Chapter 23 Applying Functions

Using apply in NumPy and Pandas is a common practice to perform element-wise operations or to apply custom functions across data structures.

Here are some examples:

23.0.1 Using apply in Pandas:

Pandas’ apply function is very powerful for applying custom functions along either axis (rows or columns) of a DataFrame or across a Series.

23.0.1.1 Applying a Function to a DataFrame Column

Imagine you have a DataFrame df with a column salary, and you want to calculate a 10% increase for all salaries.

import pandas as pd

# Sample DataFrame
data = {'employee': ['Alice', 'Bob', 'Charlie'],
        'salary': [50000, 60000, 70000]}
df = pd.DataFrame(data)

# Apply a function to increase salary by 10%
df['new_salary'] = df['salary'].apply(lambda x: x * 1.10)
print(df)

23.0.1.2 Applying a Function Across Rows

You have a DataFrame with columns x and y, and you want to create a new column z which is the sum of x and y.

# Sample DataFrame
df = pd.DataFrame({'x': [1, 2, 3], 'y': [4, 5, 6]})

# Apply a function row-wise
df['z'] = df.apply(lambda row: row['x'] + row['y'], axis=1)
print(df)

23.0.1.3 Using apply with a Custom Function

You can also use apply to apply a custom function defined outside the apply call.

# Custom function to categorize salary
def categorize_salary(salary):
    if salary > 60000:
        return 'High'
    else:
        return 'Low'

# Apply the custom function to the 'salary' column
df['salary_category'] = df['salary'].apply(categorize_salary)
print(df)

23.0.2 Using apply in NumPy:

NumPy’s apply_along_axis function is used to apply a function along a particular axis of a NumPy array.

23.0.2.1 Applying a Function to Each Row of a 2D NumPy Array

Suppose you have a 2D array and want to calculate the sum of each row.

import numpy as np

# Sample 2D array
arr = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

# Function to calculate sum of an array
def row_sum(row):
    return np.sum(row)

# Apply the function along the axis 1 (rows)
result = np.apply_along_axis(row_sum, axis=1, arr=arr)
print(result)  # Output: [ 6 15 24]

23.0.2.2 Applying a Function to Each Column of a 2D NumPy Array

Similar to the previous example, but applied along the columns.

# Function to calculate the mean of an array
def column_mean(column):
    return np.mean(column)

# Apply the function along the axis 0 (columns)
result = np.apply_along_axis(column_mean, axis=0, arr=arr)
print(result)  # Output: [4. 5. 6.]

23.0.3 Key Points to Mention in an Interview:

• Pandas’ apply function is very versatile and is generally used to apply custom functions row-wise or column-wise in a DataFrame or to a Series.

• NumPy’s apply_along_axis is more specific and used for applying a function along a specific axis (rows or columns) of an array.

• These functions are particularly useful when built-in functions are not sufficient or when custom operations are needed.

• While apply can be convenient, it’s worth mentioning that it can be slower than vectorized operations in Pandas and NumPy, so it should be used judiciously.

These examples should provide you with a solid foundation to demonstrate your understanding of apply in NumPy and Pandas during a data science interview.