[[9, 1, 4], [10, 1, 3], [0, 1, 2], [3, 1, 1], [9, 1, 1]], # Mendon
[[6, 1, 2], [10, 1, 0], [8, 1, 4], [10, 1, 0], [10, 1, 5]], # TonyGrove
[[7, 1, 0], [1, 1, 1], [10, 1, 0], [0, 1, 5], [1, 1, 3]] # WaterLab
]
pdqParam = pd.DataFrame(pdqParams, columns=sensors, index=sites)
print(pdqParam)
#########################################
# TEMPERATURE #
#########################################
# RULES BASED DETECTION #
#########################################
maximum = 18
minimum = -2
temp_df = rules_detect.range_check(temp_df, maximum, minimum)
length = 6
temp_df = rules_detect.persistence(temp_df, length)
size = rules_detect.group_size(temp_df)
temp_df = rules_detect.interpolate(temp_df)
# MODEL CREATION #
#########################################
p, d, q = pdqParam[sensor[0]][site]
print("p: " + str(p))
print("d: " + str(d))
print("q: " + str(q))
model_fit, residuals, predictions = modeling_utilities.build_arima_model(
temp_df['observed'], p, d, q, summary=True)
# DETERMINE THRESHOLD AND DETECT ANOMALIES #
def ARIMA_detect(df,
sensor,
params,
rules=False,
plots=True,
summary=True,
output=True):
"""
"""
print('\nProcessing ARIMA detections.')
# RULES BASED DETECTION #
if rules:
df = rules_detect.range_check(df, params['max_range'],
params['min_range'])
df = rules_detect.persistence(df, params['persist'])
size = rules_detect.group_size(df)
df = rules_detect.interpolate(df)
print(sensor +
' rules based detection complete. Longest detected group = ' +
str(size))
# MODEL CREATION #
[p, d, q] = params['pdq']
model_fit, residuals, predictions = modeling_utilities.build_arima_model(
df['observed'], p, d, q, summary)
print(sensor + ' ARIMA model complete.')
# DETERMINE THRESHOLD AND DETECT ANOMALIES #
threshold = anomaly_utilities.set_dynamic_threshold(
residuals[0], params['window_sz'], params['alpha'],
params['threshold_min'])
threshold.index = residuals.index
if plots:
plt.figure()
anomaly_utilities.plt_threshold(residuals, threshold, sensor)
plt.show()
print('Threshold determination complete.')
detections = anomaly_utilities.detect_anomalies(df['observed'],
predictions,
residuals,
threshold,
summary=True)
# WIDEN AND NUMBER ANOMALOUS EVENTS #
df['labeled_event'] = anomaly_utilities.anomaly_events(
df['labeled_anomaly'], params['widen'])
df['detected_anomaly'] = detections['anomaly']
df['all_anomalies'] = df.eval('detected_anomaly or anomaly')
df['detected_event'] = anomaly_utilities.anomaly_events(
df['all_anomalies'], params['widen'])
# DETERMINE METRICS #
anomaly_utilities.compare_events(df, params['widen'])
metrics = anomaly_utilities.metrics(df)
e_metrics = anomaly_utilities.event_metrics(df)
# OUTPUT RESULTS #
if output:
print('Model type: ARIMA')
print('Sensor: ' + sensor)
anomaly_utilities.print_metrics(metrics)
print('Event based calculations:')
anomaly_utilities.print_metrics(e_metrics)
print('Model report complete\n')
# GENERATE PLOTS #
if plots:
plt.figure()
anomaly_utilities.plt_results(raw=df['raw'],
predictions=detections['prediction'],
labels=df['labeled_event'],
detections=df['detected_event'],
sensor=sensor)
plt.show()
ARIMA_detect = ModelWorkflow()
ARIMA_detect.df = df
ARIMA_detect.model_fit = model_fit
ARIMA_detect.threshold = threshold
ARIMA_detect.detections = detections
ARIMA_detect.metrics = metrics
ARIMA_detect.e_metrics = e_metrics
return ARIMA_detect
def LSTM_detect_multivar(sensor_array,
sensors,
params,
LSTM_params,
model_type,
name,
rules=False,
plots=True,
summary=True,
output=True,
model_output=True,
model_save=True):
"""
"""
print('\nProcessing LSTM multivariate ' + str(model_type) + ' detections.')
