NaN or future component change dates)    \n",
    "comp_rep = comp_rep.loc[comp_rep['datetime'] > pd.to_datetime('2015-01-01')]\n",
    "\n",
    "# replace dates of most recent component change with days since most recent component change\n",
    "for comp in components:\n",
    "    comp_rep[comp] = (comp_rep['datetime'] - comp_rep[comp]) / np.timedelta64(1, 'D')\n",
    "    \n",
    "comp_rep.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>machineID</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01 06:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>214.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>169.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01 07:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>19.041667</td>\n",
       "      <td>214.041667</td>\n",
       "      <td>154.041667</td>\n",
       "      <td>169.041667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01 08:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>19.083333</td>\n",
       "      <td>214.083333</td>\n",
       "      <td>154.083333</td>\n",
       "      <td>169.083333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01 09:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>19.125000</td>\n",
       "      <td>214.125000</td>\n",
       "      <td>154.125000</td>\n",
       "      <td>169.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01 10:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>19.166667</td>\n",
       "      <td>214.166667</td>\n",
       "      <td>154.166667</td>\n",
       "      <td>169.166667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime  machineID      comp1       comp2       comp3  \\\n",
       "0 2015-01-01 06:00:00          1  19.000000  214.000000  154.000000   \n",
       "1 2015-01-01 07:00:00          1  19.041667  214.041667  154.041667   \n",
       "2 2015-01-01 08:00:00          1  19.083333  214.083333  154.083333   \n",
       "3 2015-01-01 09:00:00          1  19.125000  214.125000  154.125000   \n",
       "4 2015-01-01 10:00:00          1  19.166667  214.166667  154.166667   \n",
       "\n",
       "        comp4  \n",
       "0  169.000000  \n",
       "1  169.041667  \n",
       "2  169.083333  \n",
       "3  169.125000  \n",
       "4  169.166667  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comp_rep.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Machine Features\n",
    "The machine features can be used without further modification. These include descriptive information about the type of each machine and its age (number of years in service). If the age information had been recorded as a \"first use date\" for each machine, a transformation would have been necessary to turn those into a numeric values indicating the years in service.\n",
    "\n",
    "Lastly, we merge all the feature data sets we created earlier to get the final feature matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>voltmean_24h</th>\n",
       "      <th>rotatemean_24h</th>\n",
       "      <th>pressuremean_24h</th>\n",
       "      <th>vibrationmean_24h</th>\n",
       "      <th>voltsd_24h</th>\n",
       "      <th>rotatesd_24h</th>\n",
       "      <th>pressuresd_24h</th>\n",
       "      <th>vibrationsd_24h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 06:00:00</td>\n",
       "      <td>180.133784</td>\n",
       "      <td>440.608320</td>\n",
       "      <td>94.137969</td>\n",
       "      <td>41.551544</td>\n",
       "      <td>21.322735</td>\n",
       "      <td>48.770512</td>\n",
       "      <td>2.135684</td>\n",
       "      <td>10.037208</td>\n",
       "      <td>169.733809</td>\n",
       "      <td>445.179865</td>\n",
       "      <td>96.797113</td>\n",
       "      <td>40.385160</td>\n",
       "      <td>15.726970</td>\n",
       "      <td>39.648116</td>\n",
       "      <td>11.904700</td>\n",
       "      <td>5.601191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 09:00:00</td>\n",
       "      <td>176.364293</td>\n",
       "      <td>439.349655</td>\n",
       "      <td>101.553209</td>\n",
       "      <td>36.105580</td>\n",
       "      <td>18.952210</td>\n",
       "      <td>51.329636</td>\n",
       "      <td>13.789279</td>\n",
       "      <td>6.737739</td>\n",
       "      <td>170.614862</td>\n",
       "      <td>446.364859</td>\n",
       "      <td>96.849785</td>\n",
       "      <td>39.736826</td>\n",
       "      <td>15.635083</td>\n",
       "      <td>41.828592</td>\n",
       "      <td>11.326412</td>\n",
       "      <td>5.583521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 12:00:00</td>\n",
       "      <td>160.384568</td>\n",
       "      <td>424.385316</td>\n",
       "      <td>99.598722</td>\n",
       "      <td>36.094637</td>\n",
       "      <td>13.047080</td>\n",
       "      <td>13.702496</td>\n",
       "      <td>9.988609</td>\n",
       "      <td>1.639962</td>\n",
       "      <td>169.893965</td>\n",
       "      <td>447.009407</td>\n",
       "      <td>97.715600</td>\n",
       "      <td>39.498374</td>\n",
       "      <td>13.995465</td>\n",
       "      <td>40.843882</td>\n",
       "      <td>11.036546</td>\n",
       "      <td>5.561553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 15:00:00</td>\n",
       "      <td>170.472461</td>\n",
       "      <td>442.933997</td>\n",
       "      <td>102.380586</td>\n",
       "      <td>40.483002</td>\n",
       "      <td>16.642354</td>\n",
       "      <td>56.290447</td>\n",
       "      <td>3.305739</td>\n",
       "      <td>8.854145</td>\n",
       "      <td>171.243444</td>\n",
       "      <td>444.233563</td>\n",
       "      <td>96.666060</td>\n",
       "      <td>40.229370</td>\n",
       "      <td>13.100364</td>\n",
       "      <td>43.409841</td>\n",
       "      <td>10.972862</td>\n",
       "      <td>6.068674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 18:00:00</td>\n",
       "      <td>163.263806</td>\n",
       "      <td>468.937558</td>\n",
       "      <td>102.726648</td>\n",
       "      <td>40.921802</td>\n",
       "      <td>17.424688</td>\n",
       "      <td>38.680380</td>\n",
       "      <td>9.105775</td>\n",
       "      <td>3.060781</td>\n",
       "      <td>170.792486</td>\n",
       "      <td>448.440437</td>\n",
       "      <td>95.766838</td>\n",
       "      <td>40.055214</td>\n",
       "      <td>13.808489</td>\n",
       "      <td>43.742304</td>\n",
       "      <td>10.988704</td>\n",
       "      <td>7.286129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 21:00:00</td>\n",
       "      <td>163.278466</td>\n",
       "      <td>446.493166</td>\n",
       "      <td>104.387585</td>\n",
       "      <td>38.068116</td>\n",
       "      <td>21.580492</td>\n",
       "      <td>41.380958</td>\n",
       "      <td>20.725597</td>\n",
       "      <td>6.932127</td>\n",
       "      <td>170.556674</td>\n",
       "      <td>452.267095</td>\n",
       "      <td>98.065860</td>\n",
       "      <td>40.033247</td>\n",
       "      <td>14.187985</td>\n",
       "      <td>40.676672</td>\n",
       "      <td>11.942227</td>\n",
       "      <td>8.723238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 00:00:00</td>\n",
       "      <td>172.191198</td>\n",
       "      <td>434.214692</td>\n",
       "      <td>93.747282</td>\n",
       "      <td>39.716482</td>\n",
       "      <td>16.369836</td>\n",
       "      <td>14.636041</td>\n",
       "      <td>18.817326</td>\n",
       "      <td>3.426997</td>\n",
       "      <td>168.460525</td>\n",
       "      <td>451.031783</td>\n",
       "      <td>99.273286</td>\n",
       "      <td>38.903462</td>\n",
       "      <td>13.707794</td>\n",
       "      <td>40.509184</td>\n",
       "      <td>10.141026</td>\n",
       "      <td>8.634082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 03:00:00</td>\n",
       "      <td>175.210027</td>\n",
       "      <td>504.845430</td>\n",
       "      <td>108.512153</td>\n",
       "      <td>37.763933</td>\n",
       "      <td>5.991921</td>\n",
       "      <td>16.062702</td>\n",
       "      <td>6.382608</td>\n",
       "      <td>3.449468</td>\n",
       "      <td>169.772951</td>\n",
       "      <td>447.502464</td>\n",
       "      <td>99.005946</td>\n",
       "      <td>39.389725</td>\n",
       "      <td>11.818603</td>\n",
       "      <td>44.468516</td>\n",
       "      <td>9.444955</td>\n",
       "      <td>8.332673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 06:00:00</td>\n",
       "      <td>181.690108</td>\n",
       "      <td>472.783187</td>\n",
       "      <td>93.395164</td>\n",
       "      <td>38.621099</td>\n",
       "      <td>11.514450</td>\n",
       "      <td>47.880443</td>\n",
       "      <td>2.177029</td>\n",
       "      <td>7.670520</td>\n",
       "      <td>170.900562</td>\n",
       "      <td>453.864597</td>\n",
       "      <td>100.877342</td>\n",
       "      <td>38.696225</td>\n",
       "      <td>12.069391</td>\n",
       "      <td>46.669661</td>\n",
       "      <td>8.609526</td>\n",
       "      <td>8.089348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 09:00:00</td>\n",
       "      <td>172.382935</td>\n",
       "      <td>505.141261</td>\n",
       "      <td>98.524373</td>\n",
       "      <td>49.965572</td>\n",
       "      <td>7.065150</td>\n",
       "      <td>56.849540</td>\n",
       "      <td>5.230039</td>\n",
       "      <td>2.687565</td>\n",
       "      <td>169.533156</td>\n",
       "      <td>454.785072</td>\n",
       "      <td>100.050567</td>\n",
       "      <td>39.449734</td>\n",
       "      <td>12.755234</td>\n",
       "      <td>44.016114</td>\n",
       "      <td>9.893704</td>\n",
       "      <td>7.013132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 12:00:00</td>\n",
       "      <td>174.303858</td>\n",
       "      <td>436.182686</td>\n",
       "      <td>94.092681</td>\n",
       "      <td>50.999589</td>\n",
       "      <td>19.017196</td>\n",
       "      <td>26.420163</td>\n",
       "      <td>7.661944</td>\n",
       "      <td>3.516734</td>\n",
       "      <td>170.866013</td>\n",
       "      <td>463.871291</td>\n",
       "      <td>99.360632</td>\n",
       "      <td>40.766639</td>\n",
       "      <td>12.848646</td>\n",
       "      <td>45.090576</td>\n",
       "      <td>9.846662</td>\n",
       "      <td>5.888262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 15:00:00</td>\n",
       "      <td>176.246348</td>\n",
       "      <td>451.646684</td>\n",
       "      <td>98.102389</td>\n",
       "      <td>59.198241</td>\n",
       "      <td>12.572504</td>\n",
       "      <td>31.574383</td>\n",
       "      <td>15.559351</td>\n",
       "      <td>6.562087</td>\n",
       "      <td>171.041651</td>\n",
       "      <td>463.701291</td>\n",
       "      <td>98.965877</td>\n",
       "      <td>42.396850</td>\n",
       "      <td>14.968351</td>\n",
       "      <td>37.088898</td>\n",
       "      <td>10.133452</td>\n",
       "      <td>5.702356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 18:00:00</td>\n",
       "      <td>158.433533</td>\n",
       "      <td>453.900213</td>\n",
       "      <td>98.878129</td>\n",
       "      <td>46.851925</td>\n",
       "      <td>5.136952</td>\n",
       "      <td>21.216569</td>\n",
       "      <td>11.400650</td>\n",
       "      <td>2.688559</td>\n",
       "      <td>171.244533</td>\n",
       "      <td>464.320613</td>\n",
       "      <td>98.853189</td>\n",
       "      <td>44.608814</td>\n",
       "      <td>17.058217</td>\n",
       "      <td>36.617908</td>\n",
       "      <td>9.867174</td>\n",
       "      <td>5.743753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-03 21:00:00</td>\n",
       "      <td>162.387954</td>\n",
       "      <td>454.140377</td>\n",
       "      <td>92.651129</td>\n",
       "      <td>54.261635</td>\n",
       "      <td>4.563331</td>\n",
       "      <td>57.747656</td>\n",
       "      <td>4.754203</td>\n",
       "      <td>5.118076</td>\n",
       "      <td>171.385039</td>\n",
       "      <td>459.937314</td>\n",
       "      <td>97.292157</td>\n",
       "      <td>45.284751</td>\n",
       "      <td>18.405763</td>\n",
       "      <td>35.819938</td>\n",
       "      <td>9.743769</td>\n",
       "      <td>5.246435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 00:00:00</td>\n",
       "      <td>174.243192</td>\n",
       "      <td>394.998095</td>\n",
       "      <td>99.829845</td>\n",
       "      <td>46.930738</td>\n",
       "      <td>6.268730</td>\n",
       "      <td>29.167663</td>\n",
       "      <td>10.564287</td>\n",
       "      <td>6.822855</td>\n",
       "      <td>171.880633</td>\n",
       "      <td>461.437128</td>\n",
       "      <td>96.786742</td>\n",
       "      <td>47.311018</td>\n",
       "      <td>18.249831</td>\n",
       "      <td>42.055638</td>\n",
       "      <td>10.961128</td>\n",
       "      <td>5.093464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 03:00:00</td>\n",
       "      <td>176.443361</td>\n",
       "      <td>459.528820</td>\n",
       "      <td>111.855296</td>\n",
       "      <td>55.296056</td>\n",
       "      <td>16.330285</td>\n",
       "      <td>20.602657</td>\n",
       "      <td>7.064583</td>\n",
       "      <td>4.651468</td>\n",
       "      <td>172.513202</td>\n",
       "      <td>456.429165</td>\n",
       "      <td>97.742700</td>\n",
       "      <td>48.416442</td>\n",
       "      <td>19.141287</td>\n",
       "      <td>37.018824</td>\n",
       "      <td>10.642956</td>\n",
       "      <td>4.618287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 06:00:00</td>\n",
       "      <td>186.092896</td>\n",
       "      <td>451.641253</td>\n",
       "      <td>107.989359</td>\n",
       "      <td>55.308074</td>\n",
       "      <td>13.489090</td>\n",
       "      <td>62.185045</td>\n",
       "      <td>5.118176</td>\n",
       "      <td>4.904365</td>\n",
       "      <td>172.686245</td>\n",
       "      <td>453.387589</td>\n",
       "      <td>99.304019</td>\n",
       "      <td>51.158654</td>\n",
       "      <td>18.887033</td>\n",
       "      <td>36.997459</td>\n",
       "      <td>11.042775</td>\n",
       "      <td>5.195423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 09:00:00</td>\n",
       "      <td>166.281848</td>\n",
       "      <td>453.787824</td>\n",
       "      <td>106.187582</td>\n",
       "      <td>51.990080</td>\n",
       "      <td>24.276228</td>\n",
       "      <td>23.621315</td>\n",
       "      <td>11.176731</td>\n",
       "      <td>3.394073</td>\n",
       "      <td>172.042428</td>\n",
       "      <td>450.418764</td>\n",
       "      <td>100.284484</td>\n",
       "      <td>52.153213</td>\n",
       "      <td>20.837993</td>\n",
       "      <td>34.051825</td>\n",
       "      <td>9.654971</td>\n",
       "      <td>5.066388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 12:00:00</td>\n",
       "      <td>175.412103</td>\n",
       "      <td>445.450581</td>\n",
       "      <td>100.887363</td>\n",
       "      <td>54.251534</td>\n",
       "      <td>34.918687</td>\n",
       "      <td>11.001625</td>\n",
       "      <td>10.580336</td>\n",
       "      <td>2.921501</td>\n",
       "      <td>171.219623</td>\n",
       "      <td>443.802134</td>\n",
       "      <td>102.358897</td>\n",
       "      <td>52.854420</td>\n",
       "      <td>21.298322</td>\n",
       "      <td>36.054002</td>\n",
       "      <td>9.885781</td>\n",
       "      <td>5.246894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 15:00:00</td>\n",
       "      <td>157.347716</td>\n",
       "      <td>451.882075</td>\n",
       "      <td>101.289380</td>\n",
       "      <td>48.602686</td>\n",
       "      <td>24.617739</td>\n",
       "      <td>28.950883</td>\n",
       "      <td>9.966729</td>\n",
       "      <td>2.356486</td>\n",
       "      <td>172.013443</td>\n",
       "      <td>444.882018</td>\n",
       "      <td>102.578580</td>\n",
       "      <td>52.789794</td>\n",
       "      <td>21.200183</td>\n",
       "      <td>38.544116</td>\n",
       "      <td>10.429692</td>\n",
       "      <td>7.192434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 18:00:00</td>\n",
       "      <td>176.450550</td>\n",
       "      <td>446.033068</td>\n",
       "      <td>84.521555</td>\n",
       "      <td>47.638836</td>\n",
       "      <td>8.071400</td>\n",
       "      <td>76.511343</td>\n",
       "      <td>2.636879</td>\n",
       "      <td>4.108621</td>\n",
       "      <td>170.176321</td>\n",
       "      <td>445.069594</td>\n",
       "      <td>102.359939</td>\n",
       "      <td>51.518719</td>\n",
       "      <td>18.814679</td>\n",
       "      <td>40.547527</td>\n",
       "      <td>11.133170</td>\n",
       "      <td>7.556313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 21:00:00</td>\n",
