CO.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Karen Yu, Nick Vasios, Thibaut Perol\n",
    "\n",
    "# AM207 Final Project\n",
    "\n",
    "## Energy Disaggregation from Non-Intrusive Load Monitoring"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  DISAGGREGATION USING COMBINATORIAL OPTIMIZATION"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing Necessary Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from __future__ import print_function, division\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from os.path import join\n",
    "import pickle\n",
    "import copy\n",
    "\n",
    "from pylab import rcParams\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "rcParams['figure.figsize'] = (13, 6)\n",
    "import nilmtk\n",
    "from nilmtk import DataSet, TimeFrame, MeterGroup, HDFDataStore\n",
    "from nilmtk.utils import print_dict,find_nearest\n",
    "from nilmtk.feature_detectors import cluster\n",
    "from nilmtk.disaggregate import Disaggregator\n",
    "from nilmtk.electric import get_vampire_power\n",
    "from nilmtk.metrics import f1_score\n",
    "\n",
    "import warnings\n",
    "from warnings import warn\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "import seaborn as sns\n",
    "# sns.set_style(\"white\")\n",
    "\n",
    "# Fix the seed for repeatability of experiments\n",
    "SEED = 42\n",
    "np.random.seed(SEED)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Heart of the Notebook: The Combinatorial Optimization class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "class CombinatorialOptimisation(Disaggregator):\n",
    "    \"\"\"\n",
    "    \n",
    "    A Combinatorial Optimization Algorithm based on the implementation by NILMTK\n",
    "    \n",
    "    This class is build upon the main Dissagregator class already implemented by NILMTK\n",
    "    All the methods from Dissagregator are passed in here as well since we import the class\n",
    "    as shown above. We should note howeger that Dissagregator is nothing more than a general interface\n",
    "    class upon which all dissagregator algortihms are build. All the methods are initialized in the \n",
    "    Dissagregator class but the specific implementation is based upon the method to be implemented. \n",
    "    In other words, even though we pass in Dissagregator, all methods will be redefined again to work with \n",
    "    the Combinatorial Optimization algorithm as you can see below. \n",
    "\n",
    "    \n",
    "    Attributes\n",
    "    ----------\n",
    "    model : list of dicts\n",
    "       Each dict has these keys:\n",
    "           states : list of ints (the power (Watts) used in different states)\n",
    "           training_metadata : ElecMeter or MeterGroup object used for training\n",
    "               this set of states.  We need this information because we\n",
    "               need the appliance type (and perhaps some other metadata)\n",
    "               for each model.\n",
    "\n",
    "    state_combinations : 2D array\n",
    "        Each column is an appliance.\n",
    "        Each row is a possible combination of power demand values e.g.\n",
    "            [[0, 0,  0,   0],\n",
    "             [0, 0,  0, 100],\n",
    "             [0, 0, 50,   0],\n",
    "             [0, 0, 50, 100], ...]\n",
    "\n",
    "    MIN_CHUNK_LENGTH : int\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self):\n",
    "        self.model = []\n",
    "        self.state_combinations = None\n",
    "        self.MIN_CHUNK_LENGTH = 100\n",
    "        self.MODEL_NAME = 'Combinatorial Optimization'\n",
    "\n",
    "    def train(self, metergroup, num_states_dict=None, **load_kwargs):\n",
    "        \"\"\"\n",
    "        Train using 1D CO. Places the learnt model in the `model` attribute.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        metergroup : a nilmtk.MeterGroup object\n",
    "        num_states_dict : dict\n",
    "        **load_kwargs : keyword arguments passed to `meter.power_series()`\n",
    "\n",
    "        Notes\n",
    "        -----\n",
    "        * only uses first chunk for each meter (TODO: handle all chunks).\n",
    "        \"\"\"\n",
    "        \n",
    "        # Initializing dictionary to save the number of states\n",
    "        if num_states_dict is None:\n",
    "            num_states_dict = {}\n",
    "\n",
    "        # The CO class is only able to train in new models. We can only train once. If model exists, raise an error\n",
    "        if self.model:\n",
    "            raise RuntimeError(\n",
    "                \"This implementation of Combinatorial Optimisation\"\n",
    "                \" does not support multiple calls to `train`.\")\n",
    "\n",
    "        # How many meters do we have in the training set?\n",
    "        num_meters = len(metergroup.meters)\n",
    "        \n",
    "        # If more than 20 then reduce the number of clusters to reduce the computational cost. \n",
    "        if num_meters > 20:\n",
    "            max_num_clusters = 2\n",
    "        else:\n",
    "            max_num_clusters = 3\n",
    "        \n",
    "        print('Now training...')\n",
    "        print('Loop in all meters begins...')\n",
    "        \n",
    "        # We now loop in all meters passed in in the training data set\n",
    "        # Every time, we load the data in the meter and we call the method\n",
    "        # --> train_on_chunk. For more info about this method please see below\n",
    "        for i, meter in enumerate(metergroup.submeters().meters):\n",
    "            \n",
    "            #print('We now train for submeter {}'.format(meter))\n",
    "            \n",
    "            # Load the time series for the power consumption for this meter\n",
    "            power_series = meter.power_series(**load_kwargs)\n",
