def disaggregate_original_co(h5_input, h5_output,dataset_start_date_disag, dataset_end_date_disag, centroids=None ):
import nilmtk.disaggregate as original_nilmtk
ds = DataSet(h5_input)
elec = ds.buildings[1].elec
vampire_power_used_in_original = elec.mains().vampire_power()
#Train
plain_co = original_nilmtk.CombinatorialOptimisation()
plain_co.train(elec)
#Modify centroids manually
if centroids is not None:
for i, model in enumerate(plain_co.model):
instance = model['training_metadata'].instance()
model['states'] = centroids[instance]
#Disaggregate
ds.set_window(start=dataset_start_date_disag, end=dataset_end_date_disag)
elec = ds.buildings[1].elec
output_plain_co = HDFDataStore(h5_output, 'w')
plain_co.disaggregate(elec.mains(), output_plain_co)
output_plain_co.close()
return plain_co, vampire_power_used_in_original
def hart85(start_train, end_train, start_test, end_test, train_elec):
#Start training
data.set_window(start_train, end_train)
elec = data.buildings[1].elec
hart = hart_85.Hart85()
hart.train(train_elec, sample_period=1)
#Start disaggregating
data.set_window(start_test, end_test)
disag_filename = './build/disagg_sum_hart85_{}_k.h5'.format(
len(train_elec.meters))
output = HDFDataStore(disag_filename, 'w')
hart.disaggregate(elec.mains(), output)
output.close()
disag = DataSet(disag_filename)
disag_elec = disag.buildings[1].elec
disag_elec.plot()
plt.show()
plt.title("HART85")
#Calculate F1-Score
f1 = f1_score(disag_elec, train_elec)
f1.index = disag_elec.get_labels(f1.index)
f1.plot(kind='barh')
plt.ylabel('appliance')
plt.xlabel('f-score')
plt.title("Hart85")
plt.show()
def __init__(self, in_filepath, out_filepath):
print("Loading DataStore and Generating Dataset...")
self.km = {}
self.dataStore = HDFDataStore(in_filepath)
self.dataSet = DataSet()
self.dataSet.load(self.dataStore)
self.outDataStore = HDFDataStore(out_filepath, 'w')
self.co = CombinatorialOptimisation()
self.train_group = {}
print("Data Properly Loaded!")
def __init__ (self,in_filepath,out_filepath):
print("Loading DataStore and Generating Dataset...")
self.km = {}
self.dataStore = HDFDataStore(in_filepath)
self.dataSet = DataSet()
self.dataSet.load(self.dataStore)
self.outDataStore = HDFDataStore(out_filepath,'w')
self.co = CombinatorialOptimisation()
self.train_group = {}
print("Data Properly Loaded!")
def disaggregate_building_to_file(self,
building_idx,
filename,
model=None):
if model == None:
model = self.fit_a_model(building_idx)
elec = self.get_elec_meter_data_of_a_building(building_idx)
output = HDFDataStore(filename, 'w')
model.disaggregate(elec.mains(), output)
output.close()
def co(start_train, end_train, start_test, end_test, train_elec):
#Start training
data.set_window(start_train, end_train)
elec = data.buildings[1].elec
co = CombinatorialOptimisation()
co.train(train_elec,
ac_type='active',
physical_quantity='power',
sample_period=1)
#Start disaggregating
data.set_window(start_test, end_test)
disag_filename = './build/disagg_sum_co_{}_k.h5'.format(
len(train_elec.meters))
output = HDFDataStore(disag_filename, 'w')
co.disaggregate(elec.mains(),
output,
ac_type='active',
physical_quantity='power',
sample_period=1)
output.close()
dates_dict = {
"start_train": start_train,
"end_train": end_train,
"start_test": start_test,
"end_test": end_test
}
# write test and train timeframe into json file
with open(disag_filename + ".json", 'w') as dates_file:
json.dump(dates_dict, dates_file)
#Calulate F1-Score
disag = DataSet(disag_filename)
disag_elec = disag.buildings[1].elec
disag_elec.plot()
plt.title("CO")
plt.show()
f1 = f1_score(disag_elec, train_elec)
f1.index = disag_elec.get_labels(f1.index)
f1.plot(kind='barh')
plt.ylabel('appliance')
plt.xlabel('f-score')
plt.title("CO")
plt.show()
def mle(start_train, end_train, start_test, end_test, train_elec):
# #Start training
data.set_window(start_train, end_train)
elec = data.buildings[1].elec
mle = maximum_likelihood_estimation.MLE()
mle.sample_period = "1s"
mle.train(train_elec)
#Start disaggregating
data.set_window(start_test, end_test)
disag_filename = './build/disagg_sum_mle_{}_k.h5'.format(
len(train_elec.meters))
output = HDFDataStore(disag_filename, 'w')
mle.disaggregate(elec.mains(), output)
output.close()
dates_dict = {
"start_train": start_train,
"end_train": end_train,
"start_test": start_test,
"end_test": end_test
}
# write test and train timeframe into json file
with open(disag_filename + ".json", 'w') as dates_file:
json.dump(dates_dict, dates_file)
disag = DataSet(disag_filename)
disag_elec = disag.buildings[1].elec
disag_elec.plot()
plt.show()
plt.title("FHMM")
#Calculate F1-Score
f1 = f1_score(disag_elec, train_elec)
f1.index = disag_elec.get_labels(f1.index)
f1.plot(kind='barh')
plt.ylabel('appliance')
plt.xlabel('f-score')
plt.title("FHMM")
plt.show()
class REDD_Data(object):
'''
REDD_Data Class is an object designed to abstract the lower level commands of
the NILMTK software package, with focus on the use of REDD DataSet. Function is
designed to allow rapid experimentation and disaggregation compared to attempting
to set package up from scratch.
This class requires the following for proper usage:
- NILMTK package: https://github.com/nilmtk
- REDD Dataset (converted to .h5): redd.csail.mit.edu
- Various dependancies (that NILMTK also requires), most can be downloaded through
Anaconda: continuum.io/downloads
Parameters
-----------
in_filepath: Filepath of converted REDD dataset (in .h5 format)
out_filepath: filepath to place output disaggregation dataset (in .h5 format)
Attributes
-----------
km: Key_Map Object
initializes the key_map object which will allow for the mapping of a meters
appliance name to its specific .H5 key.
dataStore: NILMTK HDFDataStore Object
the HDFDataStore that will contain the converted REDD DataSet.
dataSet: NILMTK DataSet Object
the DataSet object that is generated from the REDD DataStore (self.dataStore)
outDataStore: NILMTK HDFDataStore Object
the HDFDataStore that will contain the disaggregated dataset.
co: NILMTK CombinatorialOptimisation object
the disaggregation model object that will be trained and will disaggregate the
working dataset
train_group: NILMTK MeterGroup object
the MeterGroup object that is used to train the disaggregation model (self.co)
'''
def __init__ (self,in_filepath,out_filepath):
print("Loading DataStore and Generating Dataset...")
self.km = {}
self.dataStore = HDFDataStore(in_filepath)
self.dataSet = DataSet()
self.dataSet.load(self.dataStore)
self.outDataStore = HDFDataStore(out_filepath,'w')
self.co = CombinatorialOptimisation()
self.train_group = {}
print("Data Properly Loaded!")
def train_disag_model(self,building_inst, use_topk = False, k = 5):
'''
Function trains the disaggregation model using a selected MeterGroup.
Parameters
-----------
building_inst: the instance # of the building that you wish to grab the
training group from.
use_topk: true if you wish to only grab the top k most energy intensive
appliance to train the model, false if you wish to use all
appliances.
k: the # of appliances you wish to use (if use_topk = True)
'''
print("Training CO Disaggregation Model using given metergroup...")
if (building_inst <= 6) & (building_inst > 0):
#Select appropiate meter group to train with
if use_topk == True:
self.train_group = self.dataSet.buildings[building_inst].elec.select_top_k(k)
else:
self.train_group = self.dataSet.buildings[building_inst].elec
self.co.train(self.train_group)
print("CO Disaggreation Model Sucessfully Trained!")
else:
print("Error: Please select a building_inst of 1-6.")
print("Model unsucessfully trained.")
def load_disag_model(self, filepath):
'''
Function loads the disaggregation model to a file.
Parameters
-----------
filepath: exact filepath of the model file.
