How Do `map`, `applymap`, and `apply` Differ in Pandas Vectorization?

Vectorization Methods in Pandas: map, applymap, and apply

Pandas offers convenient methods for applying functions to data structures. map, applymap, and apply are three such methods that facilitate data manipulation and transformation. Each method serves a specific purpose, and their usage depends on the desired outcome.

map

map is employed when applying a function element-wise to a Series. It returns a new Series with the transformed values.

Example:

import pandas as pd

series = pd.Series([1, 2, 3, 4, 5])

def square(x):
    return x ** 2

squared_series = series.map(square)

print(squared_series)
# Output:
# 0    1
# 1    4
# 2    9
# 3   16
# 4   25
# dtype: int64

applymap

applymap applies a function element-wise to a DataFrame. It creates a new DataFrame with the transformed values.

Example:

df = pd.DataFrame({
    'name': ['John', 'Jane', 'Tom'],
    'age': [20, 25, 30]
})

def capitalize(x):
    return x.capitalize()

df['name'] = df['name'].applymap(capitalize)

print(df)
# Output:
#    name  age
# 0  John   20
# 1  Jane   25
# 2  Tom    30

apply

apply allows for more complex transformations by applying a function row- or column-wise to a DataFrame. It returns a Series or DataFrame with the results.

Example:

def get_age_group(age):
    if age <= 20:
        return 'Young'
    elif age <= 40:
        return 'Middle-aged'
    else:
        return 'Senior'

df['age_group'] = df['age'].apply(get_age_group)

print(df)
# Output:
#   name  age age_group
# 0  John   20     Young
# 1  Jane   25  Middle-aged
# 2  Tom    30  Middle-aged

The above is the detailed content of How Do `map`, `applymap`, and `apply` Differ in Pandas Vectorization?. For more information, please follow other related articles on the PHP Chinese website!