    "            \n",
    "            # Note that we do not effectively load until we use the next() method\n",
    "            # We load and save into chunk. Chunk will be used in training\n",
    "            chunk = power_series.next()\n",
    "            \n",
    "            # Get the number of total states from the dictionary\n",
    "            num_total_states = num_states_dict.get(meter)\n",
    "            if num_total_states is not None:\n",
    "                num_on_states = num_total_states - 1\n",
    "            else:\n",
    "                num_on_states = None\n",
    "                \n",
    "            #print('i={},num_total_states={},num_on_states={}'.format(i,meter,num_total_states,num_on_states))    \n",
    "            \n",
    "            # The actual training happens now. We call train_on_chunk using the time series we loaded on chunk for this meter\n",
    "            self.train_on_chunk(chunk, meter, max_num_clusters, num_on_states)\n",
    "\n",
    "            # Check to see if there are any more chunks.\n",
    "            try:\n",
    "                power_series.next()\n",
    "            except StopIteration:\n",
    "                pass\n",
    "            else:\n",
    "                warn(\"The current implementation of CombinatorialOptimisation\"\n",
    "                     \" can only handle a single chunk.  But there are multiple\"\n",
    "                     \" chunks available.  So have only trained on the\"\n",
    "                     \" first chunk!\")\n",
    "\n",
    "        print(\"Done training!\")\n",
    "\n",
    "    def train_on_chunk(self, chunk, meter, max_num_clusters, num_on_states):\n",
    "        \"\"\"\n",
    "        \n",
    "        Train on chunk trains the Combinatorial Optimization Model based on the time series for the power consumption\n",
    "        passed in chunk. This method is based on the sklearn machine learning library and in particular the KMEANS \n",
    "        algorithm. It calls the cluster function which is imported in the beginning of this notebook. Cluster, prepares \n",
    "        the data in chunk so that its size is always compatible and the same and then calls the KMEANS algorithm to \n",
    "        perform the clustering. Function cluster returns only the centers of the clustered data which correspond to the\n",
    "        individual states for the given appliance/meter        \n",
    "        \"\"\"\n",
    "        \n",
    "        # Check if we've already trained on this meter. We only allow training once on each meter\n",
    "        meters_in_model = [d['training_metadata'] for d in self.model]\n",
    "        if meter in meters_in_model:\n",
    "            raise RuntimeError(\n",
    "                \"Meter {} is already in model!\"\n",
    "                \"  Can't train twice on the same meter!\"\n",
    "                .format(meter))\n",
    "\n",
    "        # Do the KMEANS clustering and return the centers\n",
    "        states = cluster(chunk, max_num_clusters, num_on_states)\n",
    "        print('\\t Now Clustering in Train on Chunk')\n",
    "        #print('\\t {}'.format(states))\n",
    "        \n",
    "        # Append the clustered data to the model\n",
    "        self.model.append({\n",
    "            'states': states,\n",
    "            'training_metadata': meter})\n",
    "\n",
    "    def _set_state_combinations_if_necessary(self):\n",
    "        \"\"\"Get centroids\"\"\"\n",
    "        # If we import sklearn at the top of the file then auto doc fails.\n",
    "        if (self.state_combinations is None or\n",
    "                self.state_combinations.shape[1] != len(self.model)):\n",
    "            \n",
    "            from sklearn.utils.extmath import cartesian\n",
    "            \n",
    "            # Saving the centroids in centroids (appliance states)\n",
    "            centroids = [model['states'] for model in self.model]\n",
    "            # Function cartesian returns all possible combinations \n",
    "            # than can be performed using centroids\n",
    "            self.state_combinations = cartesian(centroids)\n",
    "            print()\n",
    "            #print('Now printing the state combinations...')\n",
    "            #print(cartesian(centroids))\n",
    "\n",
    "    def disaggregate(self, mains, output_datastore,\n",
    "                     vampire_power=None, **load_kwargs):\n",
    "        '''Disaggregate mains according to the model learnt previously.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        mains : nilmtk.ElecMeter or nilmtk.MeterGroup\n",
    "        output_datastore : instance of nilmtk.DataStore subclass\n",
    "            For storing power predictions from disaggregation algorithm.\n",
    "        vampire_power : None or number (watts)\n",
    "            If None then will automatically determine vampire power\n",
    "            from data.  If you do not want to use vampire power then\n",
    "            set vampire_power = 0.\n",
    "        sample_period : number, optional\n",
    "            The desired sample period in seconds.  Set to 60 by default.\n",
    "        sections : TimeFrameGroup, optional\n",
    "            Set to mains.good_sections() by default.\n",
    "        **load_kwargs : key word arguments\n",
    "            Passed to `mains.power_series(**kwargs)`\n",
    "        '''\n",
    "        \n",
    "        # Performing default pre disaggregation checks. Checking meters etc..\n",
    "        load_kwargs = self._pre_disaggregation_checks(load_kwargs)\n",
    "\n",
    "        # Disaggregation defauls. Sample perios and sections\n",
    "        load_kwargs.setdefault('sample_period', 60)\n",
    "        load_kwargs.setdefault('sections', mains.good_sections())\n",
