def sim(main, query, sr=22050 // 4):
# main is good track
batch_sz = len(main) // 10
idxs, dists = mts.mass2_batch(main,
query,
batch_size=batch_sz,
top_matches=len(main) // batch_sz - 1,
n_jobs=6)
return idxs[np.argsort(dists)]
def test_mass2_batch_robotdog_multi_threaded():
"""Sanity check that compares results from UCR use case."""
robot_dog = np.loadtxt(
os.path.join(MODULE_PATH, '..', 'tests', 'robot_dog.txt'))
carpet_walk = np.loadtxt(
os.path.join(MODULE_PATH, '..', 'tests', 'carpet_query.txt'))
indices, distances = mts.mass2_batch(
robot_dog, carpet_walk, 1000, top_matches=3, n_jobs=2)
min_dist_idx = np.argmin(distances)
min_idx = indices[min_dist_idx]
assert(min_idx == 7479)
def f_ts_clustering(param_pe, param_row, param_ca_data, param_ce_data,
param_tipo, param_p_ventana, param_cores, param_batch,
param_matches):
"""
Parameters
----------
param_pe : pd.DataFrame : dataframe con precios
param_row :
param_ce_data :
param_ca_data :
param_p_ventana :
param_cores :
param_tipo :
param_batch :
param_matches :
Returns
-------
df_tabla_busqueda :
Debugging
---------
param_pe = df_precios # precios historicos para minar
param_row = 4 # renglon de iteracion de candidatos
param_ca_data = df_ind_3 # dataframe con candidatos a iterar
param_ce_data = df_ce # dataframe con calendario completo
param_tipo = 'mid'
param_p_ventana = 30 # tamano de ventana para buscar serie de tiempo
param_cores = 4 # nucleos con los cuales utilizar algoritmo
param_batch = 300
param_matches = 10
"""
# almacenar resultados
# dict_res = {'name': [], 'esc': [], 'timestamp': [],
# 'tipo_1': [], 'tipo_2': [], 'tipo_3': [], 'tipo_4': []}
# renglon con informacion de evento disparador candidato
candidate_data = param_ca_data.iloc[param_row, :]
# print('Ind disparador: ' + str(candidate_data['name']) + ' - ' + candidate_data['esc'])
# datos completos de todas las ocurrencias del evento disparador candidato
df_ancla = param_ce_data[(param_ce_data['esc'] == candidate_data['esc']) &
(param_ce_data['name'] == candidate_data['name'])]
# todos los timestamps del calendario economico completo
ts_serie_ce = list(param_ce_data['timestamp'])
# inicializar contadores de ocurrencias por escenario ancla
p1, p2, p3, p4 = 0, 0, 0, 0
# Para guardar resultados parciales
dict_res = {'ancla': df_ancla['id'].iloc[0], 'metricas': {}, 'datos': {}}
# -- ------------------------------------------------------ OCURRENCIA POR OCURRENCIA -- #
for ancla in range(0, len(df_ancla['timestamp'])):
# ancla = 31
# print(ancla)
# datos de ancla para buscar hacia el futuro
ancla_ocurr = df_ancla.iloc[ancla, ]
# print('ind: ' + ancla_ocurr['name'] + ' ' + ancla_ocurr['esc'] + ' ' +
# str(ancla_ocurr['timestamp']))
# fecha de ancla
fecha_ini = ancla_ocurr['timestamp']
# .. buscar recurrentemente la fecha mas cercana para construir serie y serie_p
while len(param_pe[param_pe['timestamp'] == fecha_ini].index) == 0:
fecha_ini = fecha_ini - timedelta(minutes=1)
# se toma el timestamp de precios igual a timestamp del primer escenario del indicador
ind_ini = param_pe[param_pe['timestamp'] == fecha_ini].index
# fecha final es la fecha inicial mas un tamano de ventana arbitrario
ind_fin = ind_ini + param_p_ventana
# se construye la serie query
df_serie_q = param_pe.copy().loc[ind_ini[0]:ind_fin[0], :]
df_serie_q = df_serie_q.reset_index(drop=True)
# se toma el mid como valor para construir series temporales
serie_q = np.array(df_serie_q[param_tipo])
# se construye la serie completa para busqueda (un array de numpy de 1 dimension)
df_serie = param_pe.copy().loc[ind_ini[0]:, :]
df_serie = df_serie.reset_index(drop=True)
# se toma el mid como valor para construir series temporales
serie = np.array(df_serie[param_tipo])
try:
# correr algoritmo y regresar los indices de coincidencias y las distancias
mass_indices, mass_dists = mass.mass2_batch(
ts=serie,
query=serie_q,
batch_size=param_batch,
n_jobs=param_cores,
top_matches=param_matches)
# Borrar inidice 0 de resultados por ser el mismo que la serie query
origen = np.where(mass_indices == 0)[0][0]
mass_indices = np.delete(mass_indices, origen)
# mass_dists = np.delete(mass_dists, origen)
# print('indices encontrados' + ' ' + str(mass_indices))
# Indice de referencia de n-esima serie similar encontrada
for indice in mass_indices:
# indice = mass_indices[0]
# print(indice)
# DataFrame de n-esima serie patron similar encontrada
df_serie_p = df_serie.copy().loc[indice:(indice +
param_p_ventana), :]
# print(df_serie_p.head())
# print('Verificando patron con f_ini: ' +
# str(list(df_serie_p['timestamp'])[0]) + ' f_fin: ' +
# str(list(df_serie_p['timestamp'])[-1]))
# Extraer el timestamp inicial para verificar si coincide con indicador
ts_serie_p = list(df_serie_p['timestamp'])[0]
# Busqueda si el timestamp inicial de cada uno de los patrones
# encontrados es igual a alguna fecha de comunicacion de toda
# la lista de indicadores que se tiene
if ts_serie_p in ts_serie_ce:
# ID del evento ancla que genero patron hacia adelante
id_ocurrencia = ancla_ocurr['id'] + '_' + ancla_ocurr['esc'] +\
'_' + str(ancla_ocurr['timestamp'])[:-6].replace(' ', '_')
match = np.where(
param_ce_data['timestamp'] == ts_serie_p)[0]
encontrados = param_ce_data.loc[match, :]
# print(' ------------------ Coincidencia encontrada ------------------')
# print('buscando en: ' + id_ocurrencia)
# print(' ----------- Se encontro el patron que empieza en: -----------')
# print(ts_serie_p)
# print('en los siguientes casos: ')
# print(encontrados)
# -- contar y sacar los datos segun tipo
# Paso 1: tener un diccionario con la llave id_ocurrencia con la encontrada
# Paso 2: dentro de la llave id_ocurrencia tener la llave datos
dict_res['datos'].update({
id_ocurrencia: {
'ocurrencias': {},
'df_serie_q': df_serie_q,
'df_serie_p': df_serie_p
}
})
# Paso 3: hacer las llaves id_sub_ocurrencia para cada sub de ocurrencia
llaves = [
encontrados['id'].iloc[j] + '_' +
encontrados['esc'].iloc[j] + '_' +
str(encontrados['timestamp'].iloc[j])[:-6].replace(
' ', '_') for j in range(0, len(encontrados['id']))
]
dict_res['datos'][id_ocurrencia]['ocurrencias'] = llaves
enc = (encontrados['name'] == ancla_ocurr['name']) & \
(encontrados['esc'] == ancla_ocurr['esc'])
# TIPO 1: name == name & esc == esc
p1 = p1 + len(encontrados.loc[enc, 'name'])
# print('tipo_1 = ' + str(p1))
# TIPO 2: name == name
p2 = p2 + len(encontrados.loc[encontrados['name'] ==
