Pandas - Standardizing a DataFrame using a custom function and applymap()

Last update on April 09 2025 12:53:03 (UTC/GMT +8 hours)

20. Standardize DataFrame Using applymap()

Write a Pandas program that uses applymap() to Standardize a DataFrame.

In this exercise, we have applied a custom function to standardize all numeric values in a DataFrame using applymap().

Sample Solution :

Code :

import pandas as pd

# Create a sample DataFrame
df = pd.DataFrame({
    'A': [10, 20, 30],
    'B': [5, 15, 25]
})

# Define a custom function to standardize each element (z-score)
def standardize(x, mean, std):
    return (x - mean) / std if std != 0 else 0  # Avoid division by zero

# Apply standardization column-wise
df_standardized = df.apply(lambda col: col.apply(standardize, args=(col.mean(), col.std())))

# Output the result
print(df_standardized)

Output:

     A    B
0 -1.0 -1.0
1  0.0  0.0
2  1.0  1.0                    

Explanation:

• Column-wise standardization: The function calculates the z-score based on the mean and standard deviation of each column, which is the correct approach for standardizing data.
• Handling division by zero: The condition if std != 0 else 0 ensures that if a column has no variance (standard deviation is zero), the function returns 0 instead of dividing by zero.
• Column-wise apply(): apply() is used to process each column independently and apply the standardization function element-wise.

For more Practice: Solve these Related Problems:

• Write a Pandas program to standardize all numeric values in a DataFrame using applymap() with a custom z-score function.
• Write a Pandas program to apply a custom standardization function to each cell of a DataFrame using applymap() and then display the transformed DataFrame.
• Write a Pandas program to use applymap() to normalize the entire DataFrame so that each cell represents its deviation from the mean.
• Write a Pandas program to benchmark the performance of applymap() for standardizing a DataFrame compared to a vectorized approach.

Python-Pandas Code Editor:

Have another way to solve this solution? Contribute your code (and comments) through Disqus.