       "      <td>190.325814</td>\n",
       "      <td>422.692565</td>\n",
       "      <td>107.393234</td>\n",
       "      <td>49.552856</td>\n",
       "      <td>8.390777</td>\n",
       "      <td>7.176553</td>\n",
       "      <td>4.262645</td>\n",
       "      <td>7.598552</td>\n",
       "      <td>172.932248</td>\n",
       "      <td>444.618018</td>\n",
       "      <td>101.425508</td>\n",
       "      <td>52.135905</td>\n",
       "      <td>16.762469</td>\n",
       "      <td>49.373445</td>\n",
       "      <td>10.443534</td>\n",
       "      <td>8.545739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 00:00:00</td>\n",
       "      <td>169.985134</td>\n",
       "      <td>458.929418</td>\n",
       "      <td>91.494362</td>\n",
       "      <td>54.882021</td>\n",
       "      <td>9.451483</td>\n",
       "      <td>12.052752</td>\n",
       "      <td>3.685906</td>\n",
       "      <td>6.621183</td>\n",
       "      <td>175.121131</td>\n",
       "      <td>443.916392</td>\n",
       "      <td>102.130179</td>\n",
       "      <td>51.653294</td>\n",
       "      <td>17.435946</td>\n",
       "      <td>43.819375</td>\n",
       "      <td>10.830449</td>\n",
       "      <td>8.809530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 03:00:00</td>\n",
       "      <td>149.082619</td>\n",
       "      <td>412.180336</td>\n",
       "      <td>93.509785</td>\n",
       "      <td>54.386079</td>\n",
       "      <td>19.075952</td>\n",
       "      <td>30.715081</td>\n",
       "      <td>3.090266</td>\n",
       "      <td>6.530610</td>\n",
       "      <td>173.407255</td>\n",
       "      <td>446.265950</td>\n",
       "      <td>100.874614</td>\n",
       "      <td>52.529450</td>\n",
       "      <td>16.661364</td>\n",
       "      <td>47.266846</td>\n",
       "      <td>11.225440</td>\n",
       "      <td>9.068824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 06:00:00</td>\n",
       "      <td>185.782709</td>\n",
       "      <td>439.531288</td>\n",
       "      <td>99.413660</td>\n",
       "      <td>51.558082</td>\n",
       "      <td>14.495664</td>\n",
       "      <td>45.663743</td>\n",
       "      <td>4.289212</td>\n",
       "      <td>7.330397</td>\n",
       "      <td>170.757841</td>\n",
       "      <td>440.958228</td>\n",
       "      <td>98.716746</td>\n",
       "      <td>51.746749</td>\n",
       "      <td>17.863934</td>\n",
       "      <td>44.895080</td>\n",
       "      <td>10.675981</td>\n",
       "      <td>7.475304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 09:00:00</td>\n",
       "      <td>169.084809</td>\n",
       "      <td>463.433785</td>\n",
       "      <td>107.678774</td>\n",
       "      <td>41.710336</td>\n",
       "      <td>12.245544</td>\n",
       "      <td>61.759107</td>\n",
       "      <td>4.400233</td>\n",
       "      <td>9.750017</td>\n",
       "      <td>171.929104</td>\n",
       "      <td>443.448775</td>\n",
       "      <td>98.675590</td>\n",
       "      <td>51.780445</td>\n",
       "      <td>15.139300</td>\n",
       "      <td>45.766081</td>\n",
       "      <td>10.959268</td>\n",
       "      <td>6.855778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 12:00:00</td>\n",
       "      <td>165.518790</td>\n",
       "      <td>449.743255</td>\n",
       "      <td>110.377851</td>\n",
       "      <td>38.952082</td>\n",
       "      <td>23.170638</td>\n",
       "      <td>45.762142</td>\n",
       "      <td>14.009473</td>\n",
       "      <td>0.797364</td>\n",
       "      <td>170.908522</td>\n",
       "      <td>443.069042</td>\n",
       "      <td>98.830333</td>\n",
       "      <td>49.679550</td>\n",
       "      <td>13.985517</td>\n",
       "      <td>42.542001</td>\n",
       "      <td>11.050133</td>\n",
       "      <td>4.842842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 15:00:00</td>\n",
       "      <td>175.989642</td>\n",
       "      <td>419.863490</td>\n",
       "      <td>112.571146</td>\n",
       "      <td>41.514254</td>\n",
       "      <td>4.028327</td>\n",
       "      <td>20.148499</td>\n",
       "      <td>5.862629</td>\n",
       "      <td>9.702498</td>\n",
       "      <td>170.416326</td>\n",
       "      <td>443.555122</td>\n",
       "      <td>100.221328</td>\n",
       "      <td>48.481038</td>\n",
       "      <td>13.344630</td>\n",
       "      <td>39.327146</td>\n",
       "      <td>10.268539</td>\n",
       "      <td>4.884344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 18:00:00</td>\n",
       "      <td>188.576444</td>\n",
       "      <td>487.336742</td>\n",
       "      <td>88.967297</td>\n",
       "      <td>36.571052</td>\n",
       "      <td>8.278605</td>\n",
       "      <td>76.534023</td>\n",
       "      <td>11.892088</td>\n",
       "      <td>1.945849</td>\n",
       "      <td>173.315167</td>\n",
       "      <td>444.049581</td>\n",
       "      <td>101.633306</td>\n",
       "      <td>47.279992</td>\n",
       "      <td>15.793146</td>\n",
       "      <td>42.984028</td>\n",
       "      <td>10.006300</td>\n",
       "      <td>4.637101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 21:00:00</td>\n",
       "      <td>166.681364</td>\n",
       "      <td>481.685320</td>\n",
       "      <td>104.154110</td>\n",
       "      <td>38.662638</td>\n",
       "      <td>11.957697</td>\n",
       "      <td>25.052743</td>\n",
       "      <td>11.999161</td>\n",
       "      <td>4.804263</td>\n",
       "      <td>173.743459</td>\n",
       "      <td>446.505202</td>\n",
       "      <td>100.540356</td>\n",
       "      <td>45.527290</td>\n",
       "      <td>16.132288</td>\n",
       "      <td>40.754154</td>\n",
       "      <td>9.744855</td>\n",
       "      <td>4.591048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291370</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 06:00:00</td>\n",
       "      <td>165.259415</td>\n",
       "      <td>432.364050</td>\n",
       "      <td>96.793097</td>\n",
       "      <td>38.697882</td>\n",
       "      <td>16.715588</td>\n",
       "      <td>9.197585</td>\n",
       "      <td>11.016730</td>\n",
       "      <td>9.167743</td>\n",
       "      <td>169.115085</td>\n",
       "      <td>459.202414</td>\n",
       "      <td>99.099044</td>\n",
       "      <td>39.342719</td>\n",
       "      <td>12.889019</td>\n",
       "      <td>60.409151</td>\n",
       "      <td>11.549081</td>\n",
       "      <td>5.671215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291371</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 09:00:00</td>\n",
       "      <td>185.907346</td>\n",
       "      <td>465.062411</td>\n",
       "      <td>94.161434</td>\n",
       "      <td>36.156060</td>\n",
       "      <td>22.822289</td>\n",
       "      <td>64.351154</td>\n",
       "      <td>6.469484</td>\n",
       "      <td>1.656610</td>\n",
       "      <td>168.838507</td>\n",
       "      <td>455.101759</td>\n",
       "      <td>98.960206</td>\n",
       "      <td>38.277576</td>\n",
       "      <td>13.917591</td>\n",
       "      <td>56.810848</td>\n",
       "      <td>11.118412</td>\n",
       "      <td>6.061118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291372</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 12:00:00</td>\n",
       "      <td>167.546991</td>\n",
       "      <td>448.203119</td>\n",
       "      <td>99.383591</td>\n",
       "      <td>39.659572</td>\n",
       "      <td>2.573507</td>\n",
       "      <td>84.299208</td>\n",
       "      <td>2.490792</td>\n",
       "      <td>2.252574</td>\n",
       "      <td>169.223690</td>\n",
       "      <td>463.630715</td>\n",
       "      <td>99.296474</td>\n",
       "      <td>38.406915</td>\n",
       "      <td>14.611939</td>\n",
       "      <td>56.534099</td>\n",
       "      <td>8.553177</td>\n",
       "      <td>5.893650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291373</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 15:00:00</td>\n",
       "      <td>175.468904</td>\n",
       "      <td>441.861941</td>\n",
       "      <td>105.814802</td>\n",
       "      <td>38.788653</td>\n",
       "      <td>9.104554</td>\n",
       "      <td>48.615069</td>\n",
       "      <td>6.004070</td>\n",
       "      <td>3.244295</td>\n",
       "      <td>172.163049</td>\n",
       "      <td>458.787617</td>\n",
       "      <td>100.063674</td>\n",
       "      <td>38.458947</td>\n",
       "      <td>15.232866</td>\n",
       "      <td>49.321412</td>\n",
       "      <td>8.345687</td>\n",
       "      <td>5.292801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291374</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 18:00:00</td>\n",
       "      <td>157.401371</td>\n",
       "      <td>459.332121</td>\n",
       "      <td>93.247465</td>\n",
       "      <td>42.236723</td>\n",
       "      <td>14.711827</td>\n",
       "      <td>45.268580</td>\n",
       "      <td>5.590642</td>\n",
       "      <td>2.204472</td>\n",
       "      <td>173.119397</td>\n",
       "      <td>456.196849</td>\n",
       "      <td>100.114215</td>\n",
       "      <td>39.063775</td>\n",
       "      <td>15.920072</td>\n",
       "      <td>48.742610</td>\n",
       "      <td>7.909495</td>\n",
       "      <td>5.249418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291375</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-02 21:00:00</td>\n",
       "      <td>168.651510</td>\n",
       "      <td>430.056138</td>\n",
       "      <td>104.487324</td>\n",
       "      <td>35.735005</td>\n",
       "      <td>16.328969</td>\n",
       "      <td>45.108180</td>\n",
       "      <td>9.103806</td>\n",
       "      <td>6.093867</td>\n",
       "      <td>168.410263</td>\n",
       "      <td>448.519301</td>\n",
       "      <td>100.049480</td>\n",
       "      <td>39.083797</td>\n",
       "      <td>14.635941</td>\n",
       "      <td>50.313046</td>\n",
       "      <td>7.582133</td>\n",
       "      <td>5.108806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291376</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 00:00:00</td>\n",
       "      <td>168.623762</td>\n",
       "      <td>497.504580</td>\n",
       "      <td>100.682235</td>\n",
       "      <td>40.610939</td>\n",
       "      <td>12.914771</td>\n",
       "      <td>25.775781</td>\n",
       "      <td>11.444951</td>\n",
       "      <td>3.359673</td>\n",
       "      <td>168.849711</td>\n",
       "      <td>450.330382</td>\n",
       "      <td>100.243957</td>\n",
       "      <td>38.469840</td>\n",
       "      <td>14.915283</td>\n",
       "      <td>50.148100</td>\n",
       "      <td>7.928488</td>\n",
       "      <td>5.518228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291377</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 03:00:00</td>\n",
       "      <td>168.537058</td>\n",
       "      <td>441.837105</td>\n",
       "      <td>87.893111</td>\n",
       "      <td>40.076219</td>\n",
       "      <td>23.866338</td>\n",
       "      <td>36.817201</td>\n",
       "      <td>16.820180</td>\n",
       "      <td>0.482728</td>\n",
       "      <td>171.513651</td>\n",
       "      <td>452.462106</td>\n",
       "      <td>98.514174</td>\n",
       "      <td>38.626944</td>\n",
       "      <td>14.750503</td>\n",
       "      <td>44.620373</td>\n",
       "      <td>8.348839</td>\n",
       "      <td>5.182752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291378</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 06:00:00</td>\n",
       "      <td>161.436752</td>\n",
       "      <td>401.579802</td>\n",
       "      <td>90.792431</td>\n",
       "      <td>35.624101</td>\n",
       "      <td>14.429283</td>\n",
       "      <td>85.801834</td>\n",
       "      <td>16.225371</td>\n",
       "      <td>1.396074</td>\n",
       "      <td>169.625319</td>\n",
       "      <td>451.508453</td>\n",
       "      <td>97.368816</td>\n",
       "      <td>38.762103</td>\n",
       "      <td>16.055332</td>\n",
       "      <td>49.218444</td>\n",
       "      <td>9.546008</td>\n",
       "      <td>5.020598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291379</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 09:00:00</td>\n",
       "      <td>188.559785</td>\n",
       "      <td>491.929571</td>\n",
       "      <td>102.821662</td>\n",
       "      <td>40.034460</td>\n",
       "      <td>12.668164</td>\n",
       "      <td>41.768856</td>\n",
       "      <td>9.448383</td>\n",
       "      <td>8.454640</td>\n",
       "      <td>167.950194</td>\n",
       "      <td>453.659486</td>\n",
       "      <td>97.292065</td>\n",
       "      <td>38.635159</td>\n",
       "      <td>15.942467</td>\n",
       "      <td>59.763569</td>\n",
       "      <td>9.431896</td>\n",
       "      <td>4.476863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291380</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 12:00:00</td>\n",
       "      <td>164.149847</td>\n",
       "      <td>466.463280</td>\n",
       "      <td>100.447025</td>\n",
       "      <td>40.694147</td>\n",
       "      <td>16.460088</td>\n",
       "      <td>31.589198</td>\n",
       "      <td>10.188202</td>\n",
       "      <td>5.086589</td>\n",
       "      <td>168.530246</td>\n",
       "      <td>448.103742</td>\n",
       "      <td>97.730594</td>\n",
       "      <td>39.228047</td>\n",
       "      <td>16.122100</td>\n",
       "      <td>57.394487</td>\n",
       "      <td>9.171658</td>\n",
       "      <td>4.866485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291381</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 15:00:00</td>\n",
       "      <td>184.727094</td>\n",
       "      <td>487.140570</td>\n",
       "      <td>100.860907</td>\n",
       "      <td>36.430834</td>\n",
       "      <td>12.401664</td>\n",
       "      <td>57.508894</td>\n",
       "      <td>18.394116</td>\n",
       "      <td>7.153156</td>\n",
       "      <td>169.765458</td>\n",
       "      <td>458.759394</td>\n",
       "      <td>98.775532</td>\n",
       "      <td>38.935773</td>\n",
       "      <td>16.994700</td>\n",
       "      <td>55.665652</td>\n",
       "      <td>9.630112</td>\n",
       "      <td>4.896106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291382</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 18:00:00</td>\n",
       "      <td>171.585447</td>\n",
       "      <td>479.021432</td>\n",
       "      <td>101.846824</td>\n",
       "      <td>49.115340</td>\n",
       "      <td>13.318878</td>\n",
       "      <td>40.471453</td>\n",
       "      <td>10.951704</td>\n",
       "      <td>0.649491</td>\n",
       "      <td>169.818254</td>\n",
       "      <td>460.880691</td>\n",
       "      <td>97.926823</td>\n",
       "      <td>39.295892</td>\n",
       "      <td>15.077105</td>\n",
       "      <td>56.097762</td>\n",
       "      <td>9.818266</td>\n",
       "      <td>4.919802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291383</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-03 21:00:00</td>\n",
       "      <td>162.144446</td>\n",
       "      <td>404.865418</td>\n",
       "      <td>98.384468</td>\n",
       "      <td>35.389856</td>\n",
       "      <td>22.516178</td>\n",
       "      <td>36.216388</td>\n",
       "      <td>11.492089</td>\n",
       "      <td>7.269283</td>\n",
       "      <td>172.199254</td>\n",
       "      <td>459.599763</td>\n",
       "      <td>98.365107</td>\n",
       "      <td>39.964091</td>\n",
       "      <td>14.414241</td>\n",
       "      <td>59.222526</td>\n",
       "      <td>10.569736</td>\n",
       "      <td>4.693342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291384</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 00:00:00</td>\n",
       "      <td>166.584930</td>\n",
       "      <td>437.980304</td>\n",
       "      <td>104.019479</td>\n",
       "      <td>43.766793</td>\n",
       "      <td>22.109109</td>\n",
       "      <td>65.256390</td>\n",
       "      <td>12.621841</td>\n",
       "      <td>1.793100</td>\n",
       "      <td>170.812061</td>\n",
       "      <td>453.681180</td>\n",
       "      <td>98.994157</td>\n",
       "      <td>40.013985</td>\n",
       "      <td>13.522301</td>\n",
       "      <td>61.197614</td>\n",
       "      <td>10.103349</td>\n",
       "      <td>4.286568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291385</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 03:00:00</td>\n",
       "      <td>173.182209</td>\n",
       "      <td>452.585928</td>\n",
       "      <td>106.572235</td>\n",
       "      <td>40.534601</td>\n",
       "      <td>16.930726</td>\n",
       "      <td>38.788180</td>\n",
       "      <td>10.747137</td>\n",
       "      <td>6.510290</td>\n",