ancla_ocurr['name'], 'name'])
# print('tipo_2 = ' + str(p2))
# TIPO 3: cualquiera en calendario
p3 = p3 + len(encontrados.loc[
encontrados['name'] != ancla_ocurr['name'], 'name'])
# print('tipo_3 = ' + str(p3))
# TIPO 4: fuera de calendario
p4 = p4 + 0
# print('tipo_4 = ' + str(p4))
else:
# TIPO 4: fuera de calendario
p4 += len(mass_indices)
# print('p4 = ' + str(p4))
# tipo_4 = Cualquier otro punto en el tiempo
except ValueError:
# print('ValueError: problemas de indices en MASS-TS')
p4 += 0
except IndexError:
# print('IndexError: problemas de indices en MASS-TS')
p4 += 0
# agregar al diccionario de resultados los casos encontrados
dict_res.update({
'metricas': {
# Mismo Indicador + Mismo Escenario que la ancla
'tipo_1': p1,
# Mismo Indicador + Cualquier Escenario
'tipo_2': p2,
# Otro Indicador en la lista
'tipo_3': p3,
# Ninguna de las anteriores
'tipo_4': p4
}
})
return dict_res
def find_cloudburst_motifs(metric, snippet, timeseries, print_output=False):
"""
Takes a snippet of timeseries, creates motifs at batch_sizes from the
snippet and searches for those motifs in the given timeseries, returns a
matched_motifs dict and a timeseries_matched dict
"""
logger = logging.getLogger(skyline_app_logger)
child_process_pid = os.getpid()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: running for process_pid - %s for %s'
% (str(child_process_pid), metric))
start = timer()
debug_logging = False
timeseries_matched = {}
timeseries_matched[metric] = {}
matched_motifs = {}
motifs_found = []
dev_null = None
for item in timeseries:
timestamp = int(item[0])
timeseries_matched[metric][timestamp] = {}
timeseries_matched[metric][timestamp]['motif_matches'] = {}
mass2_batch_times = []
mass3_times = []
exact_match_times = []
nan = np.array([np.nan])
nanj = complex(0.0, float('nan'))
empty_dists = np.array(nan + nanj)
motifs_found = []
exact_matches_found = []
motif_match_types = motif_match_types_dict()
start_full_duration = timer()
# metric_resolution = determine_data_frequency(skyline_app, timeseries, False)
logger.info(
'functions.luminosity.find_cloudburst_motifs :: looking for similar motifs in timeseries of length: %s'
% str(len(timeseries)))
exact_match_times = []
# relate_dataset = [float(item[1]) for item in fp_timeseries]
relate_dataset = [float(item[1]) for item in timeseries]
# namespace_key = 'default_inference_batch_sizes'
# for batch_size in list(IONOSPHERE_INFERENCE_MOTIFS_SETTINGS[namespace_key].keys()):
# for batch_size in [len(snippet)]:
batch_size = len(snippet)
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: checking %s at batch_size: %s'
% (metric, str(batch_size)))
# @added 20210423 - Feature #4014: Ionosphere - inference
# The convenience mass2_batch method will not work to find
# top matches if the number of top_matches to be found are
# greater than the number of indices in which a match can be
# found. In these cases such as trying to find the:
# batch_size: 1440, top_matches: 50, max_distance: 30, snippet_length: 1451
# even setting the top_matches to 1 will result in
# mass2_batch throwing the error:
# mts.mass2_batch error: kth(=1) out of bounds (1)
# So use mass3 as appropriate.
use_mass3 = False
use_mass2_batch = True
n = len(snippet)
indices = list(range(0, n - batch_size + 1, batch_size))
# mass2_batch default is 3 so if there are less than 3
# indices in which the best macthes can be found, use mass3
if len(indices) < 3:
use_mass3 = True
use_mass2_batch = False
logger.info(
'functions.luminosity.find_cloudburst_motifs :: batch_size: %s, snippet length: %s, len(indices) < 3, using mass3'
% (str(batch_size), str(n)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: batch_size: %s, snippet length: %s, len(indices) < 3, using mass3'
% (str(batch_size), str(n)))
top_matches = 1
max_distance = 1.8
find_exact_matches = True
# if use_mass2_batch:
logger.info(
'functions.luminosity.find_cloudburst_motifs :: analysis run - metric: %s, batch_size: %s, top_matches: %s, max_distance: %s, snippet_length: %s'
% (str(metric), str(batch_size), str(top_matches), str(max_distance),
str(len(snippet))))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: analysis run - metric: %s, batch_size: %s, top_matches: %s, max_distance: %s, snippet_length: %s'
% (str(metric), str(batch_size), str(top_matches),
str(max_distance), str(len(snippet))))
# Given that the snippet can be any length
if len(snippet) < batch_size:
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: skipping snippet: %s, batch_size: %s'
% (str(len(snippet)), str(batch_size)))
return matched_motifs, timeseries_matched
else:
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: checking %s, batch_size: %s'
% (metric, str(batch_size)))
# Create the subsequence that is being searched for
n = batch_size
# snippets = [snippet[i * n:(i + 1) * n] for i in range((len(snippet) + n - 1) // n)]
# snippets = [snippet]
# batch_size_anomalous_timeseries_subsequence = timeseries[-batch_size:]
# batch_size_dataset = [float(item[1]) for item in batch_size_anomalous_timeseries_subsequence]
# for i_snippet in snippets:
# batch_size_timeseries_subsequence = i_snippet[-batch_size:]
batch_size_timeseries_subsequence = snippet
batch_size_dataset = [
float(item[1]) for item in batch_size_timeseries_subsequence
]
motif_timestamp = int(batch_size_timeseries_subsequence[-1][0])
# Set defaults
current_best_indices = []
current_best_dists = []
best_indices = None
best_dists = None
# POC running all through mass3 with maximum pieces (SUPER FAST)
# and then filtering on max_distance, all_in_range and area
# percent_different
# use_mass3 = True
# use_mass2_batch = False
# POC running all through mass3 and then filtering FALIED in
# terms of time taken... due to having to run 22421 motifs
# through all_in_range and percent_different functions ...
# just these motifs checked took 62.036366 seconds, the surfacing
# and transforming of the data AND mass3 to only 2 seconds
# 2021-04-27 13:45:59 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: analysed 2 fps of full_duration 86400 in 0.330732 seconds
# 2021-04-27 13:45:59 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: 22421 distance_valid_motifs determined in 0.346807 seconds from 81432 motifs_found
# 2021-04-27 13:45:59 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: sorted_motifs from distance_valid_motifs in 0.048316 seconds
