drop_duplicates() for removal in Pandas - Time & Space Complexity

Time Complexity: drop_duplicates() for removal

O(n)

Understanding Time Complexity

When we remove duplicate rows from data using pandas, it is important to know how the time taken grows as the data gets bigger.

We want to understand how the work done changes when the number of rows increases.

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import pandas as pd

df = pd.DataFrame({
    'A': [1, 2, 2, 3, 3, 3],
    'B': ['x', 'y', 'y', 'z', 'z', 'z']
})

unique_df = df.drop_duplicates()

This code creates a small table and removes duplicate rows, keeping only unique rows.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Checking each row against others to find duplicates.
• How many times: Each row is hashed and checked once during the process.

How Execution Grows With Input

As the number of rows grows, the time to find duplicates grows roughly in a straight line.

Input Size (n)	Approx. Operations
10	About 10 checks
100	About 100 checks
1000	About 1000 checks

Pattern observation: Doubling the rows roughly doubles the work needed.

Final Time Complexity

Time Complexity: O(n)

This means the time to remove duplicates grows directly in proportion to the number of rows.

Common Mistake

[X] Wrong: "Removing duplicates takes time proportional to the square of the number of rows because every row is compared to every other row."

[OK] Correct: pandas uses efficient hashing and indexing internally, so it does not compare every row to every other row directly.

Interview Connect

Understanding how data operations scale helps you write code that works well even with large datasets, a key skill in data science roles.

Self-Check

"What if we used drop_duplicates() on only one column instead of all columns? How would the time complexity change?"

Practice

1. What does the drop_duplicates() function do in pandas?

easy

A. Removes repeated rows from a DataFrame

B. Adds new rows to a DataFrame

C. Sorts the DataFrame by column values

D. Changes the data type of columns

Solution

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

   This function is designed to remove rows that appear more than once in a DataFrame.

2. Step 2: Compare options with function behavior

   Only Removes repeated rows from a DataFrame correctly describes removing repeated rows. Other options describe different functions.

3. Final Answer:

   Removes repeated rows from a DataFrame -> Option A

4. Quick Check:

   drop_duplicates() removes repeats [OK]

Hint: Remember: drop_duplicates() cleans repeated rows fast [OK]

Common Mistakes:

• Confusing drop_duplicates() with sorting functions
• Thinking it adds or modifies data instead of removing duplicates
• Assuming it changes data types

2. Which of the following is the correct syntax to remove duplicate rows from a DataFrame df?

easy

A. df.drop_duplicate_rows()

B. df.remove_duplicates()

C. df.delete_duplicates()

D. df.drop_duplicates()

Solution

1. Step 1: Recall the exact pandas method name

   The correct method to remove duplicates is drop_duplicates().

2. Step 2: Check each option for correct syntax

   Only df.drop_duplicates() uses the correct method name and syntax. Others are invalid pandas methods.

3. Final Answer:

   df.drop_duplicates() -> Option D

4. Quick Check:

   Correct method name is drop_duplicates() [OK]

Hint: Use exact method name: drop_duplicates() [OK]

Common Mistakes:

• Using incorrect method names like remove_duplicates()
• Adding extra words like delete or rows
• Confusing with other DataFrame methods

3. Given the DataFrame df below:

import pandas as pd
data = {'A': [1, 2, 2, 3], 'B': ['x', 'y', 'y', 'z']}
df = pd.DataFrame(data)
print(df.drop_duplicates())

What will be the output?

medium

A. Rows with index 1, 2, and 3

B. Rows with index 0, 1, and 3

C. Rows with index 0, 2, and 3

D. All rows remain unchanged

Solution

1. Step 1: Identify duplicate rows in the DataFrame

   Rows 2 and 3 have the same values (A=2, B='y'). So one will be removed.

2. Step 2: Understand default behavior of drop_duplicates()

   By default, it keeps the first occurrence and removes later duplicates. So row 1 stays, row 2 is removed.

3. Final Answer:

   Rows with index 0, 1, and 3 -> Option B

4. Quick Check:

   First duplicates kept, later removed [OK]

Hint: drop_duplicates() keeps first by default [OK]

Common Mistakes:

• Assuming last duplicate is kept instead of first
• Thinking all duplicates are removed including first
• Confusing row indices after removal

4. What is wrong with this code snippet to remove duplicates from DataFrame df?

df.drop_duplicates(inplace=True)
print(df.drop_duplicates())

medium

A. Calling drop_duplicates() twice is redundant

B. Using inplace=True and then printing drop_duplicates() returns original data

C. drop_duplicates() does not have inplace parameter

D. print statement syntax is incorrect

Solution

1. Step 1: Understand inplace=True effect

   Using inplace=True modifies df directly by removing duplicates.

2. Step 2: Analyze print statement

   After the first call, df is already deduplicated. The second call to df.drop_duplicates() is redundant as it returns the same DataFrame since there are no duplicates left.

3. Final Answer:

   Calling drop_duplicates() twice is redundant -> Option A

4. Quick Check:

   Second call redundant after inplace [OK]

Hint: inplace=True changes df; extra drop_duplicates() returns copy [OK]

Common Mistakes:

• Thinking drop_duplicates() lacks inplace parameter
• Assuming print statement is wrong syntax
• Believing calling twice causes error

5. You have a DataFrame df with columns ['Name', 'Age', 'City']. You want to remove duplicates only based on Name and City, keeping the last occurrence. Which code achieves this?

hard

A. df.drop_duplicates(columns=['Name', 'City'], keep='last')

B. df.drop_duplicates(subset=['Name', 'City'], keep='first')

C. df.drop_duplicates(subset=['Name', 'City'], keep='last')

D. df.drop_duplicates(subset=['Age'], keep='last')

Solution

1. Step 1: Use subset parameter to specify columns

   To check duplicates only on 'Name' and 'City', use subset=['Name', 'City'].

2. Step 2: Set keep='last' to keep last occurrence

   keep='last' keeps the last duplicate row, removing earlier ones.

3. Final Answer:

   df.drop_duplicates(subset=['Name', 'City'], keep='last') -> Option C

4. Quick Check:

   subset selects columns; keep='last' keeps last duplicate [OK]

Hint: Use subset for columns, keep='last' to keep last duplicate [OK]

Common Mistakes:

• Using 'columns' instead of 'subset' parameter
• Forgetting to set keep='last' to keep last occurrence
• Selecting wrong columns for subset