'''
print("Loading CO Disaggreation Model...")
self.co.import_model(filepath)
print("Model Sucessfully Loaded!")
def save_disag_model(self,filepath):
'''
Function saves the disaggregation model to a file.
Parameters
-----------
filepath: exact filepath of the model file.
'''
print("Saving CO Disaggregation Model...")
self.co.export_model(filepath)
print("Model Sucessfully Saved!")
def disaggregate(self,building_inst):
'''
Function will disaggregate the mains MeterGroup of the passed building
instance, and save this to the self.outDataStore object.
Parameters
-----------
building_inst: instance # of the building mains you wish to disaggregate.
'''
print("Disaggregating Building Mains...")
self.co.disaggregate(self.dataSet.buildings[building_inst].elec.mains(),self.outDataStore)
print("Mains sucessfully disaggregated!")
def close(self):
'''
Function closes all open DataStore's being used by the program.
'''
print("Closing DataStores...")
self.dataStore.close()
self.outDataStore.close()
print("Output DataStores Sucessfully Closed")
'''
All Plot Functions below are a WORK IN PROGRESS!-----------------------------------
Documentation will be provided upon completion.------------------------------------
'''
def plot_disag_apl(self,inst,appliance,t1="",t2=""):
self.km = Key_Map(inst)
plot_series(self.outDataStore.store.get(self.km.get_key(appliance))[t1: t2])
plt.title("Disaggregated " + appliance.capitalize()+" Energy")
plt.show()
def show_plots(self):
plt.show()
def building_plot_all(self,building_inst,t1,t2):
self.dataSet.buildings[building_inst].elec.plot(t1,t2)
plt.title("Building "+str(building_inst)+" Energy per Appliance")
plt.ylabel('Power [W]')
plt.xlabel('Hour')
def plot_redd_mains_data(self, inst=1, t1 = "", t2 = ""):
self.km = Key_Map(inst)
series1 = self.dataStore.store.get(self.km.get_key('mains1'))[t1:t2]
series2 = self.dataStore.store.get(self.km.get_key('mains2'))[t1:t2]
plot_series(series1 + series2)
plt.title("Building "+str(inst)+" Mains Energy")
plt.show()
def setUpClass(cls):
filename = join(data_dir(), 'energy.h5')
cls.datastore = HDFDataStore(filename)
ElecMeter.load_meter_devices(cls.datastore)
plt.clf()
# Define the buildings to be used for training and disaggregation
train_building = 1
disag_building = 1
#~ ## Dummy training and disaggregation
### Training
dum = DummyDisaggregator()
print('\n== dum.train(dataset.buildings[%d].elec)' % (train_building))
dum.train(dataset.buildings[train_building].elec)
### Disaggregation
dum_outfile = dataset_directory / ('%s-da-co.h5' % (dataset_name.lower()))
output = HDFDataStore(str(dum_outfile), 'w')
print('\n== dum.disaggregate(dataset.buildings[%d].mains(), output)' %
(disag_building))
dum.disaggregate(dataset.buildings[disag_building].elec.mains(), output)
output.close()
### Results
print('\n== Plotting Dummy disaggregation results...')
da_data = DataSet(str(dum_outfile))
da_elec = da_data.buildings[disag_building].elec
ax = da_elec.plot()
ax.set_title("B%d Dummy disaggregation results" % (disag_building))
plt.savefig('results/%s__b%d__elec__dummy.png' %
(dataset_name, disag_building))
plt.clf()
f1 = f1_score(da_elec, dataset.buildings[disag_building].elec)
def generate_vertices():
"""
Predicts the power demand of the target appliance using the intermediate models which are exported during training.
Generates a polygon from those predictions.
"""
train = DataSet('../data/ukdale.h5')
train.clear_cache()
train.set_window(start="13-4-2013", end="31-7-2013")
test = DataSet('../data/ukdale.h5')
test.clear_cache()
test.set_window(start='7-2-2014 08:00:00', end='7-3-2014')
train_building = 1
test_building = 5
sample_period = 6
meter_key = 'kettle'
learning_rate = 1e-5
train_elec = train.buildings[train_building].elec
test_elec = test.buildings[test_building].elec
train_meter = train_elec.submeters()[meter_key]
test_mains = test_elec.mains()
results_dir = '../results/UKDALE-ACROSS-BUILDINGS-RNN-lr=1e-05-2018-02-03-11-48-12'
train_logfile = os.path.join(results_dir, 'training.log')
val_logfile = os.path.join(results_dir, 'validation.log')
rnn = RNNDisaggregator(train_logfile,
val_logfile,
learning_rate,
init=False)
verts = []
zs = [] # epochs
for z in np.arange(10, 341, 10):
# disaggregate model
model = 'UKDALE-RNN-kettle-{}epochs.h5'.format(z)
rnn.import_model(os.path.join(results_dir, model))
disag_filename = 'disag-out-{}epochs.h5'.format(z)
output = HDFDataStore(os.path.join(results_dir, disag_filename), 'w')
results_file = os.path.join(results_dir,
'results-{}epochs.txt'.format(z))
rnn.disaggregate(test_mains,
output,
results_file,
train_meter,
sample_period=sample_period)
os.remove(results_file)
output.close()
# get predicted curve for epoch=z
result = DataSet(os.path.join(results_dir, disag_filename))
res_elec = result.buildings[test_building].elec
os.remove(os.path.join(results_dir, disag_filename))
predicted = res_elec[meter_key]
predicted = predicted.power_series(sample_period=sample_period)
predicted = next(predicted)
predicted.fillna(0, inplace=True)
ys = np.array(predicted) # power
xs = np.arange(ys.shape[0]) # timestamps
verts.append(list(zip(xs, ys))) # add list of x-y-coordinates
zs.append(z)
ground_truth = test_elec[meter_key]
ground_truth = ground_truth.power_series(sample_period=sample_period)
ground_truth = next(ground_truth)
ground_truth.fillna(0, inplace=True)
ys = np.array(ground_truth) # power
xs = np.arange(ys.shape[0]) # timestamps
verts.append(list(zip(xs, ys))) # add list of x-y-coordinates
zs.append(350)
zs = np.asarray(zs)
for i in range(len(verts)):
verts[i].insert(0, [0, np.array([0])])
verts[i].append([len(verts[i]), np.array([0])])
pickle.dump(verts, open(os.path.join(results_dir, 'vertices.pkl'), 'wb'))
pickle.dump(zs, open(os.path.join(results_dir, 'zs.pkl'), 'wb'))
pickle.dump(ys, open(os.path.join(results_dir, 'ys.pkl'), 'wb'))
def get_disaggregation(device, total_aggregate):
here = os.path.dirname(os.path.abspath(__file__))
dataset_file = os.path.join(here, "dataset/iawe2.h5")
devices = ["fridge", "air conditioner", "washing machine"]
if device not in devices:
return None
total_seconds = 30 * 24 * 60
val_per_second = float(total_aggregate) / (total_seconds)
print(val_per_second)
start = 0
end = 0
with h5py.File(dataset_file, "r+") as f1:
table = f1["building1/elec/meter1/table"].value
start = int(str(table[0][0])[:10])
end = start + total_seconds
print(end - start, total_seconds)
# for i in range(total_seconds):
# # for j in range(7):
# print("Progress {:2.1%}".format(i / total_seconds), end="\r")
# table[i][1][2] = val_per_second + np.random.uniform(-1e-17,
# 1e-17, 1)
# f1["building1/elec/meter1/table"][...] = table
# print(table)
# start = datetime.fromtimestamp(start)
end = datetime.fromtimestamp(end)
# start = start.isoformat(' ', 'seconds')
end = end.isoformat(' ', 'seconds')
# print(start, end)
test = DataSet(dataset_file)
# test.set_window(start=start, end=end)
test.set_window(end=end)
test_elec = test.buildings[1].elec
test_mains = test_elec.mains()[1]
df = next(test_mains.load())
print(df)
test_meter = test_elec.submeters()[device]
disag_filename = 'disag-out.h5' # The filename of the resulting datastore
output = HDFDataStore(disag_filename, 'w')
disaggregator = ShortSeq2PointDisaggregator()
model_file = os.path.join(
here, "disag1/IAWE-RNN-h{}-{}-{}epochs.h5".format(1, device, 10))
disaggregator.import_model(model_file)
# anykey = input()
# test_mains: The aggregated signal meter
# output: The output datastore
# train_meter: This is used in order to copy the metadata of the train
# meter into the datastore
disaggregator.disaggregate(test_mains, output, test_mains, sample_period=1)
output.close()
result = DataSet(disag_filename)
res_elec = result.buildings[1].elec
# prediction = res_elec[device]
prediction = res_elec
# df = next(prediction.load())
# prediction = df["power"]["active"][0]
return prediction