    "\n",
    "        # Initializing time frames and fetching the meter for the aggregated data\n",
    "        timeframes = []\n",
    "        building_path = '/building{}'.format(mains.building())\n",
    "        mains_data_location = building_path + '/elec/meter1'\n",
    "        data_is_available = False\n",
    "        \n",
    "        # We now load the aggregated data for power consumption of the whole house in small chunks\n",
    "        # Every iteration of the following loop we perform the CO step to disaggregate\n",
    "        \n",
    "        counter = 0\n",
    "        print('Disaggregation now begins...')\n",
    "        for chunk in mains.power_series(**load_kwargs):\n",
    "            counter += 1\n",
    "            # Check that chunk is sensible size\n",
    "            if len(chunk) < self.MIN_CHUNK_LENGTH:\n",
    "                continue\n",
    "                \n",
    "            print('\\t Now processing chunk {}...'.format(counter))\n",
    "\n",
    "            # Record metadata\n",
    "            timeframes.append(chunk.timeframe)\n",
    "            measurement = chunk.name\n",
    "\n",
    "            # This is where the disaggregation happens\n",
    "            # Vampire Power is just the minimum of the power series in this chunk\n",
    "            appliance_powers = self.disaggregate_chunk(chunk, vampire_power)\n",
    "\n",
    "            # Here we save the disaggregated data for this chunk in Pandas dataframe and update the \n",
    "            # HDF5 file we created.\n",
    "            for i, model in enumerate(self.model):\n",
    "                # Fetch the disag data for this appliance\n",
    "                appliance_power = appliance_powers[i]\n",
    "                if len(appliance_power) == 0:\n",
    "                    continue\n",
    "                data_is_available = True\n",
    "                \n",
    "                # Just for saving.. Nothing major happening here\n",
    "                cols = pd.MultiIndex.from_tuples([chunk.name])\n",
    "                meter_instance = model['training_metadata'].instance()\n",
    "                df = pd.DataFrame(\n",
    "                    appliance_power.values, index=appliance_power.index,\n",
    "                    columns=cols)\n",
    "                key = '{}/elec/meter{}'.format(building_path, meter_instance)\n",
    "                output_datastore.append(key, df)\n",
    "\n",
    "            # Copy mains data to disag output\n",
    "            mains_df = pd.DataFrame(chunk, columns=cols)\n",
    "            output_datastore.append(key=mains_data_location, value=mains_df)\n",
    "\n",
    "        if data_is_available:\n",
    "            self._save_metadata_for_disaggregation(\n",
    "                output_datastore=output_datastore,\n",
    "                sample_period=load_kwargs['sample_period'],\n",
    "                measurement=measurement,\n",
    "                timeframes=timeframes,\n",
    "                building=mains.building(),\n",
    "                meters=[d['training_metadata'] for d in self.model]\n",
    "            )\n",
    "        \n",
    "        print('Disaggregation Completed Successfully...!!!')\n",
    "\n",
    "    def disaggregate_chunk(self, mains, vampire_power=None):\n",
    "        \"\"\"In-memory disaggregation.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        mains : pd.Series\n",
    "        vampire_power : None or number (watts)\n",
    "            If None then will automatically determine vampire power\n",
    "            from data.  If you do not want to use vampire power then\n",
    "            set vampire_power = 0.\n",
    "\n",
    "        Returns\n",
    "        -------\n",
    "        appliance_powers : pd.DataFrame where each column represents a\n",
    "            disaggregated appliance.  Column names are the integer index\n",
    "            into `self.model` for the appliance in question.\n",
    "        \"\"\"\n",
    "        if not self.model:\n",
    "            raise RuntimeError(\n",
    "                \"The model needs to be instantiated before\"\n",
    "                \" calling `disaggregate`.  The model\"\n",
    "                \" can be instantiated by running `train`.\")\n",
    "\n",
    "        if len(mains) < self.MIN_CHUNK_LENGTH:\n",
    "            raise RuntimeError(\"Chunk is too short.\")\n",
    "\n",
    "        # sklearn produces lots of DepreciationWarnings with PyTables\n",
    "        import warnings\n",
    "        warnings.filterwarnings(\"ignore\", category=DeprecationWarning)\n",
    "\n",
    "        # Because CombinatorialOptimisation could have been trained using\n",
    "        # either train() or train_on_chunk(), we must\n",
    "        # set state_combinations here.\n",
    "        self._set_state_combinations_if_necessary()\n",
    "\n",
    "        # Add vampire power to the model (Min of power series of the aggregated data)\n",
    "        if vampire_power is None:\n",
    "            vampire_power = get_vampire_power(mains)\n",
    "        if vampire_power > 0:\n",
    "            print()\n",
    "            #print(\"Including vampire_power = {} watts to model...\".format(vampire_power))\n",
    "            \n",
    "            # How many combinations\n",
    "            n_rows = self.state_combinations.shape[0]\n",
    "            vampire_power_array = np.zeros((n_rows, 1)) + vampire_power\n",
    "            state_combinations = np.hstack(\n",
    "                (self.state_combinations, vampire_power_array))\n",
    "        else:\n",
    "            state_combinations = self.state_combinations\n",
    "\n",
    "        summed_power_of_each_combination = np.sum(state_combinations, axis=1)\n",
    "        # summed_power_of_each_combination is now an array where each\n",
