Dropping missing values (dropna) in Data Analysis Python

Introduction

Sometimes data has empty spots called missing values. Dropping them helps clean the data so we can analyze it better.

When you want to remove rows with missing data before analysis.

When missing values in columns make calculations incorrect.

When preparing data for machine learning models that can't handle missing values.

When you want to quickly clean a dataset without filling missing spots.

When missing data is rare and dropping it won't lose much information.

Syntax

Data Analysis Python

DataFrame.dropna(axis=0, how='any', thresh=None, subset=None, inplace=False)

axis=0 drops rows; axis=1 drops columns.

how='any' drops if any missing; how='all' drops if all missing.

Examples

Drop all rows that have any missing values.

Data Analysis Python

df.dropna()

Drop columns that have any missing values.

Data Analysis Python

df.dropna(axis=1)

Drop rows only if all values are missing.

Data Analysis Python

df.dropna(how='all')

Drop rows where 'Age' or 'Salary' columns have missing values.

Data Analysis Python

df.dropna(subset=['Age', 'Salary'])

Sample Program

This code creates a table with some missing ages and salaries. Then it removes any row that has a missing value. Finally, it shows the cleaned table.

Data Analysis Python

import pandas as pd

# Create a sample DataFrame with missing values
data = {
    'Name': ['Alice', 'Bob', 'Charlie', 'David'],
    'Age': [25, None, 30, None],
    'Salary': [50000, 60000, None, 45000]
}
df = pd.DataFrame(data)

print('Original DataFrame:')
print(df)

# Drop rows with any missing values
clean_df = df.dropna()

print('\nDataFrame after dropping rows with missing values:')
print(clean_df)

Important Notes

Using inplace=True changes the original DataFrame without needing to assign it again.

Be careful dropping too many rows; you might lose important data.

You can use thresh to keep rows with a minimum number of non-missing values.

Summary

Use dropna() to remove missing data from your dataset.

You can drop rows or columns depending on your needs.

Dropping missing values helps keep your data clean for analysis.

Practice

1. What does the dropna() function do in a pandas DataFrame?

easy

A. Duplicates rows with missing values

B. Fills missing values with zeros

C. Sorts the DataFrame by index

D. Removes rows or columns with missing values

Solution

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

   The dropna() function is used to remove missing data (NaN values) from a DataFrame.

2. Step 2: Identify what is removed

   It removes rows or columns that contain missing values, depending on parameters.

3. Final Answer:

   Removes rows or columns with missing values -> Option D

4. Quick Check:

   dropna() removes missing data [OK]

Hint: Remember: dropna means drop missing data [OK]

Common Mistakes:

• Confusing dropna with fillna
• Thinking it sorts data
• Assuming it duplicates data

2. Which of the following is the correct syntax to drop rows with any missing values in a DataFrame df?

easy

A. df.dropna(inplace=False, axis=2)

B. df.dropna(axis=1)

C. df.dropna(axis=0)

D. df.dropna(axis='columns')

Solution

1. Step 1: Understand axis parameter in dropna()

   axis=0 means drop rows, axis=1 means drop columns.

2. Step 2: Identify correct axis for dropping rows

   To drop rows with missing values, use axis=0.

3. Final Answer:

   df.dropna(axis=0) -> Option C

4. Quick Check:

   axis=0 drops rows [OK]

Hint: axis=0 drops rows, axis=1 drops columns [OK]

Common Mistakes:

• Using axis=1 to drop rows
• Using invalid axis=2
• Confusing inplace parameter

3. Given the DataFrame df below, what will df.dropna() return?

import pandas as pd
data = {'A': [1, 2, None], 'B': [4, None, 6]}
df = pd.DataFrame(data)
print(df.dropna())

medium

A. Rows with index 0 only

B. Rows with index 0 and 1

C. Rows with index 1 and 2

D. Empty DataFrame

Solution

1. Step 1: Examine the DataFrame rows for missing values

   Row 0: A=1, B=4 (no missing)
   Row 1: A=2, B=NaN (missing in B)
   Row 2: A=NaN, B=6 (missing in A)

2. Step 2: Apply dropna() default behavior

   By default, dropna() drops rows with any missing values, so only row 0 remains.

3. Final Answer:

   Rows with index 0 only -> Option A

4. Quick Check:

   Only complete rows remain [OK]

Hint: dropna() keeps only rows without any NaN [OK]

Common Mistakes:

• Assuming it drops columns instead of rows
• Thinking it fills missing values
• Expecting multiple rows to remain

4. What is wrong with this code snippet if the goal is to drop columns with any missing values?

df.dropna(axis=2)

medium

A. axis=2 is invalid; axis should be 0 or 1

B. dropna cannot drop columns

C. Missing inplace=True to modify df

D. dropna only works on rows by default

Solution

1. Step 1: Check valid axis values for dropna

   Valid axis values are 0 (rows) and 1 (columns). axis=2 is invalid and causes an error.

2. Step 2: Understand error cause

   Using axis=2 will raise a ValueError because pandas does not support axis=2 for DataFrames.

3. Final Answer:

   axis=2 is invalid; axis should be 0 or 1 -> Option A

4. Quick Check:

   axis must be 0 or 1 [OK]

Hint: Only axis=0 or 1 allowed in dropna [OK]

Common Mistakes:

• Using axis=2 by mistake
• Forgetting axis parameter
• Confusing inplace with axis

5. You have a DataFrame df with some missing values. You want to drop only those rows where all columns are missing values. Which dropna() parameter should you use?

hard

A. dropna(axis=1)

B. dropna(how='all')

C. dropna(thresh=2)

D. dropna(how='any')

Solution

1. Step 1: Understand the how parameter in dropna()

   how='all' drops rows only if all values are missing.
   how='any' drops rows if any value is missing.

2. Step 2: Identify correct parameter for dropping rows with all missing

   Since the goal is to drop rows where all columns are missing, use how='all'.

3. Final Answer:

   dropna(how='all') -> Option B

4. Quick Check:

   how='all' drops rows with all NaN [OK]

Hint: Use how='all' to drop rows with all missing [OK]

Common Mistakes:

• Using how='any' which drops too many rows
• Confusing thresh with how
• Dropping columns instead of rows