presentation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "metadata": false
    }
   },
   "source": [
    "# EnBw DataBiz Challenge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "metadata": false,
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import datetime\n",
    "from utils.datetime_utils import day_of_year_to_date\n",
    "from utils.io import read_enbw_dataset\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from visualize import visualize_household\n",
    "\n",
    "filepath = 'data/hackathon_EnBW_smart_meter_data_30_hh.csv'\n",
    "data = read_enbw_dataset(filepath)\n",
    "\n",
    "data.head()\n",
    "\n",
    "households = data.groupby('id')\n",
    "household_id = 3\n",
    "# Get a random household\n",
    "group = households.get_group(household_id)\n",
    "group.set_index(group.timestampLocal, inplace=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "metadata": false,
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Analyze typical day of household"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "metadata": false,
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f2be959c5f8>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "resample = group.resample('1H').mean()\n",
    "resample['hour_idx'] = resample.index.hour\n",
    "resample.set_index('hour_idx', inplace=True)\n",
    "resample.index.name = 'hour_idx'\n",
    "resample['Hour of the day'] = resample.index \n",
    "sns.lineplot(x='Hour of the day', y='value', data=resample)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "metadata": false,
     "name": "#%% md\n"
    }
   },
   "source": [
    "Um Anomalien zu entdecken, die auf unerwartet niedrigen Stromverbrauch hinweisen, zu finden, haben wir ersteinmal\n",
    "einige Annahmen getroffen. \n",
    "\n",
    "Annahme 1: Nachts am frühen Morgen wird konsistent weniger Strom verbraucht: Benutze \"Schlafverbrauch\" \n",
    "als Referenzsignal für Anomalie."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "metadata": false,
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mean     209.917797\n",
       "min       54.777500\n",
       "max     1618.440000\n",
       "std      112.795392\n",
       "dtype: float64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "grouped_by_hour = group.groupby(lambda x: pd.to_timedelta(x.hour, unit='H'))\n",
    "average_day = grouped_by_hour.value.agg(['mean', 'min', 'max', 'std'])\n",
    "assumed_inactivity = average_day.sort_values(by='mean').iloc[0:4, :].mean()\n",
    "assumed_inactivity_std = assumed_inactivity['std']\n",
    "overall_std = group.value.std()\n",
    "assumed_inactivity\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "metadata": false,
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Seasonal Trends\n",
    "Annahme 2: Der Grundverbrauch verändert sich mittelfristig über das Jahr, bedingt durch Feiertage, \n",
    "Schulferien und Klima: der Grundverbrauch. \n",
    "\n",
    "Extrahiere saisonale Baseline um unerwartetes Verhalten besser einzuschätzen.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "metadata": false,
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/simone/.local/share/virtualenvs/enbw-challenge-0BWRMCLz/lib/python3.6/site-packages/ipykernel_launcher.py:41: RuntimeWarning:\n",
      "\n",
      "invalid value encountered in less\n",
      "\n",
      "/home/simone/.local/share/virtualenvs/enbw-challenge-0BWRMCLz/lib/python3.6/site-packages/ipykernel_launcher.py:41: RuntimeWarning:\n",
      "\n",
      "invalid value encountered in less\n",
      "\n"
     ]
    }
   ],
   "source": [