    "        # value is the total power demand for each combination of states.\n",
    "\n",
    "        # Start disaggregation\n",
    "        \n",
    "        # The following line finds the best combination from all the possible combinations\n",
    "        # Returns the index to find the best combination as well as the residual\n",
    "        # Uses the Find_Nearest algorithm\n",
    "        indices_of_state_combinations, residual_power = find_nearest(\n",
    "            summed_power_of_each_combination, mains.values)\n",
    "\n",
    "        # Now update the state for each appliance with the optimal one and return the list\n",
    "        # as Dataframe\n",
    "        appliance_powers_dict = {}\n",
    "        for i, model in enumerate(self.model):\n",
    "            #print()\n",
    "            #print(\"Estimating power demand for '{}'\".format(model['training_metadata']))\n",
    "            predicted_power = state_combinations[\n",
    "                indices_of_state_combinations, i].flatten()\n",
    "            column = pd.Series(predicted_power, index=mains.index, name=i)\n",
    "            appliance_powers_dict[i] = column\n",
    "\n",
    "        appliance_powers = pd.DataFrame(appliance_powers_dict)\n",
    "        return appliance_powers\n",
    "\n",
    "    # The current implementation of the CO does not make use of the following 2 functions.\n",
    "    #\n",
    "    #\n",
    "    # -------------------------------------------------------------------------------------\n",
    "    def import_model(self, filename):\n",
    "        imported_model = pickle.load(open(filename, 'r'))\n",
    "        self.model = imported_model.model\n",
    "        # recreate datastores from filenames\n",
    "        for pair in self.model:\n",
    "            pair['training_metadata'].store = HDFDataStore(\n",
    "                pair['training_metadata'].store)\n",
    "        self.state_combinations = imported_model.state_combinations\n",
    "        self.MIN_CHUNK_LENGTH = imported_model.MIN_CHUNK_LENGTH\n",
    "\n",
    "    def export_model(self, filename):\n",
    "        # Can't pickle datastore, so convert to filenames\n",
    "        exported_model = copy.deepcopy(self)\n",
    "        for pair in exported_model.model:\n",
    "            pair['training_metadata'].store = (\n",
    "                pair['training_metadata'].store.store.filename)\n",
    "        pickle.dump(exported_model, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing and Loading the REDD dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loaded 6 buildings\n"
     ]
    }
   ],
   "source": [
    "data_dir = '\\Users\\Nick\\Google Drive\\PhD\\Courses\\Semester 2\\AM207\\Project'\n",
    "we = DataSet(join(data_dir, 'REDD.h5'))\n",
    "print('loaded ' + str(len(we.buildings)) + ' buildings')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want to train the Combinatorial Optimization Algorithm using the data for 5 buildings and then test it against the last building. To simplify our analysis and also to enable comparison with other methods (Neural Nets, FHMM, MLE etc) we will only try to dissagregate data associated with the fridge and the microwave. However, the REDD dataset that we are using here does not contain data measurements for the fridge and microwave for all buildings. In particular, building 4 does not have measurements for the fridge. As a result, we will exclude building 4 from the dataset and we will only import the meters associated with the fridge from other buildings. \n",
    "\n",
    "The train data set will consist of meters associated with the fridge and microwave from buildings 1,2,3 and 6. We will then test the combinatorial optimization algorithm against the aggregated data for building 5. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We first plot the time window span for all buildings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Timeframe for building 1 is TimeFrame(start='2011-04-18 09:22:09-04:00', end='2011-05-24 15:57:02-04:00', empty=False)\n",
      "Timeframe for building 2 is TimeFrame(start='2011-04-17 19:18:27-04:00', end='2011-05-22 19:59:16-04:00', empty=False)\n",
      "Timeframe for building 3 is TimeFrame(start='2011-04-16 01:11:27-04:00', end='2011-05-30 20:19:54-04:00', empty=False)\n",
      "Timeframe for building 4 is TimeFrame(start='2011-04-16 21:16:19-04:00', end='2011-06-03 20:45:45-04:00', empty=False)\n",
      "Timeframe for building 5 is TimeFrame(start='2011-04-18 00:24:03-04:00', end='2011-05-31 20:20:20-04:00', empty=False)\n",
      "Timeframe for building 6 is TimeFrame(start='2011-05-21 15:39:19-04:00', end='2011-06-14 01:31:44-04:00', empty=False)\n"
     ]
    }
   ],
   "source": [
    "for i in xrange(1,7):\n",
    "    print('Timeframe for building {} is {}'.format(i,we.buildings[i].elec.get_timeframe()))    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unfortunately, due to a bug in one of the main classes of the NILMTK package the implementation of the Combinatorial Optimization do not save the meters for the disaggregated data correctly unless the building on which we test on also exists in the trainihg set. More on this issue can be found here https://github.com/nilmtk/nilmtk/issues/194"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, for us it makes no sense to use the same building for training and testing since we would like to compare this algorithm with the results from FHMM and Neural Networks. In order to circumvent this bug we do the following:\n",
    "\n",
    "The main issue is that the meter for the building we would like to disaggregate must be on the training set in order to be able to disaggregate correctly. That being said, we still want to train as less as possible on the meter we want to test on since we would like to see how the algorithm performs when a completely unknown dataset is available. In order to do that we create a metergroup comprising of the following:\n",