# 2021-04-27 13:46:01 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000590 seconds
# 2021-04-27 13:46:01 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000271 seconds
# ...
# ...
# 2021-04-27 13:46:57 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000373 seconds
# 2021-04-27 13:46:57 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000381 seconds
# 2021-04-27 13:46:58 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000363 seconds
# 2021-04-27 13:46:58 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: percent_different in 0.000348 seconds
# 2021-04-27 13:47:01 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: motifs checked in 62.036366 seconds
# 2021-04-27 13:47:01 :: 3586421 :: functions.luminosity.find_cloudburst_motifs :: 0 motif best match found from 81432 motifs_found, 4 fps where checked {604800: {'fp_count': 2}, 86400: {'fp_count': 2}} (motifs remove due to not in range 22325, percent_different 96) and it took a total of 64.761969 seconds (only mass3) to process telegraf.ssdnodes-26840.mariadb.localhost:3306.mysql.bytes_sent
# 2021-04-27 13:47:01 :: 3586421 :: inference found 0 matching similar motifs, checked 0 fps in 64.790198 seconds
if use_mass2_batch:
try:
# @added 20210419 - Feature #4014: Ionosphere - inference
# Handle top_matches being greater than possible kth that can be found
# mts.mass2_batch error: kth(=50) out of bounds (16)
use_top_matches = int(top_matches)
if (len(snippet) / int(batch_size)) <= int(top_matches):
use_top_matches = round(len(snippet) / int(batch_size)) - 2
if use_top_matches == 2:
use_top_matches = 1
if use_top_matches < 1:
use_top_matches = 1
logger.info(
'functions.luminosity.find_cloudburst_motifs :: adjusting top_matches for mass2_batch to %s (the maximum possible top - 1) as top_matches=%s will be out of bounds mts.mass2_batch'
% (str(use_top_matches), str(top_matches)))
start_mass2_batch = timer()
best_indices, best_dists = mts.mass2_batch(
relate_dataset,
batch_size_dataset,
batch_size=batch_size,
top_matches=use_top_matches)
end_mass2_batch = timer()
mass2_batch_times.append((end_mass2_batch - start_mass2_batch))
current_best_indices = best_indices.tolist()
current_best_dists = best_dists.tolist()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: mass2_batch run on batch_size: %s, top_matches: %s, in %6f seconds'
% (str(batch_size), str(use_top_matches),
(end_mass2_batch - start_mass2_batch)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: mass2_batch run on batch_size: %s, top_matches: %s, in %6f seconds'
% (str(batch_size), str(use_top_matches),
(end_mass2_batch - start_mass2_batch)))
if debug_logging:
logger.debug(
'debug :: functions.luminosity.find_cloudburst_motifs :: best_indices: %s, best_dists: %s'
% (str(current_best_indices), str(current_best_dists)))
except ValueError as e:
# If mass2_batch reports out of bounds, use mass3
if 'out of bounds' in str(e):
use_mass3 = True
best_dists = ['use_mass3']
logger.info(
'functions.luminosity.find_cloudburst_motifs :: mts.mass2_batch will be out of bounds running mass3'
)
except Exception as e:
logger.error(
'error :: functions.luminosity.find_cloudburst_motifs :: %s mts.mass2_batch error: %s'
% (str(metric), str(e)))
if print_output:
print(
'error :: functions.luminosity.find_cloudburst_motifs :: %s mts.mass2_batch error: %s'
% (str(metric), str(e)))
return matched_motifs, timeseries_matched
if not use_mass3:
try:
if str(list(best_dists)) == str(list(empty_dists)):
logger.info(
'functions.luminosity.find_cloudburst_motifs :: mts.mass2_batch no similar motif from %s - best_dists: %s'
% (str(metric), str(list(best_dists))))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: mts.mass2_batch no similar motif from %s - best_dists: %s'
% (str(metric), str(list(best_dists))))
return matched_motifs, timeseries_matched
except Exception as e:
dev_null = e
# @added 20210423 -
if use_mass3:
# pieces should be larger than the query length and as many
# as possible, a power of two would be best, but as many
# pieces as possible is the best we can achieve above 265
query_length = len(batch_size_dataset)
# if query_length < 256:
# pieces = 256
# else:
# pieces = query_length + 2
pieces = len(snippet) - query_length
if pieces < query_length:
pieces = query_length + 2
check_pieces_length = False
if check_pieces_length:
# @modified 20210504 - Feature #4014: Ionosphere - inference
# Handle the fp_timeseries being the same length (meaning
# too short) as the query length
if len(snippet) <= pieces:
logger.info(
'functions.luminosity.find_cloudburst_motifs :: skipping running mass3 with %s pieces on metric: %s, batch_size: %s because snippet length is not long enough for the query size'
% (str(pieces), str(metric), str(batch_size)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: skipping running mass3 with %s pieces on metric: %s, batch_size: %s because snippet length is not long enough for the query size'
% (str(pieces), str(metric), str(batch_size)))
return matched_motifs, timeseries_matched
# @modified 20210505 - Feature #4014: Ionosphere - inference
# Skip the batch size if the fp_timeseries is a similar
# length as the batch_size. This was specifically added to
# reduce errors were there may be missing data points in a
# timeseries and the lengths are not the same. This was
# encountered on a batch_size of 1440 with FULL_DURATION
# 86400 60 second data. A match was never found at a
# batch_size > 720 on that data, but errors were occassionally
# encountered.
ten_percent_of_batch_size = int(batch_size / 10)
if (len(snippet) - ten_percent_of_batch_size) < batch_size:
logger.info(
'functions.luminosity.find_cloudburst_motifs :: skipping running mass3 on metric: %s, batch_size: %s because the batch_size is too close to length'
% (str(metric), str(batch_size)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: skipping running mass3 on metric: %s, batch_size: %s because the batch_size is too close to length'
% (str(metric), str(batch_size)))
return matched_motifs, timeseries_matched
logger.info(
'functions.luminosity.find_cloudburst_motifs :: running mass3 with %s pieces on on metric: %s, batch_size: %s'
% (str(pieces), str(metric), str(batch_size)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: running mass3 with %s pieces on on metric: %s, batch_size: %s'
% (str(pieces), str(metric), str(batch_size)))
start_mass3 = timer()
try:
best_dists = mts.mass3(relate_dataset, batch_size_dataset, pieces)
end_mass3 = timer()
except Exception as e:
logger.error(
'error :: functions.luminosity.find_cloudburst_motifs :: metric %s mts.mass3 error: %s'
% (str(metric), str(e)))
if print_output:
print(
'error :: functions.luminosity.find_cloudburst_motifs :: metric %s mts.mass3 error: %s'
% (str(metric), str(e)))
return matched_motifs, timeseries_matched
mass3_times.append((end_mass3 - start_mass3))
current_best_dists = best_dists.tolist()
# Create current_best_indices as mass2_batch returns
current_best_indices = []
if len(relate_dataset) > batch_size:
for index in enumerate(relate_dataset):
# if index[0] >= (batch_size - 1):
# The array starts at batch_size + 1
# if index[0] >= (batch_size + 1):
# but that fails on the add_motifs comprehension
# add_motifs = [[fp_id, current_best_indices[index], best_dist.real, batch_size_timeseries_subsequence, batch_size, max_distance, max_area_percent_diff, max_y, min_y, range_padding, min_y_padded, max_y_padded] for index, best_dist in enumerate(current_best_dists)]
# IndexError: list index out of range
if index[0] >= (batch_size - 1):
current_best_indices.append(index[0])
# @modified 20210505 - Feature #4014: Ionosphere - inference
# Handle the query_length being shorter than the batch_size
if len(current_best_indices) != len(current_best_dists):
current_best_indices = []
if index[0] >= (query_length - 1):
current_best_indices.append(index[0])
if len(current_best_indices) != len(current_best_dists):
logger.info(
'functions.luminosity.find_cloudburst_motifs :: discarding mass3 results as current_best_dists length: %s, current_best_indices length: %s do not match, took %6f seconds'
% (str(len(current_best_dists)),
str(len(current_best_indices)),
(end_mass3 - start_mass3)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: discarding mass3 results as current_best_dists length: %s, current_best_indices length: %s do not match, took %6f seconds'
% (str(len(current_best_dists)),
str(len(current_best_indices)),
(end_mass3 - start_mass3)))
return matched_motifs, timeseries_matched
logger.info(
'functions.luminosity.find_cloudburst_motifs :: mass3 run, current_best_dists length: %s, current_best_indices length: %s, took %6f seconds'
% (str(len(current_best_dists)), str(len(current_best_indices)),
(end_mass3 - start_mass3)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: mass3 run, current_best_dists length: %s, current_best_indices length: %s, took %6f seconds'
%
(str(len(current_best_dists)), str(len(current_best_indices)),
(end_mass3 - start_mass3)))
if not use_mass3:
if not current_best_indices[0]:
return matched_motifs, timeseries_matched
if use_mass3 and not current_best_indices:
return matched_motifs, timeseries_matched
# All in one quicker? Yes
start_add_motifs = timer()
add_motifs = []
try:
add_motifs = [[
metric, current_best_indices[index], best_dist.real,
batch_size_timeseries_subsequence, batch_size, max_distance,
motif_timestamp
] for index, best_dist in enumerate(current_best_dists)]
if add_motifs:
motifs_found = motifs_found + add_motifs
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: functions.luminosity.find_cloudburst_motifs :: could not add_motifs to motifs_found - %s'
% (e))
end_add_motifs = timer()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: added %s motifs to motifs_found in %.6f seconds'
% (str(len(add_motifs)), (end_add_motifs - start_add_motifs)))
# Break if an exact match is found
# @modified 20210430 - Bug #4044: inference - motif distance override - exact match
# @modified 20210504 - Bug #4044: inference - motif distance override - exact match
# if len([item for item in add_motifs if item[2] == 0]) > 0:
# exact_matches_found = exact_matches_found + [item for item in add_motifs if item[2] == 0]
# break
# @modified 20210427 - Feature #4014: Ionosphere - inference
# Finding exact matches can result is more than doubling the
# runtime when used after mass2_batch runs (which do not find)
# exact matches, mass3 does. However the amount of time an
# exact match is found, is very rare
# if not use_mass3:
if not use_mass3 and find_exact_matches:
# mass3 finds exact matches, mass2_batch does not, so
# there is no need to find exacts matchs if mass3 was
# run.
# FIND EXACT MATCHES
# Seeing as I cannot reproduce finding nan+nanj which represents an
# exact match with mts.mass2_batch, do it DIY style - iterate the
# timeseries and create a batch_size subsequence for every index and
# compare the values to the anomalous_ts for an exact match.
# This takes ~0.024850 seconds on a timeseries with 10079 datapoints
try:
start_exact_match = timer()
indexed_relate_dataset = []
for index, item in enumerate(relate_dataset):
indexed_relate_dataset.append([index, item])
last_index = indexed_relate_dataset[-1][0]
current_index = 0
while current_index < last_index:
subsequence = [
value
for index, value in indexed_relate_dataset[current_index:(
current_index + batch_size)]
]
if subsequence == batch_size_dataset:
# @modified 20210419 - Feature #4014: Ionosphere - inference
# Added batch_size
exact_matches_found.append([
metric, current_index, 0.0,
batch_size_timeseries_subsequence, batch_size,
max_distance, motif_timestamp
])
motifs_found.append([
metric, current_index, 0.0,
batch_size_timeseries_subsequence, batch_size,
max_distance, motif_timestamp
])
current_index += 1
end_exact_match = timer()
exact_match_times.append((end_exact_match - start_exact_match))
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: functions.luminosity.find_cloudburst_motifs :: could not determine it any exact matches could be found in %s timeseries - %s'
% (str(metric), e))
logger.info(
'functions.luminosity.find_cloudburst_motifs :: exact matches checked in %.6f seconds'
% ((end_exact_match - start_exact_match)))
# TODO
# mass3 ALL, then evaluate, would it be quicker? No see POC
# above
logger.info(
'functions.luminosity.find_cloudburst_motifs :: mts.mass2_batch runs on %s in %.6f seconds'
% (str(metric), sum(mass2_batch_times)))
logger.info(
'functions.luminosity.find_cloudburst_motifs :: exact_match runs on %s in %.6f seconds'
% (str(metric), sum(exact_match_times)))
end_full_duration = timer()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: analysed %s in %.6f seconds'
% (str(metric), (end_full_duration - start_full_duration)))
# Patterns are sorted by distance
# The list produced with the mass3 method will include
# nans
start_distance_valid_motifs = timer()
distance_valid_motifs = [
item for item in motifs_found
if not np.isnan(item[2]) and item[2] <= item[5]
]
end_distance_valid_motifs = timer()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: %s distance_valid_motifs determined in %.6f seconds from %s motifs_found'
% (str(len(distance_valid_motifs)),
(end_distance_valid_motifs - start_distance_valid_motifs),
str(len(motifs_found))))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: %s distance_valid_motifs determined in %.6f seconds from %s motifs_found'
% (str(len(distance_valid_motifs)),
(end_distance_valid_motifs - start_distance_valid_motifs),
str(len(motifs_found))))
start_sorted_motifs = timer()
sorted_motifs = []
if motifs_found:
sorted_motifs = sorted(distance_valid_motifs, key=lambda x: x[2])
# If the areas under the curve were calculated, the
# list could be sorted by area_percent_diff then by
# distance.
# import operator
# sorted_motifs = sorted(motifs_found_in_fp, key=operator.itemgetter(2, 2))
end_sorted_motifs = timer()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: sorted_motifs from distance_valid_motifs in %.6f seconds'
% ((end_sorted_motifs - start_sorted_motifs)))
start_motifs_check = timer()
snippet_timestamps = [int(item[0]) for item in snippet]
for motif in sorted_motifs:
try:
add_match = False
metric = motif[0]
best_index = motif[1]
best_dist = motif[2]
# motif_sequence = motif[3]
# @modified 20210419 - Feature #4014: Ionosphere - inference
# Added batch_size
motif_size = motif[4]
motif_timestamp = motif[6]
add_match = True
match_type = 'distance'
if motif in exact_matches_found:
match_type = 'exact'
if debug_logging:
logger.debug(
'debug :: functions.luminosity.find_cloudburst_motifs :: exact match: %s'
% (str(motif)))
full_relate_timeseries = timeseries
relate_timeseries = [
item for index, item in enumerate(full_relate_timeseries)
if index >= best_index and index < (best_index + motif_size)
]
relate_dataset = [item[1] for item in relate_timeseries]
# relate_dataset_timestamps = [int(item[0]) for item in relate_timeseries]
matched_period_timestamps = [
int(item[0]) for item in relate_timeseries
]
in_period = False
for matched_period_timestamp in matched_period_timestamps:
if matched_period_timestamp in snippet_timestamps:
in_period = True
if not in_period:
add_match = False
if add_match:
timestamp = int(relate_timeseries[-1][0])
timeseries_matched[metric][timestamp]['motif_matches'][
motif_timestamp] = motif_size
motif_id = '%s-%s-%s' % (str(metric), str(int(
snippet[-1][0])), str(best_index))
matched_motifs[motif_id] = {}
matched_motifs[motif_id]['index'] = best_index
matched_motifs[motif_id]['distance'] = best_dist
matched_motifs[motif_id]['size'] = motif_size
matched_motifs[motif_id]['timestamp'] = timestamp
matched_motifs[motif_id][
'matched_period_timestamps'] = matched_period_timestamps
matched_motifs[motif_id]['motif_timestamp'] = motif_timestamp
matched_motifs[motif_id]['type'] = match_type
matched_motifs[motif_id]['type_id'] = motif_match_types[
match_type]
runtime_end = timer()
matched_motifs[motif_id]['runtime'] = (runtime_end - start)
# if SINGLE_MATCH:
# break
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: functions.luminosity.find_cloudburst_motifs :: metric %s and motif: %s - %s'
% (str(metric), str(motif), str(e)))
continue
end_motifs_check = timer()
logger.info(
'functions.luminosity.find_cloudburst_motifs :: motifs checked in %.6f seconds'
% ((end_motifs_check - start_motifs_check)))
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: motifs checked in %.6f seconds'
% ((end_motifs_check - start_motifs_check)))
# Sort by distance AND area_percent_diff
sorted_ordered_matched_motifs_list = []
if matched_motifs and len(matched_motifs) > 1:
ordered_matched_motifs_list = []
for motif_id in list(matched_motifs.keys()):
distance = matched_motifs[motif_id]['distance']
ordered_matched_motifs_list.append([motif_id, distance])
# If the areas under the curve were calculated, the
# list could be sorted by area_percent_diff then by
# distance.
sorted_matched_motifs = {}
sorted_ordered_matched_motifs_list = sorted(
ordered_matched_motifs_list, key=operator.itemgetter(1))
logger.info(
'functions.luminosity.find_cloudburst_motifs :: sorting %s matched_motifs by distance'
% (str(len(sorted_ordered_matched_motifs_list))))
for motif_id, distance, in sorted_ordered_matched_motifs_list:
sorted_matched_motifs[motif_id] = matched_motifs[motif_id]
# if SINGLE_MATCH:
# break
matched_motifs = sorted_matched_motifs.copy()
end = timer()
if dev_null:
del dev_null
logger.info(
'functions.luminosity.find_cloudburst_motifs :: %s motif best match found from %s motifs_found and it took a total of %.6f seconds (all mass2/mass3) to process %s'
% (
# str(len(matched_motifs)), str(len(motifs_found)), str(len(fps_checked_for_motifs)),
str(len(matched_motifs)),
str(len(motifs_found)),
(end - start),
metric))
if len(matched_motifs) > 0:
if print_output:
print(
'functions.luminosity.find_cloudburst_motifs :: %s motif best match found from %s distance valid motifs of %s motifs_found and it took a total of %.6f seconds (all mass2/mass3) to process %s'
% (
# str(len(matched_motifs)), str(len(motifs_found)), str(len(fps_checked_for_motifs)),
str(len(matched_motifs)),
str(len(distance_valid_motifs)),
str(len(motifs_found)),
(end - start),
metric))
distances = []
for match in list(matched_motifs.keys()):
distances.append(matched_motifs[match]['distance'])
distances_dict = {}
distances_dict['avg_distance'] = sum(distances) / len(distances)
distances_dict['distances'] = distances
print('%s' % str(distances_dict))
# return matched_motifs, fps_checked_for_motifs
return matched_motifs, timeseries_matched
distances = mts.mass2(ts, query)
# mass3
# distances = mts.mass3(ts, query, 256)
# mass2_batch
# start a multi-threaded batch job with all cpu cores and give me the top 5 matches.
# note that batch_size partitions your time series into a subsequence similarity search.
# even for large time series in single threaded mode, this is much more memory efficient than
# MASS2 on its own.
batch_size = 10000
top_matches = 5
n_jobs = -1
indices, distances = mts.mass2_batch(ts,
query,
batch_size,
top_matches=top_matches,
n_jobs=n_jobs)
# find minimum distance
min_idx = np.argmin(distances)
# find top 4 motif starting indices
k = 4
exclusion_zone = 25
top_motifs = mts.top_k_motifs(distances, k, exclusion_zone)
# find top 4 discord starting indices
k = 4
exclusion_zone = 25
top_discords = mts.top_k_discords(distances, k, exclusion_zone)
def on_demand_motif_analysis(metric, timestamp, similarity, batch_size,
top_matches, max_distance, range_padding,
max_area_percent_diff):
"""
Process a motif similarity search on demand
"""
import numpy as np
import mass_ts as mts
logger = logging.getLogger(skyline_app_logger)
dev_null = None
function_str = 'on_demand_motif_analysis'
logger.info(
'%s :: with parameters :: metric: %s, timestamp: %s, similarity: %s, batch_size:%s, top_matches: %s, max_distance: %s, range_padding: %s, max_area_percent_diff: %s'
% (function_str, str(metric), str(timestamp), str(similarity),
str(batch_size), str(top_matches), str(max_distance),
str(range_padding), str(max_area_percent_diff)))
trace = 'none'
fail_msg = 'none'
start = time.time()
start_timer = timer()
metric_vars_dict = {}
metric_id = 0
fp_ids = []
timeseries = []
not_similar_enough_sample = 0
not_similar_motifs = 0
similar_motifs = 0
exact_motifs = 0
distance_motifs = 0
motifs_found = []
find_exact_matches_run = False
exact_matches_found = []
fps_timeseries = {}
# A motif_analysis dict to add to and return
motif_analysis = {}
motif_analysis[metric] = {}
motif_analysis[metric]['timestamp'] = int(timestamp)
motif_analysis[metric]['started'] = start
motif_analysis[metric]['motifs'] = {}
motif_analysis[metric]['exact_motifs'] = exact_motifs
motif_analysis[metric]['similar_motifs'] = similar_motifs
motif_analysis[metric]['not_similar_motifs'] = not_similar_motifs
motif_analysis[metric][
'not_similar_enough_sample'] = not_similar_enough_sample
# @added 20210417 - Feature #4014: Ionosphere - inference
# Allow the user to define the batch_size per similarity search
motif_analysis[metric]['batch_size'] = int(batch_size)
motif_analysis[metric]['top_matches'] = int(top_matches)
motif_analysis[metric]['max_distance'] = float(max_distance)
# @added 20210425 - Feature #4014: Ionosphere - inference
# Added max_area_percent_diff for computing the area under the curve
motif_analysis[metric]['max_area_percent_diff'] = float(
max_area_percent_diff)
fps_checked_for_motifs = []
metric_dir = metric.replace('.', '/')
metric_timeseries_dir = '%s/%s/%s' % (settings.IONOSPHERE_DATA_FOLDER,
str(timestamp), metric_dir)
# @added 20210418 - Feature #4014: Ionosphere - inference
# Allow for the similarity search on saved_training_data
if 'saved_training_data' in request.args:
saved_training_data_str = request.args.get('saved_training_data',
'false')
if saved_training_data_str == 'true':
saved_metric_timeseries_dir = '%s_saved/%s/%s' % (
settings.IONOSPHERE_DATA_FOLDER, str(timestamp), metric_dir)
if path.exists(saved_metric_timeseries_dir):
metric_timeseries_dir = saved_metric_timeseries_dir
logger.info('%s :: using saved training_data dir - %s' %
(function_str, saved_metric_timeseries_dir))
metric_vars_file = '%s/%s.txt' % (metric_timeseries_dir, metric)
timeseries_json = '%s/%s.json' % (metric_timeseries_dir, metric)
full_duration_in_hours = int(settings.FULL_DURATION / 60 / 60)
full_duration_timeseries_json = '%s/%s.mirage.redis.%sh.json' % (
metric_timeseries_dir, metric, str(full_duration_in_hours))
try:
metric_vars_dict = mirage_load_metric_vars(skyline_app,
metric_vars_file, True)
except Exception as e:
logger.error(
'error :: inference :: failed to load metric variables from check file - %s - %s'
% (metric_vars_file, e))
if not metric_vars_dict:
motif_analysis[metric]['status'] = 'error'
motif_analysis[metric][
'reason'] = 'could not load training data variables'
return motif_analysis
full_duration = metric_vars_dict['metric_vars']['full_duration']
# Determine the metric details from the database
metric_id = 0
metric_db_object = {}
try:
metric_db_object = get_metrics_db_object(metric)
except Exception as e:
logger.error('error :: %s :: failed to get_metrics_db_object - %s' %
(function_str, e))
try:
metric_id = int(metric_db_object['id'])
except Exception as e:
logger.error(
'error :: %s :: failed to determine metric_id from metric_db_object %s - %s'
% (function_str, str(metric_db_object), e))
metric_id = 0
if not metric_id:
logger.error(