def dae(dataset_path, train_building, train_start, train_end, test_building,
test_start, test_end, val_building, val_start, val_end, meter_key,
sample_period, num_epochs, patience, sequence_length, optimizer,
learning_rate, loss):
# Start tracking time
start = time.time()
# Prepare dataset and options
# print("========== OPEN DATASETS ============")
dataset_path = dataset_path
train = DataSet(dataset_path)
train.set_window(start=train_start, end=train_end)
val = DataSet(dataset_path)
val.set_window(start=val_start, end=val_end)
test = DataSet(dataset_path)
test.set_window(start=test_start, end=test_end)
train_building = train_building
test_building = test_building
meter_key = meter_key
sample_period = sample_period
train_elec = train.buildings[train_building].elec
val_elec = val.buildings[val_building].elec
test_elec = test.buildings[test_building].elec
train_meter = train_elec.submeters()[meter_key]
try:
train_mains = train_elec.mains().all_meters()[0]
val_mains = val_elec.mains().all_meters()[0]
test_mains = test_elec.mains().all_meters()[0]
except AttributeError:
train_mains = train_elec.mains()
test_mains = test_elec.mains()
dae = DAEDisaggregator(sequence_length, patience, optimizer, learning_rate,
loss)
# print("========== TRAIN ============")
dae.train(train_mains,
train_meter,
epochs=num_epochs,
sample_period=sample_period)
# Get number of earlystop epochs
num_epochs = dae.stopped_epoch if dae.stopped_epoch != 0 else num_epochs
#dae.export_model("results/dae-model-{}-{}epochs.h5".format(meter_key, num_epochs))
# print("========== DISAGGREGATE ============")
# Validation
val_disag_filename = 'disag-out-val.h5'
output = HDFDataStore(val_disag_filename, 'w')
dae.disaggregate(val_mains,
output,
train_meter,
sample_period=sample_period)
output.close()
# Test
test_disag_filename = 'disag-out-test.h5'
output = HDFDataStore(test_disag_filename, 'w')
dae.disaggregate(test_mains,
output,
train_meter,
sample_period=sample_period)
output.close()
# print("========== RESULTS ============")
# Validation
result_val = DataSet(val_disag_filename)
res_elec_val = result_val.buildings[val_building].elec
rpaf_val = metrics.recall_precision_accuracy_f1(res_elec_val[meter_key],
val_elec[meter_key])
val_metrics_results_dict = {
'recall_score':
rpaf_val[0],
'precision_score':
rpaf_val[1],
'accuracy_score':
rpaf_val[2],
'f1_score':
rpaf_val[3],
'mean_absolute_error':
metrics.mean_absolute_error(res_elec_val[meter_key],
val_elec[meter_key]),
'mean_squared_error':
metrics.mean_square_error(res_elec_val[meter_key],
val_elec[meter_key]),
'relative_error_in_total_energy':
metrics.relative_error_total_energy(res_elec_val[meter_key],
val_elec[meter_key]),
'nad':
metrics.nad(res_elec_val[meter_key], val_elec[meter_key]),
'disaggregation_accuracy':
metrics.disaggregation_accuracy(res_elec_val[meter_key],
val_elec[meter_key])
}
# Test
result = DataSet(test_disag_filename)
res_elec = result.buildings[test_building].elec
rpaf = metrics.recall_precision_accuracy_f1(res_elec[meter_key],
test_elec[meter_key])
test_metrics_results_dict = {
'recall_score':
rpaf[0],
'precision_score':
rpaf[1],
'accuracy_score':
rpaf[2],
'f1_score':
rpaf[3],
'mean_absolute_error':
metrics.mean_absolute_error(res_elec[meter_key], test_elec[meter_key]),
'mean_squared_error':
metrics.mean_square_error(res_elec[meter_key], test_elec[meter_key]),
'relative_error_in_total_energy':
metrics.relative_error_total_energy(res_elec[meter_key],
test_elec[meter_key]),
'nad':
metrics.nad(res_elec[meter_key], test_elec[meter_key]),
'disaggregation_accuracy':
metrics.disaggregation_accuracy(res_elec[meter_key],
test_elec[meter_key])
}
# end tracking time
end = time.time()
time_taken = end - start # in seconds
# model_result_data = {
# 'algorithm_name': 'DAE',
# 'datapath': dataset_path,
# 'train_building': train_building,
# 'train_start': str(train_start.date()) if train_start != None else None ,
# 'train_end': str(train_end.date()) if train_end != None else None ,
# 'test_building': test_building,
# 'test_start': str(test_start.date()) if test_start != None else None ,
# 'test_end': str(test_end.date()) if test_end != None else None ,
# 'appliance': meter_key,
# 'sampling_rate': sample_period,
#
# 'algorithm_info': {
# 'options': {
# 'epochs': num_epochs
# },
# 'hyperparameters': {
# 'sequence_length': sequence_length,
# 'min_sample_split': None,
# 'num_layers': None
# },
# 'profile': {
# 'parameters': None
# }
# },
#
# 'metrics': metrics_results_dict,
#
# 'time_taken': format(time_taken, '.2f'),
# }
model_result_data = {
'val_metrics': val_metrics_results_dict,
'test_metrics': test_metrics_results_dict,
'time_taken': format(time_taken, '.2f'),
'epochs': num_epochs,
}
# Close digag_filename
result.store.close()
result_val.store.close()
# Close Dataset files
train.store.close()
val.store.close()
test.store.close()
return model_result_data
def nilmtkECOfunc(dataset_loc, train_start, train_end, test_start, test_end,
output_period):
#### configuration ####
period_s = output_period
building = 2
#### load ####
total = DataSet(dataset_loc)
train = DataSet(dataset_loc)
test = DataSet(dataset_loc)
train.set_window(start=train_start, end=train_end)
test.set_window(start=test_start, end=test_end)
print(train_start)
print(train_end)
print(test_start)
print(test_end)
#### get timeframe ####
tf_total = total.buildings[building].elec.mains().get_timeframe()
tf_train = train.buildings[building].elec.mains().get_timeframe()
tf_test = test.buildings[building].elec.mains().get_timeframe()
#### eletrical metergroup ####
total_elec = total.buildings[building].elec
train_elec = train.buildings[building].elec
test_elec = test.buildings[building].elec
#### training process ####
start = time.time()
from nilmtk.disaggregate import CombinatorialOptimisation
co = CombinatorialOptimisation()
co.train(train_elec, sample_period=period_s)
end = time.time()
print("Runtime =", end - start, "seconds.")
#### disaggregation process ####
start = time.time()
disag_filename = '../dataset/ecob-b2-kall-co-1w:11-1m.h5'
output = HDFDataStore(disag_filename, 'w')
co.disaggregate(test_elec.mains(), output, sample_period=period_s)
end = time.time()
print("Runtime =", end - start, "seconds.")
output.close()
disag_co = DataSet(disag_filename)
disag_co_elec = disag_co.buildings[building].elec
#### fraction energy assigned correctly ####
#FTE_co_all = FTE_func(disag_co_elec, test_elec);
#### total disaaggregation error ####
#Te_co_all = total_disag_err(disag_co_elec, test_elec);
#### creating dataframe from both disaggregated and ground truth metergroups
disag_co_elec_df = disag_co_elec.dataframe_of_meters()
disag_co_elec_df_nona = disag_co_elec_df.dropna()
gt_full_df = test_elec.dataframe_of_meters()
gt_full_df_nona = gt_full_df.dropna()
gt_df_nona = gt_full_df_nona.ix[disag_co_elec_df_nona.index]
#### jaccard ####
#Ja_co_all = jaccard_similarity(disag_co_elec_df_nona, gt_df_nona, disag_co_elec.submeters().instance(), test_elec.instance());
#print("FTE all", FTE_co_all);
#print("TE all", Te_co_all);
#print("Ja all", Ja_co_all);
#### output ####
# drop aggregated power
disag_co_elec_submeter_df = disag_co_elec_df.drop(
disag_co_elec_df.columns[[0]], axis=1)
# disag_co_elec_submeter_df = disag_co_elec_df
# drop the unwanted timestamp
gt_df_aligned = gt_full_df.ix[disag_co_elec_submeter_df.index]
# drop aggregated power
gt_df_sub = gt_df_aligned.drop(gt_df_aligned.columns[[0, 1, 2]], axis=1)
# train
train_elec_df = train_elec.dataframe_of_meters()
train_elec_df_aligned = train_elec_df.resample(str(period_s) +
'S').asfreq()[0:]
train_elec_df_aligned_drop = train_elec_df_aligned.drop(
train_elec_df_aligned.columns[[0, 1, 2]], axis=1)
return disag_co_elec_submeter_df, gt_df_sub, co, train_elec_df_aligned_drop
meter_key, epochs))
end = time.time()
print("Train =", end - start, "seconds.")
file = open('stackTrainSetsInfo_' + meter_key, 'r')
for line in file:
toks = line.split(',')
StackTrain = DataSet(toks[0])
print(toks[2], '-', toks[3])
StackTrain.set_window(start=toks[2], end=toks[3])
test_elec = StackTrain.buildings[int(toks[1])].elec
test_mains = test_elec.mains()
print("========== DISAGGREGATE (stackTrain)============")
disag_filename = "StackTrain-h" + toks[1] + ".h5"
output = HDFDataStore(disag_filename, 'w')
disaggregator.disaggregate(test_mains,
output,
test_elec[meter_key],
sample_period=sample_period)
output.close()
for i in test_building_list:
test_elec = test.buildings[i].elec
test_mains = test_elec.mains()
print("========== DISAGGREGATE ============")
disag_filename = "StackTest-" + str(i) + ".h5"
output = HDFDataStore(disag_filename, 'w')
disaggregator.disaggregate(test_mains,
output,
def nilmtkDREDfunc(dataset_loc, train_start, train_end, test_start, test_end,
output_period):
#### configuration ####
period_s = output_period
building = 1
#### load ####
total = DataSet(dataset_loc)
train = DataSet(dataset_loc)
test = DataSet(dataset_loc)
train.set_window(start=train_start, end=train_end)
test.set_window(start=test_start, end=test_end)
print(train_start)
print(train_end)
print(test_start)
print(test_end)
#### get timeframe ####
tf_total = total.buildings[building].elec.mains().get_timeframe()
tf_train = train.buildings[building].elec.mains().get_timeframe()
tf_test = test.buildings[building].elec.mains().get_timeframe()
#### eletrical metergroup ####
total_elec = total.buildings[building].elec
train_elec = train.buildings[building].elec
test_elec = test.buildings[building].elec
#### training process ####
start = time.time()
from nilmtk.disaggregate import CombinatorialOptimisation
co = CombinatorialOptimisation()
co.train(train_elec, sample_period=period_s)
end = time.time()
print("Runtime =", end - start, "seconds.")
#### disaggregation process ####
start = time.time()
disag_filename = dataset_loc + 'DREDapp.h5'
output = HDFDataStore(disag_filename, 'w')
co.disaggregate(test_elec.mains(), output, sample_period=period_s)
end = time.time()
print("Runtime =", end - start, "seconds.")
output.close()
disag_co = DataSet(disag_filename)
disag_co_elec = disag_co.buildings[building].elec
#### creating dataframe from both disaggregated and ground truth metergroups
disag_co_elec_df = disag_co_elec.dataframe_of_meters()
gt_full_df = test_elec.dataframe_of_meters()
# drop the NA, it might be needed (initially it is used for Ja)
disag_co_elec_df_nona = disag_co_elec_df.dropna()
gt_full_df_nona = gt_full_df.dropna()
# drop the unwanted timestamp
gt_df_nona = gt_full_df_nona.ix[disag_co_elec_df_nona.index]
#### output ####
# drop aggregated power from output
disag_co_elec_submeter_df = disag_co_elec_df.drop(
disag_co_elec_df.columns[[0]], axis=1)
# drop the unwanted timestamp on ground truth (take the sampled timestamp)
gt_df_aligned = gt_full_df.ix[disag_co_elec_submeter_df.index]
# drop aggregated power from ground truth
gt_df_sub = gt_df_aligned.drop(gt_df_aligned.columns[[0]], axis=1)
# train data frame, resample based in disaggregation period, drop the main power
train_elec_df = train_elec.dataframe_of_meters()
train_elec_df_aligned = train_elec_df.resample(str(period_s) +
'S').asfreq()[0:]
train_elec_df_aligned_drop = train_elec_df_aligned.drop(
train_elec_df_aligned.columns[[0]], axis=1)
return disag_co_elec_submeter_df, gt_df_sub, co, train_elec_df_aligned_drop
print(
"Calculating ground truth===============================================")
loc.dataset.set_window(start=dataset_start_date_disag,
end=dataset_end_date_disag)
gt = GroundTruth(loc, co, baseline=vampire_power_in_original)
gt.generate()
time_start_disag = time.time()
print("\nTotal elapsed: %s seconds ---" % (time_start_disag - start_time))
print("Section Ground truth : %s seconds ---\n" %
(time_start_disag - time_start_gt))
#DISAGREGGATION================================================================
print(
"Disaggregating=========================================================")
output = HDFDataStore(h5_disag, 'w')
loc.dataset.set_window(start=dataset_start_date_disag,
end=dataset_end_date_disag)
co.disaggregate(loc.elec.mains(),
output,
location_data=loc,
baseline=vampire_power_in_original,
resample_seconds=60)
output.close()
time_start_metrics = time.time()
print("\nTotal elapsed: %s seconds ---" % (time_start_metrics - start_time))
print("Section Disaggregation: %s seconds ---\n" %
(time_start_metrics - time_start_disag))
#METRICS=======================================================================
def test_all(path_to_directory):
'''
path_to_directory: Contains the h5 files on which the tests are supposed to be run
'''
check_directory_exists(path_to_directory)
#files=[f for f in listdir(path_to_directory) and '.h5' in f and '.swp' not in f]
files = [f for f in listdir(path_to_directory) if isfile(join(path_to_directory, f)) and
'.h5' in f and '.swp' not in f]
files.sort()
print ("Datasets collected and sorted. Processing...")
try:
for i, file in enumerate(files):
current_file=DataSet(join(path_to_directory, file))
print ("Printing metadata for current file...done.")
print_dict(current_file.metadata)
print (" Loading file # ", i, " : ", file, ". Please wait.")
for building_number in range(1, len(current_file.buildings)+1):
#Examine metadata for a single house
elec=current_file.buildings[building_number].elec
print ("The dataset being processed is : ", elec.dataset())
print ("Metadata for current file: ")
print_dict(current_file.buildings[building_number].metadata)
print ("Appliance label information: ", elec.appliance_label())
#print (elec.appliances)
print ("Appliances:- ")
for i in elec.appliances:
print (i)
print ("Examining sub-metered appliances...")
print ("Collecting stats on meters...Done.")
print (elec._collect_stats_on_all_meters)
print ("Timeframe: ", elec.get_timeframe())
print ("Available power AC types: ", elec.available_power_ac_types())
print ("Clearing cache...done.")
elec.clear_cache()
print ("Testing if there are meters from multiple buildings. Result returned by method: ", elec.contains_meters_from_multiple_buildings())
# TODO: Find a better way to test the correlation function
# print ("Testing the correlation function. ", elec.correlation(elec))
print ("List of disabled meters: ", elec.disabled_meters)
print ("Trying to determine the dominant appliance: ")
try:
elec.dominant_appliance()
except RuntimeError:
print ('''More than one dominant appliance in MeterGroup! (The dominant appliance per meter should be manually specified in the metadata. If it isn't and if there are multiple appliances for a meter then NILMTK assumes all appliances on that meter are dominant. NILMTK can't automatically distinguish between multiple appliances on the same meter (at least, not without using NILM!))''')
pass
print ("Dropout rate: ", elec.dropout_rate())
try:
print ("Calculating energy per meter:")
print (elec.energy_per_meter())
print ("Calculating total entropy")
print (elec.entropy())
print ("Calculating entropy per meter: ")
print (elec.entropy_per_meter())
except ValueError:
print ("ValueError: Total size of array must remain unchanged.")
pass
print ("Calculating fraction per meter.")
print (elec.fraction_per_meter())
#print ("Average energy per period: ", elec.average_energy_per_period())
print ("Executing functions...")