    "\n",
    "1) The meters for the Frigde and Microwave for all buildings but building 5, since building 5 is the building we would like to test on. Later we will see that building 4 needs to be excluded as well because there is no meter associated with the fridge for this building. \n",
    "\n",
    "2) The meters for the Frigde and Microwave for building 5 which is the building we would like to test on, but we limit the time window to be a very very small one. Doing that, we make sure that the meters are there and understood by the Combinatorial Optimization Class but at the same time, by limiting the time window to just a few housrd for this building do not provide enough data to overtrain. In other words, we only do this in order to be able to disaggregate correctly. \n",
    "\n",
    "After we train we will test the algorithm against the data from building 5 that werent fed into the training meters. After we disaggregate we will compare with the ground truth for the same exact window. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modifying Datasets to work with CO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Found 6 buildings in the Data Ser.. Buildings Loaded successfully.\n"
     ]
    }
   ],
   "source": [
    "# Data file directory\n",
    "data_dir = '\\Users\\Nick\\Google Drive\\PhD\\Courses\\Semester 2\\AM207\\Project'\n",
    "\n",
    "# Make the Data set\n",
    "Data = DataSet(join(data_dir, 'REDD.h5'))\n",
    "\n",
    "# Make copies of the Data Set so that local changes would not affect the global dataset\n",
    "Data_for_5 = DataSet(join(data_dir, 'REDD.h5'))\n",
    "Data_for_rest = DataSet(join(data_dir, 'REDD.h5'))\n",
    "\n",
    "# How many buildings in the data set?\n",
    "print(' Found {} buildings in the Data Ser.. Buildings Loaded successfully.'.format(len(Data.buildings)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xe1b8f98>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfMAAAFKCAYAAAAJyrb2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb789780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This is the point that we will break the data from building 5 so that we only include a small \n",
    "# portion in the training set. In fact, the line below makes sure than only a day of data is seen during training.\n",
    "break_point = '2011-04-19 02:00'\n",
    "\n",
    "# Changing the window for building 5\n",
    "Data_for_5.set_window(end=break_point)\n",
    "\n",
    "# Making a metergroup..\n",
    "e = [Data_for_5.buildings[5].elec[a] for a in ['fridge','microwave']]\n",
    "me = MeterGroup(e)\n",
    "\n",
    "# The data that we pass in for training for building 5 look like this...\n",
    "me.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating MeterGroups with the desired appliances from the desired buildings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below we define a function tha is able to create a metergroup that only includes meters for the appliances that we are interested in and is also able to exclude buildings that we don't want in the meter. Also, if an appliance is requested but a meter is not found then the meter is skipped but the metergoup is created nontheless."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def get_all_trainings(appliance, dataset, buildings_to_exclude):\n",
    "\n",
    "    # Filtering by appliances: \n",
    "    elecs = []\n",
    "    for app in appliance:\n",
    "        app_l = [app]\n",
    "        print ('Now loading data for ' + app + ' for all buildings in the data to create the metergroup')\n",
    "        print()\n",
    "        for building in dataset.buildings: \n",
    "            if building not in buildings_to_exclude:\n",
    "                print ('Processing Building ' + str(building) + '...')\n",
    "                print()\n",
    "                try:\n",
    "                    elec = dataset.buildings[building].elec[app]       \n",
    "                    \n",
    "                    elecs.append(elec)\n",
    "\n",
    "                except KeyError:\n",
    "                    print ('Appliance '+str(app)+' does not exist in this building')\n",
    "                    print ('Building skipped...')\n",
    "                    print ()\n",
    "            \n",
    "\n",
    "    metergroup = MeterGroup(elecs)\n",
    "\n",
    "    return metergroup"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we set the appliances that we want as well as the buildings to exclude and we create the metergroup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Now loading data for fridge for all buildings in the data to create the metergroup\n",
      "\n",
      "Processing Building 1...\n",
      "\n",
      "Processing Building 2...\n",
      "\n",
      "Processing Building 3...\n",
      "\n",
      "Processing Building 6...\n",
      "\n",
      "Now loading data for microwave for all buildings in the data to create the metergroup\n",
      "\n",
      "Processing Building 1...\n",
      "\n",
      "Processing Building 2...\n",
      "\n",
      "Processing Building 3...\n",
      "\n",
      "Processing Building 6...\n",
      "\n",
      "Appliance microwave does not exist in this building\n",
      "Building skipped...\n",
      "\n",
      "Now printing the Meter Group...\n",
      "\n",
      "MeterGroup(meters=\n",
      "  ElecMeter(instance=5, building=1, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=9, building=2, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=7, building=3, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=8, building=6, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=11, building=1, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      "  ElecMeter(instance=6, building=2, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      "  ElecMeter(instance=16, building=3, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "applianceName = ['fridge','microwave']\n",
    "buildings_to_exclude = [4,5]\n",
    "\n",
    "metergroup = get_all_trainings(applianceName,Data_for_rest,buildings_to_exclude)\n",
    "\n",
    "print('Now printing the Meter Group...')\n",
    "print()\n",
    "print(metergroup)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see the Metergroup was successfully created and contains all the appliances we requested (Fridge and Microwave) in all buildings that the appliances exist apart from the ones we excluded"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correcting the MeterGroup (Necessary for the CO to work)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we need to perform the trick we mentioned previously. We need to also include the meter from building 5 with the Fridge and Microwave which is the building we are going to test on but we need to make sure that only a very small portion of the data is seen for this building. We already took care of that by changing the window for the data in building 5 so now we only have to include the meters for the Fridge and Microwave for building 5 from the reduced time dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def correct_meter(Data,building,appliance,oldmeter):\n",
    "    \n",
    "    # Unpack meters from the MeterGroup\n",
    "    meters = oldmeter.all_meters()\n",
    "    \n",
    "    # Get the rest of the meters and append\n",
    "    for a in appliance:\n",
    "        meter_to_add = Data.buildings[building].elec[a]\n",
    "        meters.append(meter_to_add)\n",
    "    \n",
    "    # Group again in a single metergroup and return\n",
    "    return MeterGroup(meters)   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "corr_metergroup = correct_meter(Data_for_5,5,applianceName,metergroup)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The Modified Meter is now..\n",
      "\n",
      "MeterGroup(meters=\n",
      "  ElecMeter(instance=5, building=1, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=9, building=2, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=7, building=3, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=8, building=6, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=11, building=1, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      "  ElecMeter(instance=6, building=2, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      "  ElecMeter(instance=16, building=3, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      "  ElecMeter(instance=18, building=5, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=3, building=5, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "print('The Modified Meter is now..')\n",
    "print()\n",
    "print(corr_metergroup)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see the metergroup was updated successfully"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now need to train in the Metergroup we just created. First, let us load the class for the CO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Train\n",
    "co = CombinatorialOptimisation()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now Let's train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Now training...\n",
      "Loop in all meters begins...\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "\t Now Clustering in Train on Chunk\n",
      "Done training!\n"
     ]
    }
   ],
   "source": [
    "co.train(corr_metergroup)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparing the Testing Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that the training is done, the only thing that we have to do is to prepare the Data for Building 5 that we want to test on and call the Disaggregation. The data set is now the remaining part of building 5 that is not seen. After that, we only keep the Main meter which contains ifrormation about the aggregated data consumption and we disaggregate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "Test_Data = DataSet(join(data_dir, 'REDD.h5'))\n",
    "Test_Data.set_window(start=break_point)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# The building number on which we test\n",
    "building_for_testing = 5\n",
    "\n",
    "test = Test_Data.buildings[building_for_testing].elec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "mains = test.mains()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Disaggregating the test data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The disaggregation Begins Now"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Disaggregation now begins...\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 2...\n",
      "\n",
      "\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 7...\n",
      "\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 13...\n",
      "\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 14...\n",
      "\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 15...\n",
      "\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "Loading data for meter ElecMeterID(instance=2, building=5, dataset='REDD')     \n",
      "Done loading data all meters for this chunk.\n",
      "\t Now processing chunk 17...\n",
      "\n",
      "Disaggregation Completed Successfully...!!!\n"
     ]
    }
   ],
   "source": [
    "# Disaggregate\n",
    "disag_filename = join(data_dir, 'COMBINATORIAL_OPTIMIZATION.h5')\n",
    "\n",
    "mains = test.mains()\n",
    "\n",
    "try:\n",
    "    output = HDFDataStore(disag_filename, 'w')\n",
    "    co.disaggregate(mains, output)\n",
    "except ValueError:\n",
    "    output.close()\n",
    "    output = HDFDataStore(disag_filename, 'w')\n",
