Missing data strategies decision in Pandas - Step-by-Step Execution

Concept Flow - Missing data strategies decision

Start with DataFrame

↓

Check for missing values

↓

Decide strategy

↓

Drop

↓

Resulting DataFrame

↓

End

Start with a DataFrame, check for missing values, decide to drop, fill, or leave them, then get the final DataFrame.

Execution Sample

Pandas

import pandas as pd

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

# Strategy: fill missing with 0
df_filled = df.fillna(0)
print(df_filled)

This code creates a DataFrame with missing values and fills them with zero.

Execution Table

Step	Action	DataFrame State	Missing Count	Result
1	Create DataFrame	{'A': [1, None, 3], 'B': [4, 5, None]}	2	Initial data with 2 missing values
2	Check missing	Same as step 1	2	Found 2 missing values
3	Fill missing with 0	{'A': [1, 0, 3], 'B': [4, 5, 0]}	0	Missing values replaced by 0
4	Print filled DataFrame	Same as step 3	0	Output shows no missing values
5	End	Final DataFrame	0	Process complete

💡 All missing values handled by filling with zero, no missing values remain.

Variable Tracker

Variable	Start	After fillna	Final
df	{'A': [1, None, 3], 'B': [4, 5, None]}	Same	Same
df_filled	N/A	{'A': [1, 0, 3], 'B': [4, 5, 0]}	Same

Key Moments - 2 Insights

Why do missing values still appear in 'df' after filling in 'df_filled'?

Because 'fillna' returns a new DataFrame and does not change 'df' unless assigned back. See execution_table step 3 where 'df_filled' is created separately.

What happens if we drop missing values instead of filling?

Rows with missing values are removed, reducing data size. This is a different strategy not shown here but would change DataFrame shape.

Visual Quiz - 3 Questions

Look at the execution table, how many missing values are in the DataFrame after step 3?

A 0

B 1

C 2

D 3

💡 Hint

Check the 'Missing Count' column at step 3 in the execution_table.

According to variable_tracker, what is the value of 'df_filled' after fillna?

A {'A': [1, None, 3], 'B': [4, 5, None]}

B {'A': [1, 0, 3], 'B': [4, 5, 0]}

C N/A

D Empty DataFrame

💡 Hint

Look at the 'After fillna' column for 'df_filled' in variable_tracker.

If we changed the strategy to drop missing values, what would happen to the DataFrame size?

A It would increase

B It would stay the same

C It would decrease

D It would become empty

💡 Hint

Dropping missing values removes rows with missing data, reducing size.

Concept Snapshot

Missing data strategies:
- Check missing values with isna()
- Decide: drop rows, fill values, or leave
- fillna() fills missing values
- dropna() removes missing rows
- Choose strategy based on data and goal

Full Transcript

We start with a DataFrame that has missing values. We check how many missing values exist. Then we decide how to handle them. One way is to fill missing values with a number like zero. This creates a new DataFrame with no missing values. The original DataFrame stays the same unless we overwrite it. Another way is to drop rows with missing values, which reduces the data size. Choosing the right strategy depends on the data and what you want to do next.