Pandas: Apply a Function to each Cell of a DataFrame

# Pandas: Apply a Function to each Cell of a DataFrame

Use the DataFrame.applymap() method to apply a function to each cell of a Pandas DataFrame.

The method applies a function to a DataFrame element-wise.

main.py

Copied!
import math
import pandas as pd

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, 3, 4, 5],
    'C': [0, 0, 3, 4, 5],
})

print(df)

print('-' * 50)

print(df.applymap(math.sqrt))

The code for this article is available on GitHub

shell

Copied!
   A  B  C
0  1  1  0
1  1  2  0
2  1  3  3
3  2  4  4
4  3  5  5
--------------------------------------------------
/home/borislav/Desktop/bobbyhadz_python/main.py:14: FutureWarning: DataFrame.applymap has been deprecated. Use DataFrame.map instead.

          A         B         C
0  1.000000  1.000000  0.000000
1  1.000000  1.414214  0.000000
2  1.000000  1.732051  1.732051
3  1.414214  2.000000  2.000000
4  1.732051  2.236068  2.236068

pandas apply function to each cell of dataframe

The DataFrame.applymap() method takes a function that accepts and returns a scalar to every element of a DataFrame.

The method returns the transformed DataFrame.

However, notice that the method has been deprecated in Pandas version 2.1.0.

shell

Copied!
/home/borislav/Desktop/bobbyhadz_python/main.py:14: FutureWarning: DataFrame.applymap has been deprecated. Use DataFrame.map instead.

# Pandas: Apply a Function to each Cell of a DataFrame using `DataFrame.map()`

As the message suggests, if you use a Pandas version greater than 2.1.0, you should use the DataFrame.map method.

main.py

Copied!
import math
import pandas as pd

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, 3, 4, 5],
    'C': [0, 0, 3, 4, 5],
})

print(df)

print('-' * 50)

print(df.map(math.sqrt))

The code for this article is available on GitHub

Running the code sample produces the following output.

shell

Copied!
   A  B  C
0  1  1  0
1  1  2  0
2  1  3  3
3  2  4  4
4  3  5  5
--------------------------------------------------
          A         B         C
0  1.000000  1.000000  0.000000
1  1.000000  1.414214  0.000000
2  1.000000  1.732051  1.732051
3  1.414214  2.000000  2.000000
4  1.732051  2.236068  2.236068

apply function to each cell of dataframe using dataframe map

Starting with Pandas version 2.1.0, the DataFrame.applymap() method has been deprecated and renamed to DataFrame.map().

The function you pass to the method needs to take a single value and return a single value.

I used the math.sqrt() method in the example but you can also use a custom function.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, 3, 4, 5],
    'C': [0, 0, 3, 4, 5],
})

print(df)

print('-' * 50)


def custom_function(num):
    if num > 1:
        return num + 10
    elif num < 1:
        return num - 10
    else:
        return 1000


print(df.map(custom_function))

The code for this article is available on GitHub

Running the code sample produces the following output.

shell

Copied!
   A  B  C
0  1  1  0
1  1  2  0
2  1  3  3
3  2  4  4
4  3  5  5
--------------------------------------------------
      A     B   C
0  1000  1000 -10
1  1000    12 -10
2  1000    13  13
3    12    14  14
4    13    15  15

using dataframe map method with custom function

The function takes a single value and returns a single value.

main.py

Copied!
def custom_function(num):
    if num > 1:
        return num + 10
    elif num < 1:
        return num - 10
    else:
        return 1000

The custom function gets called with the value of each cell in the DataFrame.

The new, transformed DataFrame contains the returned values.

The DataFrame.map() method also takes a na_action argument that enables you to handle NaN values.

