Survey data analysis pattern in Data Analysis Python - Time & Space Complexity

Time Complexity: Survey data analysis pattern

O(m * n)

Understanding Time Complexity

When analyzing survey data, we often process many responses to find insights.

We want to know how the time to analyze grows as the number of survey responses increases.

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import pandas as pd

def analyze_survey(data):
    results = {}
    for question in data.columns:
        counts = data[question].value_counts()
        results[question] = counts
    return results

This code counts how many times each answer appears for every question in the survey data.

Identify Repeating Operations

• Primary operation: Looping over each question (column) and counting answers.
• How many times: Once for each question in the survey data.

How Execution Grows With Input

As the number of survey responses grows, counting answers for each question takes more time.

Input Size (n)	Approx. Operations
10 responses	Counts done on 10 answers per question
100 responses	Counts done on 100 answers per question
1000 responses	Counts done on 1000 answers per question

Pattern observation: The time grows roughly in direct proportion to the number of responses.

Final Time Complexity

Time Complexity: O(m * n)

This means the time grows with both the number of questions (m) and the number of responses (n).

Common Mistake

[X] Wrong: "Counting answers for all questions takes the same time no matter how many responses there are."

[OK] Correct: More responses mean more data to count, so the time increases with the number of responses.

Interview Connect

Understanding how data size affects analysis time helps you explain your approach clearly and shows you think about efficiency.

Self-Check

"What if we only analyzed a fixed number of questions regardless of total questions? How would the time complexity change?"

Practice

1. What does the value_counts() function do when analyzing survey data in Python?

easy

A. Calculates the average of numerical answers

B. Groups data by categories for comparison

C. Counts how many times each unique answer appears

D. Removes missing values from the dataset

Solution

1. Step 1: Understand the purpose of value_counts()

   This function counts the frequency of each unique value in a column, showing how many times each answer appears.

2. Step 2: Compare with other functions

   Calculating averages is done by mean(), grouping is done by groupby(), and removing missing values uses dropna().

3. Final Answer:

   Counts how many times each unique answer appears -> Option C

4. Quick Check:

   value_counts() = count unique answers [OK]

Hint: Remember: value_counts counts answers, not averages [OK]

Common Mistakes:

• Confusing value_counts with mean()
• Thinking value_counts groups data
• Assuming it removes missing data

2. Which of the following is the correct syntax to get the average rating from a survey column named rating in a pandas DataFrame df?

easy

A. df['rating'].mean()

B. df['rating'].count()

C. df.groupby('rating').sum()

D. df['rating'].value_counts()

Solution

1. Step 1: Identify the function for average calculation

   The mean() function calculates the average of numerical values in a column.

2. Step 2: Check other options

   count() counts non-missing values, groupby().sum() sums grouped data, and value_counts() counts unique values.

3. Final Answer:

   df['rating'].mean() -> Option A

4. Quick Check:

   Average = mean() [OK]

Hint: Use mean() to get average of numeric columns [OK]

Common Mistakes:

• Using count() instead of mean()
• Confusing value_counts() with average
• Trying to sum without grouping

3. Given the following code, what will be the output?

import pandas as pd
data = {'gender': ['M', 'F', 'F', 'M', 'F'], 'score': [5, 3, 4, 2, 5]}
df = pd.DataFrame(data)
result = df.groupby('gender')['score'].mean()
print(result)

medium

A. gender
F 3.5
M 4.0
Name: score, dtype: float64

B. gender
F 4.0
M 3.5
Name: score, dtype: float64

C. gender
M 4.0
F 3.5
Name: score, dtype: float64

D. gender
F 3.0
M 4.0
Name: score, dtype: float64

Solution

1. Step 1: Group scores by gender

   Group 'score' values by 'gender': For 'F' scores are [3,4,5], for 'M' scores are [5,2].

2. Step 2: Calculate mean for each group

   Mean for 'F' = (3+4+5)/3 = 4.0; Mean for 'M' = (5+2)/2 = 3.5.

3. Final Answer:

   gender F 4.0 M 3.5 Name: score, dtype: float64 -> Option B

4. Quick Check:

   Group mean = correct averages [OK]

Hint: Calculate group means by averaging grouped values [OK]

Common Mistakes:

• Mixing up group labels and values
• Calculating sums instead of means
• Ignoring group sizes

4. The following code is intended to count how many times each answer appears in the 'response' column, but it raises an error. What is the error and how to fix it?

import pandas as pd
df = pd.DataFrame({'response': ['Yes', 'No', 'Yes', 'Maybe']})
counts = df.response.value_count()
print(counts)

medium

A. KeyError because 'response' column does not exist

B. SyntaxError due to missing parentheses in print statement

C. TypeError because value_count() requires an argument

D. AttributeError because the method name should be value_counts()

Solution

1. Step 1: Identify the method name error

   The method value_count() is incorrect; the correct method is value_counts() with an 's' at the end.

2. Step 2: Correct the method name

   Changing value_count() to value_counts() fixes the AttributeError and allows counting unique values.

3. Final Answer:

   AttributeError because the method name should be value_counts() -> Option D

4. Quick Check:

   Correct method is value_counts() [OK]

Hint: Remember the method ends with 's': value_counts() [OK]

Common Mistakes:

• Misspelling method name without 's'
• Confusing print syntax errors
• Assuming column name is missing

5. You have a survey DataFrame df with columns age_group and satisfaction_score. How would you calculate the average satisfaction score for each age group and then find which age group has the highest average score?

hard

A. Use df.groupby('age_group')['satisfaction_score'].mean().idxmax()

B. Use df['satisfaction_score'].max() directly

C. Use df.value_counts('age_group') and pick max

D. Use df.groupby('satisfaction_score')['age_group'].mean()

Solution

1. Step 1: Group by age group and calculate average satisfaction

   Use groupby('age_group')['satisfaction_score'].mean() to get average scores per group.

2. Step 2: Find the age group with the highest average

   Use idxmax() on the resulting Series to get the group label with the highest average score.

3. Final Answer:

   Use df.groupby('age_group')['satisfaction_score'].mean().idxmax() -> Option A

4. Quick Check:

   Group mean + idxmax() finds highest average [OK]

Hint: Chain mean() and idxmax() to find top group quickly [OK]

Common Mistakes:

• Using max() on scores without grouping
• Confusing value_counts() with averages
• Grouping by wrong column