Tests using nilm_gjw_data

see gjw_converter_test to generate initial datastore from .csv files

First import the modules needed



In [1]:

    
import numpy as np
import pandas as pd
from os.path import join

from pylab import rcParams
import matplotlib.pyplot as plt
%matplotlib inline
rcParams['figure.figsize'] = (14, 6)

plt.style.use('ggplot')
import nilmtk
from nilmtk import DataSet, TimeFrame, MeterGroup, HDFDataStore
from nilmtk.disaggregate import CombinatorialOptimisation
from nilmtk.utils import print_dict
from nilmtk.metrics import f1_score

import warnings
warnings.filterwarnings("ignore")

Load the dataset



In [2]:

    
gjw = DataSet("C:/Users/GJWood/nilm_gjw_data/HDF5/nilm_gjw_data.hdf5")
print('loaded ' + str(len(gjw.buildings)) + ' buildings')









    



loaded 1 buildings

Examine metadata for a single house



In [3]:

    
building_number =1
print_dict(gjw.buildings[building_number].metadata)









    




description: Family home
ownership: bought
description_of_occupants: 2 adults and 1 teenager.
original_name: house_1
heating: natural gas
instance: 1
building_type: detached
energy_improvements: loft insulation
solid wall insulation
double glazing
rooms: {'instance': 1, 'name': 'lounge', 'floor': 0}
{'instance': 1, 'name': 'hall', 'floor': 0}
{'instance': 1, 'name': 'kitchen', 'floor': 0}
{'instance': 1, 'name': 'utility', 'floor': 0}
{'instance': 1, 'name': 'dining room', 'floor': 0}
{'instance': 1, 'name': 'family room', 'floor': 0}
{'instance': 1, 'name': 'wc', 'floor': 0}
{'instance': 1, 'name': 'garage', 'floor': 0}
{'instance': 1, 'name': 'shed', 'floor': 0}
{'instance': 1, 'name': 'greenhouse', 'floor': 0}
{'instance': 1, 'floor': 1, 'description': 'master bedroom', 'name': 'bedroom'}
{'instance': 2, 'floor': 1, 'description': 'guest bedroom', 'name': 'bedroom'}
{'instance': 3, 'floor': 1, 'description': 'teenagers bedroom', 'name': 'bedroom'}
{'instance': 4, 'floor': 1, 'description': 'spare bedroom', 'name': 'bedroom'}
{'instance': 1, 'name': 'study', 'floor': 1}
{'instance': 1, 'floor': 1, 'description': 'ensuite', 'name': 'bathroom'}
{'instance': 2, 'floor': 1, 'description': 'family bathroom', 'name': 'bathroom'}
{'instance': 1, 'name': 'landing', 'floor': 1}
timeframe: start: 2013-11-12 09:28:27+00:00
end: 2015-05-27 14:41:08+01:00
construction_year: 1970
dataset: gjwdata
n_occupants: 3

List Appliances



In [5]:

    
elec = gjw.buildings[building_number].elec
elec.appliances









    Out[5]:





[Appliance(type='tumble dryer', instance=1),
 Appliance(type='wireless phone charger', instance=1),
 Appliance(type='light', instance=15),
 Appliance(type='CE appliance', instance=1),
 Appliance(type='fan', instance=1),
 Appliance(type='light', instance=11),
 Appliance(type='television', instance=2),
 Appliance(type='electric oven', instance=1),
 Appliance(type='light', instance=7),
 Appliance(type='food steamer', instance=1),
 Appliance(type='server computer', instance=1),
 Appliance(type='light', instance=14),
 Appliance(type='blu-ray player', instance=1),
 Appliance(type='plate warmer', instance=1),
 Appliance(type='vacuum cleaner', instance=2),
 Appliance(type='light', instance=3),
 Appliance(type='light', instance=10),
 Appliance(type='water pump', instance=1),
 Appliance(type='light', instance=6),
 Appliance(type='coffee maker', instance=1),
 Appliance(type='light', instance=13),
 Appliance(type='vacuum cleaner', instance=1),
 Appliance(type='fridge', instance=1),
 Appliance(type='light', instance=2),
 Appliance(type='light', instance=9),
 Appliance(type='microwave', instance=1),
 Appliance(type='toaster', instance=1),
 Appliance(type='waste disposal unit', instance=1),
 Appliance(type='light', instance=5),
 Appliance(type='light', instance=12),
 Appliance(type='audio system', instance=1),
 Appliance(type='dish washer', instance=1),
 Appliance(type='light', instance=1),
 Appliance(type='washing machine', instance=1),
 Appliance(type='boiler', instance=1),
 Appliance(type='clothes iron', instance=1),
 Appliance(type='electric space heater', instance=1),
 Appliance(type='CE appliance', instance=2),
 Appliance(type='freezer', instance=1),
 Appliance(type='television', instance=1),
 Appliance(type='kettle', instance=1),
 Appliance(type='CD player', instance=1),
 Appliance(type='water filter', instance=1),
 Appliance(type='light', instance=4),
 Appliance(type='mobile phone charger', instance=1),
 Appliance(type='motor', instance=1),
 Appliance(type='light', instance=8)]

Show the meter's relationships



In [6]:

    
elec









    Out[6]:





MeterGroup(meters=
  ElecMeter(instance=1, building=1, dataset='gjwdata', site_meter, appliances=[Appliance(type='washing machine', instance=1), Appliance(type='tumble dryer', instance=1), Appliance(type='freezer', instance=1), Appliance(type='boiler', instance=1), Appliance(type='light', instance=1), Appliance(type='light', instance=2), Appliance(type='clothes iron', instance=1), Appliance(type='food steamer', instance=1), Appliance(type='kettle', instance=1), Appliance(type='toaster', instance=1), Appliance(type='coffee maker', instance=1), Appliance(type='microwave', instance=1), Appliance(type='electric oven', instance=1), Appliance(type='fridge', instance=1), Appliance(type='dish washer', instance=1), Appliance(type='waste disposal unit', instance=1), Appliance(type='water filter', instance=1), Appliance(type='fan', instance=1), Appliance(type='light', instance=3), Appliance(type='mobile phone charger', instance=1), Appliance(type='television', instance=1), Appliance(type='server computer', instance=1), Appliance(type='wireless phone charger', instance=1), Appliance(type='light', instance=4), Appliance(type='light', instance=5), Appliance(type='light', instance=6), Appliance(type='vacuum cleaner', instance=1), Appliance(type='vacuum cleaner', instance=2), Appliance(type='television', instance=2), Appliance(type='blu-ray player', instance=1), Appliance(type='light', instance=7), Appliance(type='light', instance=8), Appliance(type='audio system', instance=1), Appliance(type='CD player', instance=1), Appliance(type='light', instance=9), Appliance(type='light', instance=10), Appliance(type='light', instance=11), Appliance(type='plate warmer', instance=1), Appliance(type='light', instance=12), Appliance(type='motor', instance=1), Appliance(type='light', instance=13), Appliance(type='electric space heater', instance=1), Appliance(type='light', instance=14), Appliance(type='light', instance=15), Appliance(type='water pump', instance=1), Appliance(type='CE appliance', instance=1), Appliance(type='CE appliance', instance=2)])
)

List available data for an appliance



In [7]:

    
fridge = elec['fridge']
fridge.available_columns()









    Out[7]:





[('power', 'apparent'), ('power', 'reactive')]

List coherent blocks of meter readings



In [8]:

    
mains = elec.mains()
mains_good_sections = elec.mains().good_sections()
elec.mains().good_sections()









    Out[8]:





[TimeFrame(start='2013-10-17 18:19:07+01:00', end='2014-01-09 16:04:49+00:00', empty=False),
 TimeFrame(start='2014-01-12 13:02:22+00:00', end='2014-02-09 13:29:49+00:00', empty=False),
 TimeFrame(start='2014-03-13 21:59:56+00:00', end='2014-03-26 11:19:40+00:00', empty=False),
 TimeFrame(start='2014-04-27 00:17:57+01:00', end='2014-05-26 20:19:31+01:00', empty=False),
 TimeFrame(start='2015-04-17 12:48:40+01:00', end='2015-06-10 17:41:16+01:00', empty=False)]



In [9]:

    
mains_good_sections.plot()









    Out[9]:





<matplotlib.axes._subplots.AxesSubplot at 0x17f9aba8>

Find out what power types are available



In [10]:

    
mains.available_power_ac_types()









    Out[10]:





['apparent', 'reactive']



In [11]:

    
mains_energy = mains.total_energy(sections=mains_good_sections)
mains_energy









    Out[11]:





reactive     203.194441
apparent    3387.119111
dtype: float64



In [16]:

    
whole_house = nilmtk.global_meter_group.select()
whole_house.select(building=1).total_energy()









    



Calculating total_energy for ElecMeterID(instance=1, building=1, dataset='gjwdata') ...   





    Out[16]:





reactive     203.202124
apparent    3387.235051
dtype: float64



In [17]:

    
whole_house.plot()









    Out[17]:





<matplotlib.axes._subplots.AxesSubplot at 0x17ff0438>



In [26]:

    
gjw.set_window(start='2013-11-13 00:00:00', end='2013-11-14 00:00:00')
elec.plot();
plt.xlabel("Time");



In [13]:

    
mains = elec.mains()



In [15]:

    
m = [mains.meters[1]]









    



---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-15-aa104cbab7ea> in <module>()
----> 1 m = [mains.meters[1]]

AttributeError: 'ElecMeter' object has no attribute 'meters'



In [22]:

    
m.power_series_all_data()









    Out[22]:





2013-05-24 05:30:00+05:30    282.744995
2013-05-24 05:30:01+05:30    288.757996
2013-05-24 05:30:02+05:30    288.757996
2013-05-24 05:30:03+05:30    297.394012
2013-05-24 05:30:04+05:30    303.722992
2013-05-24 05:30:05+05:30    297.273987
2013-05-24 05:30:06+05:30    297.330994
2013-05-24 05:30:07+05:30    297.330994
2013-05-24 05:30:08+05:30    282.475006
2013-05-24 05:30:09+05:30    282.580994
2013-05-24 05:30:10+05:30    282.899994
2013-05-24 05:30:11+05:30    282.915985
2013-05-24 05:30:12+05:30    283.238007
2013-05-24 05:30:13+05:30    283.238007
2013-05-24 05:30:14+05:30    283.174011
...
2013-08-05 21:02:46+05:30    696.109985
2013-08-05 21:02:49+05:30    696.047974
2013-08-05 21:02:50+05:30    695.791992
2013-08-05 21:02:53+05:30    695.333008
2013-08-05 21:02:54+05:30    695.291992
2013-08-05 21:02:57+05:30    695.656982
2013-08-05 21:02:58+05:30    695.742004
2013-08-05 21:03:01+05:30    696.257019
2013-08-05 21:03:04+05:30    696.877991
2013-08-05 21:03:07+05:30    696.685974
2013-08-05 21:03:09+05:30    696.762024
2013-08-05 21:03:10+05:30    696.864990
2013-08-05 21:03:11+05:30    696.864990
2013-08-05 21:03:12+05:30    697.007996
2013-08-05 21:03:15+05:30    697.237976
Name: (power, active), Length: 4391004



In [23]:

    
m.total_energy()









    Out[23]:





active      1005.572193
reactive      96.815441
apparent    1043.941205
dtype: float64



In [10]:

    
import time
start  =  time.time()
elec.fraction_per_meter()
end  =  time.time()









    