       "      <td>170.439828</td>\n",
       "      <td>455.036021</td>\n",
       "      <td>99.830401</td>\n",
       "      <td>40.129183</td>\n",
       "      <td>14.551122</td>\n",
       "      <td>65.430315</td>\n",
       "      <td>9.389132</td>\n",
       "      <td>4.484563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291386</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 06:00:00</td>\n",
       "      <td>155.554082</td>\n",
       "      <td>464.175866</td>\n",
       "      <td>102.615428</td>\n",
       "      <td>36.003311</td>\n",
       "      <td>7.678204</td>\n",
       "      <td>24.248612</td>\n",
       "      <td>6.064152</td>\n",
       "      <td>5.007039</td>\n",
       "      <td>172.264492</td>\n",
       "      <td>453.800429</td>\n",
       "      <td>102.163329</td>\n",
       "      <td>40.050108</td>\n",
       "      <td>13.103400</td>\n",
       "      <td>62.190761</td>\n",
       "      <td>9.128160</td>\n",
       "      <td>4.502950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291387</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 09:00:00</td>\n",
       "      <td>163.814555</td>\n",
       "      <td>433.614467</td>\n",
       "      <td>114.798438</td>\n",
       "      <td>36.454615</td>\n",
       "      <td>5.259901</td>\n",
       "      <td>40.947023</td>\n",
       "      <td>10.677648</td>\n",
       "      <td>8.252193</td>\n",
       "      <td>170.243198</td>\n",
       "      <td>455.333806</td>\n",
       "      <td>102.290708</td>\n",
       "      <td>40.197103</td>\n",
       "      <td>13.120654</td>\n",
       "      <td>57.910021</td>\n",
       "      <td>9.238228</td>\n",
       "      <td>4.803393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291388</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 12:00:00</td>\n",
       "      <td>169.196188</td>\n",
       "      <td>403.488184</td>\n",
       "      <td>94.199431</td>\n",
       "      <td>39.189491</td>\n",
       "      <td>22.977467</td>\n",
       "      <td>27.176467</td>\n",
       "      <td>9.430194</td>\n",
       "      <td>13.841831</td>\n",
       "      <td>167.765844</td>\n",
       "      <td>451.355029</td>\n",
       "      <td>103.465520</td>\n",
       "      <td>40.252524</td>\n",
       "      <td>13.315211</td>\n",
       "      <td>59.466979</td>\n",
       "      <td>9.471171</td>\n",
       "      <td>4.442337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291389</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 15:00:00</td>\n",
       "      <td>165.814250</td>\n",
       "      <td>446.765824</td>\n",
       "      <td>99.334107</td>\n",
       "      <td>44.464271</td>\n",
       "      <td>1.457549</td>\n",
       "      <td>58.086715</td>\n",
       "      <td>1.622380</td>\n",
       "      <td>3.173978</td>\n",
       "      <td>167.312650</td>\n",
       "      <td>439.435929</td>\n",
       "      <td>102.009695</td>\n",
       "      <td>40.435349</td>\n",
       "      <td>11.768963</td>\n",
       "      <td>60.646198</td>\n",
       "      <td>9.136786</td>\n",
       "      <td>5.149517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291390</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 18:00:00</td>\n",
       "      <td>167.848340</td>\n",
       "      <td>438.393471</td>\n",
       "      <td>90.054937</td>\n",
       "      <td>40.301288</td>\n",
       "      <td>15.958412</td>\n",
       "      <td>40.168662</td>\n",
       "      <td>11.238292</td>\n",
       "      <td>3.633503</td>\n",
       "      <td>166.963446</td>\n",
       "      <td>438.276090</td>\n",
       "      <td>101.637723</td>\n",
       "      <td>40.172602</td>\n",
       "      <td>14.141910</td>\n",
       "      <td>58.372646</td>\n",
       "      <td>8.463179</td>\n",
       "      <td>5.221115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291391</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-04 21:00:00</td>\n",
       "      <td>173.508300</td>\n",
       "      <td>439.917848</td>\n",
       "      <td>93.063793</td>\n",
       "      <td>38.750136</td>\n",
       "      <td>7.633479</td>\n",
       "      <td>44.399657</td>\n",
       "      <td>11.019912</td>\n",
       "      <td>4.952713</td>\n",
       "      <td>165.979125</td>\n",
       "      <td>435.138421</td>\n",
       "      <td>102.005617</td>\n",
       "      <td>39.497908</td>\n",
       "      <td>14.000620</td>\n",
       "      <td>59.820047</td>\n",
       "      <td>6.856021</td>\n",
       "      <td>5.392136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291392</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 00:00:00</td>\n",
       "      <td>182.432617</td>\n",
       "      <td>497.264899</td>\n",
       "      <td>95.443869</td>\n",
       "      <td>40.594815</td>\n",
       "      <td>9.940475</td>\n",
       "      <td>77.558997</td>\n",
       "      <td>4.707020</td>\n",
       "      <td>3.106529</td>\n",
       "      <td>168.142646</td>\n",
       "      <td>447.915202</td>\n",
       "      <td>99.620102</td>\n",
       "      <td>39.635003</td>\n",
       "      <td>14.372707</td>\n",
       "      <td>59.563942</td>\n",
       "      <td>7.988174</td>\n",
       "      <td>5.256284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291393</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 03:00:00</td>\n",
       "      <td>158.783988</td>\n",
       "      <td>438.405164</td>\n",
       "      <td>100.420803</td>\n",
       "      <td>40.153025</td>\n",
       "      <td>10.849108</td>\n",
       "      <td>72.556330</td>\n",
       "      <td>2.576581</td>\n",
       "      <td>4.504970</td>\n",
       "      <td>168.398872</td>\n",
       "      <td>448.148851</td>\n",
       "      <td>99.351099</td>\n",
       "      <td>39.518646</td>\n",
       "      <td>15.763140</td>\n",
       "      <td>57.682117</td>\n",
       "      <td>8.088214</td>\n",
       "      <td>5.301986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291394</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 06:00:00</td>\n",
       "      <td>183.150826</td>\n",
       "      <td>426.209117</td>\n",
       "      <td>98.880399</td>\n",
       "      <td>34.418557</td>\n",
       "      <td>20.539063</td>\n",
       "      <td>29.605169</td>\n",
       "      <td>13.588936</td>\n",
       "      <td>7.168643</td>\n",
       "      <td>168.651040</td>\n",
       "      <td>446.075986</td>\n",
       "      <td>98.741443</td>\n",
       "      <td>39.840623</td>\n",
       "      <td>15.331755</td>\n",
       "      <td>60.839923</td>\n",
       "      <td>7.891711</td>\n",
       "      <td>5.269038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291395</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 09:00:00</td>\n",
       "      <td>188.267556</td>\n",
       "      <td>407.256175</td>\n",
       "      <td>108.931184</td>\n",
       "      <td>36.553233</td>\n",
       "      <td>9.599915</td>\n",
       "      <td>40.722980</td>\n",
       "      <td>1.639521</td>\n",
       "      <td>5.724500</td>\n",
       "      <td>171.826650</td>\n",
       "      <td>441.278667</td>\n",
       "      <td>98.311919</td>\n",
       "      <td>39.196175</td>\n",
       "      <td>16.429023</td>\n",
       "      <td>62.147934</td>\n",
       "      <td>7.475540</td>\n",
       "      <td>5.448962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291396</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 12:00:00</td>\n",
       "      <td>167.859576</td>\n",
       "      <td>465.992407</td>\n",
       "      <td>107.953155</td>\n",
       "      <td>42.708899</td>\n",
       "      <td>14.190347</td>\n",
       "      <td>92.277799</td>\n",
       "      <td>9.577243</td>\n",
       "      <td>0.735339</td>\n",
       "      <td>174.657123</td>\n",
       "      <td>444.147310</td>\n",
       "      <td>98.520388</td>\n",
       "      <td>38.820190</td>\n",
       "      <td>17.019808</td>\n",
       "      <td>64.730136</td>\n",
       "      <td>8.961444</td>\n",
       "      <td>5.833191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291397</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 15:00:00</td>\n",
       "      <td>170.348099</td>\n",
       "      <td>434.234744</td>\n",
       "      <td>104.514343</td>\n",
       "      <td>38.607950</td>\n",
       "      <td>10.232598</td>\n",
       "      <td>49.524471</td>\n",
       "      <td>12.445345</td>\n",
       "      <td>2.596743</td>\n",
       "      <td>173.787879</td>\n",
       "      <td>448.842085</td>\n",
       "      <td>100.028549</td>\n",
       "      <td>39.375067</td>\n",
       "      <td>17.096392</td>\n",
       "      <td>64.718132</td>\n",
       "      <td>9.420879</td>\n",
       "      <td>5.738756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291398</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 18:00:00</td>\n",
       "      <td>152.265370</td>\n",
       "      <td>459.557611</td>\n",
       "      <td>103.536524</td>\n",
       "      <td>40.718426</td>\n",
       "      <td>6.758667</td>\n",
       "      <td>27.051145</td>\n",
       "      <td>12.824247</td>\n",
       "      <td>2.752883</td>\n",
       "      <td>172.496791</td>\n",
       "      <td>442.086577</td>\n",
       "      <td>100.361794</td>\n",
       "      <td>38.943434</td>\n",
       "      <td>15.119775</td>\n",
       "      <td>65.929509</td>\n",
       "      <td>8.836617</td>\n",
       "      <td>6.139142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291399</th>\n",
       "      <td>100</td>\n",
       "      <td>2015-10-05 21:00:00</td>\n",
       "      <td>162.887965</td>\n",
       "      <td>481.415205</td>\n",
       "      <td>96.687092</td>\n",
       "      <td>37.162591</td>\n",
       "      <td>20.541773</td>\n",
       "      <td>55.057460</td>\n",
       "      <td>11.713728</td>\n",
       "      <td>3.539798</td>\n",
       "      <td>170.782713</td>\n",
       "      <td>448.188498</td>\n",
       "      <td>100.794970</td>\n",
       "      <td>38.980896</td>\n",
       "      <td>15.573014</td>\n",
       "      <td>61.859239</td>\n",
       "      <td>9.942610</td>\n",
       "      <td>6.191276</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>290601 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        machineID            datetime  voltmean_3h  rotatemean_3h  \\\n",
       "7               1 2015-01-02 06:00:00   180.133784     440.608320   \n",
       "8               1 2015-01-02 09:00:00   176.364293     439.349655   \n",
       "9               1 2015-01-02 12:00:00   160.384568     424.385316   \n",
       "10              1 2015-01-02 15:00:00   170.472461     442.933997   \n",
       "11              1 2015-01-02 18:00:00   163.263806     468.937558   \n",
       "12              1 2015-01-02 21:00:00   163.278466     446.493166   \n",
       "13              1 2015-01-03 00:00:00   172.191198     434.214692   \n",
       "14              1 2015-01-03 03:00:00   175.210027     504.845430   \n",
       "15              1 2015-01-03 06:00:00   181.690108     472.783187   \n",
       "16              1 2015-01-03 09:00:00   172.382935     505.141261   \n",
       "17              1 2015-01-03 12:00:00   174.303858     436.182686   \n",
       "18              1 2015-01-03 15:00:00   176.246348     451.646684   \n",
       "19              1 2015-01-03 18:00:00   158.433533     453.900213   \n",
       "20              1 2015-01-03 21:00:00   162.387954     454.140377   \n",
       "21              1 2015-01-04 00:00:00   174.243192     394.998095   \n",
       "22              1 2015-01-04 03:00:00   176.443361     459.528820   \n",
       "23              1 2015-01-04 06:00:00   186.092896     451.641253   \n",
       "24              1 2015-01-04 09:00:00   166.281848     453.787824   \n",
       "25              1 2015-01-04 12:00:00   175.412103     445.450581   \n",
       "26              1 2015-01-04 15:00:00   157.347716     451.882075   \n",
       "27              1 2015-01-04 18:00:00   176.450550     446.033068   \n",
       "28              1 2015-01-04 21:00:00   190.325814     422.692565   \n",
       "29              1 2015-01-05 00:00:00   169.985134     458.929418   \n",
       "30              1 2015-01-05 03:00:00   149.082619     412.180336   \n",
       "31              1 2015-01-05 06:00:00   185.782709     439.531288   \n",
       "32              1 2015-01-05 09:00:00   169.084809     463.433785   \n",
       "33              1 2015-01-05 12:00:00   165.518790     449.743255   \n",
       "34              1 2015-01-05 15:00:00   175.989642     419.863490   \n",
       "35              1 2015-01-05 18:00:00   188.576444     487.336742   \n",
       "36              1 2015-01-05 21:00:00   166.681364     481.685320   \n",
       "...           ...                 ...          ...            ...   \n",
       "291370        100 2015-10-02 06:00:00   165.259415     432.364050   \n",
       "291371        100 2015-10-02 09:00:00   185.907346     465.062411   \n",
       "291372        100 2015-10-02 12:00:00   167.546991     448.203119   \n",
       "291373        100 2015-10-02 15:00:00   175.468904     441.861941   \n",
       "291374        100 2015-10-02 18:00:00   157.401371     459.332121   \n",
       "291375        100 2015-10-02 21:00:00   168.651510     430.056138   \n",
       "291376        100 2015-10-03 00:00:00   168.623762     497.504580   \n",
       "291377        100 2015-10-03 03:00:00   168.537058     441.837105   \n",
       "291378        100 2015-10-03 06:00:00   161.436752     401.579802   \n",
       "291379        100 2015-10-03 09:00:00   188.559785     491.929571   \n",
       "291380        100 2015-10-03 12:00:00   164.149847     466.463280   \n",
       "291381        100 2015-10-03 15:00:00   184.727094     487.140570   \n",
       "291382        100 2015-10-03 18:00:00   171.585447     479.021432   \n",
       "291383        100 2015-10-03 21:00:00   162.144446     404.865418   \n",
       "291384        100 2015-10-04 00:00:00   166.584930     437.980304   \n",
       "291385        100 2015-10-04 03:00:00   173.182209     452.585928   \n",
       "291386        100 2015-10-04 06:00:00   155.554082     464.175866   \n",
       "291387        100 2015-10-04 09:00:00   163.814555     433.614467   \n",
       "291388        100 2015-10-04 12:00:00   169.196188     403.488184   \n",
       "291389        100 2015-10-04 15:00:00   165.814250     446.765824   \n",
       "291390        100 2015-10-04 18:00:00   167.848340     438.393471   \n",
       "291391        100 2015-10-04 21:00:00   173.508300     439.917848   \n",
       "291392        100 2015-10-05 00:00:00   182.432617     497.264899   \n",
       "291393        100 2015-10-05 03:00:00   158.783988     438.405164   \n",
       "291394        100 2015-10-05 06:00:00   183.150826     426.209117   \n",
       "291395        100 2015-10-05 09:00:00   188.267556     407.256175   \n",
       "291396        100 2015-10-05 12:00:00   167.859576     465.992407   \n",
       "291397        100 2015-10-05 15:00:00   170.348099     434.234744   \n",
       "291398        100 2015-10-05 18:00:00   152.265370     459.557611   \n",
       "291399        100 2015-10-05 21:00:00   162.887965     481.415205   \n",
       "\n",
       "        pressuremean_3h  vibrationmean_3h  voltsd_3h  rotatesd_3h  \\\n",
       "7             94.137969         41.551544  21.322735    48.770512   \n",
       "8            101.553209         36.105580  18.952210    51.329636   \n",
       "9             99.598722         36.094637  13.047080    13.702496   \n",
       "10           102.380586         40.483002  16.642354    56.290447   \n",
       "11           102.726648         40.921802  17.424688    38.680380   \n",
       "12           104.387585         38.068116  21.580492    41.380958   \n",
       "13            93.747282         39.716482  16.369836    14.636041   \n",
       "14           108.512153         37.763933   5.991921    16.062702   \n",
       "15            93.395164         38.621099  11.514450    47.880443   \n",
       "16            98.524373         49.965572   7.065150    56.849540   \n",
       "17            94.092681         50.999589  19.017196    26.420163   \n",
       "18            98.102389         59.198241  12.572504    31.574383   \n",
       "19            98.878129         46.851925   5.136952    21.216569   \n",
       "20            92.651129         54.261635   4.563331    57.747656   \n",
       "21            99.829845         46.930738   6.268730    29.167663   \n",
       "22           111.855296         55.296056  16.330285    20.602657   \n",
       "23           107.989359         55.308074  13.489090    62.185045   \n",
       "24           106.187582         51.990080  24.276228    23.621315   \n",
       "25           100.887363         54.251534  34.918687    11.001625   \n",
