Utils.py 6.85 KB
   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
# -*- coding: utf-8 -*-
import sqlite3
import re
import math
import logging
import os

try:
import cPickle as pickle
except:
import pickle

from sqlite3 import Error

logger = logging.getLogger(__name__)


def create_sqlite_connection(db_file):
conn = None
try:
conn = sqlite3.connect(db_file)
except Error as e:
logger.exception("Can't connect to SQLite")
return conn


def select_model_attribute(conn, trainID, attribute):
time_seg = ""
cur = conn.cursor()
cur.execute(
"SELECT " + attribute + " FROM models WHERE model_id=?", (int(trainID),)
)
rows = cur.fetchone()
if rows != None:
if len(rows) > 0:
time_seg = rows[0]
return time_seg

def check_pid(pid):
""" Check For the existence of a unix pid. """
try:
os.kill(pid, 0)
except OSError:
return False
else:
return True


# def load_data(train_id, train_dir=DEFAULT_TRAIN_DIR, only_aem=False):
# filename = "train_" + str(train_id) + ".dat"
# train_path = os.path.join(train_dir, filename)
# infile = open(train_path, "rb")
# weak_bins_mapping = pickle.load(infile)
# profile_index = pickle.load(infile)
# mp = pickle.load(infile)
# train_dataset_values = np.array(pickle.load(infile))
# art_events_dataset = pickle.load(infile)
# processed_art_events_dataset = ""
# regr = ""
# feature_importance = ""
# if not only_aem:
# processed_art_events_dataset = pickle.load(infile)
# regr = pickle.load(infile)
# feature_importance = pickle.load(infile)
# infile.close()
# return (
# weak_bins_mapping,
# profile_index,
# mp,
# train_dataset_values,
# art_events_dataset,
# processed_art_events_dataset,
# regr,
# feature_importance,
# )


def create_table(conn, create_table_sql):
try:
c = conn.cursor()
c.execute(create_table_sql)
except Error as e:
print(e)


def sigmoid_risk(s, midpoint, t):
try:
return 1 / (1 + math.exp(s * (t - midpoint)))
except OverflowError as ex:
logger.warning(f"Timedelta to big: {t}. " f"Returning 0 risk")
return 0


def sigmoid_mins(v, s, midpoint, hours_before):
if v > 1:
logger.error(f"SIGMOID Risk can't be more than one {v}")
raise
elif v == 1:
return 0
elif v == 0:
return convert_hours_to_mins(strtime_to_hours(hours_before))
return (math.log(1 / v - 1) / s) + midpoint


def convert_hours_to_mins(hours):
return hours * 60


def strtime_to_hours(time_segments):
time_segments_digits = float(re.findall(r"\d+\.\d+|\d+", time_segments)[0])
time_segments_hours = 0.0
if "H" in time_segments:
time_segments_hours = time_segments_digits
elif "T" in time_segments or "min" in time_segments:
time_segments_hours = time_segments_digits / 60.0
elif "S" in time_segments:
time_segments_hours = time_segments_digits / 60.0 / 60.0
return time_segments_hours


def inBufferWindow(actual_hours, buffer_time):
"""
Converts the buffer_size, which is in hours, into segments and checks
whether t is higher than ep_length-buffer

Parameters
----------
actual_hours : int
the current segment
buffer_time : str
the size of the buffer window in hours

Returns
-------
Returns true if t inside the buffer window

"""
if buffer_time is not None:
buffer_time_mins = convert_hours_to_mins(strtime_to_hours(buffer_time))
return strtime_to_hours(convert_hours_to_mins(actual_hours)) <= buffer_time_mins
return False

time_segments_hours = time_segments_to_hours(chunks_type)
last_segments_count = math.ceil(buffer_time / time_segments_hours)
return ep_length - last_segments_count <= t


def time_segments_to_hours(time_segments):
time_segments_digits = float(re.findall(r"\d+", time_segments)[0])
time_segments_hours = 0.0
if "H" in time_segments:
time_segments_hours = time_segments_digits
elif "T" in time_segments or "min" in time_segments:
time_segments_hours = time_segments_digits / 60.0
elif "S" in time_segments:
time_segments_hours = time_segments_digits / 60.0 / 60.0
return time_segments_hours


def fetch_influxdb_data(
inf_client, start_date, end_date, dbname, measurement, multidimensional_data
):
dataset_dates = []
dataset_values = []

if multidimensional_data:
query = (
"SELECT * FROM "
+ measurement
+ " WHERE "
+ "time>='"
+ start_date
+ "' AND time <= '"
+ end_date
+ "'"
)
result = inf_client.query(query, database=dbname)

if len(result.items()) > 0:
for c in result.items()[0][1]:
ae = [0] * (len(c.keys()) - 1)
for key in c.keys():
if key == "time":
time = c["time"]
else:
ae[int(key)] = c[key]
if len(ae) == 1 or sum(ae) == 0:
continue
dataset_dates.append(time)
dataset_values.append(ae)
else:
query = (
"SELECT * FROM "
+ measurement
+ " WHERE "
+ "time>='"
+ start_date
+ "' AND time <= '"
+ end_date
+ "'"
)
result = inf_client.query(query, database=dbname)

if len(result.items()) > 0:
for c in result.items()[0][1]:
ae = [0] * (len(c.keys()) - 1)
for key in c.keys():
if key == "time":
time = c["time"]
else:
ae[int(re.findall(r"\d+\.\d+|\d+", key)[0])] = c[key]
if len(ae) == 1 or sum(ae) == 0:
continue
dataset_dates.append(time)
dataset_values.append(max(ae))
return dataset_dates, dataset_values


def compute_thresholds(
warning_hours_thres, prediction_threshold, s, midpoint, hours_before, buffer_time
):
buffer_time_mins = 0
if buffer_time is not None:
buffer_time_mins = convert_hours_to_mins(strtime_to_hours(buffer_time))
if warning_hours_thres is not None:
prediction_threshold = sigmoid_risk(
s,
convert_hours_to_mins(strtime_to_hours(midpoint)),
convert_hours_to_mins(strtime_to_hours(warning_hours_thres))
- buffer_time_mins,
)
else:
warning_hours_thres = (
sigmoid_mins(
prediction_threshold,
s,
convert_hours_to_mins(strtime_to_hours(midpoint)),
hours_before,
)
/ 60.0
)
return warning_hours_thres, prediction_threshold