Apply a Function to a Pandas DataFrame

Pandas dataframes allow you the flexibility of applying a function along a particular axis of a dataframe. In this tutorial, we’ll look at how to apply a function to a pandas dataframe through some examples.

The pandas DataFrame apply() function

The pandas dataframe apply() function is used to apply a function along a particular axis of a dataframe. The following is the syntax:

result = df.apply(func, axis=0)

We pass the function to be applied and the axis along which to apply it as arguments. To apply the function to each column, pass 0 or 'index' to the axis parameter which is 0 by default. And to apply the function to each row, pass 1 or 'columns' to the axis parameter. The examples below illustrate the difference.

Examples

Let’s look at some of the use-cases of the apply() function through examples.

1. Apply a function to each column of the dataframe

Let’s say you want to apply a function to each column of a dataframe, that is, along the index axis. For instance, you’re working with a dataframe having all numerical columns and you want to find the mean for each of those columns.

import pandas as pd

# create a sample dataframe
df = pd.DataFrame({
    'History': [76, 84, 68, 94],
    'Math': [81, 67, 91, 86],
    'English': [72, 93, 84, 76]
})

# print the dataframe
print("The original dataframe:\n")
print(df)

# function to be applied
def get_mean(scores):
    return sum(scores)/len(scores)

# get the mean score for each subject
result = df.apply(get_mean)
print("\nThe result of applying the function on the dataframe:\n")
print(result)

Output:

The original dataframe:

   History  Math  English
0       76    81       72
1       84    67       93
2       68    91       84
3       94    86       76

The result of applying the function on the dataframe:

History    80.50
Math       81.25
English    81.25
dtype: float64

In the above example, the dataframe df contains scores of students in three subjects. Rows represent the students whereas columns represent the subjects. Here, the apply() function is used to get the average score for each of the subjects across all students. Note that, since the function get_mean() is applied to each column, we didn’t need to explicitly pass 0 to the axis parameter since it is its default value.

2. Apply a function to each row of the dataframe

Now, let’s say you want to apply a function to each row of a dataframe, that is, along the columns axis. For instance, you’re working with a dataframe having all numerical rows and you want to find the mean for each of those rows.

import pandas as pd

# create a sample dataframe
df = pd.DataFrame({
    'History': [76, 84, 68, 94],
    'Math': [81, 67, 91, 86],
    'English': [72, 93, 84, 76]
}, index=['Sam', 'Greta', 'Mike', 'Emma'])

# print the dataframe
print("The original dataframe:\n")
print(df)

# function to be applied
def get_mean(scores):
    return sum(scores)/len(scores)

# get the mean score for each student
result = df.apply(get_mean, axis=1)
print("\nThe result of applying the function on the dataframe:\n")
print(result)

Output:

The original dataframe:

       History  Math  English
Sam         76    81       72
Greta       84    67       93
Mike        68    91       84
Emma        94    86       76

The result of applying the function on the dataframe:

Sam      76.333333
Greta    81.333333
Mike     81.000000
Emma     85.333333
dtype: float64

In the above example, the dataframe df contains scores of students in three subjects. Rows represent the students whereas columns represent the subjects. For more clarity, we give each row the name of its corresponding student. Here, the apply() function is used to get the average score for each student across all the three subjects. Note that, we had to pass axis=1 since we wanted the get_mean() function to be applied to each row.

For more on the apply() function, refer to its official documentation.

With this, we come to the end of this tutorial. The code examples and results presented in this tutorial have been implemented in a Jupyter Notebook with a python (version 3.8.3) kernel having pandas version 1.0.5