       "26           101.289380         48.602686  24.617739    28.950883   \n",
       "27            84.521555         47.638836   8.071400    76.511343   \n",
       "28           107.393234         49.552856   8.390777     7.176553   \n",
       "29            91.494362         54.882021   9.451483    12.052752   \n",
       "30            93.509785         54.386079  19.075952    30.715081   \n",
       "31            99.413660         51.558082  14.495664    45.663743   \n",
       "32           107.678774         41.710336  12.245544    61.759107   \n",
       "33           110.377851         38.952082  23.170638    45.762142   \n",
       "34           112.571146         41.514254   4.028327    20.148499   \n",
       "35            88.967297         36.571052   8.278605    76.534023   \n",
       "36           104.154110         38.662638  11.957697    25.052743   \n",
       "...                 ...               ...        ...          ...   \n",
       "291370        96.793097         38.697882  16.715588     9.197585   \n",
       "291371        94.161434         36.156060  22.822289    64.351154   \n",
       "291372        99.383591         39.659572   2.573507    84.299208   \n",
       "291373       105.814802         38.788653   9.104554    48.615069   \n",
       "291374        93.247465         42.236723  14.711827    45.268580   \n",
       "291375       104.487324         35.735005  16.328969    45.108180   \n",
       "291376       100.682235         40.610939  12.914771    25.775781   \n",
       "291377        87.893111         40.076219  23.866338    36.817201   \n",
       "291378        90.792431         35.624101  14.429283    85.801834   \n",
       "291379       102.821662         40.034460  12.668164    41.768856   \n",
       "291380       100.447025         40.694147  16.460088    31.589198   \n",
       "291381       100.860907         36.430834  12.401664    57.508894   \n",
       "291382       101.846824         49.115340  13.318878    40.471453   \n",
       "291383        98.384468         35.389856  22.516178    36.216388   \n",
       "291384       104.019479         43.766793  22.109109    65.256390   \n",
       "291385       106.572235         40.534601  16.930726    38.788180   \n",
       "291386       102.615428         36.003311   7.678204    24.248612   \n",
       "291387       114.798438         36.454615   5.259901    40.947023   \n",
       "291388        94.199431         39.189491  22.977467    27.176467   \n",
       "291389        99.334107         44.464271   1.457549    58.086715   \n",
       "291390        90.054937         40.301288  15.958412    40.168662   \n",
       "291391        93.063793         38.750136   7.633479    44.399657   \n",
       "291392        95.443869         40.594815   9.940475    77.558997   \n",
       "291393       100.420803         40.153025  10.849108    72.556330   \n",
       "291394        98.880399         34.418557  20.539063    29.605169   \n",
       "291395       108.931184         36.553233   9.599915    40.722980   \n",
       "291396       107.953155         42.708899  14.190347    92.277799   \n",
       "291397       104.514343         38.607950  10.232598    49.524471   \n",
       "291398       103.536524         40.718426   6.758667    27.051145   \n",
       "291399        96.687092         37.162591  20.541773    55.057460   \n",
       "\n",
       "        pressuresd_3h  vibrationsd_3h  voltmean_24h  rotatemean_24h  \\\n",
       "7            2.135684       10.037208    169.733809      445.179865   \n",
       "8           13.789279        6.737739    170.614862      446.364859   \n",
       "9            9.988609        1.639962    169.893965      447.009407   \n",
       "10           3.305739        8.854145    171.243444      444.233563   \n",
       "11           9.105775        3.060781    170.792486      448.440437   \n",
       "12          20.725597        6.932127    170.556674      452.267095   \n",
       "13          18.817326        3.426997    168.460525      451.031783   \n",
       "14           6.382608        3.449468    169.772951      447.502464   \n",
       "15           2.177029        7.670520    170.900562      453.864597   \n",
       "16           5.230039        2.687565    169.533156      454.785072   \n",
       "17           7.661944        3.516734    170.866013      463.871291   \n",
       "18          15.559351        6.562087    171.041651      463.701291   \n",
       "19          11.400650        2.688559    171.244533      464.320613   \n",
       "20           4.754203        5.118076    171.385039      459.937314   \n",
       "21          10.564287        6.822855    171.880633      461.437128   \n",
       "22           7.064583        4.651468    172.513202      456.429165   \n",
       "23           5.118176        4.904365    172.686245      453.387589   \n",
       "24          11.176731        3.394073    172.042428      450.418764   \n",
       "25          10.580336        2.921501    171.219623      443.802134   \n",
       "26           9.966729        2.356486    172.013443      444.882018   \n",
       "27           2.636879        4.108621    170.176321      445.069594   \n",
       "28           4.262645        7.598552    172.932248      444.618018   \n",
       "29           3.685906        6.621183    175.121131      443.916392   \n",
       "30           3.090266        6.530610    173.407255      446.265950   \n",
       "31           4.289212        7.330397    170.757841      440.958228   \n",
       "32           4.400233        9.750017    171.929104      443.448775   \n",
       "33          14.009473        0.797364    170.908522      443.069042   \n",
       "34           5.862629        9.702498    170.416326      443.555122   \n",
       "35          11.892088        1.945849    173.315167      444.049581   \n",
       "36          11.999161        4.804263    173.743459      446.505202   \n",
       "...               ...             ...           ...             ...   \n",
       "291370      11.016730        9.167743    169.115085      459.202414   \n",
       "291371       6.469484        1.656610    168.838507      455.101759   \n",
       "291372       2.490792        2.252574    169.223690      463.630715   \n",
       "291373       6.004070        3.244295    172.163049      458.787617   \n",
       "291374       5.590642        2.204472    173.119397      456.196849   \n",
       "291375       9.103806        6.093867    168.410263      448.519301   \n",
       "291376      11.444951        3.359673    168.849711      450.330382   \n",
       "291377      16.820180        0.482728    171.513651      452.462106   \n",
       "291378      16.225371        1.396074    169.625319      451.508453   \n",
       "291379       9.448383        8.454640    167.950194      453.659486   \n",
       "291380      10.188202        5.086589    168.530246      448.103742   \n",
       "291381      18.394116        7.153156    169.765458      458.759394   \n",
       "291382      10.951704        0.649491    169.818254      460.880691   \n",
       "291383      11.492089        7.269283    172.199254      459.599763   \n",
       "291384      12.621841        1.793100    170.812061      453.681180   \n",
       "291385      10.747137        6.510290    170.439828      455.036021   \n",
       "291386       6.064152        5.007039    172.264492      453.800429   \n",
       "291387      10.677648        8.252193    170.243198      455.333806   \n",
       "291388       9.430194       13.841831    167.765844      451.355029   \n",
       "291389       1.622380        3.173978    167.312650      439.435929   \n",
       "291390      11.238292        3.633503    166.963446      438.276090   \n",
       "291391      11.019912        4.952713    165.979125      435.138421   \n",
       "291392       4.707020        3.106529    168.142646      447.915202   \n",
       "291393       2.576581        4.504970    168.398872      448.148851   \n",
       "291394      13.588936        7.168643    168.651040      446.075986   \n",
       "291395       1.639521        5.724500    171.826650      441.278667   \n",
       "291396       9.577243        0.735339    174.657123      444.147310   \n",
       "291397      12.445345        2.596743    173.787879      448.842085   \n",
       "291398      12.824247        2.752883    172.496791      442.086577   \n",
       "291399      11.713728        3.539798    170.782713      448.188498   \n",
       "\n",
       "        pressuremean_24h  vibrationmean_24h  voltsd_24h  rotatesd_24h  \\\n",
       "7              96.797113          40.385160   15.726970     39.648116   \n",
       "8              96.849785          39.736826   15.635083     41.828592   \n",
       "9              97.715600          39.498374   13.995465     40.843882   \n",
       "10             96.666060          40.229370   13.100364     43.409841   \n",
       "11             95.766838          40.055214   13.808489     43.742304   \n",
       "12             98.065860          40.033247   14.187985     40.676672   \n",
       "13             99.273286          38.903462   13.707794     40.509184   \n",
       "14             99.005946          39.389725   11.818603     44.468516   \n",
       "15            100.877342          38.696225   12.069391     46.669661   \n",
       "16            100.050567          39.449734   12.755234     44.016114   \n",
       "17             99.360632          40.766639   12.848646     45.090576   \n",
       "18             98.965877          42.396850   14.968351     37.088898   \n",
       "19             98.853189          44.608814   17.058217     36.617908   \n",
       "20             97.292157          45.284751   18.405763     35.819938   \n",
       "21             96.786742          47.311018   18.249831     42.055638   \n",
       "22             97.742700          48.416442   19.141287     37.018824   \n",
       "23             99.304019          51.158654   18.887033     36.997459   \n",
       "24            100.284484          52.153213   20.837993     34.051825   \n",
       "25            102.358897          52.854420   21.298322     36.054002   \n",
       "26            102.578580          52.789794   21.200183     38.544116   \n",
       "27            102.359939          51.518719   18.814679     40.547527   \n",
       "28            101.425508          52.135905   16.762469     49.373445   \n",
       "29            102.130179          51.653294   17.435946     43.819375   \n",
       "30            100.874614          52.529450   16.661364     47.266846   \n",
       "31             98.716746          51.746749   17.863934     44.895080   \n",
       "32             98.675590          51.780445   15.139300     45.766081   \n",
       "33             98.830333          49.679550   13.985517     42.542001   \n",
       "34            100.221328          48.481038   13.344630     39.327146   \n",
       "35            101.633306          47.279992   15.793146     42.984028   \n",
       "36            100.540356          45.527290   16.132288     40.754154   \n",
       "...                  ...                ...         ...           ...   \n",
       "291370         99.099044          39.342719   12.889019     60.409151   \n",
       "291371         98.960206          38.277576   13.917591     56.810848   \n",
       "291372         99.296474          38.406915   14.611939     56.534099   \n",
       "291373        100.063674          38.458947   15.232866     49.321412   \n",
       "291374        100.114215          39.063775   15.920072     48.742610   \n",
       "291375        100.049480          39.083797   14.635941     50.313046   \n",
       "291376        100.243957          38.469840   14.915283     50.148100   \n",
       "291377         98.514174          38.626944   14.750503     44.620373   \n",
       "291378         97.368816          38.762103   16.055332     49.218444   \n",
       "291379         97.292065          38.635159   15.942467     59.763569   \n",
       "291380         97.730594          39.228047   16.122100     57.394487   \n",
       "291381         98.775532          38.935773   16.994700     55.665652   \n",
       "291382         97.926823          39.295892   15.077105     56.097762   \n",
       "291383         98.365107          39.964091   14.414241     59.222526   \n",
       "291384         98.994157          40.013985   13.522301     61.197614   \n",
       "291385         99.830401          40.129183   14.551122     65.430315   \n",
       "291386        102.163329          40.050108   13.103400     62.190761   \n",
       "291387        102.290708          40.197103   13.120654     57.910021   \n",
       "291388        103.465520          40.252524   13.315211     59.466979   \n",
       "291389        102.009695          40.435349   11.768963     60.646198   \n",
       "291390        101.637723          40.172602   14.141910     58.372646   \n",
       "291391        102.005617          39.497908   14.000620     59.820047   \n",
       "291392         99.620102          39.635003   14.372707     59.563942   \n",
       "291393         99.351099          39.518646   15.763140     57.682117   \n",
       "291394         98.741443          39.840623   15.331755     60.839923   \n",
       "291395         98.311919          39.196175   16.429023     62.147934   \n",
       "291396         98.520388          38.820190   17.019808     64.730136   \n",
       "291397        100.028549          39.375067   17.096392     64.718132   \n",
       "291398        100.361794          38.943434   15.119775     65.929509   \n",
       "291399        100.794970          38.980896   15.573014     61.859239   \n",
       "\n",
       "        pressuresd_24h  vibrationsd_24h  \n",
       "7            11.904700         5.601191  \n",
       "8            11.326412         5.583521  \n",
       "9            11.036546         5.561553  \n",
       "10           10.972862         6.068674  \n",
       "11           10.988704         7.286129  \n",
       "12           11.942227         8.723238  \n",
       "13           10.141026         8.634082  \n",
       "14            9.444955         8.332673  \n",
       "15            8.609526         8.089348  \n",
       "16            9.893704         7.013132  \n",
       "17            9.846662         5.888262  \n",
       "18           10.133452         5.702356  \n",
       "19            9.867174         5.743753  \n",
       "20            9.743769         5.246435  \n",
       "21           10.961128         5.093464  \n",
       "22           10.642956         4.618287  \n",
       "23           11.042775         5.195423  \n",
       "24            9.654971         5.066388  \n",
       "25            9.885781         5.246894  \n",
       "26           10.429692         7.192434  \n",
       "27           11.133170         7.556313  \n",
       "28           10.443534         8.545739  \n",
       "29           10.830449         8.809530  \n",
       "30           11.225440         9.068824  \n",
       "31           10.675981         7.475304  \n",
       "32           10.959268         6.855778  \n",
       "33           11.050133         4.842842  \n",
       "34           10.268539         4.884344  \n",
       "35           10.006300         4.637101  \n",
       "36            9.744855         4.591048  \n",
       "...                ...              ...  \n",
       "291370       11.549081         5.671215  \n",
       "291371       11.118412         6.061118  \n",
       "291372        8.553177         5.893650  \n",
       "291373        8.345687         5.292801  \n",
       "291374        7.909495         5.249418  \n",
       "291375        7.582133         5.108806  \n",
       "291376        7.928488         5.518228  \n",
       "291377        8.348839         5.182752  \n",
       "291378        9.546008         5.020598  \n",
       "291379        9.431896         4.476863  \n",
       "291380        9.171658         4.866485  \n",
       "291381        9.630112         4.896106  \n",
       "291382        9.818266         4.919802  \n",
       "291383       10.569736         4.693342  \n",
       "291384       10.103349         4.286568  \n",
       "291385        9.389132         4.484563  \n",
       "291386        9.128160         4.502950  \n",
       "291387        9.238228         4.803393  \n",
       "291388        9.471171         4.442337  \n",
       "291389        9.136786         5.149517  \n",
       "291390        8.463179         5.221115  \n",
       "291391        6.856021         5.392136  \n",
       "291392        7.988174         5.256284  \n",
       "291393        8.088214         5.301986  \n",
       "291394        7.891711         5.269038  \n",
       "291395        7.475540         5.448962  \n",
       "291396        8.961444         5.833191  \n",
       "291397        9.420879         5.738756  \n",
       "291398        8.836617         6.139142  \n",
       "291399        9.942610         6.191276  \n",
       "\n",
       "[290601 rows x 18 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "telemetry_feat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   machineID            datetime  voltmean_3h  rotatemean_3h  pressuremean_3h  \\\n",
      "0          1 2015-01-02 06:00:00   180.133784     440.608320        94.137969   \n",
      "1          1 2015-01-02 09:00:00   176.364293     439.349655       101.553209   \n",
      "2          1 2015-01-02 12:00:00   160.384568     424.385316        99.598722   \n",
      "3          1 2015-01-02 15:00:00   170.472461     442.933997       102.380586   \n",
      "4          1 2015-01-02 18:00:00   163.263806     468.937558       102.726648   \n",
      "\n",
      "   vibrationmean_3h  voltsd_3h  rotatesd_3h  pressuresd_3h  vibrationsd_3h  \\\n",
      "0         41.551544  21.322735    48.770512       2.135684       10.037208   \n",
      "1         36.105580  18.952210    51.329636      13.789279        6.737739   \n",
      "2         36.094637  13.047080    13.702496       9.988609        1.639962   \n",
      "3         40.483002  16.642354    56.290447       3.305739        8.854145   \n",
      "4         40.921802  17.424688    38.680380       9.105775        3.060781   \n",
      "\n",
      "  ...   error2count  error3count  error4count  error5count   comp1    comp2  \\\n",
      "0 ...           0.0          0.0          0.0          0.0  20.000  215.000   \n",
      "1 ...           0.0          0.0          0.0          0.0  20.125  215.125   \n",
      "2 ...           0.0          0.0          0.0          0.0  20.250  215.250   \n",
      "3 ...           0.0          0.0          0.0          0.0  20.375  215.375   \n",
      "4 ...           0.0          0.0          0.0          0.0  20.500  215.500   \n",
      "\n",
      "     comp3    comp4   model  age  \n",
      "0  155.000  170.000  model3   18  \n",
      "1  155.125  170.125  model3   18  \n",
      "2  155.250  170.250  model3   18  \n",
      "3  155.375  170.375  model3   18  \n",
      "4  155.500  170.500  model3   18  \n",
      "\n",
      "[5 rows x 29 columns]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>voltmean_24h</th>\n",
       "      <th>...</th>\n",
       "      <th>error1count</th>\n",
       "      <th>error2count</th>\n",
       "      <th>error3count</th>\n",
       "      <th>error4count</th>\n",
       "      <th>error5count</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "      <td>290601.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.380935</td>\n",
       "      <td>170.774427</td>\n",
       "      <td>446.609386</td>\n",
       "      <td>100.858340</td>\n",
       "      <td>40.383609</td>\n",
       "      <td>13.300173</td>\n",
       "      <td>44.453951</td>\n",
       "      <td>8.885780</td>\n",
       "      <td>4.440575</td>\n",
       "      <td>170.775661</td>\n",
       "      <td>...</td>\n",
       "      <td>0.027560</td>\n",
       "      <td>0.027058</td>\n",
       "      <td>0.022846</td>\n",
       "      <td>0.019955</td>\n",
       "      <td>0.009780</td>\n",
       "      <td>53.382610</td>\n",
       "      <td>51.256589</td>\n",
       "      <td>52.536687</td>\n",
       "      <td>53.679601</td>\n",
       "      <td>11.345226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>28.798424</td>\n",
       "      <td>9.498824</td>\n",
       "      <td>33.119738</td>\n",
       "      <td>7.411701</td>\n",
       "      <td>3.475512</td>\n",
       "      <td>6.966389</td>\n",
       "      <td>23.214291</td>\n",
       "      <td>4.656364</td>\n",
       "      <td>2.319989</td>\n",
       "      <td>4.720237</td>\n",
       "      <td>...</td>\n",
       "      <td>0.166026</td>\n",
       "      <td>0.164401</td>\n",
       "      <td>0.151266</td>\n",
       "      <td>0.140998</td>\n",
       "      <td>0.098931</td>\n",
       "      <td>62.478424</td>\n",
       "      <td>59.156008</td>\n",
       "      <td>58.822946</td>\n",
       "      <td>59.658975</td>\n",
       "      <td>5.826345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>125.532506</td>\n",
       "      <td>211.811184</td>\n",
       "      <td>72.118639</td>\n",
       "      <td>26.569635</td>\n",
       "      <td>0.025509</td>\n",
       "      <td>0.078991</td>\n",
       "      <td>0.027417</td>\n",
       "      <td>0.015278</td>\n",
       "      <td>155.812721</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25.000000</td>\n",
       "      <td>164.447794</td>\n",
       "      <td>427.564793</td>\n",
       "      <td>96.239534</td>\n",
       "      <td>38.147458</td>\n",
       "      <td>8.028675</td>\n",
       "      <td>26.906319</td>\n",
       "      <td>5.369959</td>\n",
       "      <td>2.684556</td>\n",
       "      <td>168.072275</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>13.250000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>12.875000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.000000</td>\n",
       "      <td>170.432407</td>\n",
       "      <td>448.380260</td>\n",
       "      <td>100.235357</td>\n",
       "      <td>40.145874</td>\n",
       "      <td>12.495542</td>\n",
       "      <td>41.793798</td>\n",
       "      <td>8.345801</td>\n",
       "      <td>4.173704</td>\n",
       "      <td>170.212704</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>32.625000</td>\n",
       "      <td>29.500000</td>\n",
       "      <td>32.125000</td>\n",
       "      <td>32.375000</td>\n",
       "      <td>12.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>75.000000</td>\n",
       "      <td>176.610017</td>\n",
       "      <td>468.443933</td>\n",
       "      <td>104.406534</td>\n",
       "      <td>42.226898</td>\n",
       "      <td>17.688520</td>\n",
       "      <td>59.092354</td>\n",
       "      <td>11.789358</td>\n",
       "      <td>5.898512</td>\n",
       "      <td>172.462228</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>68.500000</td>\n",
       "      <td>65.875000</td>\n",
       "      <td>67.125000</td>\n",
       "      <td>70.250000</td>\n",
       "      <td>16.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>100.000000</td>\n",
       "      <td>241.420717</td>\n",
       "      <td>586.682904</td>\n",
       "      <td>162.309656</td>\n",
       "      <td>69.311324</td>\n",
       "      <td>58.444332</td>\n",
       "      <td>179.903039</td>\n",
       "      <td>35.659369</td>\n",
       "      <td>18.305595</td>\n",
       "      <td>220.782618</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>491.875000</td>\n",
       "      <td>348.875000</td>\n",
       "      <td>370.875000</td>\n",
       "      <td>394.875000</td>\n",
       "      <td>20.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           machineID    voltmean_3h  rotatemean_3h  pressuremean_3h  \\\n",
       "count  290601.000000  290601.000000  290601.000000    290601.000000   \n",
       "mean       50.380935     170.774427     446.609386       100.858340   \n",
       "std        28.798424       9.498824      33.119738         7.411701   \n",
       "min         1.000000     125.532506     211.811184        72.118639   \n",
       "25%        25.000000     164.447794     427.564793        96.239534   \n",
       "50%        50.000000     170.432407     448.380260       100.235357   \n",
       "75%        75.000000     176.610017     468.443933       104.406534   \n",
       "max       100.000000     241.420717     586.682904       162.309656   \n",
       "\n",
       "       vibrationmean_3h      voltsd_3h    rotatesd_3h  pressuresd_3h  \\\n",
       "count     290601.000000  290601.000000  290601.000000  290601.000000   \n",
       "mean          40.383609      13.300173      44.453951       8.885780   \n",
       "std            3.475512       6.966389      23.214291       4.656364   \n",
       "min           26.569635       0.025509       0.078991       0.027417   \n",
       "25%           38.147458       8.028675      26.906319       5.369959   \n",
       "50%           40.145874      12.495542      41.793798       8.345801   \n",
       "75%           42.226898      17.688520      59.092354      11.789358   \n",
       "max           69.311324      58.444332     179.903039      35.659369   \n",
       "\n",
       "       vibrationsd_3h   voltmean_24h      ...          error1count  \\\n",
       "count   290601.000000  290601.000000      ...        290601.000000   \n",
       "mean         4.440575     170.775661      ...             0.027560   \n",
       "std          2.319989       4.720237      ...             0.166026   \n",
       "min          0.015278     155.812721      ...             0.000000   \n",
       "25%          2.684556     168.072275      ...             0.000000   \n",
       "50%          4.173704     170.212704      ...             0.000000   \n",
       "75%          5.898512     172.462228      ...             0.000000   \n",
       "max         18.305595     220.782618      ...             2.000000   \n",
       "\n",
       "         error2count    error3count    error4count    error5count  \\\n",
       "count  290601.000000  290601.000000  290601.000000  290601.000000   \n",
       "mean        0.027058       0.022846       0.019955       0.009780   \n",
       "std         0.164401       0.151266       0.140998       0.098931   \n",
       "min         0.000000       0.000000       0.000000       0.000000   \n",
       "25%         0.000000       0.000000       0.000000       0.000000   \n",
       "50%         0.000000       0.000000       0.000000       0.000000   \n",
       "75%         0.000000       0.000000       0.000000       0.000000   \n",
       "max         2.000000       2.000000       2.000000       2.000000   \n",
       "\n",
       "               comp1          comp2          comp3          comp4  \\\n",
       "count  290601.000000  290601.000000  290601.000000  290601.000000   \n",
       "mean       53.382610      51.256589      52.536687      53.679601   \n",
       "std        62.478424      59.156008      58.822946      59.658975   \n",
       "min         0.000000       0.000000       0.000000       0.000000   \n",
       "25%        13.250000      12.000000      13.000000      12.875000   \n",
       "50%        32.625000      29.500000      32.125000      32.375000   \n",
       "75%        68.500000      65.875000      67.125000      70.250000   \n",
       "max       491.875000     348.875000     370.875000     394.875000   \n",
       "\n",
       "                 age  \n",
       "count  290601.000000  \n",
       "mean       11.345226  \n",
       "std         5.826345  \n",
       "min         0.000000  \n",
       "25%         7.000000  \n",
       "50%        12.000000  \n",
       "75%        16.000000  \n",
       "max        20.000000  \n",
       "\n",
       "[8 rows x 27 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_feat = telemetry_feat.merge(error_count, on=['datetime', 'machineID'], how='left')\n",
    "final_feat = final_feat.merge(comp_rep, on=['datetime', 'machineID'], how='left')\n",
    "final_feat = final_feat.merge(machines, on=['machineID'], how='left')\n",
    "\n",
    "print(final_feat.head())\n",
    "final_feat.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Label Construction\n",
    "When using multi-class classification for predicting failure due to a problem, labelling is done by taking a time window prior to the failure of an asset and labelling the feature records that fall into that window as \"about to fail due to a problem\" while labelling all other records as \"Â€Âœnormal.\" This time window should be picked according to the business case: in some situations it may be enough to predict failures hours in advance, while in others days or weeks may be needed to allow e.g. for arrival of replacement parts.\n",
    "\n",
    "The prediction problem for this example scenerio is to estimate the probability that a machine will fail in the near future due to a failure of a certain component. More specifically, the goal is to compute the probability that a machine will fail in the next 24 hours due to a certain component failure (component 1, 2, 3, or 4). Below, a categorical failure feature is created to serve as the label. All records within a 24 hour window before a failure of component 1 have failure=comp1, and so on for components 2, 3, and 4; all records not within 24 hours of a component failure have failure=none."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>...</th>\n",
       "      <th>error3count</th>\n",
       "      <th>error4count</th>\n",
       "      <th>error5count</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "      <th>model</th>\n",
       "      <th>age</th>\n",
       "      <th>failure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 06:00:00</td>\n",
       "      <td>180.133784</td>\n",
       "      <td>440.608320</td>\n",
       "      <td>94.137969</td>\n",
       "      <td>41.551544</td>\n",
       "      <td>21.322735</td>\n",
       "      <td>48.770512</td>\n",
       "      <td>2.135684</td>\n",
       "      <td>10.037208</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.000</td>\n",
       "      <td>215.000</td>\n",
       "      <td>155.000</td>\n",
       "      <td>170.000</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 09:00:00</td>\n",
       "      <td>176.364293</td>\n",
       "      <td>439.349655</td>\n",
       "      <td>101.553209</td>\n",
       "      <td>36.105580</td>\n",
       "      <td>18.952210</td>\n",
       "      <td>51.329636</td>\n",
       "      <td>13.789279</td>\n",
       "      <td>6.737739</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.125</td>\n",
       "      <td>215.125</td>\n",
       "      <td>155.125</td>\n",
       "      <td>170.125</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 12:00:00</td>\n",
       "      <td>160.384568</td>\n",
       "      <td>424.385316</td>\n",
       "      <td>99.598722</td>\n",
       "      <td>36.094637</td>\n",
       "      <td>13.047080</td>\n",
       "      <td>13.702496</td>\n",
       "      <td>9.988609</td>\n",
       "      <td>1.639962</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.250</td>\n",
       "      <td>215.250</td>\n",
       "      <td>155.250</td>\n",
       "      <td>170.250</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 15:00:00</td>\n",
       "      <td>170.472461</td>\n",
       "      <td>442.933997</td>\n",
       "      <td>102.380586</td>\n",
       "      <td>40.483002</td>\n",
       "      <td>16.642354</td>\n",
       "      <td>56.290447</td>\n",
       "      <td>3.305739</td>\n",
       "      <td>8.854145</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.375</td>\n",
       "      <td>215.375</td>\n",
       "      <td>155.375</td>\n",
       "      <td>170.375</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-02 18:00:00</td>\n",
       "      <td>163.263806</td>\n",
       "      <td>468.937558</td>\n",
       "      <td>102.726648</td>\n",
       "      <td>40.921802</td>\n",
       "      <td>17.424688</td>\n",
       "      <td>38.680380</td>\n",
       "      <td>9.105775</td>\n",
       "      <td>3.060781</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.500</td>\n",
       "      <td>215.500</td>\n",
       "      <td>155.500</td>\n",
       "      <td>170.500</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   machineID            datetime  voltmean_3h  rotatemean_3h  pressuremean_3h  \\\n",
       "0          1 2015-01-02 06:00:00   180.133784     440.608320        94.137969   \n",
       "1          1 2015-01-02 09:00:00   176.364293     439.349655       101.553209   \n",
       "2          1 2015-01-02 12:00:00   160.384568     424.385316        99.598722   \n",
       "3          1 2015-01-02 15:00:00   170.472461     442.933997       102.380586   \n",
       "4          1 2015-01-02 18:00:00   163.263806     468.937558       102.726648   \n",
       "\n",
       "   vibrationmean_3h  voltsd_3h  rotatesd_3h  pressuresd_3h  vibrationsd_3h  \\\n",
       "0         41.551544  21.322735    48.770512       2.135684       10.037208   \n",
       "1         36.105580  18.952210    51.329636      13.789279        6.737739   \n",
       "2         36.094637  13.047080    13.702496       9.988609        1.639962   \n",
       "3         40.483002  16.642354    56.290447       3.305739        8.854145   \n",
       "4         40.921802  17.424688    38.680380       9.105775        3.060781   \n",
       "\n",
       "    ...     error3count  error4count  error5count   comp1    comp2    comp3  \\\n",
       "0   ...             0.0          0.0          0.0  20.000  215.000  155.000   \n",
       "1   ...             0.0          0.0          0.0  20.125  215.125  155.125   \n",