# RULES BASED DETECTION #
if rules:
size = dict()
for snsr in sensors:
sensor_array[snsr], r_c = rules_detect.range_check(
sensor_array[snsr], params[snsr].max_range,
params[snsr].min_range)
sensor_array[snsr], p_c = rules_detect.persistence(
sensor_array[snsr], params[snsr].persist)
size[snsr] = rules_detect.group_size(sensor_array[snsr])
sensor_array[snsr] = rules_detect.interpolate(sensor_array[snsr])
print(snsr + ' maximum detected group length = ' + str(size[snsr]))
print('Rules based detection complete.\n')
# Create new data frames with raw and observed (after applying rules) and preliminary anomaly detections for selected sensors
df_raw = pd.DataFrame(index=sensor_array[sensors[0]].index)
df_observed = pd.DataFrame(index=sensor_array[sensors[0]].index)
df_anomaly = pd.DataFrame(index=sensor_array[sensors[0]].index)
for snsr in sensors:
df_raw[snsr + '_raw'] = sensor_array[snsr]['raw']
df_observed[snsr + '_obs'] = sensor_array[snsr]['observed']
df_anomaly[snsr + '_anom'] = sensor_array[snsr]['anomaly']
print('Raw data shape: ' + str(df_raw.shape))
print('Observed data shape: ' + str(df_observed.shape))
print('Initial anomalies data shape: ' + str(df_anomaly.shape))
# MODEL CREATION #
if model_type == 'vanilla':
model = modeling_utilities.LSTM_multivar(df_observed, df_anomaly,
df_raw, LSTM_params, summary,
name, model_output,
model_save)
else:
model = modeling_utilities.LSTM_multivar_bidir(df_observed, df_anomaly,
df_raw, LSTM_params,
summary, name,
model_output,
model_save)
print('multivariate ' + str(model_type) + ' LSTM model complete.\n')
# Plot Metrics and Evaluate the Model
if plots:
plt.figure()
plt.plot(model.history.history['loss'], label='Training Loss')
plt.plot(model.history.history['val_loss'], label='Validation Loss')
plt.legend()
plt.show()
# DETERMINE THRESHOLD AND DETECT ANOMALIES #
ts = LSTM_params['time_steps']
residuals = pd.DataFrame(model.test_residuals)
residuals.columns = sensors
predictions = pd.DataFrame(model.predictions)
predictions.columns = sensors
if model_type == 'vanilla':
residuals.index = df_observed[ts:].index
predictions.index = df_observed[ts:].index
observed = df_observed[ts:]
else:
residuals.index = df_observed[ts:-ts].index
predictions.index = df_observed[ts:-ts].index
observed = df_observed[ts:-ts]
threshold = dict()
detections = dict()
for snsr in sensors:
threshold[snsr] = anomaly_utilities.set_dynamic_threshold(
residuals[snsr], params[snsr]['window_sz'], params[snsr]['alpha'],
params[snsr]['threshold_min'])
threshold[snsr].index = residuals.index
detections[snsr] = anomaly_utilities.detect_anomalies(
observed[snsr + '_obs'],
predictions[snsr],
residuals[snsr],
threshold[snsr],
summary=True)
if plots:
plt.figure()
anomaly_utilities.plt_threshold(residuals[snsr], threshold[snsr],
sensors[snsr])
plt.show()
print('Threshold determination complete.')
# WIDEN AND NUMBER ANOMALOUS EVENTS #
all_data = dict()
for snsr in sensors:
if model_type == 'vanilla':
all_data[snsr] = sensor_array[snsr].iloc[ts:]
else:
all_data[snsr] = sensor_array[snsr].iloc[ts:-ts]
all_data[snsr]['labeled_event'] = anomaly_utilities.anomaly_events(
all_data[snsr]['labeled_anomaly'], params[snsr]['widen'])
all_data[snsr]['detected_anomaly'] = detections[snsr]['anomaly']
all_data[snsr]['all_anomalies'] = all_data[snsr].eval(
'detected_anomaly or anomaly')
all_data[snsr]['detected_event'] = anomaly_utilities.anomaly_events(
all_data[snsr]['all_anomalies'], params[snsr]['widen'])
# DETERMINE METRICS #
metrics = dict()
e_metrics = dict()
for snsr in sensors:
anomaly_utilities.compare_events(all_data[snsr], params[snsr]['widen'])
metrics[snsr] = anomaly_utilities.metrics(all_data[snsr])
e_metrics[snsr] = anomaly_utilities.event_metrics(all_data[snsr])
# OUTPUT RESULTS #
if output:
for snsr in sensors:
print('\nModel type: LSTM multivariate ' + str(model_type))
print('Sensor: ' + snsr)
anomaly_utilities.print_metrics(metrics[snsr])
print('Event based calculations:')
anomaly_utilities.print_metrics(e_metrics[snsr])
print('Model report complete\n')
# GENERATE PLOTS #
if plots:
for snsr in sensors:
plt.figure()
anomaly_utilities.plt_results(
raw=sensor_array[snsr]['raw'],
predictions=detections[snsr]['prediction'],
labels=sensor_array[snsr]['labeled_event'],
detections=all_data[snsr]['detected_event'],
sensor=snsr)
plt.show()
LSTM_detect_multivar = ModelWorkflow()
LSTM_detect_multivar.sensor_array = sensor_array
LSTM_detect_multivar.df_observed = df_observed
LSTM_detect_multivar.df_raw = df_raw
LSTM_detect_multivar.df_anomaly = df_anomaly
LSTM_detect_multivar.model = model
LSTM_detect_multivar.threshold = threshold
LSTM_detect_multivar.detections = detections
LSTM_detect_multivar.all_data = all_data
LSTM_detect_multivar.metrics = metrics
LSTM_detect_multivar.e_metrics = e_metrics
return LSTM_detect_multivar
def LSTM_detect_univar(df,
sensor,
params,
LSTM_params,
model_type,
name,
rules=False,
plots=True,
summary=True,
output=True,
model_output=True,
model_save=True):
"""
"""
print('\nProcessing LSTM univariate ' + str(model_type) + ' detections.')
# RULES BASED DETECTION #
if rules:
df = rules_detect.range_check(df, params['max_range'],
params['min_range'])
df = rules_detect.persistence(df, params['persist'])
size = rules_detect.group_size(df)
df = rules_detect.interpolate(df)
print(
sensor +
' rules based detection complete. Maximum detected group length = '
+ str(size))
# MODEL CREATION #
if model_type == 'vanilla':
model = modeling_utilities.LSTM_univar(df, LSTM_params, summary, name,
model_output, model_save)
else:
model = modeling_utilities.LSTM_univar_bidir(df, LSTM_params, summary,
name, model_output,
model_save)
print(sensor + ' ' + str(model_type) + ' LSTM model complete.')
if plots:
plt.figure()
plt.plot(model.history.history['loss'], label='Training Loss')
plt.plot(model.history.history['val_loss'], label='Validation Loss')
plt.legend()
plt.show()
# DETERMINE THRESHOLD AND DETECT ANOMALIES #
ts = LSTM_params['time_steps']
threshold = anomaly_utilities.set_dynamic_threshold(
model.test_residuals[0], params['window_sz'], params['alpha'],
params['threshold_min'])
if model_type == 'vanilla':
threshold.index = df[ts:].index
else:
threshold.index = df[ts:-ts].index
residuals = pd.DataFrame(model.test_residuals)
residuals.index = threshold.index
if plots:
plt.figure()
anomaly_utilities.plt_threshold(residuals, threshold, sensor)
plt.show()
if model_type == 'vanilla':
observed = df[['observed']][ts:]
else:
observed = df[['observed']][ts:-ts]
print('Threshold determination complete.')
detections = anomaly_utilities.detect_anomalies(observed,
model.predictions,
model.test_residuals,
threshold,
summary=True)
# WIDEN AND NUMBER ANOMALOUS EVENTS #
if model_type == 'vanilla':
df_anomalies = df.iloc[ts:]
else:
df_anomalies = df.iloc[ts:-ts]
df_anomalies['labeled_event'] = anomaly_utilities.anomaly_events(
df_anomalies['labeled_anomaly'], params['widen'])
df_anomalies['detected_anomaly'] = detections['anomaly']
df_anomalies['all_anomalies'] = df_anomalies.eval(
'detected_anomaly or anomaly')
df_anomalies['detected_event'] = anomaly_utilities.anomaly_events(
df_anomalies['all_anomalies'], params['widen'])
# DETERMINE METRICS #
anomaly_utilities.compare_events(df_anomalies, params['widen'])
metrics = anomaly_utilities.metrics(df_anomalies)
e_metrics = anomaly_utilities.event_metrics(df_anomalies)
# OUTPUT RESULTS #
if output:
print('Model type: LSTM univariate ' + str(model_type))
print('Sensor: ' + sensor)
anomaly_utilities.print_metrics(metrics)
print('Event based calculations:')
anomaly_utilities.print_metrics(e_metrics)
print('Model report complete\n')
# GENERATE PLOTS #
if plots:
plt.figure()
anomaly_utilities.plt_results(
raw=df['raw'],
predictions=detections['prediction'],
labels=df['labeled_event'],
detections=df_anomalies['detected_event'],
sensor=sensor)
plt.show()
LSTM_detect_univar = ModelWorkflow()
LSTM_detect_univar.df = df
LSTM_detect_univar.model = model
LSTM_detect_univar.threshold = threshold
LSTM_detect_univar.detections = detections
LSTM_detect_univar.df_anomalies = df_anomalies
LSTM_detect_univar.metrics = metrics
LSTM_detect_univar.e_metrics = e_metrics
return LSTM_detect_univar
year,
path="LRO_data/")
# RULES DETECTION PARAMETERS #
#########################################
maximum = [20, 2700, 9.5, 15]
minimum = [-2, 150, 7.5, 5]
length = [10, 10, 10, 20]
# RULES BASED ANOMALY DETECTION #
#########################################
# size = []
range_count = []
persist_count = []
for i in range(0, len(sensor_array)):
sensor_array[sensor[i]], r_c = rules_detect.range_check(
sensor_array[sensor[i]], maximum[i], minimum[i])
range_count.append(r_c)
sensor_array[sensor[i]], p_c = rules_detect.persistence(
sensor_array[sensor[i]], length[i])
persist_count.append(p_c)
# s = rules_detect.group_size(sensor_array[sensor[i]])
# size.append(s)
sensor_array[sensor[i]] = rules_detect.add_labels(sensor_array[sensor[i]],
-9999)
sensor_array[sensor[i]] = rules_detect.interpolate(sensor_array[sensor[i]])
# print(str(sensor[i]) + ' maximum detected group length = ' + str(size[i]))
print('Rules based detection complete.\n')
# ANOMALY DETECTION PARAMETERS #
#########################################
window_sz = [30, 40, 20, 30]
# - Use a subset of raw data that is in decent shape without NaNs, -9999 values, or data gaps. # df_sub = df.loc['2017-01-01 00:00':'2017-07-01 00:00'] # - Use raw data that have been preprocessed to filter out extreme values and have drift correction applied. # - Either with raw or corrected data for training, use data that are not labeled as anomalous/corrected. # df_cor = df_cor.replace(-9999, np.NaN) # RULES BASED DETECTION # ######################################### # General sensor ranges for LRO data: # Temp min: -5, max: 30 # SpCond min: 100, max: 900 # pH min: 7.5, max: 9.0 # do min: 2, max: 16 maximum = 900 minimum = 150 df = rules_detect.range_check(df, maximum, minimum) length = 6 df = rules_detect.persistence(df, length) size = rules_detect.group_size(df) df = rules_detect.interpolate(df) ######################################### # LSTM Univariate Vanilla Model # ######################################### # MODEL CREATION # ######################################### # scales data, reshapes data, builds and trains model, evaluates model results time_steps = 10 samples = 5000 cells = 128
"\n\n###########################################\n#Processing data for site: "
+ sites[j] + ".\n###########################################")
df_full, sensor_array = anomaly_utilities.get_data(sites[j],
sensor,
year,
path="LRO_data/")
# RULES BASED ANOMALY DETECTION #
#########################################
range_count = []
persist_count = []
methods_output.rules_metrics = []
# size = []
for i in range(0, len(sensor_array)):
sensor_array[sensor[i]], r_c = rules_detect.range_check(
sensor_array[sensor[i]], site_params[j][i].max_range,
site_params[j][i].min_range)
range_count.append(r_c)
sensor_array[sensor[i]], p_c = rules_detect.persistence(
sensor_array[sensor[i]], site_params[j][i].persist)
persist_count.append(p_c)
# s = rules_detect.group_size(sensor_array[sensor[i]])
# size.append(s)
sensor_array[sensor[i]] = rules_detect.add_labels(
sensor_array[sensor[i]], -9999)
sensor_array[sensor[i]] = rules_detect.interpolate(
sensor_array[sensor[i]])
# print(str(sensor[i]) + ' longest detected group = ' + str(size[i]))
# metrics for rules based detection #
df_rules_metrics = sensor_array[sensor[i]]
sensors = ['temp', 'cond', 'ph', 'do']
years = [2014, 2015, 2016, 2017, 2018, 2019]
df_full, sensor_array = anomaly_utilities.get_data(site,
sensors,
years,
path="LRO_data/")
#### Rules Based Anomaly Detection
#########################################
range_count = dict()
persist_count = dict()
rules_metrics = dict()
for snsr in sensor_array:
sensor_array[snsr], range_count[snsr] = \
rules_detect.range_check(sensor_array[snsr], site_params[site][snsr]['max_range'], site_params[site][snsr]['min_range'])
sensor_array[snsr], persist_count[snsr] = \
rules_detect.persistence(sensor_array[snsr], site_params[site][snsr]['persist'], output_grp=True)
sensor_array[snsr] = rules_detect.add_labels(sensor_array[snsr], -9999)
sensor_array[snsr] = rules_detect.interpolate(sensor_array[snsr])
# s = rules_detect.group_size(sensor_array[snsr])
# size.append(s)
# print(str(snsr) + ' longest detected group = ' + str(size))
# metrics for rules based detection #
df_rules_metrics = sensor_array[snsr]
df_rules_metrics['labeled_event'] = anomaly_utilities.anomaly_events(
df_rules_metrics['labeled_anomaly'], wf=0)
df_rules_metrics['detected_event'] = anomaly_utilities.anomaly_events(
df_rules_metrics['anomaly'], wf=0)
anomaly_utilities.compare_events(df_rules_metrics, wf=0)