    "inactivity_anomaly_vectors = []\n",
    "is_nocturnal_activity_vectors = []\n",
    "\n",
    "# for day, day_group in grouped_by_day:\n",
    "day_manual = 0\n",
    "# day = 229\n",
    "\n",
    "seasonal_mean = group.resample('1W').mean()\n",
    "grouped_by_day = group.groupby(lambda x: x.dayofyear)\n",
    "\n",
    "# Parameters :)\n",
    "alpha = 0.005\n",
    "beta = 1 / 2\n",
    "m = np.reshape([1, 2], (2, 1))\n",
    "\n",
    "for day, day_group in grouped_by_day:\n",
    "    try:\n",
    "        if day_manual > 0:\n",
    "            day_group = grouped_by_day.get_group(day_manual)\n",
    "        day_statistics = day_group.resample('1H').value.agg(['mean', 'min', 'max', 'std'])\n",
    "        day_statistics['offset'] = pd.to_timedelta(day_statistics.index.hour, unit='H')\n",
    "\n",
    "        date = day_of_year_to_date(day, 2014)\n",
    "        seasonal_mean['sort_val'] = abs((seasonal_mean.index - date).days)\n",
    "        unexpected_signal_threshold = seasonal_mean.sort_values('sort_val').iloc[0]['value'] + beta * assumed_inactivity_std\n",
    "\n",
    "        merged_statistics = day_statistics.merge(average_day, how='left', left_on=day_statistics.offset,\n",
    "                                                 right_on=average_day.index)\n",
    "\n",
    "        expectations_to_inactivity = (\n",
    "                    (average_day['mean'] - assumed_inactivity['mean']) / assumed_inactivity['std']).values\n",
    "        is_unexpected = average_day['mean'] > unexpected_signal_threshold\n",
    "\n",
    "        deviation_from_inactivity = (\n",
    "                (merged_statistics['mean_x'] - assumed_inactivity['mean']) / assumed_inactivity['std']).values\n",
    "        # comparison_to_night_activity.index.name = 'index'\n",
    "\n",
    "        deviation_from_expectation = ((merged_statistics['mean_x'] - merged_statistics['mean_y']) / merged_statistics[\n",
    "            'mean_y']).values\n",
    "\n",
    "        is_inactivity = deviation_from_inactivity < alpha * overall_std\n",
    "        is_inactivity_indices, = np.where(deviation_from_inactivity)\n",
    "\n",
    "        inactivity_anomaly_vector = np.sum(np.vstack((is_unexpected, is_inactivity)).astype(np.int32) * m, axis=0)\n",
    "        inactivity_anomaly_vectors.append(inactivity_anomaly_vector)\n",
    "\n",
    "        # 2. anomaly: nocturnal activity\n",
    "        # nocturnal_activity_vector = np.logical_and(deviation_from_expectation > gamma, night_hours).astype(np.int)\n",
    "        # is_nocturnal_activity_vectors.append(nocturnal_activity_vector)\n",
    "        if day_manual > 0:\n",
    "            break\n",
    "    except ValueError:\n",
    "        continue\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "metadata": false
    }
   },
   "source": [
    "# Ergebnisse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "pycharm": {
     "is_executing": false,
     "metadata": false,
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/simone/projects/enbw-challenge/enbw-challenge/visualize.py:29: SettingWithCopyWarning:\n",
      "\n",
      "\n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "\n",
      "/home/simone/.local/share/virtualenvs/enbw-challenge-0BWRMCLz/lib/python3.6/site-packages/pandas/plotting/_converter.py:129: FutureWarning:\n",
      "\n",
      "Using an implicitly registered datetime converter for a matplotlib plotting method. The converter was registered by pandas on import. Future versions of pandas will require you to explicitly register matplotlib converters.\n",