lis=[]
func=""
'''for function in dir(elec):
try:
start=time.time()
if ("__" not in function or "dataframe_of_meters" not in function):
func=getattr(elec, function)
print ("Currently executing ", function, ". Please wait...")
print (func())
# print ("cProfile stats - printed")
# cProfile.run("func")
end=time.time()
print ("Time taken for the entire process : ", (end - start))
except AttributeError:
print ("Attribute error occured. ")
except TypeError:
lis.append(function)
print ("Warning: TypeError")
pass'''
print ("Plotting wiring hierarchy of meters....")
elec.draw_wiring_graph()
## DISAGGREGATION STARTS HERE
appliance_type="unknown"
#TODO : appliance_type should cycle through all appliances and check for each of them. For this, use a list.
selected_appliance=nilmtk.global_meter_group.select_using_appliances(type=appliance_type)
appliance_restricted = MeterGroup(selected_appliance.meters)
if ((appliance_restricted.proportion_of_upstream_total_per_meter()) is not None):
proportion_per_appliance = appliance_restricted.proportion_of_upstream_total_per_meter()
proportion_per_appliance.plot(kind='bar');
plt.title('Appliance energy as proportion of total building energy');
plt.ylabel('Proportion');
plt.xlabel('Appliance (<appliance instance>, <building instance>, <dataset name>)');
selected_appliance.select(building=building_number).total_energy()
selected_appliance.select(building=1).plot();
appliance_restricted = MeterGroup(selected_appliance.meters)
daily_energy = pd.DataFrame([meter.average_energy_per_period(offset_alias='D')
for meter in appliance_restricted.meters])
daily_energy.plot(kind='hist');
plt.title('Histogram of daily energy');
plt.xlabel('energy (kWh)');
plt.ylabel('Occurences');
plt.legend().set_visible(False)
current_file.store.window=TimeFrame(start='2012-04-01 00:00:00-05:00', end='2012-04-02 00:00:00-05:00')
#elec.plot();
fraction = elec.submeters().fraction_per_meter().dropna()
labels = elec.get_appliance_labels(fraction.index)
plt.figure(figsize=(8,8))
fraction.plot(kind='pie', labels=labels);
elec.select_using_appliances(category='heating')
elec.select_using_appliances(category='single-phase induction motor')
co = CombinatorialOptimisation()
co.train(elec)
for model in co.model:
print_dict(model)
disag_filename = join(data_dir, 'ampds-disag.h5')
output = HDFDataStore(disag_filename, 'w')
co.disaggregate(elec.mains(), output)
output.close()
disag = DataSet(disag_filename)
disag_elec = disag.buildings[building_number].elec
f1 = f1_score(disag_elec, elec)
f1.index = disag_elec.get_appliance_labels(f1.index)
f1.plot(kind='bar')
plt.xlabel('appliance');
plt.ylabel('f-score');
disag_elec.plot()
disag.store.close()
except AttributeError:
print ("AttributeError occured while executing. This means that the value returned by proportion_per_appliance = appliance_restricted.proportion_of_upstream_total_per_meter() is None")
pass
def runExperiment(experiment: experimentInfo, metricsResFileName,
clearMetricsFile):
dsPathsList_Test = experiment.dsList
outFileName = experiment.outName
test_building = experiment.building
meter_key = experiment.meter_key
pathOrigDS = experiment.pathOrigDS
meterTH = experiment.meterTH
print('House ', test_building)
# Load a "complete" dataset to have the test's timerange
test = DataSet(dsPathsList_Test[0])
test_elec = test.buildings[test_building].elec
testRef_meter = test_elec.submeters(
)[meter_key] # will be used as reference to align all meters based on this
# Align every test meter with testRef_meter as master
test_series_list = []
for path in dsPathsList_Test:
test = DataSet(path)
test_elec = test.buildings[test_building].elec
test_meter = test_elec.submeters()[meter_key]
# print('Stack test: ', test_meter.get_timeframe().start.date(), " - ", test_meter.get_timeframe().end.date())
aligned_meters = align_two_meters(testRef_meter, test_meter)
test_series_list.append(aligned_meters)
# Init vars for the output
MIN_CHUNK_LENGTH = 300 # Depends on the basemodels of the ensemble
timeframes = []
building_path = '/building{}'.format(test_meter.building())
mains_data_location = building_path + '/elec/meter1'
data_is_available = False
disag_filename = outFileName
output_datastore = HDFDataStore(disag_filename, 'w')
run = True
chunkDataForOutput = None
# -- Used to hold necessary data for saving the results using NILMTK (e.g. timeframes).
# -- (in case where chunks have different size (not in current implementation), must use the chunk whose windowsSize is the least (to have all the data))
while run:
try:
testX = []
columnInd = 0
# Get Next chunk of each series
for testXGen in test_series_list:
chunkALL = next(testXGen)
chunk = chunkALL[
'slave'] # slave is the meter needed (master is only for aligning)
chunk.fillna(0, inplace=True)
if (columnInd == 0):
chunkDataForOutput = chunk # Use 1st found chunk for it's metadata
if (testX == []):
testX = np.zeros(
[len(chunk), len(test_series_list)]
) # Initialize the array that will hold all of the series as columns
testX[:, columnInd] = chunk[:]
columnInd += 1
testX = scaler.transform(testX)
except:
run = False
break
if len(chunkDataForOutput) < MIN_CHUNK_LENGTH:
continue
# print("New sensible chunk: {}".format(len(chunk)))
startTime = chunkDataForOutput.index[0]
endTime = chunkDataForOutput.index[
-1] # chunkDataForOutput.shape[0] - 1
# print('Start:',startTime,'End:',endTime)
timeframes.append(TimeFrame(
startTime, endTime)) #info needed for output for use with NILMTK
measurement = ('power', 'active')
pred = clf.predict(testX)
column = pd.Series(pred, index=chunkDataForOutput.index, name=0)
appliance_powers_dict = {}
appliance_powers_dict[0] = column
appliance_power = pd.DataFrame(appliance_powers_dict)
appliance_power[appliance_power < 0] = 0
# Append prediction to output
data_is_available = True
cols = pd.MultiIndex.from_tuples([measurement])
meter_instance = test_meter.instance()
df = pd.DataFrame(appliance_power.values,
index=appliance_power.index,
columns=cols,
dtype="float32")
key = '{}/elec/meter{}'.format(building_path, meter_instance)
output_datastore.append(key, df)
# Append aggregate data to output
mains_df = pd.DataFrame(chunkDataForOutput,
columns=cols,
dtype="float32")
# Note (For later): not 100% right. Should be mains. But it won't be used anywhere, so it doesn't matter in this case
output_datastore.append(key=mains_data_location, value=mains_df)
# Save metadata to output
if data_is_available:
disagr = Disaggregator()
disagr.MODEL_NAME = 'Stacked model'
disagr._save_metadata_for_disaggregation(
output_datastore=output_datastore,
sample_period=sample_period,
measurement=measurement,
timeframes=timeframes,
building=test_meter.building(),
meters=[test_meter])
#======================== Calculate Metrics =====================================
testYDS = DataSet(pathOrigDS)
testYDS.set_window(start=test_meter.get_timeframe().start.date(),
end=test_meter.get_timeframe().end.date())
testY_elec = testYDS.buildings[test_building].elec
testY_meter = testY_elec.submeters()[meter_key]
test_mains = testY_elec.mains()
result = DataSet(disag_filename)
res_elec = result.buildings[test_building].elec
rpaf = metrics.recall_precision_accuracy_f1(res_elec[meter_key],
testY_meter, meterTH, meterTH)
relError = metrics.relative_error_total_energy(res_elec[meter_key],
testY_meter)
MAE = metrics.mean_absolute_error(res_elec[meter_key], testY_meter)
RMSE = metrics.RMSE(res_elec[meter_key], testY_meter)
print("============ Recall: {}".format(rpaf[0]))
print("============ Precision: {}".format(rpaf[1]))
print("============ Accuracy: {}".format(rpaf[2]))
print("============ F1 Score: {}".format(rpaf[3]))
print("============ Relative error in total energy: {}".format(relError))
print("============ Mean absolute error(in Watts): {}".format(MAE))
print("=== For docs: {:.4}\t{:.4}\t{:.4}\t{:.4}\t{:.4}\t{:.4}".format(
rpaf[0], rpaf[1], rpaf[2], rpaf[3], relError, MAE))
# print("============ RMSE: {}".format(RMSE))
# print("============ TECA: {}".format(metrics.TECA([res_elec[meter_key]],[testY_meter],test_mains)))
resDict = {
'model': 'TEST',
'building': test_building,
'Appliance': meter_key,
'Appliance_Type': 2,
'Recall': rpaf[0],
'Precision': rpaf[1],
'Accuracy': rpaf[2],
'F1': rpaf[3],
'relError': relError,
'MAE': MAE,
'RMSE': RMSE
}
metrics.writeResultsToCSV(resDict, metricsResFileName, clearMetricsFile)
co7.train(training_set7)
print("set 7 trained")
co8.train(training_set8)
print("set 8 trained")
co9.train(training_set9)
print("set 9 trained")
print("Algorithms trained!")
'''
Create 9 output files to hold disaggregated data.
'''
print("Creating output files...")
outData1 = HDFDataStore("C:/NILM/Data/Model_Train/output1.h5", 'w')
outData2 = HDFDataStore("C:/NILM/Data/Model_Train/output2.h5", 'w')
outData3 = HDFDataStore("C:/NILM/Data/Model_Train/output3.h5", 'w')
outData4 = HDFDataStore("C:/NILM/Data/Model_Train/output4.h5", 'w')
outData5 = HDFDataStore("C:/NILM/Data/Model_Train/output5.h5", 'w')
outData6 = HDFDataStore("C:/NILM/Data/Model_Train/output6.h5", 'w')
outData7 = HDFDataStore("C:/NILM/Data/Model_Train/output7.h5", 'w')
outData8 = HDFDataStore("C:/NILM/Data/Model_Train/output8.h5", 'w')
outData9 = HDFDataStore("C:/NILM/Data/Model_Train/output9.h5", 'w')
print("output files created!")
'''
Disaggregate building 1 data using each training set
'''
print("Disaggregating building 1 mains using each trained model...")
def plot_zoomed_new_predicted_energy_consumption():
"""
Predicts a new short window (of the given test set).
"""
train = DataSet('../data/ukdale.h5')
train.clear_cache()
train.set_window(start="13-4-2013", end="31-7-2013")
test = DataSet('../data/ukdale.h5')
test.clear_cache()
test.set_window(start='16-9-2013 17:00:00', end='16-9-2013 18:00:00')
train_building = 1
test_building = 1
sample_period = 6
meter_key = 'kettle'
learning_rate = 1e-5
best_epoch = 140
train_elec = train.buildings[train_building].elec
test_elec = test.buildings[test_building].elec
train_meter = train_elec.submeters()[meter_key]
test_mains = test_elec.mains()
results_dir = '../results/UKDALE-RNN-lr=1e-05-2018-02-16-18-52-34'
train_logfile = os.path.join(results_dir, 'training.log')
val_logfile = os.path.join(results_dir, 'validation.log')
rnn = RNNDisaggregator(train_logfile,
val_logfile,
learning_rate,
init=False)
model = 'UKDALE-RNN-kettle-{}epochs.h5'.format(best_epoch)
rnn.import_model(os.path.join(results_dir, model))
disag_filename = 'disag-out-{}epochs.h5'.format(best_epoch)
output = HDFDataStore(os.path.join(results_dir, disag_filename), 'w')
results_file = os.path.join(results_dir,
'results-{}epochs.txt'.format(best_epoch))
rnn.disaggregate(test_mains,
output,
results_file,
train_meter,
sample_period=sample_period)
os.remove(results_file)
output.close()
# get predicted curve for the best epoch
result = DataSet(os.path.join(results_dir, disag_filename))
res_elec = result.buildings[test_building].elec
os.remove(os.path.join(results_dir, disag_filename))
predicted = res_elec[meter_key]
predicted = predicted.power_series(sample_period=sample_period)
predicted = next(predicted)
predicted.fillna(0, inplace=True)
y1 = np.array(predicted) # power
x1 = np.arange(y1.shape[0]) # timestamps
ground_truth = test_elec[meter_key]
ground_truth = ground_truth.power_series(sample_period=sample_period)
ground_truth = next(ground_truth)
ground_truth.fillna(0, inplace=True)
y2 = np.array(ground_truth) # power
x2 = np.arange(y2.shape[0]) # timestamps
fig, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, sharey=True)
ax1.plot(x1, y1, color='r', label='predicted')
ax1.plot(x2, y2, color='b', label='ground truth')
ax2.plot(x1, y1, color='r')
ax3.plot(x2, y2, color='b')
ax1.set_title('Appliance: {}'.format(meter_key))
fig.legend()
fig.savefig(
os.path.join(results_dir, 'zoomed_new_predicted_vs_ground_truth.png'))
class REDD_Data(object):
'''
REDD_Data Class is an object designed to abstract the lower level commands of
the NILMTK software package, with focus on the use of REDD DataSet. Function is
designed to allow rapid experimentation and disaggregation compared to attempting
to set package up from scratch.
This class requires the following for proper usage:
- NILMTK package: https://github.com/nilmtk
- REDD Dataset (converted to .h5): redd.csail.mit.edu
- Various dependancies (that NILMTK also requires), most can be downloaded through
Anaconda: continuum.io/downloads
Parameters
-----------
in_filepath: Filepath of converted REDD dataset (in .h5 format)
out_filepath: filepath to place output disaggregation dataset (in .h5 format)
Attributes
-----------
km: Key_Map Object
initializes the key_map object which will allow for the mapping of a meters
appliance name to its specific .H5 key.
dataStore: NILMTK HDFDataStore Object
the HDFDataStore that will contain the converted REDD DataSet.
dataSet: NILMTK DataSet Object
the DataSet object that is generated from the REDD DataStore (self.dataStore)
outDataStore: NILMTK HDFDataStore Object
the HDFDataStore that will contain the disaggregated dataset.
co: NILMTK CombinatorialOptimisation object
the disaggregation model object that will be trained and will disaggregate the
working dataset
train_group: NILMTK MeterGroup object
the MeterGroup object that is used to train the disaggregation model (self.co)
'''
def __init__(self, in_filepath, out_filepath):
print("Loading DataStore and Generating Dataset...")
self.km = {}
self.dataStore = HDFDataStore(in_filepath)
self.dataSet = DataSet()
self.dataSet.load(self.dataStore)
self.outDataStore = HDFDataStore(out_filepath, 'w')
self.co = CombinatorialOptimisation()
self.train_group = {}
print("Data Properly Loaded!")
def train_disag_model(self, building_inst, use_topk=False, k=5):
'''
Function trains the disaggregation model using a selected MeterGroup.
Parameters
-----------
building_inst: the instance # of the building that you wish to grab the
training group from.
use_topk: true if you wish to only grab the top k most energy intensive
appliance to train the model, false if you wish to use all
appliances.
k: the # of appliances you wish to use (if use_topk = True)
'''
print("Training CO Disaggregation Model using given metergroup...")
if (building_inst <= 6) & (building_inst > 0):
#Select appropiate meter group to train with
if use_topk == True:
self.train_group = self.dataSet.buildings[
building_inst].elec.select_top_k(k)
else:
self.train_group = self.dataSet.buildings[building_inst].elec
self.co.train(self.train_group)
print("CO Disaggreation Model Sucessfully Trained!")
else:
print("Error: Please select a building_inst of 1-6.")
print("Model unsucessfully trained.")
def load_disag_model(self, filepath):
'''
Function loads the disaggregation model to a file.
Parameters
-----------
filepath: exact filepath of the model file.
'''
print("Loading CO Disaggreation Model...")
self.co.import_model(filepath)
print("Model Sucessfully Loaded!")
def save_disag_model(self, filepath):
'''
Function saves the disaggregation model to a file.
Parameters
-----------
filepath: exact filepath of the model file.
'''
print("Saving CO Disaggregation Model...")
self.co.export_model(filepath)
print("Model Sucessfully Saved!")
def disaggregate(self, building_inst):
'''
Function will disaggregate the mains MeterGroup of the passed building
instance, and save this to the self.outDataStore object.
Parameters
-----------
building_inst: instance # of the building mains you wish to disaggregate.
'''
print("Disaggregating Building Mains...")
self.co.disaggregate(
self.dataSet.buildings[building_inst].elec.mains(),
self.outDataStore)
print("Mains sucessfully disaggregated!")
def close(self):
'''
Function closes all open DataStore's being used by the program.
