Web analytics data pattern in Data Analysis Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ data_output

intermediate

Identify the number of unique users per day

You have a web analytics dataset with columns user_id and date. Which code snippet correctly calculates the number of unique users for each day?

Data Analysis Python

import pandas as pd

data = pd.DataFrame({
    'user_id': [1, 2, 1, 3, 2, 4],
    'date': ['2024-06-01', '2024-06-01', '2024-06-02', '2024-06-02', '2024-06-02', '2024-06-03']
})

A data.groupby('date')['user_id'].nunique()

B data.groupby('user_id')['date'].nunique()

C data.groupby('date').count()['user_id']

D data['user_id'].unique().groupby(data['date'])

💡 Hint

Think about grouping by date and counting unique users, not total visits.

Grouping by date and applying nunique() on user_id counts unique users per day. Option A groups by user, not date. Option A is invalid syntax. Option A counts all visits, not unique users.

❓ Predict Output

intermediate

Output of session duration calculation

Given a DataFrame df with columns session_id and timestamp, what is the output of this code?

Data Analysis Python

import pandas as pd

df = pd.DataFrame({
    'session_id': [1, 1, 2, 2, 2],
    'timestamp': pd.to_datetime(['2024-06-01 10:00', '2024-06-01 10:05', '2024-06-01 11:00', '2024-06-01 11:10', '2024-06-01 11:15'])
})
session_duration = df.groupby('session_id')['timestamp'].agg(lambda x: (x.max() - x.min()).seconds)
print(session_duration.to_dict())

A {1: 5, 2: 15}

B {1: 300, 2: 900}

C {1: '00:05:00', 2: '00:15:00'}

D {1: 600, 2: 900}

💡 Hint

Remember that .seconds returns total seconds as an integer.

Session 1 duration is 5 minutes = 300 seconds. Session 2 duration is 15 minutes = 900 seconds. Option B shows minutes, not seconds. Option B shows strings, not integers. Option B has wrong value for session 1.

❓ visualization

advanced

Correct plot for hourly page views

You want to plot the number of page views per hour from a DataFrame df with a timestamp column. Which code produces a correct bar plot of hourly counts?

Data Analysis Python

import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({
    'timestamp': pd.to_datetime([
        '2024-06-01 10:15', '2024-06-01 10:45', '2024-06-01 11:00',
        '2024-06-01 11:30', '2024-06-01 12:00', '2024-06-01 12:15'
    ])
})

A

df['hour'] = df['timestamp'].dt.hour
counts = df.groupby('hour').size()
counts.plot(kind='bar')
plt.show()

B

df['timestamp'].hour
counts = df.groupby('hour').count()
counts.plot(kind='bar')
plt.show()

C

counts = df.groupby(df['timestamp'].dt.hour).count()
counts.plot(kind='line')
plt.show()

D

df['hour'] = df['timestamp'].dt.hour
counts = df.groupby('hour')['timestamp'].count()
counts.plot(kind='bar')
plt.show()

💡 Hint

Use dt.hour to extract hour and count rows per hour.

Option D correctly extracts hour, counts timestamps per hour, and plots a bar chart. Option D misses counting column and may plot size series but less explicit. Option D uses wrong attribute .hour on Series. Option D plots line instead of bar.

🔧 Debug

advanced

Error in calculating bounce rate

Given a DataFrame df with columns session_id and page_views, this code tries to calculate bounce rate (percentage of sessions with only 1 page view). What error or problem will occur?

Data Analysis Python

import pandas as pd

df = pd.DataFrame({
    'session_id': [1, 2, 3, 4],
    'page_views': [1, 3, 1, 2]
})
bounce_rate = df[df['page_views'] == 1].count() / df.count()
print(bounce_rate)

A Raises ZeroDivisionError due to division by zero

B Raises KeyError because 'page_views' is missing

C Returns a Series with ratio per column, not a single bounce rate value

D Returns a single float value representing bounce rate

💡 Hint

Check what count() returns on a DataFrame.

count() returns counts per column, so dividing two DataFrames returns a Series with ratios per column. This is not a single bounce rate value. No error occurs. Option C is incorrect because output is not a single float.

🧠 Conceptual

expert

Interpreting user session patterns from data

You analyze web sessions and find that the average session duration is very low, but the number of page views per session is high. What is the most likely explanation?

A Users quickly navigate many pages but spend little time on each, indicating possible confusion or fast scanning

B Users have very few sessions but each session lasts a long time

C Data collection has errors causing session durations to be overestimated

D Users spend a long time on the first page and then leave, causing low average duration

💡 Hint

Think about what many page views but short duration means for user behavior.

High page views with low session duration suggests users move quickly through pages, possibly scanning or confused. Option A contradicts high page views. Option A is about data errors, not behavior. Option A contradicts the given data.