anatolykopyl/DINS-test-assignment
1
0
Fork 0
mirror of https://github.com/anatolykopyl/DINS-test-assignment.git synced 2026-03-26 12:55:05 +00:00
Code Issues Projects Releases Wiki Activity

Commented and removed some useless code

Browse Source
This commit is contained in:
Anatoly Kopyl
2017-11-12 23:18:05 +03:00
parent 5d2daba08c
commit 25a437e283
2 changed files with 12 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
anomaly.py
View File

@@ -2,10 +2,8 @@ import numpy as np
def is_anomaly(result): def is_anomaly(result):
data = [item[1] for name in result for item in result[name]] data = np.array([item[1] for name in result for item in result[name]])
x = np.array(data)
for name in result: for name in result:
for item in result[name]: for item in result[name]:
item[2] = item[1] > x.mean() + 3 * x.std() item[2] = item[1] > data.mean() + 3 * data.std()

22
main.py
View File

@@ -5,33 +5,31 @@ import MySQLdb
from anomaly import is_anomaly from anomaly import is_anomaly
TIME_SPAN = 900 TIME_SPAN = 900 # 15 minutes
result = {} result = {}
with open('raw_data.csv', 'rt', encoding="UTF-8") as csvfile: with open('raw_data2.csv', 'rt', encoding="UTF-8") as csvfile:
reader = csv.reader(csvfile, quotechar='"') reader = csv.reader(csvfile, quotechar='"')
for row in reader: for row in reader:
if row[0] != "ts": if row[0] != "ts":
timestamp = time.mktime(datetime.datetime.strptime(row[0], "%Y-%m-%d %H:%M:%S,%f").timetuple()) timestamp = time.mktime(datetime.datetime.strptime(row[0], "%Y-%m-%d %H:%M:%S,%f").timetuple()) # generating a unix timestamp
name = row[1] + "*" + row[2] name = row[1] + "*" + row[2] # combining api_name and http_method into a pair name
if 'start_time' not in locals(): if 'start_time' not in locals():
start_time = timestamp start_time = timestamp
start_time = int(min(start_time, timestamp)) start_time = int(min(start_time, timestamp)) # earliest request
if 'end_time' not in locals(): if 'end_time' not in locals():
end_time = timestamp end_time = timestamp
end_time = int(max(end_time, timestamp)) end_time = int(max(end_time, timestamp)) # latest request
if name not in result: if name not in result:
result[name] = [] result[name] = []
with open('raw_data.csv', 'rt', encoding="UTF-8") as csvfile:
reader = csv.reader(csvfile, quotechar='"')
name = row[1] + "*" + row[2]
for t in range(start_time, end_time, TIME_SPAN): for t in range(start_time, end_time, TIME_SPAN):
result[name].append([t, 0, 0]) result[name].append([t, 0, 0]) # [timestamp, num of requests, is anomaly]
csvfile.seek(0)
for row in reader: for row in reader:
if (row[3])[:1] == "5": if (row[3])[:1] == "5":