For example, if the argument is set to 'ignore', the NaN values are propagated without passing them to the supplied function.

main.py

Copied!
import pandas as pd

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, None, 4, 5],
    'C': [0, 0, None, None, 5],
})

print(df)

print('-' * 50)


def custom_function(num):
    if num > 1:
        return num + 10
    elif num < 1:
        return num - 10
    else:
        return 1000


print(df.map(custom_function, na_action='ignore'))

The code for this article is available on GitHub

Running the code sample produces the following output.

shell

Copied!
   A    B    C
0  1  1.0  0.0
1  1  2.0  0.0
2  1  NaN  NaN
3  2  4.0  NaN
4  3  5.0  5.0
--------------------------------------------------
      A       B     C
0  1000  1000.0 -10.0
1  1000    12.0 -10.0
2  1000     NaN   NaN
3    12    14.0   NaN
4    13    15.0  15.0

setting na action argument

Notice that the NaN values weren't handled.

# Pandas: Apply a Function to each Cell of a DataFrame using `numpy.vectorize()`

You can also use the numpy.vectorize method with DataFrame.apply to apply a function to each cell of a DataFrame.

First, make sure you have the numpy module installed.

shell

Copied!
pip install numpy

# or with pip3
pip3 install numpy

Now, import the module and use the numpy.vectorize() method.

main.py

Copied!
import pandas as pd
import numpy as np

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, 3, 4, 5],
    'C': [0, 0, 3, 4, 5],
})

print(df)

print('-' * 50)


def custom_function(num):
    if num > 1:
        return num + 10
    elif num < 1:
        return num - 10
    else:
        return 1000


print(df.apply(np.vectorize(custom_function)))

The code for this article is available on GitHub

Running the code sample produces the following output.

shell

Copied!
   A  B  C
0  1  1  0
1  1  2  0
2  1  3  3
3  2  4  4
4  3  5  5
--------------------------------------------------
      A     B   C
0  1000  1000 -10
1  1000    12 -10
2  1000    13  13
3    12    14  14
4    13    15  15

using numpy vectorize to apply function to each cell

The numpy.vectorize() method returns an object that acts like the supplied function but takes arrays as input.

main.py

Copied!
# <numpy.vectorize object at 0x7fe31020e110>
print(np.vectorize(custom_function))

The DataFrame.apply() method applies the supplied function along an axis of the DataFrame.

You can also use the numpy.vectorize() method in a more manual manner to achieve the same result.

main.py

Copied!
import pandas as pd
import numpy as np

df = pd.DataFrame({
    'A': [1, 1, 1, 2, 3],
    'B': [1, 2, 3, 4, 5],
    'C': [0, 0, 3, 4, 5],
})

print(df)

print('-' * 50)


def custom_function(num):
    if num > 1:
        return num + 10
    elif num < 1:
        return num - 10
    else:
        return 1000


df[:] = np.vectorize(custom_function)(df)
print(df)

The code for this article is available on GitHub

Running the code sample produces the following output.

shell

Copied!
   A  B  C
0  1  1  0
1  1  2  0
2  1  3  3
3  2  4  4
4  3  5  5
--------------------------------------------------
      A     B   C
0  1000  1000 -10
1  1000    12 -10
2  1000    13  13
3    12    14  14
4    13    15  15

We called the function that is returned from numpy.vectorize() with the DataFrame object.

# Additional Resources

You can learn more about the related topics by checking out the following tutorials:

I wrote a book in which I share everything I know about how to become a better, more efficient programmer.

You can use the search field on my Home Page to filter through all of my articles.

Pandas: Apply a Function to each Cell of a DataFrame

# Pandas: Apply a Function to each Cell of a DataFrame

# Pandas: Apply a Function to each Cell of a DataFrame using DataFrame.map()

# Pandas: Apply a Function to each Cell of a DataFrame using numpy.vectorize()

# Additional Resources

Borislav Hadzhiev

# Pandas: Apply a Function to each Cell of a DataFrame using `DataFrame.map()`

# Pandas: Apply a Function to each Cell of a DataFrame using `numpy.vectorize()`