/home/nipun/git/nilmtk/nilmtk/preprocessing/clip.py:50: RuntimeWarning: No measurement limits for ('frequency', '').
  warn('No measurement limits for {}.'.format(measurement), RuntimeWarning)
/home/nipun/git/nilmtk/nilmtk/preprocessing/clip.py:50: RuntimeWarning: No measurement limits for ('voltage', '').
  warn('No measurement limits for {}.'.format(measurement), RuntimeWarning)
/home/nipun/git/nilmtk/nilmtk/preprocessing/clip.py:50: RuntimeWarning: No measurement limits for ('pf', '').
  warn('No measurement limits for {}.'.format(measurement), RuntimeWarning)
/home/nipun/git/nilmtk/nilmtk/preprocessing/clip.py:50: RuntimeWarning: No measurement limits for ('current', '').
  warn('No measurement limits for {}.'.format(measurement), RuntimeWarning)



In [14]:

    
print elec.fraction_per_meter()









    



12/12 ElecMeter(instance=12, building=1, dataset='iAWE', appliances=[Appliance(type='motor', instance=1)])1                     NaN
2                     NaN
3                     NaN
4                     NaN
5                     NaN
6                     NaN
7                     NaN
8                     NaN
9                     NaN
10                    NaN
11                    NaN
12                    NaN
(1, 1, iAWE)     0.192552
(2, 1, iAWE)     0.519530
(3, 1, iAWE)     0.049750
(4, 1, iAWE)     0.103877
(5, 1, iAWE)     0.103186
(6, 1, iAWE)     0.001082
(7, 1, iAWE)     0.021384
(8, 1, iAWE)     0.000453
(9, 1, iAWE)     0.000403
(10, 1, iAWE)    0.006101
(11, 1, iAWE)    0.001146
(12, 1, iAWE)    0.000536
dtype: float64



In [12]:

    
print end - start









    



7.8085038662



In [13]:

    
elec.select_top_k(k=4)









    Out[13]:






  
    
      
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
    
  
  
    
      active
       360.496060
       1005.572193
       82.958314
       203.569523
       203.889213
       2.152638
       24.351495
       0.885766
       0.722616
       11.495919
       0.558716
       1.077918
    
    
      apparent
       386.911386
       1043.941205
       99.966983
       208.730314
       207.341398
       2.173996
       42.969043
       0.911080
       0.810098
       12.258396
       2.302620
            NaN
    
    
      reactive
        30.660539
         96.815441
       53.965398
        42.482991
        35.209142
       0.018759
        0.002313
       0.000528
       0.249570
        0.000176
       2.087380
            NaN



In [14]:

    
elec.energy_per_meter()









    Out[14]:






  
    
      
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
    
  
  
    
      active
       360.496060
       1005.572193
       82.958314
       203.569523
       203.889213
       2.152638
       24.351495
       0.885766
       0.722616
       11.495919
       0.558716
       1.077918
    
    
      apparent
       386.911386
       1043.941205
       99.966983
       208.730314
       207.341398
       2.173996
       42.969043
       0.911080
       0.810098
       12.258396
       2.302620
            NaN
    
    
      reactive
        30.660539
         96.815441
       53.965398
        42.482991
        35.209142
       0.018759
        0.002313
       0.000528
       0.249570
        0.000176
       2.087380
            NaN



In [16]:

    
elec.total_energy?



In [ ]:

	1	2	3	4	5	6	7	8	9	10	11	12
active	360.496060	1005.572193	82.958314	203.569523	203.889213	2.152638	24.351495	0.885766	0.722616	11.495919	0.558716	1.077918
apparent	386.911386	1043.941205	99.966983	208.730314	207.341398	2.173996	42.969043	0.911080	0.810098	12.258396	2.302620	NaN
reactive	30.660539	96.815441	53.965398	42.482991	35.209142	0.018759	0.002313	0.000528	0.249570	0.000176	2.087380	NaN