Filling missing values with fillna() in Pandas - Time & Space Complexity

Time Complexity: Filling missing values with fillna()

O(n)

Understanding Time Complexity

We want to understand how the time needed to fill missing values changes as the data grows.

How does the work grow when we use fillna() on bigger data?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import pandas as pd

df = pd.DataFrame({
    'A': [1, None, 3, None, 5],
    'B': [None, 2, None, 4, 5]
})

filled_df = df.fillna(0)

This code replaces all missing values in the DataFrame with zero.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Checking each cell in the DataFrame to see if it is missing and replacing it if so.
• How many times: Once for every cell in the DataFrame (rows x columns).

How Execution Grows With Input

As the number of rows or columns grows, the number of cells to check grows too.

Input Size (rows x columns)	Approx. Operations
10 x 2 = 20	About 20 checks and replacements
100 x 2 = 200	About 200 checks and replacements
1000 x 2 = 2000	About 2000 checks and replacements

Pattern observation: The work grows directly with the number of cells; doubling the data doubles the work.

Final Time Complexity

Time Complexity: O(n)

This means the time to fill missing values grows linearly with the total number of cells in the DataFrame.

Common Mistake

[X] Wrong: "fillna() only checks missing values, so it runs faster than looking at every cell."

[OK] Correct: The method must look at every cell to know if it is missing or not, so it still touches all data points.

Interview Connect

Understanding how data size affects operations like filling missing values helps you write efficient data cleaning steps in real projects.

Self-Check

"What if we used fillna() only on one column instead of the whole DataFrame? How would the time complexity change?"

Practice

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

easy

A. It changes the data type of columns.

B. It removes all rows with missing values.

C. It sorts the DataFrame by column values.

D. It replaces missing values with specified values.

Solution

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

   The fillna() function is designed to handle missing data by replacing NaN values with a specified value or method.

2. Step 2: Compare options with fillna() behavior

   Options B, C, and D describe other DataFrame operations, but only It replaces missing values with specified values. correctly describes filling missing values.

3. Final Answer:

   It replaces missing values with specified values. -> Option D

4. Quick Check:

   fillna() = replace missing values [OK]

Hint: fillna() fills missing spots with your chosen value [OK]

Common Mistakes:

• Confusing fillna() with dropna() which removes rows
• Thinking fillna() sorts data
• Assuming fillna() changes data types

2. Which of the following is the correct syntax to fill missing values in a DataFrame df with zero using fillna()?

easy

A. df.fillna(na=0)

B. df.fillna = 0

C. df.fillna(0)

D. df.fillna(0, value=0)

Solution

1. Step 1: Recall fillna() method syntax

   The correct way to fill missing values is calling fillna() with the value as the first argument, like df.fillna(0).

2. Step 2: Evaluate each option

   df.fillna = 0 tries to assign to fillna which is incorrect. df.fillna(na=0) uses a wrong keyword argument. df.fillna(0, value=0) passes 0 both positionally (as value) and explicitly as value=0 keyword, causing TypeError: multiple values for argument 'value'. Only C is correct.

3. Final Answer:

   df.fillna(0) -> Option C

4. Quick Check:

   fillna(value) fills missing with value [OK]

Hint: Use fillna(value) to fill missing values simply [OK]

Common Mistakes:

• Assigning fillna instead of calling it as a method
• Using wrong argument names like na=0
• Confusing inplace parameter usage

3. Given the DataFrame:

import pandas as pd
df = pd.DataFrame({'A': [1, None, 3], 'B': [None, 2, None]})
filled = df.fillna({'A': 0, 'B': 5})
print(filled)

What is the output?

medium

A. A B
0 1 5
1 0 2
2 3 5

B. A B
0 1 NaN
1 0 2
2 3 NaN

C. A B
0 1 0
1 0 2
2 3 0

D. A B
0 0 5
1 0 2
2 0 5

Solution

1. Step 1: Understand fillna() with dict argument

   Using a dictionary with fillna() fills missing values in each column with the corresponding value from the dict. Here, column 'A' missing values become 0, and column 'B' missing values become 5.

2. Step 2: Apply fillna() to each column

   Original df has NaN at df.loc[1, 'A'] and df.loc[0, 'B'], df.loc[2, 'B']. After fillna, these become 0 for 'A' and 5 for 'B'. The rest of the values remain unchanged.

3. Final Answer:

   A B 0 1 5 1 0 2 2 3 5 -> Option A

4. Quick Check:

   fillna(dict) fills columns separately [OK]

Hint: Use dict in fillna() to fill columns differently [OK]

Common Mistakes:

• Expecting fillna() to fill all columns with same value
• Confusing NaN replacement with zero by default
• Misreading which values are replaced

4. What is wrong with this code snippet?

import pandas as pd
df = pd.DataFrame({'X': [None, 2, None]})
df.fillna(inplace=True, 0)
print(df)

medium

A. fillna() keyword arguments are incorrectly used.

B. The arguments to fillna() are in wrong order.

C. The value argument should come before inplace.

D. fillna() does not have an inplace parameter.

Solution

1. Step 1: Check fillna() parameter names and order

   In Python functions, all positional arguments must come before any keyword arguments.

2. Step 2: Identify incorrect argument order

   The code uses inplace=True (keyword) followed by 0 (positional for value), which raises SyntaxError: positional argument follows keyword argument. Correct is df.fillna(0, inplace=True).

3. Final Answer:

   The arguments to fillna() are in wrong order. -> Option B

4. Quick Check:

   Positional args before keywords [OK]

Hint: Pass value first, then inplace as keyword in fillna() [OK]

Common Mistakes:

• Swapping argument order without keywords
• Using inplace without value argument
• Confusing positional and keyword arguments

5. You have a DataFrame df with columns 'Score' and 'Grade'. 'Score' has missing values. You want to fill missing 'Score' values with the mean of 'Score' and then fill missing 'Grade' values with 'Unknown'. Which code correctly does this?

hard

A. df.fillna({'Score': df['Score'].mean(), 'Grade': 'Unknown'}, inplace=True)

B. df['Score'].fillna(df['Score'].mean())
df['Grade'].fillna('Unknown')

C. df.fillna({'Score': 'Unknown', 'Grade': df['Grade'].mean()}, inplace=True)

D. df.fillna(df.mean(), inplace=True)
df['Grade'].fillna('Unknown')

Solution

1. Step 1: Calculate mean and prepare fill dictionary

   We calculate the mean of 'Score' using df['Score'].mean() and want to fill missing 'Score' with this mean. For 'Grade', missing values should be filled with 'Unknown'.

2. Step 2: Use fillna() with a dictionary and inplace=True

   Using df.fillna({'Score': mean_value, 'Grade': 'Unknown'}, inplace=True) fills missing values in both columns correctly in one step.

3. Step 3: Evaluate other options

   df['Score'].fillna(df['Score'].mean()) df['Grade'].fillna('Unknown') does not assign back the filled columns, so missing values remain. df.fillna({'Score': 'Unknown', 'Grade': df['Grade'].mean()}, inplace=True) swaps fill values incorrectly. df.fillna(df.mean(), inplace=True) df['Grade'].fillna('Unknown') fills 'Score' with its column mean (assuming 'Score' numeric) but the second line returns a new Series without assignment or inplace=True, leaving 'Grade' unchanged.

4. Final Answer:

   df.fillna({'Score': df['Score'].mean(), 'Grade': 'Unknown'}, inplace=True) -> Option A

5. Quick Check:

   Use dict with fillna() and inplace=True for multiple columns [OK]

Hint: Fill multiple columns at once with dict and inplace=True [OK]

Common Mistakes:

• Not assigning back after fillna() without inplace
• Mixing fill values between columns
• Using fillna() separately without inplace or assignment