    "    co.disaggregate(mains, output)\n",
    "\n",
    "for meter in range(1, 2):\n",
    "    df1 = output.store.get('/building5/elec/meter{}'.format(meter))\n",
    "    df2 = we.store.store.get('/building5/elec/meter{}'.format(meter))\n",
    "\n",
    "    \n",
    "output.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK.. Now we are all done. All that remains is to interpret the results and plot the scores.."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Post Processing & Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MeterGroup(meters=\n",
      "  ElecMeter(instance=18, building=5, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])\n",
      "  ElecMeter(instance=3, building=5, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# Opening the Dataset with the Disaggregated data\n",
    "disag = DataSet(disag_filename)\n",
    "\n",
    "# Getting electric appliances and meters\n",
    "disag_elec = disag.buildings[building_for_testing].elec\n",
    "\n",
    "# We also get the electric appliances and meters for the ground truth data to compare\n",
    "elec = Test_Data.buildings[building_for_testing].elec\n",
    "\n",
    "e = [test[a] for a in applianceName]\n",
    "me = MeterGroup(e)\n",
    "print(me)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Resampling to align meters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before we are able to calculate and plot the metrics we need to align the ground truth meter with the disaggregated meters. Why so? If you notice in the dissagregation method of the CO class above, you may see that by default the time sampling is changed from 3s which is the raw data to 60s. This has to happen in order to make the disaggregation more efficient computationally but also because it is impossible to disaggregate using the actual time step. So in order to compare now we have to resample the meter for the ground truth and align it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def align_two_meters(master, slave, func='when_on'):\n",
    "    \"\"\"Returns a generator of 2-column pd.DataFrames.  The first column is from\n",
    "    `master`, the second from `slave`.\n",
    "\n",
    "    Takes the sample rate and good_periods of `master` and applies to `slave`.\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    master, slave : ElecMeter or MeterGroup instances\n",
    "    \"\"\"\n",
    "    sample_period = master.sample_period()\n",
    "    period_alias = '{:d}S'.format(sample_period)\n",
    "    sections = master.good_sections()\n",
    "    master_generator = getattr(master, func)(sections=sections)\n",
    "    for master_chunk in master_generator:\n",
    "        if len(master_chunk) < 2:\n",
    "            return\n",
    "        chunk_timeframe = TimeFrame(master_chunk.index[0],\n",
    "                                    master_chunk.index[-1])\n",
    "        slave_generator = getattr(slave, func)(sections=[chunk_timeframe])\n",
    "        slave_chunk = next(slave_generator)\n",
    "\n",
    "        # TODO: do this resampling in the pipeline?\n",
    "        slave_chunk = slave_chunk.resample(period_alias)\n",
    "        if slave_chunk.empty:\n",
    "            continue\n",
    "        master_chunk = master_chunk.resample(period_alias)\n",
    "\n",
    "    return master_chunk,slave_chunk\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we just plot the disaggregated data alongside the ground truth for the Fridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1727aba8>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfMAAAFUCAYAAAAwFtUdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb95d1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "disag_elec.select(instance=18).plot()\n",
    "me.select(instance=18).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Aligning meters, Converting to Numpy and Computing Metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this part of the Notebook, we call the function we previously defined to align the meters and then we convert the meters to pandas and ultimately to numpy arrays. We check if any NaN's exist (which is something possible after resmplilng.. Resampling errors may occur) and replace them with 0's if they do. We also compute the following metrics for each appliance:\n",
    "\n",
    "1) True Positive, False Positive, False Negative, True Negative\n",
    "\n",
    "2) Precision and Recall\n",
    "\n",
    "3) Accuracy and F1-Score\n",
    "\n",
    "For more information about these metrics please refer to the report."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "appliances_scores = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing Fridge...\n",
      "\t Predictions array seems to be fine...\n",
      "\t No Nans detected\n",
      "\n",
      "Processing Microwave...\n",
      "\t Predictions array seems to be fine...\n",
      "\t No Nans detected\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for m in me.meters:\n",
    "    print('Processing {}...'.format(m.label()))\n",
    "    ground_truth = m\n",
    "    inst = m.instance()\n",
    "    prediction = disag_elec.select(instance=inst)\n",
    "    \n",
    "    a = prediction.meters[0]\n",
    "    \n",
    "    b = a.power_series_all_data()\n",
    "    \n",
    "    pr_a,gt_a = align_two_meters(prediction.meters[0],ground_truth)\n",
    "    \n",
    "    gt = gt_a.as_matrix()\n",
    "    pr = pr_a.as_matrix()\n",
    "    \n",
    "    if np.all(np.isnan(pr)==False):\n",
    "        print('\\t Predictions array seems to be fine...')\n",
    "        print('\\t No Nans detected')\n",
    "        print()\n",
    "    else:\n",
    "        print('\\t Serious error in Predictions...')\n",
    "        print('\\t The resampled array contains Nans')\n",
    "        print()\n",
    "    \n",