'error :: %s :: failed to get metric id for %s from the database' %
(function_str, str(metric)))
fail_msg = 'failed to get metric id'
motif_analysis[metric]['status'] = 'error'
motif_analysis[metric]['reason'] = 'could not determine metric id'
return motif_analysis, fail_msg, trace
# @modified 20210419 - Feature #4014: Ionosphere - inference
# Create a unique dir for each batch_size max_distance
# motif_images_dir = '%s/motifs' % metric_timeseries_dir
motif_images_dir = '%s/motifs/batch_size.%s/top_matches.%s/max_distance.%s' % (
metric_timeseries_dir, str(batch_size), str(top_matches),
str(max_distance))
if not path.exists(motif_images_dir):
# provision motifs image resources
mkdir_p(motif_images_dir)
full_durations = [full_duration]
if path.isfile(full_duration_timeseries_json):
full_durations = [full_duration, settings.FULL_DURATION]
logger.info('%s :: full_durations - %s' %
(function_str, str(full_durations)))
# Loop through analysis per full_duration
for full_duration in full_durations:
start_full_duration = timer()
fp_ids = []
try:
query = 'SELECT id,last_matched from ionosphere WHERE metric_id=%s AND full_duration=%s AND enabled=1 ORDER BY last_matched DESC' % (
str(metric_id), str(full_duration))
results = mysql_select(skyline_app, query)
for row in results:
fp_ids.append(int(row[0]))
except Exception as e:
logger.error(
'error :: %s :: failed to get fp ids via mysql_select from %s - %s'
% (function_str, metric, e))
logger.info('%s :: metric_id: %s, full_duration: %s, fp_ids: %s' %
(function_str,
(metric_id), str(full_duration), str(fp_ids)))
if not fp_ids:
continue
# Now there are known fps, load the timeseries
if full_duration == settings.FULL_DURATION:
timeseries_json_file = full_duration_timeseries_json
else:
timeseries_json_file = timeseries_json
try:
with open((timeseries_json_file), 'r') as f:
raw_timeseries = f.read()
timeseries_array_str = str(raw_timeseries).replace('(',
'[').replace(
')', ']')
del raw_timeseries
timeseries = literal_eval(timeseries_array_str)
del timeseries_array_str
except Exception as e:
logger.error(
'error :: %s :: failed to load timeseries for %s from %s - %s'
% (function_str, metric, timeseries_json_file, e))
continue
anomalous_timeseries_subsequence = []
for timestamp_float, value in timeseries[-int(batch_size):]:
anomalous_timeseries_subsequence.append(
[int(timestamp_float), value])
logger.info(
'%s :: looking for motif in trained fps of full_duration: %s' %
(function_str, (full_duration)))
dataset = [float(item[1]) for item in anomalous_timeseries_subsequence]
max_y = max(dataset)
min_y = min(dataset)
# full_y_range = max_y - min_y
# range_padding_percent = range_padding
# This was just a test that did not have the desired results
# if full_y_range < 10:
# range_padding_percent = 35
# if full_y_range < 5:
# range_padding_percent = 75
# if full_y_range < 2:
# range_padding_percent = 100
use_range_padding = ((max_y - min_y) / 100) * range_padding
if min_y > 0 and (min_y - use_range_padding) > 0:
min_y_padded = min_y - use_range_padding
else:
min_y_padded = min_y
max_y_padded = max_y + use_range_padding
if min_y_padded == max_y_padded:
min_y_padded = min_y_padded - (
(min_y_padded / 100) * range_padding)
max_y_padded = max_y_padded + (
(max_y_padded / 100) * range_padding)
# anomalous_ts = np.array(dataset)
anomalous_ts = dataset
mass2_batch_times = []
exact_match_times = []
nan = np.array([np.nan])
nanj = complex(0.0, float('nan'))
empty_dists = np.array(nan + nanj)
# plotted = False
count = 0
# fp_ids = [fp_id for index, fp_id in enumerate(fp_ids) if index == 0]
# motifs_found = []
# exact_matches_found = []
# fps_timeseries = {}
for fp_id in fp_ids:
if (time.time() - start) >= 20:
break
# Attempt to surface the fp timeseries from memcache and/or db
# @modified 20210424 - Feature #4014: Ionosphere - inference
# Task #4030: refactoring
fp_timeseries = None
try:
fp_timeseries = get_fp_timeseries(skyline_app, metric_id,
fp_id)
except Exception as e:
logger.error(
'inference :: did not get fp timeseries with get_fp_timeseries(%s, %s, %s) - %s'
% (skyline_app, str(metric_id), str(fp_id), e))
if not fp_timeseries:
continue
relate_dataset = [float(item[1]) for item in fp_timeseries]
fps_timeseries[fp_id] = fp_timeseries
current_best_indices = []
current_best_dists = []
best_indices = None
best_dists = None
try:
logger.info(
'%s :: running mts.mass2_batch fp_id: %s, full_duration: %s, batch_size: %s, top_matches: %s, max_distance: %s, motif_size: %s'
% (function_str, str(fp_id), str(full_duration),
str(batch_size), str(top_matches), str(max_distance),
str(len(anomalous_ts))))
# @added 20210418 - Feature #4014: Ionosphere - inference
# Handle top_matches being greater than possible kth that can be found
# mts.mass2_batch error: kth(=50) out of bounds (16)
use_top_matches = int(top_matches)
if (len(fp_timeseries) / int(batch_size)) <= int(top_matches):
use_top_matches = round(
len(fp_timeseries) / int(batch_size)) - 1
if use_top_matches == 2:
use_top_matches = 1
logger.info(
'%s :: adjusting top_matches to %s (the maximum possible top - 1) as kth(=%s) will be out of bounds mts.mass2_batch'
%
(function_str, str(use_top_matches), str(top_matches)))
start_mass2_batch = timer()
# @modified 20210418 - Feature #4014: Ionosphere - inference
# Handle top_matches being greater than possible kth that can be found
# best_indices, best_dists = mts.mass2_batch(relate_dataset, anomalous_ts, batch_size=int(batch_size), top_matches=int(top_matches))
best_indices, best_dists = mts.mass2_batch(
relate_dataset,
anomalous_ts,
batch_size=int(batch_size),
top_matches=int(use_top_matches))
end_mass2_batch = timer()
mass2_batch_times.append((end_mass2_batch - start_mass2_batch))
current_best_indices = best_indices.tolist()
current_best_dists = best_dists.tolist()
# @added 20210412 - Feature #4014: Ionosphere - inference
# Branch #3590: inference
# Add fp_id to fps_checked_for_motifs to enable ionosphere to update the
# motif related columns in the ionosphere database table
fps_checked_for_motifs.append(fp_id)
except Exception as e:
logger.error('error :: %s :: %s mts.mass2_batch error: %s' %
(function_str, (fp_id), str(e)))
continue
try:
if str(list(best_dists)) == str(list(empty_dists)):
logger.info(
'%s :: mts.mass2_batch no similar motif from fp id %s - best_dists: %s'
% (function_str, (fp_id), str(list(best_dists))))
continue
except Exception as e:
dev_null = e
if not current_best_indices[0]:
continue
# if list(best_indices)[0] != anomalous_index:
# continue
# If the best_dists is > 1 they are not very similar
# if list(best_dists)[0].real > 1.0:
# continue
# if list(best_indices)[0] and best_dists:
for index, best_dist in enumerate(current_best_dists):