      "\n",
      "To register the converters:\n",
      "\t>>> from pandas.plotting import register_matplotlib_converters\n",
      "\t>>> register_matplotlib_converters()\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            timestampLocal  id    value  sun  holiday      time        date\n",
      "67683  2014-01-01 00:15:00   3   453.52    0        1  00:15:00  2014-01-01\n",
      "67684  2014-01-01 00:30:00   3   500.44    0        1  00:30:00  2014-01-01\n",
      "67685  2014-01-01 00:45:00   3   621.51    0        1  00:45:00  2014-01-01\n",
      "67686  2014-01-01 01:00:00   3   699.91    0        1  01:00:00  2014-01-01\n",
      "67687  2014-01-01 01:15:00   3   467.25    0        1  01:15:00  2014-01-01\n",
      "67688  2014-01-01 01:30:00   3   437.98    0        1  01:30:00  2014-01-01\n",
      "67689  2014-01-01 01:45:00   3   498.41    0        1  01:45:00  2014-01-01\n",
      "67690  2014-01-01 02:00:00   3   434.74    0        1  02:00:00  2014-01-01\n",
      "67691  2014-01-01 02:15:00   3   304.03    0        1  02:15:00  2014-01-01\n",
      "67692  2014-01-01 02:30:00   3   274.08    0        1  02:30:00  2014-01-01\n",
      "67693  2014-01-01 02:45:00   3   232.53    0        1  02:45:00  2014-01-01\n",
      "67694  2014-01-01 03:00:00   3   199.32    0        1  03:00:00  2014-01-01\n",
      "67695  2014-01-01 03:15:00   3   179.88    0        1  03:15:00  2014-01-01\n",
      "67696  2014-01-01 03:30:00   3   133.97    0        1  03:30:00  2014-01-01\n",
      "67697  2014-01-01 03:45:00   3   106.53    0        1  03:45:00  2014-01-01\n",
      "67698  2014-01-01 04:00:00   3   125.41    0        1  04:00:00  2014-01-01\n",
      "67699  2014-01-01 04:15:00   3   199.98    0        1  04:15:00  2014-01-01\n",
      "67700  2014-01-01 04:30:00   3   263.67    0        1  04:30:00  2014-01-01\n",
      "67701  2014-01-01 04:45:00   3   267.50    0        1  04:45:00  2014-01-01\n",
      "67702  2014-01-01 05:00:00   3   309.24    0        1  05:00:00  2014-01-01\n",
      "67703  2014-01-01 05:15:00   3   365.94    0        1  05:15:00  2014-01-01\n",
      "67704  2014-01-01 05:30:00   3   388.11    0        1  05:30:00  2014-01-01\n",
      "67705  2014-01-01 05:45:00   3   315.28    0        1  05:45:00  2014-01-01\n",
      "67706  2014-01-01 06:00:00   3   170.54    0        1  06:00:00  2014-01-01\n",
      "67707  2014-01-01 06:15:00   3   145.03    0        1  06:15:00  2014-01-01\n",
      "67708  2014-01-01 06:30:00   3   170.71    0        1  06:30:00  2014-01-01\n",
      "67709  2014-01-01 06:45:00   3   248.43    0        1  06:45:00  2014-01-01\n",
      "67710  2014-01-01 07:00:00   3   284.65    0        1  07:00:00  2014-01-01\n",
      "67711  2014-01-01 07:15:00   3   246.48    0        1  07:15:00  2014-01-01\n",
      "67712  2014-01-01 07:30:00   3   209.21    0        1  07:30:00  2014-01-01\n",
      "...                    ...  ..      ...  ...      ...       ...         ...\n",
      "102672 2014-12-31 16:30:00   3   669.18    1        0  16:30:00  2014-12-31\n",
      "102673 2014-12-31 16:45:00   3  1033.89    1        0  16:45:00  2014-12-31\n",
      "102674 2014-12-31 17:00:00   3   989.35    1        0  17:00:00  2014-12-31\n",
      "102675 2014-12-31 17:15:00   3  1306.60    1        0  17:15:00  2014-12-31\n",
      "102676 2014-12-31 17:30:00   3   525.88    0        0  17:30:00  2014-12-31\n",
      "102677 2014-12-31 17:45:00   3   385.56    0        0  17:45:00  2014-12-31\n",
      "102678 2014-12-31 18:00:00   3   348.56    0        0  18:00:00  2014-12-31\n",
      "102679 2014-12-31 18:15:00   3   317.29    0        0  18:15:00  2014-12-31\n",
      "102680 2014-12-31 18:30:00   3   374.47    0        0  18:30:00  2014-12-31\n",
      "102681 2014-12-31 18:45:00   3   285.44    0        0  18:45:00  2014-12-31\n",
      "102682 2014-12-31 19:00:00   3   398.24    0        0  19:00:00  2014-12-31\n",
      "102683 2014-12-31 19:15:00   3   388.28    0        0  19:15:00  2014-12-31\n",
      "102684 2014-12-31 19:30:00   3   288.02    0        0  19:30:00  2014-12-31\n",
      "102685 2014-12-31 19:45:00   3   275.67    0        0  19:45:00  2014-12-31\n",
      "102686 2014-12-31 20:00:00   3   264.34    0        0  20:00:00  2014-12-31\n",
      "102687 2014-12-31 20:15:00   3   273.96    0        0  20:15:00  2014-12-31\n",
      "102688 2014-12-31 20:30:00   3   490.94    0        0  20:30:00  2014-12-31\n",
      "102689 2014-12-31 20:45:00   3   394.36    0        0  20:45:00  2014-12-31\n",
      "102690 2014-12-31 21:00:00   3   343.57    0        0  21:00:00  2014-12-31\n",
      "102691 2014-12-31 21:15:00   3   274.19    0        0  21:15:00  2014-12-31\n",
      "102692 2014-12-31 21:30:00   3   271.92    0        0  21:30:00  2014-12-31\n",
      "102693 2014-12-31 21:45:00   3   440.63    0        0  21:45:00  2014-12-31\n",
      "102694 2014-12-31 22:00:00   3   288.50    0        0  22:00:00  2014-12-31\n",
      "102695 2014-12-31 22:15:00   3   210.77    0        0  22:15:00  2014-12-31\n",
      "102696 2014-12-31 22:30:00   3   294.08    0        0  22:30:00  2014-12-31\n",
      "102697 2014-12-31 22:45:00   3   266.02    0        0  22:45:00  2014-12-31\n",
      "102698 2014-12-31 23:00:00   3   501.91    0        0  23:00:00  2014-12-31\n",
      "102699 2014-12-31 23:15:00   3   468.01    0        0  23:15:00  2014-12-31\n",
      "102700 2014-12-31 23:30:00   3   292.85    0        0  23:30:00  2014-12-31\n",
      "102701 2014-12-31 23:45:00   3   226.91    0        0  23:45:00  2014-12-31\n",
      "\n",
      "[34923 rows x 7 columns]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Y = np.hstack(inactivity_anomaly_vectors)\n",
    "unexpected_anomalies_count = 0\n",
    "current_anomaly = []\n",
    "unexpected_anomalies = []\n",
    "min_anomaly_length = 5\n",
    "year = 2014\n",
    "for k, y in enumerate(Y):\n",
    "    if y >= 2:\n",
    "        if y >= 3:\n",
    "            unexpected_anomalies_count += 1\n",
    "            current_anomaly.append(k)\n",
    "    else:\n",
    "        if unexpected_anomalies_count >= min_anomaly_length:\n",
    "            anomaly_range = \\\n",
    "                [\n",
    "                    datetime.datetime(year, 1, 1) + datetime.timedelta(hours=int(current_anomaly[0])),\n",
    "                    datetime.datetime(year, 1, 1) + datetime.timedelta(hours=int(current_anomaly[-1])),\n",
    "                ]\n",
    "\n",
    "            unexpected_anomalies.append(anomaly_range)\n",
    "\n",
    "        current_anomaly = []\n",
    "        unexpected_anomalies_count = 0\n",
    "\n",
    "first = '2014-01-01 00:00:00'\n",
    "last = '2014-12-31 23:59:59'\n",
    "unexpected_anomalies_str = [[x.strftime('%Y-%m-%d %H:%M:%S') for x in r] for r in unexpected_anomalies]\n",
    "visualize_household(household_id, first, last, unexpected_anomalies_str)"
   ]
  }
 ],
 "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": "3.6.7"
  },
  "stem_cell": {
   "cell_type": "raw",
   "metadata": {
    "pycharm": {
     "metadata": false
    }
   },
   "source": "#EnBw Challenge\n"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}