In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
In [2]:
internet_usage = pd.read_csv('C:/Users/gsaladi/OneDrive - FactSet/Desktop/jntuh/ML DL Assignment 1/internet_session.csv', parse_dates=['start_time'])
In [3]:
internet_usage
Out[3]:
name start_time usage_time IP MAC upload download total_transfer seession_break_reason
0 user1 2022-05-10 02:59:32 00:00:36:28 10.55.14.222 48:E7:DA:58:22:E9 15861.76 333168.64 349030.40 Idle-Timeout
1 user1 2022-05-10 18:53:27 00:01:49:56 10.55.2.253 48:E7:DA:58:22:E9 16957.44 212152.32 229109.76 Idle-Timeout
2 user1 2022-05-10 21:20:44 00:01:35:00 10.55.2.253 48:E7:DA:58:22:E9 14080.0 195153.92 209233.92 Idle-Timeout
3 user1 2022-05-11 00:37:42 00:00:26:00 10.55.2.253 48:E7:DA:58:22:E9 5242.88 40806.4 46049.28 Idle-Timeout
4 user1 2022-05-11 02:59:38 00:00:11:52 10.55.2.253 48:E7:DA:58:22:E9 22067.2 10772.48 32839.68 Idle-Timeout
... ... ... ... ... ... ... ... ... ...
4707 user9 2022-11-04 01:11:34 00:06:54:32 10.55.4.189 DA:2F:97:0E:B7:D0 107960.32 2390753.28 2495610.88 Idle-Timeout
4708 user9 2022-11-04 10:26:09 00:00:23:49 10.55.4.59 DA:2F:97:0E:B7:D0 11407.36 209674.24 221081.60 Idle-Timeout
4709 user9 2022-11-04 20:41:42 00:01:24:13 10.55.15.186 DA:2F:97:0E:B7:D0 18995.2 373657.6 392652.80 Idle-Timeout
4710 user9 2022-11-05 00:21:06 00:08:49:43 10.55.4.159 DA:2F:97:0E:B7:D0 46602.24 593766.4 640368.64 Idle-Timeout
4711 user9 2022-11-05 20:55:37 00:01:06:20 10.55.2.33 DA:2F:97:0E:B7:D0 21237.76 298536.96 319774.72 NaN

4712 rows × 9 columns

In [4]:
internet_usage.columns
Out[4]:
Index(['name', 'start_time', 'usage_time', 'IP', 'MAC', 'upload', 'download',
       'total_transfer', 'seession_break_reason'],
      dtype='object')
In [5]:
internet_usage.isna().sum()
Out[5]:
name                     0
start_time               0
usage_time               0
IP                       0
MAC                      0
upload                   0
download                 0
total_transfer           0
seession_break_reason    9
dtype: int64
In [6]:
internet_usage = internet_usage.dropna().copy()
internet_usage.isna().sum()
Out[6]:
name                     0
start_time               0
usage_time               0
IP                       0
MAC                      0
upload                   0
download                 0
total_transfer           0
seession_break_reason    0
dtype: int64
In [7]:
device = []
basename = 'device'
mac = internet_usage['MAC'][0]
device_number = 1
for i in internet_usage['MAC']:
    if i == mac:
        device.append(basename + str(device_number))
    else:
        device_number += 1
        device.append(basename + str(device_number))
        mac = i
internet_usage['device'] = device

internet_usage.dtypes
Out[7]:
name                             object
start_time               datetime64[ns]
usage_time                       object
IP                               object
MAC                              object
upload                           object
download                         object
total_transfer                  float64
seession_break_reason            object
device                           object
dtype: object

What is the most frequent internet activity time of the day ?¶

In [8]:
internet_usage['hour'] = pd.to_datetime(internet_usage['start_time']).dt.hour
frequent_activity_time_of_day = internet_usage['hour'].value_counts().sort_index()
plt.figure(figsize=(18, 9))
sns.lineplot(data=frequent_activity_time_of_day)
plt.xticks(np.linspace(start=0, stop=24, num=25))
plt.show()
plt.clf()

print ("Most frequent activity time of day is" , internet_usage.usage_time.max())