try:
motif_added = False
"""
Note: mass_ts finds similar motifs NOT the same motif, the same motif
will result in the best_dists being a nan+nanj
So it is DIYed
"""
try:
# @modified 20210414 - Feature #4014: Ionosphere - inference
# Branch #3590: inference
# Store the not anomalous motifs
# motif = [fp_id, current_best_indices[index], best_dist.real]
motif = [
fp_id, current_best_indices[index], best_dist.real,
anomalous_timeseries_subsequence, full_duration
]
except Exception as e:
dev_null = e
motif = []
# if list(best_indices)[0] and best_dists:
# If it is greater than 1.0 it is not similar
# if best_dist.real > 1.0:
# if best_dist.real > IONOSPHERE_INFERENCE_MASS_TS_MAX_DISTANCE:
if best_dist.real > float(max_distance):
continue
else:
if motif:
count += 1
motifs_found.append(motif)
motif_added = True
if not motif_added:
if best_dist == nanj:
count += 1
motifs_found.append(motif)
motif_added = True
if not motif_added:
if str(best_dist) == 'nan+nanj':
count += 1
motifs_found.append([
fp_id, current_best_indices[index], 0.0,
anomalous_timeseries_subsequence, full_duration
])
motif_added = True
if not motif_added:
if best_dist == empty_dists:
count += 1
motifs_found.append(motif)
motif_added = True
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: %s :: could not determine is if fp id %s timeseries at index %s was a match - %s'
% (function_str, str(fp_id),
str(current_best_indices[index]), e))
continue
# FIND EXACT MATCHES
# Seeing as I cannot reproduce finding nan+nanj which represents an
# exact match with mts.mass2_batch, do it DIY style - iterate the
# timeseries and create a batch_size subsequence for every index and
# compare the values to the anomalous_ts for an exact match.
# This takes ~0.024850 seconds on a timeseries with 10079 datapoints
# @modified 20210418 - Feature #4014: Ionosphere - inference
# However fiding exact matches can add ~2.5 seconds on 90 minute
# batch_size and with a proproptionally scaled max_distance of say 15
# finding an exact match in a longer sequence is less important,
# the greater the batch_size the most likely greater the variability
# and the chance of an exact match decreases. So save 2.5 seconds.
# UNLESS
# At a 5 (to 10) batch_size and max_distance of 1.0 an exact match
# can be found. Exact matches are quite frequent and sometimes with
# such little variability, similar matchs may not be found.
# Therefore find find_exact_matches has its place. MASS
# A CAVEAT here is that boring metrics and that change and have a
# low variability even at a larger batch_size could also benefit and
# possibly achieve better accruracy from the use of find_exact_matches
# as they can be shapelets resemble a batch_size 5 shapelet.
# It would perhaps be possible to use one or more of the features
# profile tsfresh values to identify these types of shapelets, if
# you knew which feature/s were most descriptive of this type of
# shapelet, e.g. 'value__skewness': 3.079477685394873, etc (maybe)
# However I predict that this method will perform worst on these
# types of shapelets.
# find_exact_matches = False
# exact matches can be found in batch sizes of 500 and similar not
# So actually always run it.
find_exact_matches = True
find_exact_matches_run = True
if int(batch_size) < 10:
find_exact_matches = True
find_exact_matches_run = True
if find_exact_matches:
try:
start_exact_match = timer()
indexed_relate_dataset = []
for index, item in enumerate(relate_dataset):
indexed_relate_dataset.append([index, item])
last_index = indexed_relate_dataset[-1][0]
current_index = 0
while current_index < last_index:
subsequence = [
value for index, value in
indexed_relate_dataset[current_index:(
current_index + int(batch_size))]
]
if subsequence == anomalous_ts:
exact_matches_found.append([
fp_id, current_index, 0.0,
anomalous_timeseries_subsequence, full_duration
])
motifs_found.append([
fp_id, current_index, 0.0,
anomalous_timeseries_subsequence, full_duration
])
current_index += 1
end_exact_match = timer()
exact_match_times.append(
(end_exact_match - start_exact_match))
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: %s :: could not determine it any exact matches could be found in fp id %s timeseries - %s'
% (function_str, str(fp_id), e))
logger.info(
'%s :: mts.mass2_batch runs on %s fps of full_duration %s in %.6f seconds'
% (function_str, str(len(mass2_batch_times)), str(full_duration),
sum(mass2_batch_times)))
if find_exact_matches_run:
logger.info(
'%s :: exact_match runs on %s fps of full_duration %s in %.6f seconds'
% (function_str, str(len(exact_match_times)),
str(full_duration), sum(exact_match_times)))
end_full_duration = timer()
logger.info(
'%s :: analysed %s fps of full_duration %s in %.6f seconds' %
(function_str, str(len(fp_ids)), str(full_duration),
(end_full_duration - start_full_duration)))
# Patterns are sorted
sorted_motifs = []
motifs_found_in_fps = []
if motifs_found:
sorted_motifs = sorted(motifs_found, key=lambda x: x[2])
for item in sorted_motifs:
motifs_found_in_fps.append(item[0])
logger.info('%s :: %s motifs found' %
(function_str, str(len(sorted_motifs))))
for motif in sorted_motifs:
if (time.time() - start) >= 25:
break
try:
add_match = False
all_in_range = False
fp_id = motif[0]
best_index = motif[1]
best_dist = motif[2]
# @added 20210414 - Feature #4014: Ionosphere - inference
# Branch #3590: inference
# Store the not anomalous motifs
motif_sequence = motif[3]
motif_full_duration = motif[4]
match_type = 'not_similar_enough'
if motif in exact_matches_found:
add_match = True
match_type = 'exact'
all_in_range = True
exact_motifs += 1
full_relate_timeseries = fps_timeseries[fp_id]
# full_relate_dataset = [float(item[1]) for item in full_relate_timeseries]
relate_timeseries = [
item for index, item in enumerate(full_relate_timeseries)
if index >= best_index and index < (best_index +
int(batch_size))
]
relate_dataset = [item[1] for item in relate_timeseries]
if not add_match:
all_in_range = True
for value in relate_dataset:
if value < min_y_padded:
all_in_range = False
break
if value > max_y_padded:
all_in_range = False
break
if all_in_range:
related_max_y = max(relate_dataset)
if related_max_y < (max_y - range_padding):
all_in_range = False
if related_max_y > (max_y + range_padding):
all_in_range = False
related_min_y = min(relate_dataset)
if related_min_y < (min_y - range_padding):
all_in_range = False
if related_min_y > (min_y + range_padding):
all_in_range = False
if all_in_range:
logger.info(
'%s :: ALL IN RANGE - all_in_range: %s, motif: %s'
% (function_str, str(all_in_range),
str(relate_dataset[0:2])))
add_match = True
match_type = 'all_in_range'
