Python - How to fill NAN values with mean in Pandas?

When working with datasets in Pandas, missing values (NaN) are common. You can fill these NaN values with the mean of the column using mean() and fillna() functions.

Creating a DataFrame with NaN Values

Let us first import the required libraries and create a sample DataFrame ?

import pandas as pd
import numpy as np

# Create DataFrame with NaN values
dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Lexus', 'Mustang', 'Bentley', 'Mustang'],
    "Units": [100, 150, np.NaN, 80, np.NaN, np.NaN]
})

print("Original DataFrame:")
print(dataFrame)

Original DataFrame:
       Car  Units
0      BMW  100.0
1    Lexus  150.0
2    Lexus    NaN
3  Mustang   80.0
4  Bentley    NaN
5  Mustang    NaN

Calculating the Mean

Calculate the mean of the column containing NaN values. The mean() function automatically ignores NaN values ?

import pandas as pd
import numpy as np

dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Lexus', 'Mustang', 'Bentley', 'Mustang'],
    "Units": [100, 150, np.NaN, 80, np.NaN, np.NaN]
})

# Calculate mean (ignores NaN values automatically)
meanVal = dataFrame['Units'].mean()
print(f"Mean of Units column: {meanVal}")

Mean of Units column: 110.0

Filling NaN Values with Mean

Use fillna() to replace all NaN values in the column with the calculated mean ?

import pandas as pd
import numpy as np

# Create DataFrame
dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Lexus', 'Mustang', 'Bentley', 'Mustang'],
    "Units": [100, 150, np.NaN, 80, np.NaN, np.NaN]
})

print("Original DataFrame:")
print(dataFrame)

# Calculate mean and fill NaN values
meanVal = dataFrame['Units'].mean()
dataFrame['Units'].fillna(value=meanVal, inplace=True)

print(f"\nDataFrame after filling NaN with mean ({meanVal}):")
print(dataFrame)

Original DataFrame:
       Car  Units
0      BMW  100.0
1    Lexus  150.0
2    Lexus    NaN
3  Mustang   80.0
4  Bentley    NaN
5  Mustang    NaN

DataFrame after filling NaN with mean (110.0):
       Car  Units
0      BMW  100.0
1    Lexus  150.0
2    Lexus  110.0
3  Mustang   80.0
4  Bentley  110.0
5  Mustang  110.0

Alternative Methods

You can also fill NaN values directly without storing the mean in a variable ?

import pandas as pd
import numpy as np

dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Lexus', 'Mustang', 'Bentley', 'Mustang'],
    "Units": [100, 150, np.NaN, 80, np.NaN, np.NaN]
})

# Direct method - fill NaN with mean in one line
dataFrame['Units'] = dataFrame['Units'].fillna(dataFrame['Units'].mean())
print(dataFrame)

       Car  Units
0      BMW  100.0
1    Lexus  150.0
2    Lexus  110.0
3  Mustang   80.0
4  Bentley  110.0
5  Mustang  110.0

Conclusion

Use fillna(dataFrame['column'].mean()) to replace NaN values with the column mean. The mean() function automatically excludes NaN values from the calculation, making this approach reliable for data cleaning.