       "2   ...             0.0          0.0          0.0  20.250  215.250  155.250   \n",
       "3   ...             0.0          0.0          0.0  20.375  215.375  155.375   \n",
       "4   ...             0.0          0.0          0.0  20.500  215.500  155.500   \n",
       "\n",
       "     comp4   model  age  failure  \n",
       "0  170.000  model3   18     none  \n",
       "1  170.125  model3   18     none  \n",
       "2  170.250  model3   18     none  \n",
       "3  170.375  model3   18     none  \n",
       "4  170.500  model3   18     none  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labeled_features = final_feat.merge(failures, on=['datetime', 'machineID'], how='left')\n",
    "labeled_features = labeled_features.fillna(method='bfill', limit=7) # fill backward up to 24h\n",
    "labeled_features = labeled_features.fillna('none')\n",
    "labeled_features.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below is an example of records that are labeled as failure=comp4 in the failure column. Notice that the first 8 records all occur in the 24-hour window before the first recorded failure of component 4. The next 8 records are within the 24 hour window before another failure of component 4."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>machineID</th>\n",
       "      <th>datetime</th>\n",
       "      <th>voltmean_3h</th>\n",
       "      <th>rotatemean_3h</th>\n",
       "      <th>pressuremean_3h</th>\n",
       "      <th>vibrationmean_3h</th>\n",
       "      <th>voltsd_3h</th>\n",
       "      <th>rotatesd_3h</th>\n",
       "      <th>pressuresd_3h</th>\n",
       "      <th>vibrationsd_3h</th>\n",
       "      <th>...</th>\n",
       "      <th>error3count</th>\n",
       "      <th>error4count</th>\n",
       "      <th>error5count</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "      <th>model</th>\n",
       "      <th>age</th>\n",
       "      <th>failure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 09:00:00</td>\n",
       "      <td>166.281848</td>\n",
       "      <td>453.787824</td>\n",
       "      <td>106.187582</td>\n",
       "      <td>51.990080</td>\n",
       "      <td>24.276228</td>\n",
       "      <td>23.621315</td>\n",
       "      <td>11.176731</td>\n",
       "      <td>3.394073</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.125</td>\n",
       "      <td>217.125</td>\n",
       "      <td>157.125</td>\n",
       "      <td>172.125</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 12:00:00</td>\n",
       "      <td>175.412103</td>\n",
       "      <td>445.450581</td>\n",
       "      <td>100.887363</td>\n",
       "      <td>54.251534</td>\n",
       "      <td>34.918687</td>\n",
       "      <td>11.001625</td>\n",
       "      <td>10.580336</td>\n",
       "      <td>2.921501</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.250</td>\n",
       "      <td>217.250</td>\n",
       "      <td>157.250</td>\n",
       "      <td>172.250</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 15:00:00</td>\n",
       "      <td>157.347716</td>\n",
       "      <td>451.882075</td>\n",
       "      <td>101.289380</td>\n",
       "      <td>48.602686</td>\n",
       "      <td>24.617739</td>\n",
       "      <td>28.950883</td>\n",
       "      <td>9.966729</td>\n",
       "      <td>2.356486</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.375</td>\n",
       "      <td>217.375</td>\n",
       "      <td>157.375</td>\n",
       "      <td>172.375</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 18:00:00</td>\n",
       "      <td>176.450550</td>\n",
       "      <td>446.033068</td>\n",
       "      <td>84.521555</td>\n",
       "      <td>47.638836</td>\n",
       "      <td>8.071400</td>\n",
       "      <td>76.511343</td>\n",
       "      <td>2.636879</td>\n",
       "      <td>4.108621</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.500</td>\n",
       "      <td>217.500</td>\n",
       "      <td>157.500</td>\n",
       "      <td>172.500</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-04 21:00:00</td>\n",
       "      <td>190.325814</td>\n",
       "      <td>422.692565</td>\n",
       "      <td>107.393234</td>\n",
       "      <td>49.552856</td>\n",
       "      <td>8.390777</td>\n",
       "      <td>7.176553</td>\n",
       "      <td>4.262645</td>\n",
       "      <td>7.598552</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.625</td>\n",
       "      <td>217.625</td>\n",
       "      <td>157.625</td>\n",
       "      <td>172.625</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 00:00:00</td>\n",
       "      <td>169.985134</td>\n",
       "      <td>458.929418</td>\n",
       "      <td>91.494362</td>\n",
       "      <td>54.882021</td>\n",
       "      <td>9.451483</td>\n",
       "      <td>12.052752</td>\n",
       "      <td>3.685906</td>\n",
       "      <td>6.621183</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.750</td>\n",
       "      <td>217.750</td>\n",
       "      <td>157.750</td>\n",
       "      <td>172.750</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 03:00:00</td>\n",
       "      <td>149.082619</td>\n",
       "      <td>412.180336</td>\n",
       "      <td>93.509785</td>\n",
       "      <td>54.386079</td>\n",
       "      <td>19.075952</td>\n",
       "      <td>30.715081</td>\n",
       "      <td>3.090266</td>\n",
       "      <td>6.530610</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>22.875</td>\n",
       "      <td>217.875</td>\n",
       "      <td>157.875</td>\n",
       "      <td>172.875</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-05 06:00:00</td>\n",
       "      <td>185.782709</td>\n",
       "      <td>439.531288</td>\n",
       "      <td>99.413660</td>\n",
       "      <td>51.558082</td>\n",
       "      <td>14.495664</td>\n",
       "      <td>45.663743</td>\n",
       "      <td>4.289212</td>\n",
       "      <td>7.330397</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>218.000</td>\n",
       "      <td>158.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1337</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-18 09:00:00</td>\n",
       "      <td>169.324639</td>\n",
       "      <td>453.923471</td>\n",
       "      <td>101.313249</td>\n",
       "      <td>53.092274</td>\n",
       "      <td>28.155693</td>\n",
       "      <td>42.557599</td>\n",
       "      <td>7.688674</td>\n",
       "      <td>2.488851</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.125</td>\n",
       "      <td>29.125</td>\n",
       "      <td>14.125</td>\n",
       "      <td>134.125</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1338</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-18 12:00:00</td>\n",
       "      <td>190.691297</td>\n",
       "      <td>441.577271</td>\n",
       "      <td>97.192512</td>\n",
       "      <td>44.025425</td>\n",
       "      <td>6.296827</td>\n",
       "      <td>47.271008</td>\n",
       "      <td>7.577957</td>\n",
       "      <td>4.648336</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.250</td>\n",
       "      <td>29.250</td>\n",
       "      <td>14.250</td>\n",
       "      <td>134.250</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1339</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-18 15:00:00</td>\n",
       "      <td>163.602957</td>\n",
       "      <td>433.781185</td>\n",
       "      <td>93.173047</td>\n",
       "      <td>43.051368</td>\n",
       "      <td>18.147449</td>\n",
       "      <td>30.242516</td>\n",
       "      <td>10.870615</td>\n",
       "      <td>2.740922</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.375</td>\n",
       "      <td>29.375</td>\n",
       "      <td>14.375</td>\n",
       "      <td>134.375</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1340</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-18 18:00:00</td>\n",
       "      <td>178.587550</td>\n",
       "      <td>427.300815</td>\n",
       "      <td>118.643186</td>\n",
       "      <td>50.958609</td>\n",
       "      <td>2.229649</td>\n",
       "      <td>17.168087</td>\n",
       "      <td>15.714144</td>\n",
       "      <td>5.669003</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.500</td>\n",
       "      <td>29.500</td>\n",
       "      <td>14.500</td>\n",
       "      <td>134.500</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1341</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-18 21:00:00</td>\n",
       "      <td>158.851795</td>\n",
       "      <td>520.113831</td>\n",
       "      <td>101.974559</td>\n",
       "      <td>44.156671</td>\n",
       "      <td>14.554854</td>\n",
       "      <td>77.101968</td>\n",
       "      <td>4.788908</td>\n",
       "      <td>5.468742</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.625</td>\n",
       "      <td>29.625</td>\n",
       "      <td>14.625</td>\n",
       "      <td>134.625</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1342</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-19 00:00:00</td>\n",
       "      <td>162.191516</td>\n",
       "      <td>453.545010</td>\n",
       "      <td>101.521779</td>\n",
       "      <td>49.136659</td>\n",
       "      <td>12.553190</td>\n",
       "      <td>33.332139</td>\n",
       "      <td>5.983913</td>\n",
       "      <td>1.893250</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.750</td>\n",
       "      <td>29.750</td>\n",
       "      <td>14.750</td>\n",
       "      <td>134.750</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1343</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-19 03:00:00</td>\n",
       "      <td>166.732741</td>\n",
       "      <td>485.036994</td>\n",
       "      <td>100.284288</td>\n",
       "      <td>44.587560</td>\n",
       "      <td>11.099161</td>\n",
       "      <td>57.308864</td>\n",
       "      <td>3.052958</td>\n",
       "      <td>3.062215</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>89.875</td>\n",
       "      <td>29.875</td>\n",
       "      <td>14.875</td>\n",
       "      <td>134.875</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1344</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-06-19 06:00:00</td>\n",
       "      <td>172.059069</td>\n",
       "      <td>463.242610</td>\n",
       "      <td>96.905050</td>\n",
       "      <td>53.701413</td>\n",
       "      <td>14.757880</td>\n",
       "      <td>55.874000</td>\n",
       "      <td>3.204981</td>\n",
       "      <td>2.329615</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>30.000</td>\n",
       "      <td>15.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>model3</td>\n",
       "      <td>18</td>\n",
       "      <td>comp4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      machineID            datetime  voltmean_3h  rotatemean_3h  \\\n",
       "17            1 2015-01-04 09:00:00   166.281848     453.787824   \n",
       "18            1 2015-01-04 12:00:00   175.412103     445.450581   \n",
       "19            1 2015-01-04 15:00:00   157.347716     451.882075   \n",
       "20            1 2015-01-04 18:00:00   176.450550     446.033068   \n",
       "21            1 2015-01-04 21:00:00   190.325814     422.692565   \n",
       "22            1 2015-01-05 00:00:00   169.985134     458.929418   \n",
       "23            1 2015-01-05 03:00:00   149.082619     412.180336   \n",
       "24            1 2015-01-05 06:00:00   185.782709     439.531288   \n",
       "1337          1 2015-06-18 09:00:00   169.324639     453.923471   \n",
       "1338          1 2015-06-18 12:00:00   190.691297     441.577271   \n",
       "1339          1 2015-06-18 15:00:00   163.602957     433.781185   \n",
       "1340          1 2015-06-18 18:00:00   178.587550     427.300815   \n",
       "1341          1 2015-06-18 21:00:00   158.851795     520.113831   \n",
       "1342          1 2015-06-19 00:00:00   162.191516     453.545010   \n",
       "1343          1 2015-06-19 03:00:00   166.732741     485.036994   \n",
       "1344          1 2015-06-19 06:00:00   172.059069     463.242610   \n",
       "\n",
       "      pressuremean_3h  vibrationmean_3h  voltsd_3h  rotatesd_3h  \\\n",
       "17         106.187582         51.990080  24.276228    23.621315   \n",
       "18         100.887363         54.251534  34.918687    11.001625   \n",
       "19         101.289380         48.602686  24.617739    28.950883   \n",
       "20          84.521555         47.638836   8.071400    76.511343   \n",
       "21         107.393234         49.552856   8.390777     7.176553   \n",
       "22          91.494362         54.882021   9.451483    12.052752   \n",
       "23          93.509785         54.386079  19.075952    30.715081   \n",
       "24          99.413660         51.558082  14.495664    45.663743   \n",
       "1337       101.313249         53.092274  28.155693    42.557599   \n",
       "1338        97.192512         44.025425   6.296827    47.271008   \n",
       "1339        93.173047         43.051368  18.147449    30.242516   \n",
       "1340       118.643186         50.958609   2.229649    17.168087   \n",
       "1341       101.974559         44.156671  14.554854    77.101968   \n",
       "1342       101.521779         49.136659  12.553190    33.332139   \n",
       "1343       100.284288         44.587560  11.099161    57.308864   \n",
       "1344        96.905050         53.701413  14.757880    55.874000   \n",
       "\n",
       "      pressuresd_3h  vibrationsd_3h   ...     error3count  error4count  \\\n",
       "17        11.176731        3.394073   ...             1.0          0.0   \n",
       "18        10.580336        2.921501   ...             1.0          0.0   \n",
       "19         9.966729        2.356486   ...             1.0          0.0   \n",
       "20         2.636879        4.108621   ...             1.0          0.0   \n",
       "21         4.262645        7.598552   ...             1.0          0.0   \n",
       "22         3.685906        6.621183   ...             0.0          0.0   \n",
       "23         3.090266        6.530610   ...             0.0          0.0   \n",
       "24         4.289212        7.330397   ...             0.0          0.0   \n",
       "1337       7.688674        2.488851   ...             0.0          0.0   \n",
       "1338       7.577957        4.648336   ...             0.0          0.0   \n",
       "1339      10.870615        2.740922   ...             0.0          0.0   \n",
       "1340      15.714144        5.669003   ...             0.0          0.0   \n",
       "1341       4.788908        5.468742   ...             0.0          0.0   \n",
       "1342       5.983913        1.893250   ...             0.0          0.0   \n",
       "1343       3.052958        3.062215   ...             0.0          0.0   \n",
       "1344       3.204981        2.329615   ...             0.0          0.0   \n",
       "\n",
       "      error5count   comp1    comp2    comp3    comp4   model  age  failure  \n",
       "17            1.0  22.125  217.125  157.125  172.125  model3   18    comp4  \n",
       "18            1.0  22.250  217.250  157.250  172.250  model3   18    comp4  \n",
       "19            1.0  22.375  217.375  157.375  172.375  model3   18    comp4  \n",
       "20            1.0  22.500  217.500  157.500  172.500  model3   18    comp4  \n",
       "21            1.0  22.625  217.625  157.625  172.625  model3   18    comp4  \n",
       "22            1.0  22.750  217.750  157.750  172.750  model3   18    comp4  \n",
       "23            1.0  22.875  217.875  157.875  172.875  model3   18    comp4  \n",
       "24            1.0   0.000  218.000  158.000    0.000  model3   18    comp4  \n",
       "1337          1.0  89.125   29.125   14.125  134.125  model3   18    comp4  \n",
       "1338          1.0  89.250   29.250   14.250  134.250  model3   18    comp4  \n",
       "1339          1.0  89.375   29.375   14.375  134.375  model3   18    comp4  \n",
       "1340          1.0  89.500   29.500   14.500  134.500  model3   18    comp4  \n",
       "1341          1.0  89.625   29.625   14.625  134.625  model3   18    comp4  \n",
       "1342          1.0  89.750   29.750   14.750  134.750  model3   18    comp4  \n",
       "1343          1.0  89.875   29.875   14.875  134.875  model3   18    comp4  \n",
       "1344          1.0   0.000   30.000   15.000    0.000  model3   18    comp4  \n",
       "\n",
       "[16 rows x 30 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labeled_features.loc[labeled_features['failure'] == 'comp4'][:16]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modelling\n",
    "After the feature engineering and labelling steps, either Azure Machine Learning Studio or this notebook can be used to create a predictive model. The recommend Azure Machine Learning Studio experiment can be found in the Cortana Intelligence Gallery: Predictive Maintenance Modelling Guide Experiment. Below, we describe the modelling process and provide an example Python model.\n",
    "\n",
    "# Training, Validation and Testing\n",
    "When working with time-stamped data as in this example, record partitioning into training, validation, and test sets should be performed carefully to prevent overestimating the performance of the models. In predictive maintenance, the features are usually generated using lagging aggregates: records in the same time window will likely have identical labels and similar feature values. These correlations can give a model an \"unfair advantage\" when predicting on a test set record that shares its time window with a training set record. We therefore partition records into training, validation, and test sets in large chunks, to minimize the number of time intervals shared between them.\n",
    "\n",
    "Predictive models have no advance knowledge of future chronological trends: in practice, such trends are likely to exist and to adversely impact the model's performance. To obtain an accurate assessment of a predictive model's performance, we recommend training on older records and validating/testing using newer records.\n",
    "\n",
    "For both of these reasons, a time-dependent record splitting strategy is an excellent choice for predictive maintenace models. The split is effected by choosing a point in time based on the desired size of the training and test sets: all records before the timepoint are used for training the model, and all remaining records are used for testing. (If desired, the timeline could be further divided to create validation sets for parameter selection.) To prevent any records in the training set from sharing time windows with the records in the test set, we remove any records at the boundary -- in this case, by ignoring 24 hours' worth of data prior to the timepoint."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "# make test and training splits\n",
    "threshold_dates = [[pd.to_datetime('2015-07-31 01:00:00'), pd.to_datetime('2015-08-01 01:00:00')],\n",
    "                   [pd.to_datetime('2015-08-31 01:00:00'), pd.to_datetime('2015-09-01 01:00:00')],\n",
    "                   [pd.to_datetime('2015-09-30 01:00:00'), pd.to_datetime('2015-10-01 01:00:00')]]\n",
    "\n",
    "test_results = []\n",
    "models = []\n",
    "for last_train_date, first_test_date in threshold_dates:\n",
    "    # split out training and test data\n",
    "    train_y = labeled_features.loc[labeled_features['datetime'] < last_train_date, 'failure']\n",
    "    train_X = pd.get_dummies(labeled_features.loc[labeled_features['datetime'] < last_train_date].drop(['datetime',\n",
    "                                                                                                        'machineID',\n",
    "                                                                                                        'failure'], 1))\n",
    "    test_X = pd.get_dummies(labeled_features.loc[labeled_features['datetime'] > first_test_date].drop(['datetime',\n",
    "                                                                                                       'machineID',\n",
    "                                                                                                       'failure'], 1))\n",
    "    # train and predict using the model, storing results for later\n",
    "    my_model = GradientBoostingClassifier(random_state=42)\n",
    "    my_model.fit(train_X, train_y)\n",
    "    test_result = pd.DataFrame(labeled_features.loc[labeled_features['datetime'] > first_test_date])\n",
    "    test_result['predicted_failure'] = my_model.predict(test_X)\n",
    "    test_results.append(test_result)\n",
    "    models.append(my_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x256510e55c0>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmUAAAGyCAYAAAClXLMtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x256506cf0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(10, 6))\n",
    "labels, importances = zip(*sorted(zip(test_X.columns, models[0].feature_importances_), reverse=True, key=lambda x: x[1]))\n",
    "plt.xticks(range(len(labels)), labels)\n",
    "_, labels = plt.xticks()\n",
    "plt.setp(labels, rotation=90)\n",
    "plt.bar(range(len(importances)), importances)\n",
    "plt.ylabel('Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation\n",
    "In predictive maintenance, machine failures are usually rare occurrences in the lifetime of the assets compared to normal operation. This causes an imbalance in the label distribution which usually causes poor performance as algorithms tend to classify majority class examples better at the expense of minority class examples as the total misclassification error is much improved when majority class is labeled correctly. This causes low recall rates although accuracy can be high and becomes a larger problem when the cost of false alarms to the business is very high. To help with this problem, sampling techniques such as oversampling of the minority examples are usually used along with more sophisticated techniques which are not covered in this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x25600910f60>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAEbCAYAAABHtoc8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xt8VPWd//H3ZGZCa2YCiYTdsiZ4zQracAsXNwMs1JYu\nwhZFxUTjtiAoDUFwiUEIIYjcZEO1hEi1rd2mkggNpdg8tnULQqQB0qaNFGLQB15ABbkKmQAZkjm/\nP/pz1qwYwOXMZM55PR8PH5JvTmY+520pb75n5ozDMAxDAADAtmIiPQAAAIgsygAAADZHGQAAwOYo\nAwAA2BxlAAAAm6MMAABgc65IDxApR482RXqEy5aQcJVOnjwT6TEsjYzNR8bmI+PwiLack5K8X/g9\ndgaiiMvljPQIlkfG5iNj85FxeFgpZ8oAAAA2RxkAAMDmKAMAANgcZQAAAJujDAAAYHOUAQAAbI4y\nAACAzVEGAACwOcoAAAA2Z9vbEZtl0rItkR7hsv10zqhIjwAAiCB2BgAAsDnKAAAANkcZAADA5igD\nAADYHGUAAACbowwAAGBzlAEAAGyOMgAAgM1RBgAAsDnKAAAANkcZAADA5igDAADYHGUAAACbowwA\nAGBzlAEAAGyOMgAAgM1RBgAAsDnKAAAANkcZAADA5igDAADYnMuMBz1//rzmzp2rDz/8UIFAQNOm\nTdPXvvY1Pfzww7r22mslSZmZmRozZozWrVuniooKuVwuTZs2TSNHjtS5c+eUl5en48ePKy4uTsuX\nL1diYqLq6+u1ePFiOZ1O+Xw+TZ8+XZJUUlKirVu3yuVyae7cuUpLSzPjtAAAsCRTysCmTZvUrVs3\nrVixQp988onGjx+vnJwcfe9739OkSZNCxx09elRlZWWqrKxUS0uLsrKylJGRofLycqWmpio3N1dV\nVVUqLS1VQUGBFixYoFWrVik5OVlTp05VQ0ODDMNQbW2t1q9fr0OHDik3N1eVlZVmnBYAAJZkShn4\n9re/rdGjR0uSDMOQ0+nUnj179O6772rz5s3q1auX5s6dq927d6t///6KjY1VbGysUlJS1NjYqLq6\nOj300EOSpOHDh6u0tFR+v1+BQEApKSmSJJ/Pp5qaGsXGxsrn88nhcKhnz55qa2vTiRMnlJiYaMap\nAQBgOaa8ZiAuLk4ej0d+v18zZszQzJkzlZaWpscff1wvvfSSkpOTtXr1avn9fnm93nY/5/f7263H\nxcWpqalJfr9fHo+n3bEdrQMAgEtjys6AJB06dEg5OTnKysrSuHHjdPr0acXHx0uSvvnNb2rRokVK\nT09Xc3Nz6Geam5vl9Xrl8XhC683NzYqPj2+39tl1t9t9wce4mISEq+RyOa/U6Ua1pKSL52Un5GE+\nMjYfGYeHVXI2pQwcO3ZMkyZNUmFhoW677TZJ0uTJkzV//nylpaVpx44duuWWW5SWlqZnnnlGLS0t\nCgQC2r9/v1JTUzVgwABt27ZNaWlpqq6u1sCBA+XxeOR2u3XgwAElJydr+/btmj59upxOp1asWKHJ\nkyfr8OHDCgaDl3SJ4OTJM2acelQ6epSdlE8lJXnJw2RkbD4yDo9oy7mj4mJKGVizZo1Onz6t0tJS\nlZaWSpLmzJmjJUuWyO12q3v37lq0aJE8Ho+ys7OVlZUlwzA0a9YsdenSRZmZmcrPz1dmZqbcbreK\ni4slSQsXLtTs2bPV1tYmn8+nvn37SpLS09M1ceJEBYNBFRYWmnFKAABYlsMwDCPSQ0SCWW1u0rIt\npjyumX46Z1SkR+g0oq3pRyMyNh8Zh0e05dzRzgA3HQIAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABs\njjIAAIDNUQYAALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYAALA5\nygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYAALA5ygAAADZHGQAAwOYo\nAwAA2BxlAAAAm6MMAABgc5QBAABszmXGg54/f15z587Vhx9+qEAgoGnTpunGG2/UnDlz5HA4dNNN\nN2nBggWKiYnRunXrVFFRIZfLpWnTpmnkyJE6d+6c8vLydPz4ccXFxWn58uVKTExUfX29Fi9eLKfT\nKZ/Pp+nTp0uSSkpKtHXrVrlcLs2dO1dpaWlmnBYAAJZkShnYtGmTunXrphUrVuiTTz7R+PHjdfPN\nN2vmzJkaMmSICgsLtXnzZvXr109lZWWqrKxUS0uLsrKylJGRofLycqWmpio3N1dVVVUqLS1VQUGB\nFixYoFWrVik5OVlTp05VQ0ODDMNQbW2t1q9fr0OHDik3N1eVlZVmnBYAAJZkShn49re/rdGjR0uS\nDMOQ0+nU3r17NXjwYEnS8OHD9Yc//EExMTHq37+/YmNjFRsbq5SUFDU2Nqqurk4PPfRQ6NjS0lL5\n/X4FAgGlpKRIknw+n2pqahQbGyufzyeHw6GePXuqra1NJ06cUGJiohmnBgCA5ZhSBuLi4iRJfr9f\nM2bM0MyZM7V8+XI5HI7Q95uamuT3++X1etv9nN/vb7f+2WM9Hk+7Yw8ePKguXbqoW7du7dabmpou\nWgYSEq6Sy+W8YucczZKSvBc/yEbIw3xkbD4yDg+r5GxKGZCkQ4cOKScnR1lZWRo3bpxWrFgR+l5z\nc7Pi4+Pl8XjU3Nzcbt3r9bZb7+jY+Ph4ud3uCz7GxZw8eeZKnKYlHD3aFOkROo2kJC95mIyMzUfG\n4RFtOXdUXEx5N8GxY8c0adIk5eXl6e6775Yk9enTR7t27ZIkVVdXKz09XWlpaaqrq1NLS4uampq0\nf/9+paamasCAAdq2bVvo2IEDB8rj8cjtduvAgQMyDEPbt29Xenq6BgwYoO3btysYDOqjjz5SMBjk\nEgEAAJfBlJ2BNWvW6PTp0yotLVVpaakkad68eXrqqae0cuVKXX/99Ro9erScTqeys7OVlZUlwzA0\na9YsdenSRZmZmcrPz1dmZqbcbreKi4slSQsXLtTs2bPV1tYmn8+nvn37SpLS09M1ceJEBYNBFRYW\nmnFKAABYlsMwDCPSQ0SCWVs7k5ZtMeVxzfTTOaMiPUKnEW3bftGIjM1HxuERbTmH/TIBAACIHpQB\nAABsjjIAAIDNUQYAALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYA\nALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNXVIZePvttz+3Vl9ff8WH\nAQAA4efq6Jt1dXUKBoMqKCjQ4sWLZRiGJKm1tVVFRUX63e9+F5YhAQCAeTosAzU1NaqtrdWRI0f0\n7LPP/s8PuVyaOHGi6cMBAADzdVgGcnNzJUkbN27U+PHjwzIQAAAIrw7LwKcGDRqk5cuX69SpU6FL\nBZK0dOlS0wYDAADhcUllYObMmUpPT1d6erocDofZMwEAgDC6pDLQ2tqq/Px8s2cBAAARcElvLRw4\ncKC2bNmiQCBg9jwAACDMLmln4Le//a1+8YtftFtzOBx68803TRkKAACEzyWVge3bt3+pB3/jjTf0\nH//xHyorK1NDQ4MefvhhXXvttZKkzMxMjRkzRuvWrVNFRYVcLpemTZumkSNH6ty5c8rLy9Px48cV\nFxen5cuXKzExUfX19Vq8eLGcTqd8Pp+mT58uSSopKdHWrVvlcrk0d+5cpaWlfal5AQCwo0sqAyUl\nJRdc//QP4wt54YUXtGnTJn31q1+VJO3du1ff+973NGnSpNAxR48eVVlZmSorK9XS0qKsrCxlZGSo\nvLxcqampys3NVVVVlUpLS1VQUKAFCxZo1apVSk5O1tSpU9XQ0CDDMFRbW6v169fr0KFDys3NVWVl\n5eVkAACArV32ZxOcP39eW7Zs0fHjxzs8LiUlRatWrQp9vWfPHm3dulX333+/5s6dK7/fr927d6t/\n//6KjY2V1+tVSkqKGhsbVVdXp2HDhkmShg8frh07dsjv9ysQCCglJUUOh0M+n081NTWqq6uTz+eT\nw+FQz5491dbWphMnTlzuaQEAYFuXtDPwv3cAcnJy2v0N/0JGjx6tDz74IPR1Wlqa7rnnHt166616\n7rnntHr1at18883yer2hY+Li4uT3++X3+0PrcXFxampqkt/vl8fjaXfswYMH1aVLF3Xr1q3delNT\nkxITEy/l1AAAsL1LKgP/W3Nzsz766KPL+plvfvObio+PD/160aJFSk9PV3Nzc7vH9Xq98ng8ofXm\n5mbFx8e3W/vsutvtvuBjXExCwlVyuZyXdQ5WlZR08bzshDzMR8bmI+PwsErOl1QGRo0aFbrZkGEY\nOn36tCZPnnxZTzR58mTNnz9faWlp2rFjh2655RalpaXpmWeeUUtLiwKBgPbv36/U1FQNGDBA27Zt\nU1pamqqrqzVw4EB5PB653W4dOHBAycnJ2r59u6ZPny6n06kVK1Zo8uTJOnz4sILB4CXtCpw8eeay\n5reyo0ebIj1Cp5GU5CUPk5Gx+cg4PKIt546KyyWVgbKystCvHQ5H6G/ql6OoqEiLFi2S2+1W9+7d\ntWjRInk8HmVnZysrK0uGYWjWrFnq0qWLMjMzlZ+fr8zMTLndbhUXF0uSFi5cqNmzZ6utrU0+n099\n+/aVJKWnp2vixIkKBoMqLCy8rLkAALA7h/HZDxv4AoZhqLy8XDt37lRra6uGDh2qBx54QDExl/36\nw07DrDY3adkWUx7XTD+dMyrSI3Qa0db0oxEZm4+MwyPacv4/7ww8/fTTev/99zVhwgQZhqENGzbo\n4MGDmjdv3hUbEgAARMYllYE//OEP2rhxY2gn4J//+Z81btw4UwcDAADhcUn7/G1tbWptbW33tdPJ\nK/EBALCCS9oZGDdunB588EHdcccdkqSqqiqNHTvW1MEAAEB4XLQMnDp1Svfee6969+6tnTt3ateu\nXXrwwQc1fvz4cMwHAABM1uFlgoaGBt1xxx3as2ePRowYofz8fPl8PhUXF6uxsTFcMwIAABN1WAaW\nL1+u4uJiDR8+PLT22GOPacmSJVq2bJnpwwEAAPN1WAZOnz6tIUOGfG592LBhOnnypGlDAQCA8Omw\nDLS2tioYDH5uPRgM6vz586YNBQAAwqfDMjBo0CCVlJR8br20tFS33nqraUMBAIDw6fDdBI899pim\nTp2qV155RV//+tdlGIYaGhqUmJio5557LlwzAgAAE3VYBjwej1566SXt3LlTb775pmJiYnT//fcr\nPT09XPMBAACTXfQ+Aw6HQ7fddptuu+22cMwDAADCLHo/dhAAAFwRlAEAAGyOMgAAgM1RBgAAsDnK\nAAAANkcZAADA5igDAADYHGUAAACbowwAAGBzlAEAAGyOMgAAgM1RBgAAsDnKAAAANkcZAADA5igD\nAADYnKll4I033lB2drYk6f3331dmZqaysrK0YMECBYNBSdK6det011136d5779Vrr70mSTp37pxy\nc3OVlZWlKVOm6MSJE5Kk+vp63XPPPbrvvvtUUlISep6SkhLdfffduu+++7R7924zTwkAAMsxrQy8\n8MILKigoUEtLiyRp6dKlmjlzptauXSvDMLR582YdPXpUZWVlqqio0E9+8hOtXLlSgUBA5eXlSk1N\n1dq1azV+/HiVlpZKkhYsWKDi4mKVl5frjTfeUENDg/bu3ava2lqtX79eK1eu1MKFC806JQAALMm0\nMpCSkqJVq1aFvt67d68GDx4sSRo+fLhqamq0e/du9e/fX7GxsfJ6vUpJSVFjY6Pq6uo0bNiw0LE7\nduyQ3+9XIBBQSkqKHA6HfD6fampqVFdXJ5/PJ4fDoZ49e6qtrS20kwAAAC7OtDIwevRouVyu0NeG\nYcjhcEiS4uLi1NTUJL/fL6/XGzomLi5Ofr+/3fpnj/V4PO2O7WgdAABcGtfFD7kyYmL+p3c0Nzcr\nPj5eHo9Hzc3N7da9Xm+79Y6OjY+Pl9vtvuBjXExCwlVyuZxX4tSiXlLSxfOyE/IwHxmbj4zDwyo5\nh60M9OnTR7t27dKQIUNUXV2toUOHKi0tTc8884xaWloUCAS0f/9+paamasCAAdq2bZvS0tJUXV2t\ngQMHyuPxyO1268CBA0pOTtb27ds1ffp0OZ1OrVixQpMnT9bhw4cVDAaVmJh40XlOnjwThrOODkeP\nspPyqaQkL3mYjIzNR8bhEW05d1RcwlYG8vPzNX/+fK1cuVLXX3+9Ro8eLafTqezsbGVlZckwDM2a\nNUtdunRRZmam8vPzlZmZKbfbreLiYknSwoULNXv2bLW1tcnn86lv376SpPT0dE2cOFHBYFCFhYXh\nOiUAACzBYRiGEekhIsGsNjdp2RZTHtdMP50zKtIjdBrR1vSjERmbj4zDI9py7mhngJsOAQBgc5QB\nAABsjjIAAIDNUQYAALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYA\nALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYAALA5ygAAADZHGQAA\nwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYAALA5ygAAADbnCvcT3nnnnfJ4PJKka665\nRo888ojmzJkjh8Ohm266SQsWLFBMTIzWrVuniooKuVwuTZs2TSNHjtS5c+eUl5en48ePKy4uTsuX\nL1diYqLq6+u1ePFiOZ1O+Xw+TZ8+PdynBQBA1AprGWhpaZFhGCorKwutPfLII5o5c6aGDBmiwsJC\nbd68Wf369VNZWZkqKyvV0tKirKwsZWRkqLy8XKmpqcrNzVVVVZVKS0tVUFCgBQsWaNWqVUpOTtbU\nqVPV0NCgPn36hPPUAACIWmG9TNDY2KizZ89q0qRJevDBB1VfX6+9e/dq8ODBkqThw4erpqZGu3fv\nVv/+/RUbGyuv16uUlBQ1Njaqrq5Ow4YNCx27Y8cO+f1+BQIBpaSkyOFwyOfzqaamJpynBQBAVAvr\nzsBXvvIVTZ48Wffcc4/ee+89TZkyRYZhyOFwSJLi4uLU1NQkv98vr9cb+rm4uDj5/f5265899tPL\nDp+uHzx48KKzJCRcJZfLeYXPMDolJXkvfpCNkIf5yNh8ZBweVsk5rGXguuuuU69eveRwOHTdddep\nW7du2rt3b+j7zc3Nio+Pl8fjUXNzc7t1r9fbbr2jY+Pj4y86y8mTZ67gmUW3o0ebIj1Cp5GU5CUP\nk5Gx+cg4PKIt546KS1gvE/zyl7/UsmXLJEkff/yx/H6/MjIytGvXLklSdXW10tPTlZaWprq6OrW0\ntKipqUn79+9XamqqBgwYoG3btoWOHThwoDwej9xutw4cOCDDMLR9+3alp6eH87QAAIhqYd0ZuPvu\nu/XEE08oMzNTDodDS5YsUUJCgubPn6+VK1fq+uuv1+jRo+V0OpWdna2srCwZhqFZs2apS5cuyszM\nVH5+vjIzM+V2u1VcXCxJWrhwoWbPnq22tjb5fD717ds3nKcFAEBUcxiGYUR6iEgwa2tn0rItpjyu\nmX46Z1SkR+g0om3bLxqRsfnIODyiLedOc5kAAAB0PpQBAABsjjIAAIDNUQYAALA5ygAAADZHGQAA\nwOYoAwAA2BxlAAAAm6MMAABgc5QBAABsjjIAAIDNUQYAALA5ygAAADZHGQAAwOYoAwAA2BxlAAAA\nm6MMAABgc5QBAABsjjIAAIDNUQYAALA5ygAAADZHGQAAwOYoAwAA2BxlAAAAm6MMAABgc5QBAABs\njjIAAIDNUQYAALA5V6QHuFKCwaCKioq0b98+xcbG6qmnnlKvXr0iPRYAAJ2eZXYGfv/73ysQCOjl\nl1/Wv//7v2vZsmWRHgkAgKhgmZ2Buro6DRs2TJLUr18/7dmzJ8ITwUyTlm2J9AiX5adzRkV6BAD4\nQpYpA36/Xx6PJ/S10+lUa2urXC7LnCIQVhQu80VbxlL05UzGl8ZhGIYR9mc1wdKlS9W3b1+NGTNG\nkjR8+HBVV1dHeCoAADo/y7xmYMCAAaE//Ovr65WamhrhiQAAiA6W2Rn49N0Eb731lgzD0JIlS3TD\nDTdEeiwAADo9y5QBAADw5VjmMgEAAPhyKAMAANgcZQAAAJujDAAAYHOUAQAAbI4y0Mm99dZbysrK\n0tixY/X888/rtddei/RIltHS0qJf/OIXWr9+vQKBQGi9oqIiglMB/zcvvPBCpEewrMbGRh04cECB\nQEAlJSVavXq1zp49G+mxrgjeWtjJ/du//ZuefPJJFRQU6Nlnn9VDDz2kDRs2RHosS3j00UfVq1cv\ntba2qra2Vj/5yU/UtWtXPfjgg/r5z38e6fEsYfv27V/4PZ/PF8ZJrOuxxx6Tw+GQJBmGoV27dmno\n0KGSpOLi4kiOZinFxcV644035Pf7lZSUpN69eysuLk6NjY2WyJkb90eBXr16yeFwKDExUXFxcZEe\nxzJOnDihZ599VpL06quvatq0afrZz34m+vGVs27dOu3Zs0dDhgz53PcoA1dGamqqtm3bphkzZigm\nJkbvvPOOJk6cGOmxLOePf/yjKioq1NzcrHHjxulHP/qRJCk7OzvCk10ZlIFOrmvXrqqoqNDZs2dV\nVVWl+Pj4SI9kGefPn9eJEyeUmJiob33rW/roo480e/ZsnT9/PtKjWcYPfvADPfDAA5oyZYquv/76\nSI9jSY888oh69+6tl156SU8++aTi4+M1ePDgSI9lOcFgUB999JF69uypH/zgB5Kk06dPt7vEGM14\nzUAnt2TJEn3wwQdKSEjQnj17tHjx4kiPZBmPPvqo7r//fh07dkyS9N3vfle9e/fm46+vIKfTqaef\nfpqCZbIRI0bo8ccfV15enk6fPh3pcSzp8ccfV25uroLBoPr27StJmjZtmh5++OEIT3Zl8JqBKHD8\n+HG1tLSEvu7Zs2cEp7GmYDCoEydO6Oqrrw79G1fWZzP+9Bo3rqwzZ85oy5YtuuOOO8jYRFb83zJl\noJMrKipSdXW1evToIcMw5HA4eLX7Ffbqq69q2bJlio+PV3Nzs4qKipSRkRHpsSyFjM1HxuFh2ZwN\ndGp33nmn0dbWFukxLO073/mOcezYMcMwDOPo0aPGhAkTIjyR9ZCx+cg4PKyaM68Z6OR69erV7hIB\nrrxu3bqFLgt0795dHo8nwhNZDxmbj4zDw6o5c5mgk7vvvvv03nvvqVevXpLEZQIT5OTk6Ny5cxo0\naJD27NmjY8eOhV6N/dhjj0V4OmsgY/ORcXhYNWfeWtjJWeFmFp3d7bffHvr13/3d30VwEusiY/OR\ncXhYNWd2Bjq5w4cPa8mSJdq/f7+uvfZaPfHEE7rmmmsiPZalNDU1qba2tt3lmDFjxkRwIushY/OR\ncXhYNWd2Bjq5goICZWZmatCgQaqtrdW8efP0n//5n5Eey1ImTZqkG2+8UV6vV9LfLsVY4Td3Z0LG\n5iPj8LBqzpSBTq6lpUXf+MY3JP1te+rFF1+M8ETW4/V6tXTp0kiPYWlkbD4yDg+r5kwZ6OTa2tq0\nb98+/eM//qP27dtnmRtcdCY+n0/l5eW68cYbQ2uDBg2K4ETWQ8bmI+PwsGrOlIFObv78+Zo3b56O\nHDmiHj166Kmnnor0SJbzpz/9SYFAQH/84x8l/W3bzwq/uTsTMjYfGYeHVXOmDHRyjY2Nam5ulsvl\n0okTJ5STk6PNmzdHeixLOXPmjH72s59FegxLI2PzkXF4WDVnykAn9+Mf/1hr1qzR1772tUiPYlk3\n3XSTqqqq1Lt379BlmOuuuy7CU1kLGZuPjMPDqjlTBjq55OTk0A2HYI7GxkY1NjaGvnY4HPr5z38e\nwYmsh4zNR8bhYdWcuc9AJzdz5kz5/f52LTSa73LVWZ08eVIHDx7UNddco8TExEiPY0lkbD4yDg8r\n5uwsKioqivQQ+GJnz55VamqqEhISQv/07t070mNZyn/9139p9uzZeuedd/SjH/1IXbt21c033xzp\nsSyFjM1HxuFh2Zwj+SlJQGdw7733Gn6/3zAMw2hqajLuuuuuCE9kPWRsPjIOD6vmzKcWwvYcDofi\n4uIkSR6PR126dInwRNZDxuYj4/Cwas68gBC2l5ycrGXLlik9PV11dXVKSUmJ9EiWQ8bmI+PwsGrO\n7AzA9iZOnKiuXbuqpqZGGzZs0P333x/pkSyHjM1HxuFh2ZwjfZ0CiLS77rrLeP/99w3DMIwDBw4Y\nWVlZEZ7IesjYfGQcHlbNmZ0B2J7b7Q5t9SUnJysmht8WVxoZm4+Mw8OqOfOaAdhez549tXLlSvXr\n10+7d+9Wjx49Ij2S5ZCx+cg4PKyaMzcdgu21tLSovLxc7777rm644Qbdd999io2NjfRYlkLG5iPj\n8LBqzpQBAABszhoXOwAAwJdGGQAAwOYoA4AF+P1+LVy4UGPHjtV3vvMdZWdna+/evZEe6/+kqalJ\n3//+9y/4vR/+8If6xje+oRdffPELf37KlCn6+OOPtWHDBs2ZM6fdGoD2eDcBEOWCwaCmTJmiIUOG\naOPGjXK5XNq5c6emTJmiqqoqJSQkRHrEL+XUqVPtPir2s37961/rxz/+cYefI//CCy9c0hoAygAQ\n9Xbt2qUjR45oxowZofc8Dx06VEuXLlUwGJQkrVmzRps2bZLT6VRGRoby8vJ06NAh5eTkKDk5WW+9\n9ZZuvfVWDR48WL/61a906tQprV69WjfccINGjRqlUaNG6U9/+pMkacmSJerTp4/effddFRYW6pNP\nPtFVV12lefPmKS0tTXPmzJHH49HevXv18ccfKycnRxMmTFBzc7OefPJJvf3222pra9OUKVM0duxY\nbdiwQa+//rpOnTqlgwcPKiMjQ0VFRXrqqad05MgR5eTkaPXq1aHzLSwsDD1ucXGx6urq9Otf/1pn\nz56Vw+HQM888E5r7f3/O/KdrtbW1F3xOSSouLtbvfvc7JSQkKCkpSaNGjdJdd90Vhv+SQORwmQCI\ncg0NDfr617/+uZufjBgxQldffbW2bdumLVu2aMOGDfrVr36l999/XxUVFZKkffv26fvf/75++9vf\n6q9//asBdgFpAAAEDklEQVQ+/PBDvfzyyxo7dqxefvnl0GN169ZNGzdu1IwZM5Sfny9JysvLU3Z2\ntl555RU98cQTevTRRxUIBCRJhw8f1tq1a/Xcc8/p6aefliQ999xzuuWWW7Rhwwa99NJLWrNmjQ4e\nPChJ+stf/qIf/vCH2rRpk1577TXt27dPBQUF6tGjR7siIElPPvmkevTooeeff17Jycn6/e9/r7Ky\nMv3mN7/R7bffrrVr115Sbhd6zi1btqiurk6/+c1v9Pzzz6uhoeFL/BcBog87A0CUi4mJUUfvEN65\nc6fuuOMOfeUrX5EkTZgwQRs3btSIESPUvXt39enTR5L093//97rtttsk/e3GKh988EHoMe69915J\nf/ub9Zw5c3T48GEdOHBA3/rWtyRJ/fr1U9euXfXOO+9IkjIyMuRwOJSamqpPPvlEklRTU6Nz586p\nsrJSknTmzBm9/fbbkqT+/fvL4/FI+ttd3U6dOhX6ZLiOeDweFRcXq6qqSu+9955ef/119e7d+5Jy\nu9Bz1tTU6F/+5V8UGxur2NhY3X777Zf0WEC0owwAUe7WW2/V2rVrZRiGHA5HaH3lypX6p3/6p9Cl\ngs9qbW2VpM/dLMXpdF7wOVyu//m/imAwqLa2ts8VEMMw1NbWJkmhj3X97DzBYFArVqzQLbfcIkk6\nduyYunbtqldeeaXdx8A6HI4Oy81nHTp0SNnZ2XrggQc0fPhwde/eXW+++eYl/eyFnjMmJuaCeQFW\nx2UCIMqlp6fr6quvVklJSegP49dff10bNmzQjTfeqKFDh6qqqkrnzp1Ta2urKisrNXTo0Mt6jqqq\nKknSf//3f+uGG27QP/zDPyg5OVmvvvqqJKm+vl7Hjh3TTTfd9IWPMXToUJWXl0uSjhw5on/913/V\noUOHvvB4l8sVKi1f5K9//at69eql7373u+rbt6+qq6tDGXwZGRkZevXVVxUIBOT3+7V169Z2hQaw\nKnYGgCjncDhUWlqqpUuXauzYsXK5XEpISNDzzz+v7t27a+TIkXrzzTc1YcIEtba2atiwYXrggQd0\n+PDhS36OP//5z/rlL3+pr371q1q2bJkkacWKFSoqKtKqVavkdru1atWqDm/LOn36dBUVFWns2LFq\na2tTXl6eUlJSQi9M/N+uvvpq9ezZU9nZ2SorK7vgMRkZGSovL9eYMWMUGxurtLS00KWHL2PEiBH6\n85//rDvvvFNdu3ZVjx492u0gAFbF7YgBdOjTV+Bfc801kR7FdH/5y1/03nvv6c4779T58+c1ceJE\nLVmyRDfffHOkRwNMxc4AAPx/1113nUpKSvTiiy/KMAyNHz+eIgBbYGcAAACb4wWEAADYHGUAAACb\nowwAAGBzlAEAAGyOMgAAgM1RBgAAsLn/BztoZ4LLGwd8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2564fe6e6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(8, 4))\n",