similar_motifs += 1
# @added 20210425 - Feature #4014: Ionosphere - inference
# Compute the area using the composite trapezoidal rule.
motif_area = None
fp_motif_area = None
percent_different = None
try:
batch_size_dataset = [
float(item[1]) for item in motif_sequence
]
y_motif = np.array(batch_size_dataset)
motif_area = np.trapz(y_motif, dx=1)
except Exception as e:
logger.error(
'error :: %s :: failed to get motif_area with np.trapz - %s'
% (function_str, e))
try:
y_fp_motif = np.array(relate_dataset)
fp_motif_area = np.trapz(y_fp_motif, dx=1)
except Exception as e:
logger.error(
'error :: %s :: failed to get fp_motif_area with np.trapz - %s'
% (function_str, e))
# Determine the percentage difference (as a
# positive value) of the areas under the
# curves.
if motif_area and fp_motif_area:
percent_different = get_percent_different(
fp_motif_area, motif_area, True)
if percent_different > max_area_percent_diff:
if add_match:
logger.info(
'%s :: AREA TOO DIFFERENT - not adding all_in_range match'
% (function_str))
add_match = False
# BUT ...
if best_dist < 3 and not add_match:
logger.info(
'%s :: DISTANCE VERY SIMILAR - adding match even though area_percent_diff is greater than max_area_percent_diff because best_dist: %s'
% (function_str, str(best_dist)))
add_match = True
match_type = 'distance'
distance_motifs += 1
if similarity == 'all':
if not add_match:
not_similar_motifs += 1
if not_similar_enough_sample >= 10:
continue
not_similar_enough_sample += 1
add_match = True
match_type = 'not_similar_enough'
if add_match:
generation = 0
fp_id_row = None
try:
fp_id_row = get_ionosphere_fp_db_row(
skyline_app, int(fp_id))
except Exception as e:
logger.error(
'error :: %s :: failed to get_ionosphere_fp_db_row for fp_id %s - %s'
% (function_str, str(fp_id), e))
if fp_id_row:
try:
generation = fp_id_row['generation']
except Exception as e:
logger.error(
'error :: %s :: failed to generation from fp_id_row for fp_id %s - %s'
% (function_str, str(fp_id), e))
if generation == 0:
generation_str = 'trained'
else:
generation_str = 'LEARNT'
motif_match_types = motif_match_types_dict()
type_id = motif_match_types[match_type]
motif_id = '%s-%s' % (str(fp_id), str(best_index))
motif_analysis[metric]['motifs'][motif_id] = {}
motif_analysis[metric]['motifs'][motif_id][
'metric_id'] = metric_id
motif_analysis[metric]['motifs'][motif_id]['fp_id'] = fp_id
motif_analysis[metric]['motifs'][motif_id][
'generation'] = generation
motif_analysis[metric]['motifs'][motif_id][
'index'] = best_index
motif_analysis[metric]['motifs'][motif_id][
'distance'] = best_dist
motif_analysis[metric]['motifs'][motif_id]['size'] = int(
batch_size)
motif_analysis[metric]['motifs'][motif_id][
'max_distance'] = float(max_distance)
motif_analysis[metric]['motifs'][motif_id][
'timestamp'] = timestamp
motif_analysis[metric]['motifs'][motif_id][
'type_id'] = type_id
motif_analysis[metric]['motifs'][motif_id][
'type'] = match_type
motif_analysis[metric]['motifs'][motif_id][
'full_duration'] = motif_full_duration
# @added 20210414 - Feature #4014: Ionosphere - inference
# Branch #3590: inference
# Store the not anomalous motifs
motif_analysis[metric]['motifs'][motif_id][
'motif_timeseries'] = anomalous_timeseries_subsequence
motif_analysis[metric]['motifs'][motif_id][
'motif_sequence'] = motif_sequence
not_anomalous_timestamp = int(
anomalous_timeseries_subsequence[-1][0])
graph_period_seconds = not_anomalous_timestamp - int(
anomalous_timeseries_subsequence[0][0])
motif_analysis[metric]['motifs'][motif_id][
'motif_period_seconds'] = graph_period_seconds
motif_analysis[metric]['motifs'][motif_id][
'motif_period_minutes'] = round(graph_period_seconds /
60)
motif_analysis[metric]['motifs'][motif_id]['image'] = None
motif_analysis[metric]['motifs'][motif_id][
'motif_area'] = motif_area
motif_analysis[metric]['motifs'][motif_id][
'fp_motif_area'] = fp_motif_area
motif_analysis[metric]['motifs'][motif_id][
'area_percent_diff'] = percent_different
motif_analysis[metric]['motifs'][motif_id][
'max_area_percent_diff'] = max_area_percent_diff
if (time.time() - start) >= 25:
continue
graph_image_file = '%s/motif.%s.%s.%s.with_max_distance.%s.png' % (
motif_images_dir, motif_id, match_type,
str(batch_size), str(max_distance))
plotted_image = False
on_demand_motif_analysis = True
if not path.isfile(graph_image_file):
plotted_image, plotted_image_file = plot_motif_match(
skyline_app, metric, timestamp, fp_id,
full_duration, generation_str, motif_id,
best_index, int(batch_size), best_dist, type_id,
relate_dataset, anomalous_timeseries_subsequence,
graph_image_file, on_demand_motif_analysis)
else:
plotted_image = True
logger.info('%s :: plot already exists - %s' %
(function_str, str(graph_image_file)))
if plotted_image:
motif_analysis[metric]['motifs'][motif_id][
'image'] = graph_image_file
else:
logger.error('failed to plot motif match plot')
graph_image_file = None
except Exception as e:
logger.error(traceback.format_exc())
logger.error(
'error :: inference :: with fp id %s proceesing motif at index: %s - %s'
% (str(fp_id), str(motif[0]), str(e)))
continue
end_timer = timer()
motif_analysis[metric]['fps_checked'] = fps_checked_for_motifs
motif_analysis[metric]['exact_motifs'] = exact_motifs
motif_analysis[metric]['similar_motifs'] = similar_motifs
motif_analysis[metric]['distance_motifs'] = distance_motifs
motif_analysis[metric]['not_similar_motifs'] = not_similar_motifs
motif_analysis[metric][
'not_similar_enough_sample'] = not_similar_enough_sample
motif_analysis_file = '%s/motif.analysis.similarity_%s.batch_size_%s.top_matches_%s.max_distance_%s.dict' % (
motif_images_dir, similarity, str(batch_size), str(top_matches),
str(max_distance))
try:
write_data_to_file(skyline_app, motif_analysis_file, 'w',
str(motif_analysis))
except Exception as e:
trace = traceback.format_exc()
logger.error('%s' % trace)
fail_msg = '%s :: error :: failed to write motif_analysis_file - %s' % (
function_str, motif_analysis_file)
logger.error('%s' % fail_msg)
dev_null = e
motif_ids = list(motif_analysis[metric]['motifs'].keys())
logger.info(
'%s :: %s motif matches found, %s fps where checked and motifs plotted in %.6f seconds for %s'
% (function_str, str(len(motif_ids)), str(len(fps_checked_for_motifs)),
(end_timer - start_timer), metric))
if dev_null:
del dev_null
return motif_analysis, fail_msg, trace