Forward fill and backward fill in Data Analysis Python

Introduction

Sometimes data has missing values. Forward fill and backward fill help fill these gaps using nearby known values.

You have a time series with missing sensor readings and want to fill gaps using the last known value.

You want to fill missing daily sales data by carrying forward the previous day's sales.

You have survey data with missing answers and want to fill them using the next available answer.

You want to prepare data for analysis by replacing missing values with nearby known values.

You want to smooth data by filling missing spots without guessing new values.

Syntax

Data Analysis Python

DataFrame.fillna(method='ffill')  # forward fill
DataFrame.fillna(method='bfill')  # backward fill

ffill means fill missing values using the last known value above (forward fill).

bfill means fill missing values using the next known value below (backward fill).

Examples

Fill missing values by copying the last known value forward.

Data Analysis Python

df.fillna(method='ffill')

Fill missing values by copying the next known value backward.

Data Analysis Python

df.fillna(method='bfill')

Forward fill but only fill up to 1 consecutive missing value.

Data Analysis Python

df.fillna(method='ffill', limit=1)

Sample Program

This code creates a small table with missing temperatures. It shows how forward fill and backward fill replace missing values using nearby known temperatures.

Data Analysis Python

import pandas as pd

# Create example data with missing values
data = {'Day': [1, 2, 3, 4, 5],
        'Temperature': [20.0, None, None, 22.0, None]}
df = pd.DataFrame(data)

# Forward fill missing values
ffill_df = df.fillna(method='ffill')

# Backward fill missing values
bfill_df = df.fillna(method='bfill')

print('Original DataFrame:')
print(df)
print('\nAfter Forward Fill:')
print(ffill_df)
print('\nAfter Backward Fill:')
print(bfill_df)

Important Notes

Forward fill works well when missing data should carry the last known value forward in time.

Backward fill is useful when you want to fill missing data using the next known value.

Be careful: filling too many missing values blindly can hide real gaps or errors.

Summary

Forward fill copies the last known value forward to fill missing spots.

Backward fill copies the next known value backward to fill missing spots.

Both help prepare data for analysis by handling missing values simply.

Practice

1. What does forward fill do when handling missing data in a dataset?

easy

A. It removes rows with missing values.

B. It copies the next known value backward to fill missing spots.

C. It replaces missing values with zero.

D. It copies the last known value forward to fill missing spots.

Solution

1. Step 1: Understand forward fill concept

   Forward fill means filling missing data by copying the last known value forward to the next missing spot.

2. Step 2: Compare options with definition

   "It copies the last known value forward to fill missing spots." matches this definition exactly, while others describe different methods.

3. Final Answer:

   It copies the last known value forward to fill missing spots. -> Option D

4. Quick Check:

   Forward fill = copy last known value forward [OK]

Hint: Forward fill copies previous value forward to fill gaps [OK]

Common Mistakes:

• Confusing forward fill with backward fill
• Thinking forward fill removes missing data
• Assuming forward fill replaces with zero

2. Which of the following is the correct syntax to apply forward fill on a pandas DataFrame df?

easy

A. df.fillna(method='ffill')

B. df.fillna(method='bfill')

C. df.fillna(forward=True)

D. df.fillna(direction='forward')

Solution

1. Step 1: Recall pandas fillna syntax

   In pandas, forward fill is done by fillna(method='ffill').

2. Step 2: Check options for correct syntax

   df.fillna(method='ffill') matches the correct method name 'ffill'. df.fillna(method='bfill') performs backward fill. The remaining options use invalid syntax.

3. Final Answer:

   df.fillna(method='ffill') -> Option A

4. Quick Check:

   Forward fill method = 'ffill' [OK]

Hint: Use method='ffill' for forward fill in pandas [OK]

Common Mistakes:

• Using 'bfill' instead of 'ffill' for forward fill
• Passing invalid parameters like forward=True
• Confusing method names in fillna

3. Given the DataFrame df below, what will be the result of df.bfill()?

import pandas as pd
df = pd.DataFrame({'A': [1, None, None, 4, None]})
print(df.bfill())

medium

A. A
0 1.0
1 4.0
2 4.0
3 4.0
4 NaN

B. A
0 1.0
1 1.0
2 1.0
3 4.0
4 NaN

C. A
0 1.0
1 NaN
2 NaN
3 4.0
4 NaN

D. A
0 1.0
1 4.0
2 4.0
3 4.0
4 4.0

Solution

1. Step 1: Understand backward fill on given data

   Backward fill copies the next known value backward to fill missing spots. Here, missing values at index 1 and 2 get filled with 4.0 from index 3. The last missing at index 4 has no next value, so remains NaN.

2. Step 2: Match output with options

   A 0 1.0 1 4.0 2 4.0 3 4.0 4 NaN matches this result exactly. A 0 1.0 1 4.0 2 4.0 3 4.0 4 4.0 incorrectly fills last NaN with 4.0, others do not match.

3. Final Answer:

   A 0 1.0 1 4.0 2 4.0 3 4.0 4 NaN -> Option A

4. Quick Check:

   Backward fill fills missing with next known value [OK]

Hint: Backward fill uses next known value to fill missing [OK]

Common Mistakes:

• Assuming last NaN gets filled when no next value exists
• Confusing forward and backward fill results
• Misreading index positions in output

4. You run the code below but get an error. What is the mistake?

import pandas as pd
df = pd.DataFrame({'A': [None, 2, None]})
df.fillna(method='forward')

medium

A. fillna requires inplace=True to work.

B. DataFrame cannot have None values.

C. The method name 'forward' is invalid; should be 'ffill'.

D. fillna does not accept method parameter.

Solution

1. Step 1: Check fillna method parameter

   The valid method names are 'ffill' for forward fill and 'bfill' for backward fill. 'forward' is not valid.

2. Step 2: Identify cause of error

   Using 'forward' causes a ValueError. DataFrames can have None values, fillna does not require inplace=True, and it accepts the method parameter.

3. Final Answer:

   The method name 'forward' is invalid; should be 'ffill'. -> Option C

4. Quick Check:

   Use 'ffill' for forward fill method [OK]

Hint: Use 'ffill' not 'forward' as method name in fillna [OK]

Common Mistakes:

• Using 'forward' instead of 'ffill'
• Thinking inplace=True is mandatory
• Believing fillna can't handle None values

5. You have a DataFrame df with missing values scattered in column 'B'. You want to fill missing values by first forward filling, then backward filling any remaining missing values. Which code correctly does this?

hard

A. df['B'].fillna(method='ffill', method='bfill')

B. df['B'].fillna(method='ffill').fillna(method='bfill')

C. df['B'].fillna(method='bfill').fillna(method='ffill')

D. df['B'].fillna(method=['ffill', 'bfill'])

Solution

1. Step 1: Understand chaining fillna calls

   To fill missing values first forward then backward, chain fillna calls: first with method='ffill', then on result with method='bfill'.

2. Step 2: Evaluate options for correct chaining

   df['B'].fillna(method='ffill').fillna(method='bfill') correctly chains fillna twice. df['B'].fillna(method='bfill').fillna(method='ffill') reverses order. Attempting multiple methods in one call (duplicate kwargs or list) uses invalid syntax.

3. Final Answer:

   df['B'].fillna(method='ffill').fillna(method='bfill') -> Option B

4. Quick Check:

   Chain fillna with 'ffill' then 'bfill' [OK]

Hint: Chain fillna with 'ffill' then 'bfill' to fill all gaps [OK]

Common Mistakes:

• Trying to pass multiple methods in one fillna call
• Reversing forward and backward fill order
• Not chaining calls, so some NaNs remain