'''
print("Closing DataStores...")
self.dataStore.close()
self.outDataStore.close()
print("Output DataStores Sucessfully Closed")
'''
All Plot Functions below are a WORK IN PROGRESS!-----------------------------------
Documentation will be provided upon completion.------------------------------------
'''
def plot_disag_apl(self, inst, appliance, t1="", t2=""):
self.km = Key_Map(inst)
plot_series(
self.outDataStore.store.get(self.km.get_key(appliance))[t1:t2])
plt.title("Disaggregated " + appliance.capitalize() + " Energy")
plt.show()
def show_plots(self):
plt.show()
def building_plot_all(self, building_inst, t1, t2):
self.dataSet.buildings[building_inst].elec.plot(t1, t2)
plt.title("Building " + str(building_inst) + " Energy per Appliance")
plt.ylabel('Power [W]')
plt.xlabel('Hour')
def plot_redd_mains_data(self, inst=1, t1="", t2=""):
self.km = Key_Map(inst)
series1 = self.dataStore.store.get(self.km.get_key('mains1'))[t1:t2]
series2 = self.dataStore.store.get(self.km.get_key('mains2'))[t1:t2]
plot_series(series1 + series2)
plt.title("Building " + str(inst) + " Mains Energy")
plt.show()
#same day but smaller scale for more refined data plots
t1 ="2011-05-1 6:00"
t2 ="2011-05-1 7:00"
#initialize key map for building 1
kmap = Key_map()
#set the disaggregated plot here
disag_apl = 'fridge'
disag_key = kmap.get_key(disag_apl)
#First we must load in the converted REDD Dataset
print ("Loading DataSet.....")
#declare datastore and load converted HDF that stores REDD data
r_datastore = HDFDataStore("C:/NILM/Data_Sets/redd_data.h5")
#declare dataset object to work with and load datastore into Dataset
r_dataset = DataSet()
r_dataset.load(r_datastore)
#output sucessfull loading of data to user
print("DataSet Sucessfully Loaded!")
#now we take the data and elminate all sections with no samples
print("Conditioning Data... \n")
#load the metergroup from building one (house1 in REDD)
r_elec = r_dataset.buildings[1].elec
https://github.com/nilmtk/nilmtk/issues/376
"""
data_dir = '/data/REDD'
building_number = 3
disag_filename = join(data_dir, 'disag-fhmm' + str(building_number) + '.h5')
data = DataSet(join(data_dir, 'redd.h5'))
print("Loading building " + str(building_number))
elec = data.buildings[building_number].elec
top_train_elec = elec.submeters().select_top_k(k=5)
fhmm = fhmm_exact.FHMM()
fhmm.train(top_train_elec)
output = HDFDataStore(disag_filename, 'w')
fhmm.disaggregate(elec.mains(), output)
output.close()
### f1score fhmm
disag = DataSet(disag_filename)
disag_elec = disag.buildings[building_number].elec
f1 = f1_score(disag_elec, elec)
f1.index = disag_elec.get_labels(f1.index)
f1.plot(kind='barh')
plt.ylabel('appliance');
plt.xlabel('f-score');
plt.title("FHMM");
plt.savefig(join(data_dir, 'f1-fhmm' + str(building_number) + '.png'))
disag.store.close()
if km.is_in_map(disag_appliance) == False:
sys.exit(
"An incorrect appliance name has been entered. Please ensure the entered name is exactly correct."
)
redd_data = DataSet("C:/NILM/Data/REDD/redd.h5")
# load mains of the building
building_mains = redd_data.buildings[redd_building].elec.mains()
#train disaggregation set
co = CombinatorialOptimisation()
training_set = redd_data.buildings[redd_building].elec
co.train(training_set)
#set output datastore
outputData = HDFDataStore("C:/NILM/Data/Output/output.h5", 'w')
#disaggregate
co.disaggregate(building_mains, outputData)
# to add:
# 1) get the meter instance # of the appliance selected
# 2) export the meter instance series of the output datastore to database using SQL, within t1-t2 parameters*
#
# *Cannot be implemented until database is setup in environment
#Close open datastores
redd_data.store.close()
outputData.store.close()
print(middleTimeStr) train.set_window(end=middleTimeStr) test.set_window(start=middleTimeStr) train_elec = train.buildings[building_number].elec test_elec = test.buildings[building_number].elec top_train_elec = train_elec.submeters().select_top_k(k=5) fhmm = fhmm_exact.FHMM() #mk change this later to default fhmm.train(top_train_elec, sample_period=60, resample=True) outputAddress = "/nilmtk/data/iawe_449_3.h5" output = HDFDataStore(outputAddress, 'w') fhmm.disaggregate(test_elec.mains(), output, sample_period=60, resample=True) output.close() disag = DataSet(outputAddress) #load FHMM prediction disag_elec = disag.buildings[building_number].elec #disag_elec.plot() # plot all disaggregated data f1 = f1_score(disag_elec, test_elec) f1.index = disag_elec.get_labels(f1.index) f1.plot(kind='barh') disag.store.window = TimeFrame(start='2013-07-10 18:00:00-05:00', end='2013-07-17 04:00:00-05:00') disag.buildings[building_number].elec.plot() # plot all disaggregated data
# verify a real appliance has been entered
#if km.is_in_map(disag_appliance) == False:
# sys.exit("An incorrect appliance name has been entered. Please ensure the entered name is exactly correct.")
redd_data = DataSet(redd_fp)
# load mains of the building
building_mains = redd_data.buildings[1].elec.mains()
#train disaggregation set
co = CombinatorialOptimisation()
training_set = redd_data.buildings[1].elec.select_top_k(15)
co.train(training_set)
#set output datastore
outputData = HDFDataStore(output_fp, 'w')
#disaggregate
co.disaggregate(building_mains, outputData)
print(outputData.store.keys())
#set sub-datastore for CSV outputs
output_csv_store = outputData.store.__getitem__(km.get_key(disag_appliance))
mains1 = redd_data.store.__getitem__(km.get_key("mains1"))
mains2 = redd_data.store.__getitem__(km.get_key("mains2"))
mains1 = mains1.fillna(value=0)
mains1 = mains1.resample("1min")
sys.exit(
"An incorrect appliance name has been entered. Please ensure the entered name is exactly correct."
)
redd_data = DataSet("/home/mike/workspace/data/redd_data.h5")
# load mains of the building
building_mains = redd_data.buildings[1].elec.mains()
#train disaggregation set
co = CombinatorialOptimisation()
training_set = redd_data.buildings[1].elec
co.train(training_set)
#set output datastore
outputData = HDFDataStore("/home/mike/workspace/data/redd_output.h5", 'w')
#disaggregate
co.disaggregate(building_mains, outputData)
#set sub-datastore for CSV output
output_csv_store = outputData.store.__getitem__(km.get_key(disag_appliance))
#set date parameters
output_csv_store = output_csv_store[t1:t2]
#fill NA values with 0 for graphing
output_csv_store = output_csv_store.fillna(value=0)
# #metrics processing ----------------------------------------------------------
# #create dict to hold energy metrics
def plot_prediction_over_epochs_ploty():
"""
Predicts the power demand of the target appliance using the intermediate models which are exported during training.
Plots the prediction curves using plotly.
"""
train = DataSet('../data/ukdale.h5')
train.clear_cache()
train.set_window(start="13-4-2013", end="31-7-2013")
test = DataSet('../data/ukdale.h5')
test.clear_cache()
test.set_window(start="23-7-2014 10:00:00", end="23-7-2014 11:00:00")
train_building = 1
test_building = 5
sample_period = 6
meter_key = 'kettle'
learning_rate = 1e-5
train_elec = train.buildings[train_building].elec
test_elec = test.buildings[test_building].elec
train_meter = train_elec.submeters()[meter_key]
test_mains = test_elec.mains()
results_dir = '../results/UKDALE-ACROSS-BUILDINGS-RNN-lr=1e-05-2018-02-03-11-48-12'
train_logfile = os.path.join(results_dir, 'training.log')
val_logfile = os.path.join(results_dir, 'validation.log')
rnn = RNNDisaggregator(train_logfile,
val_logfile,
learning_rate,
init=False)
data = []
for i in range(10, 401, 10):
# disaggregate model
model = 'UKDALE-RNN-kettle-{}epochs.h5'.format(i)
rnn.import_model(os.path.join(results_dir, model))
disag_filename = 'disag-out-{}epochs.h5'.format(i)
output = HDFDataStore(os.path.join(results_dir, disag_filename), 'w')
results_file = os.path.join(results_dir,
'results-{}epochs.txt'.format(i))
rnn.disaggregate(test_mains,
output,
results_file,
train_meter,
sample_period=sample_period)
os.remove(results_file)
output.close()
# plot predicted curve for epoch=i
result = DataSet(os.path.join(results_dir, disag_filename))
res_elec = result.buildings[test_building].elec
os.remove(os.path.join(results_dir, disag_filename))
predicted = res_elec[meter_key]
predicted = predicted.power_series(sample_period=sample_period)
predicted = next(predicted)
predicted.fillna(0, inplace=True)
power = predicted.tolist()
length = len(power)
timestamps = list(range(length))
x = []
y = []
z = []
ci = int(255 / 420 * i) # ci = "color index"
for j in range(length):
x.append([timestamps[j], timestamps[j]]) # timestamps
y.append([i, i + 5]) # epochs
z.append([power[j], power[j]]) # power
data.append(
dict(
z=z,
x=x,
y=y,
colorscale=[[i, 'rgb(%d,%d,255)' % (ci, ci)]
for i in np.arange(0, 1.1, 0.1)],
showscale=False,
type='surface',
))
# plot ground truth curve as the last curve
ground_truth = test_elec[meter_key]
ground_truth = ground_truth.power_series(sample_period=sample_period)
ground_truth = next(ground_truth)
ground_truth.fillna(0, inplace=True)
power = ground_truth.tolist()
length = len(power)
timestamps = list(range(length))
i = 410
x = []
y = []
z = []
ci = int(255 / 410 * i) # ci = "color index"
for j in range(length):
x.append([timestamps[j], timestamps[j]]) # timestamps
y.append([i, i + 5]) # epochs
z.append([power[j], power[j]]) # power
data.append(
dict(
z=z,
x=x,
y=y,
colorscale=[[i, 'rgb(%d,%d,255)' % (ci, ci)]
for i in np.arange(0, 1.1, 0.1)],
showscale=False,
type='surface',
))
layout = dict(title='prediction over epochs',
showlegend=False,
scene=dict(xaxis=dict(title='timestamps'),
yaxis=dict(title='epochs'),
zaxis=dict(title='power'),
camera=dict(eye=dict(x=-1.7, y=-1.7, z=0.5))))
fig = dict(data=data, layout=layout)
plotly.offline.plot(fig, filename='filled-3d-lines')
def fhmm(dataset_path, train_building, train_start, train_end, val_building,
val_start, val_end, test_building, test_start, test_end, meter_key,
sample_period):
# Start tracking time
start = time.time()
# Prepare dataset and options
# print("========== OPEN DATASETS ============")
dataset_path = dataset_path
train = DataSet(dataset_path)
train.set_window(start=train_start, end=train_end)
val = DataSet(dataset_path)
val.set_window(start=val_start, end=val_end)
test = DataSet(dataset_path)
test.set_window(start=test_start, end=test_end)
train_building = train_building
test_building = test_building
meter_key = meter_key
sample_period = sample_period
train_elec = train.buildings[train_building].elec
val_elec = val.buildings[val_building].elec
test_elec = test.buildings[test_building].elec
appliances = [meter_key]
selected_meters = [train_elec[app] for app in appliances]
selected_meters.append(train_elec.mains())
selected = MeterGroup(selected_meters)
fhmm = FHMM()
# print("========== TRAIN ============")
fhmm.train(selected, sample_period=sample_period)
# print("========== DISAGGREGATE ============")
# Validation
val_disag_filename = 'disag-out-val.h5'
output = HDFDataStore(val_disag_filename, 'w')
fhmm.disaggregate(val_elec.mains(), output_datastore=output)
output.close()
# Test
test_disag_filename = 'disag-out-test.h5'
output = HDFDataStore(test_disag_filename, 'w')
fhmm.disaggregate(test_elec.mains(), output_datastore=output)
output.close()
# print("========== RESULTS ============")
# Validation
result_val = DataSet(val_disag_filename)
res_elec_val = result_val.buildings[val_building].elec
rpaf_val = metrics.recall_precision_accuracy_f1(res_elec_val[meter_key],
val_elec[meter_key])
val_metrics_results_dict = {
'recall_score':
rpaf_val[0],
'precision_score':
rpaf_val[1],
'accuracy_score':
rpaf_val[2],
'f1_score':
rpaf_val[3],
'mean_absolute_error':
metrics.mean_absolute_error(res_elec_val[meter_key],
val_elec[meter_key]),
'mean_squared_error':
metrics.mean_square_error(res_elec_val[meter_key],
val_elec[meter_key]),
'relative_error_in_total_energy':
metrics.relative_error_total_energy(res_elec_val[meter_key],
val_elec[meter_key]),
'nad':
metrics.nad(res_elec_val[meter_key], val_elec[meter_key]),
'disaggregation_accuracy':
metrics.disaggregation_accuracy(res_elec_val[meter_key],
val_elec[meter_key])
}
# Test
result = DataSet(test_disag_filename)
res_elec = result.buildings[test_building].elec
rpaf = metrics.recall_precision_accuracy_f1(res_elec[meter_key],
test_elec[meter_key])
test_metrics_results_dict = {
'recall_score':
rpaf[0],
'precision_score':
rpaf[1],
'accuracy_score':
rpaf[2],
'f1_score':
rpaf[3],
'mean_absolute_error':
metrics.mean_absolute_error(res_elec[meter_key], test_elec[meter_key]),
'mean_squared_error':
metrics.mean_square_error(res_elec[meter_key], test_elec[meter_key]),
'relative_error_in_total_energy':
metrics.relative_error_total_energy(res_elec[meter_key],
test_elec[meter_key]),
'nad':
metrics.nad(res_elec[meter_key], test_elec[meter_key]),
'disaggregation_accuracy':
metrics.disaggregation_accuracy(res_elec[meter_key],
test_elec[meter_key])
}
# end tracking time
end = time.time()
time_taken = end - start # in seconds
# model_result_data = {
# 'algorithm_name': 'FHMM',
# 'datapath': dataset_path,
# 'train_building': train_building,
# 'train_start': str(train_start.date()) if train_start != None else None ,
# 'train_end': str(train_end.date()) if train_end != None else None ,
# 'test_building': test_building,
# 'test_start': str(test_start.date()) if test_start != None else None ,
# 'test_end': str(test_end.date()) if test_end != None else None ,
# 'appliance': meter_key,
# 'sampling_rate': sample_period,
#
# 'algorithm_info': {
# 'options': {
# 'epochs': None
# },
# 'hyperparameters': {
# 'sequence_length': None,
# 'min_sample_split': None,
# 'num_layers': None
# },
# 'profile': {
# 'parameters': None
# }
# },
#
# 'metrics': metrics_results_dict,
#
# 'time_taken': format(time_taken, '.2f'),
# }
model_result_data = {
'val_metrics': val_metrics_results_dict,
'test_metrics': test_metrics_results_dict,
'time_taken': format(time_taken, '.2f'),
'epochs': None,
}
# Close digag_filename
result.store.close()
result_val.store.close()
# Close Dataset files
train.store.close()
val.store.close()
test.store.close()
return model_result_data
validation = pd.read_csv(val_logfile)
epochs = np.array(validation.as_matrix()[:, 0], dtype='int')
loss = np.array(validation.as_matrix()[:, 1], dtype='float32')
argmin = np.argmin(loss)
best_epoch = epochs[argmin] + 1
rnn.import_model(
os.path.join(results_dir,
"UKDALE-RNN-{}-{}epochs.h5".format(meter_key, best_epoch)))
test_loss = rnn.evaluate(test_mains, test_meter, sample_period=sample_period)
line = 'Test loss: {}'.format(test_loss)
with open(results_file, "a") as text_file:
text_file.write(line + '\n')
print(line)
disag_filename = 'disag-out.h5'
output = HDFDataStore(os.path.join(results_dir, disag_filename), 'w')
rnn.disaggregate(test_mains,
output,
results_file,
train_meterlist[0],
sample_period=sample_period)
output.close()
print("========== PLOTS ============")
# plot train, validation and test loss
plot_loss(train_logfile, val_logfile, results_dir, best_epoch, test_loss)
# plot predicted energy consumption
result = DataSet(os.path.join(results_dir, disag_filename))
res_elec = result.buildings[test_building].elec
predicted = res_elec[meter_key]