    "labeled_features['failure'].value_counts().plot(kind='bar')\n",
    "plt.xlabel('Component failing')\n",
    "plt.ylabel('Count')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also, due to the class imbalance problem, it is important to look at evaluation metrics other than accuracy alone and compare those metrics to the baseline metrics which are computed when random chance is used to make predictions rather than a machine learning model. The comparison will bring out the value and benefits of using a machine learning model better.\n",
    "\n",
    "In the following, we use an evaluation function that computes many important evaluation metrics along with baseline metrics for classification problems. For a detailed explanation of the metrics, please refer to the scikit-learn documentation and a companion blog post (with examples in R, not Python),"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix, recall_score, accuracy_score, precision_score\n",
    "\n",
    "def Evaluate(predicted, actual, labels):\n",
    "    output_labels = []\n",
    "    output = []\n",
    "    \n",
    "    # Calculate and display confusion matrix\n",
    "    cm = confusion_matrix(actual, predicted, labels=labels)\n",
    "    print('Confusion matrix\\n- x-axis is true labels (none, comp1, etc.)\\n- y-axis is predicted labels')\n",
    "    print(cm)\n",
    "    \n",
    "    # Calculate precision, recall, and F1 score\n",
    "    accuracy = np.array([float(np.trace(cm)) / np.sum(cm)] * len(labels))\n",
    "    precision = precision_score(actual, predicted, average=None, labels=labels)\n",
    "    recall = recall_score(actual, predicted, average=None, labels=labels)\n",
    "    f1 = 2 * precision * recall / (precision + recall)\n",
    "    output.extend([accuracy.tolist(), precision.tolist(), recall.tolist(), f1.tolist()])\n",
    "    output_labels.extend(['accuracy', 'precision', 'recall', 'F1'])\n",
    "    \n",
    "    # Calculate the macro versions of these metrics\n",
    "    output.extend([[np.mean(precision)] * len(labels),\n",
    "                   [np.mean(recall)] * len(labels),\n",
    "                   [np.mean(f1)] * len(labels)])\n",
    "    output_labels.extend(['macro precision', 'macro recall', 'macro F1'])\n",
    "    \n",
    "    # Find the one-vs.-all confusion matrix\n",
    "    cm_row_sums = cm.sum(axis = 1)\n",
    "    cm_col_sums = cm.sum(axis = 0)\n",
    "    s = np.zeros((2, 2))\n",
    "    for i in range(len(labels)):\n",
    "        v = np.array([[cm[i, i],\n",
    "                       cm_row_sums[i] - cm[i, i]],\n",
    "                      [cm_col_sums[i] - cm[i, i],\n",
    "                       np.sum(cm) + cm[i, i] - (cm_row_sums[i] + cm_col_sums[i])]])\n",
    "        s += v\n",
    "    s_row_sums = s.sum(axis = 1)\n",
    "    \n",
    "    # Add average accuracy and micro-averaged  precision/recall/F1\n",
    "    avg_accuracy = [np.trace(s) / np.sum(s)] * len(labels)\n",
    "    micro_prf = [float(s[0,0]) / s_row_sums[0]] * len(labels)\n",
    "    output.extend([avg_accuracy, micro_prf])\n",
    "    output_labels.extend(['average accuracy',\n",
    "                          'micro-averaged precision/recall/F1'])\n",
    "    \n",
    "    # Compute metrics for the majority classifier\n",
    "    mc_index = np.where(cm_row_sums == np.max(cm_row_sums))[0][0]\n",
    "    cm_row_dist = cm_row_sums / float(np.sum(cm))\n",
    "    mc_accuracy = 0 * cm_row_dist; mc_accuracy[mc_index] = cm_row_dist[mc_index]\n",
    "    mc_recall = 0 * cm_row_dist; mc_recall[mc_index] = 1\n",
    "    mc_precision = 0 * cm_row_dist\n",
    "    mc_precision[mc_index] = cm_row_dist[mc_index]\n",
    "    mc_F1 = 0 * cm_row_dist;\n",
    "    mc_F1[mc_index] = 2 * mc_precision[mc_index] / (mc_precision[mc_index] + 1)\n",
    "    output.extend([mc_accuracy.tolist(), mc_recall.tolist(),\n",
    "                   mc_precision.tolist(), mc_F1.tolist()])\n",
    "    output_labels.extend(['majority class accuracy', 'majority class recall',\n",
    "                          'majority class precision', 'majority class F1'])\n",
    "        \n",
    "    # Random accuracy and kappa\n",
    "    cm_col_dist = cm_col_sums / float(np.sum(cm))\n",
    "    exp_accuracy = np.array([np.sum(cm_row_dist * cm_col_dist)] * len(labels))\n",
    "    kappa = (accuracy - exp_accuracy) / (1 - exp_accuracy)\n",
    "    output.extend([exp_accuracy.tolist(), kappa.tolist()])\n",
    "    output_labels.extend(['expected accuracy', 'kappa'])\n",
    "    \n",
    "\n",
    "    # Random guess\n",
    "    rg_accuracy = np.ones(len(labels)) / float(len(labels))\n",
    "    rg_precision = cm_row_dist\n",
    "    rg_recall = np.ones(len(labels)) / float(len(labels))\n",
    "    rg_F1 = 2 * cm_row_dist / (len(labels) * cm_row_dist + 1)\n",
    "    output.extend([rg_accuracy.tolist(), rg_precision.tolist(),\n",
    "                   rg_recall.tolist(), rg_F1.tolist()])\n",
    "    output_labels.extend(['random guess accuracy', 'random guess precision',\n",
    "                          'random guess recall', 'random guess F1'])\n",
    "    \n",
    "    # Random weighted guess\n",
    "    rwg_accuracy = np.ones(len(labels)) * sum(cm_row_dist**2)\n",
    "    rwg_precision = cm_row_dist\n",
    "    rwg_recall = cm_row_dist\n",
    "    rwg_F1 = cm_row_dist\n",
    "    output.extend([rwg_accuracy.tolist(), rwg_precision.tolist(),\n",
    "                   rwg_recall.tolist(), rwg_F1.tolist()])\n",
    "    output_labels.extend(['random weighted guess accuracy',\n",
    "                          'random weighted guess precision',\n",
    "                          'random weighted guess recall',\n",
    "                          'random weighted guess F1'])\n",
    "\n",
    "    output_df = pd.DataFrame(output, columns=labels)\n",
    "    output_df.index = output_labels\n",
    "                  \n",
    "    return output_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Split 1:\n",
      "Confusion matrix\n",
      "- x-axis is true labels (none, comp1, etc.)\n",
      "- y-axis is predicted labels\n",
      "[[119594     21      0      4      3]\n",
      " [    18    515      3      5      1]\n",
      " [     0      0    860      0      1]\n",
      " [    13      0      2    372      1]\n",
      " [     2      2      6      0    497]]\n",
      "\n",
      "Split 2:\n",
      "Confusion matrix\n",
      "- x-axis is true labels (none, comp1, etc.)\n",
      "- y-axis is predicted labels\n",
      "[[95266    13     0     4     3]\n",
      " [   19   399     2     1     1]\n",
      " [    0     0   700     0     0]\n",
      " [   12     0     2   291     1]\n",
      " [    2     2     4     0   392]]\n",
      "\n",
      "Split 3:\n",
      "Confusion matrix\n",
      "- x-axis is true labels (none, comp1, etc.)\n",
      "- y-axis is predicted labels\n",
      "[[71724     7     0     4     3]\n",
      " [   17   300     1     1     1]\n",
      " [    0     1   547     0     0]\n",
      " [   11     0     0   212     1]\n",
      " [    2     1     3     0   274]]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>none</th>\n",
       "      <th>comp1</th>\n",
       "      <th>comp2</th>\n",
       "      <th>comp3</th>\n",
       "      <th>comp4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>accuracy</th>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.999724</td>\n",
       "      <td>0.957249</td>\n",
       "      <td>0.987371</td>\n",
       "      <td>0.976378</td>\n",
       "      <td>0.988072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.999766</td>\n",
       "      <td>0.950185</td>\n",
       "      <td>0.998839</td>\n",
       "      <td>0.958763</td>\n",
       "      <td>0.980276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>F1</th>\n",
       "      <td>0.999745</td>\n",
       "      <td>0.953704</td>\n",
       "      <td>0.993072</td>\n",
       "      <td>0.967490</td>\n",
       "      <td>0.984158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>macro precision</th>\n",
       "      <td>0.981759</td>\n",
       "      <td>0.981759</td>\n",
       "      <td>0.981759</td>\n",
       "      <td>0.981759</td>\n",
       "      <td>0.981759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>macro recall</th>\n",
       "      <td>0.977566</td>\n",
       "      <td>0.977566</td>\n",
       "      <td>0.977566</td>\n",
       "      <td>0.977566</td>\n",
       "      <td>0.977566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>macro F1</th>\n",
       "      <td>0.979634</td>\n",
       "      <td>0.979634</td>\n",
       "      <td>0.979634</td>\n",
       "      <td>0.979634</td>\n",
       "      <td>0.979634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average accuracy</th>\n",
       "      <td>0.999731</td>\n",
       "      <td>0.999731</td>\n",
       "      <td>0.999731</td>\n",
       "      <td>0.999731</td>\n",
       "      <td>0.999731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>micro-averaged precision/recall/F1</th>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "      <td>0.999327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>majority class accuracy</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>majority class recall</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>majority class precision</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>majority class F1</th>\n",
       "      <td>0.990486</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>expected accuracy</th>\n",
       "      <td>0.962796</td>\n",
       "      <td>0.962796</td>\n",
       "      <td>0.962796</td>\n",
       "      <td>0.962796</td>\n",
       "      <td>0.962796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kappa</th>\n",
       "      <td>0.981922</td>\n",
       "      <td>0.981922</td>\n",
       "      <td>0.981922</td>\n",
       "      <td>0.981922</td>\n",
       "      <td>0.981922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random guess accuracy</th>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random guess precision</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.004446</td>\n",
       "      <td>0.007062</td>\n",
       "      <td>0.003182</td>\n",
       "      <td>0.004158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random guess recall</th>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random guess F1</th>\n",
       "      <td>0.332269</td>\n",
       "      <td>0.008698</td>\n",
       "      <td>0.013642</td>\n",
       "      <td>0.006265</td>\n",
       "      <td>0.008148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random weighted guess accuracy</th>\n",
       "      <td>0.962755</td>\n",
       "      <td>0.962755</td>\n",
       "      <td>0.962755</td>\n",
       "      <td>0.962755</td>\n",
       "      <td>0.962755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random weighted guess precision</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.004446</td>\n",
       "      <td>0.007062</td>\n",
       "      <td>0.003182</td>\n",
       "      <td>0.004158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random weighted guess recall</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.004446</td>\n",
       "      <td>0.007062</td>\n",
       "      <td>0.003182</td>\n",
       "      <td>0.004158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>random weighted guess F1</th>\n",
       "      <td>0.981152</td>\n",
       "      <td>0.004446</td>\n",
       "      <td>0.007062</td>\n",
       "      <td>0.003182</td>\n",
       "      <td>0.004158</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        none     comp1     comp2     comp3  \\\n",
       "accuracy                            0.999327  0.999327  0.999327  0.999327   \n",
       "precision                           0.999724  0.957249  0.987371  0.976378   \n",
       "recall                              0.999766  0.950185  0.998839  0.958763   \n",
       "F1                                  0.999745  0.953704  0.993072  0.967490   \n",
       "macro precision                     0.981759  0.981759  0.981759  0.981759   \n",
       "macro recall                        0.977566  0.977566  0.977566  0.977566   \n",
       "macro F1                            0.979634  0.979634  0.979634  0.979634   \n",
       "average accuracy                    0.999731  0.999731  0.999731  0.999731   \n",
       "micro-averaged precision/recall/F1  0.999327  0.999327  0.999327  0.999327   \n",
       "majority class accuracy             0.981152  0.000000  0.000000  0.000000   \n",
       "majority class recall               1.000000  0.000000  0.000000  0.000000   \n",
       "majority class precision            0.981152  0.000000  0.000000  0.000000   \n",
       "majority class F1                   0.990486  0.000000  0.000000  0.000000   \n",
       "expected accuracy                   0.962796  0.962796  0.962796  0.962796   \n",
       "kappa                               0.981922  0.981922  0.981922  0.981922   \n",
       "random guess accuracy               0.200000  0.200000  0.200000  0.200000   \n",
       "random guess precision              0.981152  0.004446  0.007062  0.003182   \n",
       "random guess recall                 0.200000  0.200000  0.200000  0.200000   \n",
       "random guess F1                     0.332269  0.008698  0.013642  0.006265   \n",
       "random weighted guess accuracy      0.962755  0.962755  0.962755  0.962755   \n",
       "random weighted guess precision     0.981152  0.004446  0.007062  0.003182   \n",
       "random weighted guess recall        0.981152  0.004446  0.007062  0.003182   \n",
       "random weighted guess F1            0.981152  0.004446  0.007062  0.003182   \n",
       "\n",
       "                                       comp4  \n",
       "accuracy                            0.999327  \n",
       "precision                           0.988072  \n",
       "recall                              0.980276  \n",
       "F1                                  0.984158  \n",
       "macro precision                     0.981759  \n",
       "macro recall                        0.977566  \n",
       "macro F1                            0.979634  \n",
       "average accuracy                    0.999731  \n",
       "micro-averaged precision/recall/F1  0.999327  \n",
       "majority class accuracy             0.000000  \n",
       "majority class recall               0.000000  \n",
       "majority class precision            0.000000  \n",
       "majority class F1                   0.000000  \n",
       "expected accuracy                   0.962796  \n",
       "kappa                               0.981922  \n",
       "random guess accuracy               0.200000  \n",
       "random guess precision              0.004158  \n",
       "random guess recall                 0.200000  \n",
       "random guess F1                     0.008148  \n",
       "random weighted guess accuracy      0.962755  \n",
       "random weighted guess precision     0.004158  \n",
       "random weighted guess recall        0.004158  \n",
       "random weighted guess F1            0.004158  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluation_results = []\n",
    "for i, test_result in enumerate(test_results):\n",
    "    print('\\nSplit %d:' % (i+1))\n",
    "    evaluation_result = Evaluate(actual = test_result['failure'],\n",
    "                                 predicted = test_result['predicted_failure'],\n",
    "                                 labels = ['none', 'comp1', 'comp2', 'comp3', 'comp4'])\n",
    "    evaluation_results.append(evaluation_result)\n",
    "evaluation_results[0]  # show full results for first split only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": 1
}