In [1]:

    

from __future__ import print_function, division
import pandas as pd
import numpy as np
from copy import deepcopy
from os.path import join, isdir, isfile, dirname, abspath
from os import listdir, getcwd
import re
from sys import stdout, getfilesystemencoding
from nilmtk.datastore import Key
from nilmtk.timeframe import TimeFrame
from nilmtk.measurement import LEVEL_NAMES
from nilm_metadata import convert_yaml_to_hdf5
from inspect import currentframe, getfile, getsourcefile
from copy import copy


    

In [2]:

    

import matplotlib.pyplot as plt


    

In [3]:

    

from nilmtk import DataSet, TimeFrame
from nilmtk.disaggregate import CombinatorialOptimisation


    

In [4]:

    

train = DataSet("/home/nipun/Desktop/redd.h5")
test = copy(train)


    

In [5]:

    

elec = train.buildings[2].elec


    

In [6]:

    

building_2 = train.buildings[2]


    

In [7]:

    

elec.map_meter_instances_to_appliance_ids()


    

    
Out[7]:





{1: [],
 2: [],
 3: [Appliance(type='sockets', instance=1)],
 4: [Appliance(type='light', instance=1)],
 5: [Appliance(type='electric stove', instance=1)],
 6: [Appliance(type='microwave', instance=1)],
 7: [Appliance(type='washer dryer', instance=1)],
 8: [Appliance(type='sockets', instance=2)],
 9: [Appliance(type='fridge', instance=1)],
 10: [Appliance(type='dish washer', instance=1)],
 11: [Appliance(type='waste disposal unit', instance=1)]}

In [25]:

    

top_k = elec.select_top_k(k=6)


    

In [26]:

    

top_k


    

    
Out[26]:





MeterGroup(meters=
  ElecMeter(instance=3, building=2, dataset='REDD', appliances=[Appliance(type='sockets', instance=1)])
  ElecMeter(instance=4, building=2, dataset='REDD', appliances=[Appliance(type='light', instance=1)])
  ElecMeter(instance=6, building=2, dataset='REDD', appliances=[Appliance(type='microwave', instance=1)])
  ElecMeter(instance=8, building=2, dataset='REDD', appliances=[Appliance(type='sockets', instance=2)])
  ElecMeter(instance=9, building=2, dataset='REDD', appliances=[Appliance(type='fridge', instance=1)])
  ElecMeter(instance=10, building=2, dataset='REDD', appliances=[Appliance(type='dish washer', instance=1)])
)

In [27]:

    

elec.energy_per_meter().T


    

    
Out[27]:






  

    

      

      
active
      
apparent
    
  
  

    

      
1 
      
       NaN
      
 14.358204
    
    

      
2 
      
       NaN
      
 62.864776
    
    

      
3 
      
  2.066441
      
       NaN
    
    

      
4 
      
  8.886603
      
       NaN
    
    

      
5 
      
  0.483484
      
       NaN
    
    

      
6 
      
  5.141312
      
       NaN
    
    

      
7 
      
  0.714265
      
       NaN
    
    

      
8 
      
  3.431987
      
       NaN
    
    

      
9 
      
 26.634318
      
       NaN
    
    

      
10
      
  2.992981
      
       NaN
    
    

      
11
      
  0.026742
      
       NaN
    
  

In [28]:

    

split_point = elec.train_test_split(0.5)


    

In [29]:

    

split_point


    

    
Out[29]:





Timestamp('2011-04-25 03:22:04-0400', tz='US/Eastern')

In [30]:

    

elec.good_sections()


    

    
Out[30]:





[TimeFrame(start='2011-04-17 19:18:27-04:00', end='2011-04-17 19:22:04-04:00', empty=False),
 TimeFrame(start='2011-04-17 19:25:09-04:00', end='2011-04-17 19:26:14-04:00', empty=False),
 TimeFrame(start='2011-04-17 21:45:01-04:00', end='2011-04-17 21:45:25-04:00', empty=False),
 TimeFrame(start='2011-04-17 22:18:13-04:00', end='2011-04-17 22:18:26-04:00', empty=False),
 TimeFrame(start='2011-04-17 23:09:35-04:00', end='2011-04-17 23:09:43-04:00', empty=False),
 TimeFrame(start='2011-04-17 23:15:29-04:00', end='2011-04-17 23:15:37-04:00', empty=False),
 TimeFrame(start='2011-04-17 23:30:57-04:00', end='2011-04-17 23:31:21-04:00', empty=False),
 TimeFrame(start='2011-04-17 23:42:54-04:00', end='2011-04-17 23:43:03-04:00', empty=False),
 TimeFrame(start='2011-04-18 00:00:53-04:00', end='2011-04-18 00:01:02-04:00', empty=False),
 TimeFrame(start='2011-04-18 00:09:34-04:00', end='2011-04-18 01:21:53-04:00', empty=False),
 TimeFrame(start='2011-04-18 01:22:41-04:00', end='2011-04-18 01:29:10-04:00', empty=False),
 TimeFrame(start='2011-04-18 01:31:36-04:00', end='2011-04-19 12:13:37-04:00', empty=False),
 TimeFrame(start='2011-04-19 12:14:16-04:00', end='2011-04-19 18:45:17-04:00', empty=False),
 TimeFrame(start='2011-04-19 20:20:07-04:00', end='2011-04-21 17:45:11-04:00', empty=False),
 TimeFrame(start='2011-04-21 19:41:34-04:00', end='2011-04-22 02:08:11-04:00', empty=False),
 TimeFrame(start='2011-04-22 02:11:36-04:00', end='2011-04-22 22:47:03-04:00', empty=False),
 TimeFrame(start='2011-04-22 22:48:39-04:00', end='2011-04-23 12:23:17-04:00', empty=False),
 TimeFrame(start='2011-04-23 12:24:06-04:00', end='2011-04-23 13:36:09-04:00', empty=False),
 TimeFrame(start='2011-04-23 13:36:50-04:00', end='2011-04-23 17:02:16-04:00', empty=False),
 TimeFrame(start='2011-04-23 17:04:59-04:00', end='2011-04-24 20:46:50-04:00', empty=False),
 TimeFrame(start='2011-04-24 20:47:30-04:00', end='2011-04-25 19:27:17-04:00', empty=False),
 TimeFrame(start='2011-04-25 19:32:14-04:00', end='2011-04-27 03:17:55-04:00', empty=False),
 TimeFrame(start='2011-04-27 03:21:47-04:00', end='2011-04-29 17:48:41-04:00', empty=False),
 TimeFrame(start='2011-04-29 17:48:42-04:00', end='2011-05-01 01:03:59-04:00', empty=False),
 TimeFrame(start='2011-05-01 01:06:15-04:00', end='2011-05-01 23:04:44-04:00', empty=False),
 TimeFrame(start='2011-05-22 15:57:03-04:00', end='2011-05-22 19:59:16-04:00', empty=False)]

In [31]:

    

train.store.window = TimeFrame(start=None, end = split_point)
test.store.window = TimeFrame(start=split_point, end=None)


    

In [32]:

    

from nilmtk import DataSet, TimeFrame
from nilmtk.disaggregate import CombinatorialOptimisation


    

In [33]:

    

co = CombinatorialOptimisation()


    

In [34]:

    

co.train(top_k)


    

In [35]:

    

co.model


    

    
Out[35]:





{3: array([  0,  14, 777], dtype=int32),
 4: array([  0,  20, 155], dtype=int32),
 6: array([   0,   49, 1826], dtype=int32),
 8: array([   0,  191, 1054], dtype=int32),
 9: array([  0, 163, 410], dtype=int32),
 10: array([   0,  208, 1199], dtype=int32)}

In [36]:

    

from nilmtk import HDFDataStore
output = HDFDataStore('output.h5', 'w')
test_mains = test.buildings[2].elec.mains()
co.disaggregate(test_mains, output)


    

In [37]:

    

output.store.keys()


    

    
Out[37]:





['/building2/elec/meter10',
 '/building2/elec/meter1_2',
 '/building2/elec/meter3',
 '/building2/elec/meter4',
 '/building2/elec/meter6',
 '/building2/elec/meter8',
 '/building2/elec/meter9']

In [38]:

    

%matplotlib inline


    

In [39]:

    

fig, ax = plt.subplots(nrows=2, sharex=True)
output.store['/building2/elec/meter9'][:'May 3 2011'].plot(label='Predicted', ax=ax[0])
train.buildings[2].elec['fridge'].power_series_all_data()[:'May 3 2011'].plot(ax=ax[1])
plt.legend()


    

    
Out[39]:





<matplotlib.legend.Legend at 0x7f7c6e114a90>






    

In [41]:

    

output.store['fridge']


    

    




---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-41-2aad1e896a87> in <module>()
----> 1 output.store['fridge']

/home/nipun/anaconda/lib/python2.7/site-packages/pandas/io/pytables.pyc in __getitem__(self, key)
    414 
    415     def __getitem__(self, key):
--> 416         return self.get(key)
    417 
    418     def __setitem__(self, key, value):

/home/nipun/anaconda/lib/python2.7/site-packages/pandas/io/pytables.pyc in get(self, key)
    617         group = self.get_node(key)
    618         if group is None:
--> 619             raise KeyError('No object named %s in the file' % key)
    620         return self._read_group(group)
    621 

KeyError: 'No object named fridge in the file'

In [ ]: