What is Pandas fillna() and how to use it

The Python pandas DataFrame.fillna() function is used to replace missing values in a DataFrame. This can help to simplify data cleaning processes or be a useful tool when performing analyses.

What is the syntax for pandas fillna()?

The fillna() function accepts up to five parameters and is structured as follows:

DataFrame.fillna(value=None, method=None, axis=None, inplace=False, limit=None)

Important parameters for fillna()

The behaviour of the DataFrame.fillna() function can be adjusted using various parameters:

How to use Pandas DataFrame.fillna()

The Pandas fillna() function can be used in several different ways:

Replacing NaN values with a fixed value

First, let’s create a DataFrame:

import pandas as pd
# Sample DataFrame with different values
data = {
    'A': [1, 2, None, 4],
    'B': [None, 2, 3, 4],
    'C': [1, None, 3, 4]
}
df = pd.DataFrame(data)
print(df)

The DataFrame looks like this:

A    B    C
0  1.0  NaN  1.0
1  2.0  2.0  NaN
2  NaN  3.0  3.0
3  4.0  4.0  4.0

To replace the missing values with 0, you can use the pandas fillna() function:

# Replacing missing values with zero
df_filled = df.fillna(0)
print(df_filled)

The result is that every NaN value has been replaced with 0:

A    B    C
0  1.0  0.0  1.0
1  2.0  2.0  0.0
2  0.0  3.0  3.0
3  4.0  4.0  4.0

Using the forward filling method ffill

If you want to fill NaN values with the value that directly precedes them in the column where they are located, you can pass the ffill method as a parameter:

# Replace all NaN values with the value that precedes them
df_ffill = df.fillna(method='ffill')
print(df_ffill)

In this example, the NaN values in columns A and C have been filled with the preceding values in the same column. Since there was no preceding value in column B for row 0, the NaN value is retained:

A    B    C
0  1.0  NaN  1.0
1  2.0  2.0  1.0
2  2.0  3.0  3.0
3  4.0  4.0  4.0

Using the backward filling method bfill for rows

NaN values can also be filled with succeeding values based on their row position. To do this, you need to use the bfill method and set the axis parameter to 1:

df_bfill = df.fillna(method='bfill', axis=1)
print(df_bfill)

The result shows that the NaN values in rows 0 and 2 have been replaced by the values that follow them in the same row. The NaN value in the first row, however, remains the same because it’s the last value in that row:

A    B    C
0  1.0  1.0  1.0
1  2.0  2.0  NaN
2  3.0  3.0  3.0
3  4.0  4.0  4.0