    "    gt_states_on = gt > 0.1\n",
    "    pr_states_on = pr > 0.1\n",
    "    \n",
    "    TP = np.sum(np.logical_and(gt_states_on==True,pr_states_on[1:]==True))\n",
    "    FP = np.sum(np.logical_and(gt_states_on==True,pr_states_on[1:]==False))\n",
    "    FN = np.sum(np.logical_and(gt_states_on==False,pr_states_on[1:]==True))\n",
    "    TN = np.sum(np.logical_and(gt_states_on==False,pr_states_on[1:]==False))\n",
    "    P = np.sum(gt_states_on==True)\n",
    "    N = np.sum(gt_states_on==False)\n",
    "    \n",
    "    recall = TP/float(TP+FN)\n",
    "    precision = TP/float(TP+FP)\n",
    "    f1 = 2*precision*recall/(precision+recall)\n",
    "    accuracy = (TP+TN)/float(P+N)\n",
    "    \n",
    "    result = {'F1-Score':f1,\n",
    "             'Precision':precision,\n",
    "             'Recall':recall,\n",
    "             'Accuracy':accuracy}\n",
    "    appliances_scores[m.label()] = result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Fridge': {'Recall': 0.51383399209486169, 'F1-Score': 0.2810810810810811, 'Precision': 0.19345238095238096, 'Accuracy': 0.52363896848137537}, 'Microwave': {'Recall': 0.11328976034858387, 'F1-Score': 0.18118466898954702, 'Precision': 0.45217391304347826, 'Accuracy': 0.66332378223495703}}\n"
     ]
    }
   ],
   "source": [
    "print(appliances_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "Names = ['Fridge','Microwave']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we just plot the scores for both the Fridge and the Microwave in order to be able to visualize what is going on. We do not comment on the results in this notebook since we do this in the report. There is a separate notebook where all these results are combined along with the corresponding results from the Neural Network and the FHMM method and the total results are reported side by side to ease comparison. We plot them here as well for housekeeping although it is redundant."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### F1-Score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "x = np.arange(2)\n",
    "y = np.array([appliances_scores[i]['F1-Score'] for i in Names])\n",
    "y[np.isnan(y)] = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLQAAAIeCAYAAACiDFYRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noKUQ0AIAAAAAAEBL+S/TUBL/xrqwHwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xbc1bf98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(18,8))\n",
    "plt.rc('font', size=20, **{'family': 'serif', 'serif': ['Computer Modern']})\n",
    "plt.rc('text', usetex=True)\n",
    "ax = f.add_axes([0.2,0.2,0.8,0.8])\n",
    "ax.bar(x,y,align='center')\n",
    "ax.set_xticks(x)\n",
    "ax.set_yticks(y)\n",
    "ax.set_yticklabels(y,fontsize=20)\n",
    "ax.set_xticklabels(Names,fontsize=20)\n",
    "ax.set_xlim([min(x)-0.5,max(x)+0.5])\n",
    "plt.xlabel('Appliances',fontsize=20)\n",
    "plt.ylabel('F1-Score',fontsize=20)\n",
    "plt.title('Combinatorial Optimization',fontsize=22)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Precision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "x = np.arange(2)\n",
    "y = np.array([appliances_scores[i]['Precision'] for i in Names])\n",
    "y[np.isnan(y)] = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLQAAAIeCAYAAACiDFYRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x17264898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(18,8))\n",
    "plt.rc('font', size=20, **{'family': 'serif', 'serif': ['Computer Modern']})\n",
    "plt.rc('text', usetex=True)\n",
    "ax = f.add_axes([0.2,0.2,0.8,0.8])\n",
    "ax.bar(x,y,align='center')\n",
    "ax.set_xticks(x)\n",
    "ax.set_yticks(y)\n",
    "ax.set_yticklabels(y,fontsize=20)\n",
    "ax.set_xticklabels(Names,fontsize=20)\n",
    "ax.set_xlim([min(x)-0.5,max(x)+0.5])\n",
    "plt.xlabel('Appliances',fontsize=20)\n",
    "plt.ylabel('Precision',fontsize=20)\n",
    "plt.title('Combinatorial Optimization',fontsize=22)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x = np.arange(2)\n",
    "y = np.array([appliances_scores[i]['Recall'] for i in Names])\n",
    "y[np.isnan(y)] = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLQAAAIeCAYAAACiDFYRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x123a1240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(18,8))\n",
    "plt.rc('font', size=20, **{'family': 'serif', 'serif': ['Computer Modern']})\n",
    "plt.rc('text', usetex=True)\n",
    "ax = f.add_axes([0.2,0.2,0.8,0.8])\n",
    "ax.bar(x,y,align='center')\n",
    "ax.set_xticks(x)\n",
    "ax.set_yticks(y)\n",
    "ax.set_yticklabels(y,fontsize=20)\n",
    "ax.set_xticklabels(['Fridge','Sockets','Lights'],fontsize=20)\n",
    "ax.set_xlim([min(x)-0.5,max(x)+0.5])\n",
    "plt.xlabel('Appliances',fontsize=20)\n",
    "plt.ylabel('Recall',fontsize=20)\n",
    "plt.title('Combinatorial Optimization',fontsize=22)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x = np.arange(2)\n",
    "y = np.array([appliances_scores[i]['Accuracy'] for i in Names])\n",
    "y[np.isnan(y)] = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLQAAAIeCAYAAACiDFYRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11d75b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(18,8))\n",
    "plt.rc('font', size=20, **{'family': 'serif', 'serif': ['Computer Modern']})\n",
    "plt.rc('text', usetex=True)\n",
    "ax = f.add_axes([0.2,0.2,0.8,0.8])\n",
    "ax.bar(x,y,align='center')\n",
    "ax.set_xticks(x)\n",
    "ax.set_yticks(y)\n",
    "ax.set_yticklabels(y,fontsize=20)\n",
    "ax.set_xticklabels(['Fridge','Sockets','Lights'],fontsize=20)\n",
    "ax.set_xlim([min(x)-0.5,max(x)+0.5])\n",
    "plt.xlabel('Appliances',fontsize=20)\n",
    "plt.ylabel('Accuracy',fontsize=20)\n",
    "plt.title('Combinatorial Optimization',fontsize=22)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}