Most frequent activity time of day is 01:00:21:07

<Figure size 432x288 with 0 Axes>

How often the ip changes ?¶

In [9]:
base_ip = '48:E7:DA:58:22:E9'
ip_count = 0
for i in range(1, internet_usage.shape[0]):
    if internet_usage.iloc[i]['IP'] != base_ip:
        ip_count +=1
        base_ip = internet_usage.iloc[i]['IP']

print('The IP Adress changed ' + str(ip_count) + ' times') 

The IP Adress changed 2303 times

How often the device changed?¶

In [10]:
base_device = 'device1'
device_count = 0
for i in range(1, internet_usage.shape[0]):
    if internet_usage.iloc[i]['device'] != base_device:
        device_count +=1
        base_device = internet_usage.iloc[i]['device']

print('The device changed ' + str(device_count) + ' times')

The device changed 1223 times

What is the average usage per hour , per day and per month ?¶

In [18]:
internet_usage['day'] = internet_usage['start_time'].dt.day
internet_usage['month'] = internet_usage['start_time'].dt.month
hourly_average = internet_usage.groupby('hour').total_transfer.mean()
print('The Average usage per hour is:\n ' +  str(round(hourly_average, 2)))

The Average usage per hour is:
 hour
0     464530.44
1     530880.86
2     431576.11
3     345303.34
4     359809.44
5     275960.91
6     468959.59
7     292886.83
8     366681.92
9     377480.64
10    393259.12
11    309492.45
12    310137.98
13    335270.58
14    472403.71
15    517005.11
16    403919.40
17    525423.69
18    666590.76
19    389841.79
20    355862.80
21    474038.34
22    449600.50
23    407785.08
Name: total_transfer, dtype: float64
In [13]:
plt.figure(figsize=(15, 7))
sns.barplot(x='hour', y='total_transfer' , data=internet_usage, ci=None, estimator=np.mean)
plt.title("Average usage per hour")
plt.show()
plt.clf()

<Figure size 432x288 with 0 Axes>
In [14]:
daily_average = internet_usage.groupby('day').total_transfer.mean()
print('The Average usage per day is:\n ' +  str(round(daily_average, 2)))

The Average usage per day is:
 day
1     396705.04
2     494496.48
3     445865.63
4     676332.03
5     652195.66
6     396261.75
7     402259.89
8     301859.57
9     393521.97
10    350665.02
11    729857.65
12    346695.95
13    501906.70
14    352701.10
15    521520.51
16    426719.39
17    475795.71
18    337490.93
19    301941.32
20    365130.12
21    462211.69
22    486595.37
23    383153.93
24    320598.94
25    443689.47
26    463432.02
27    324318.12
28    494576.34
29    363645.61
30    361418.88
31    369118.01
Name: total_transfer, dtype: float64
In [15]:
plt.figure(figsize=(15, 7))
sns.barplot(x='day', y='total_transfer' , data=internet_usage, ci=None, estimator=np.mean)
plt.title("Average usage per day")
plt.show()
plt.clf()

<Figure size 432x288 with 0 Axes>
In [19]:
monthly_average = internet_usage.groupby('month').total_transfer.mean()
print('The Average usage per month is:\n ' +  str(round(monthly_average, 2)))

The Average usage per month is:
 month
5     311177.16
6     338418.08
7     418583.99
8     479042.44
9     482955.52
10    549467.63
11    399804.11
Name: total_transfer, dtype: float64
In [17]:
plt.figure(figsize=(15, 7))
sns.barplot(x='month', y='total_transfer' , data=internet_usage, ci=None, estimator=np.mean)
plt.title("Average usage per month")
plt.show()
plt.clf()

<Figure size 432x288 with 0 Axes>
In [ ]: