apply() 用於在DataFrame 的行或列或Series 的元素上應(yīng)用函數(shù)，支持靈活操作，可返回標(biāo)量、Series 或DataFrame；2. map() 僅用於Series，通過字典、Series 或函數(shù)進(jìn)行元素級映射，常用於值替換；3. applymap() 曾用於DataFrame 的每個元素應(yīng)用函數(shù)，但自pandas 2.1.0 起已棄用，應(yīng)改用df.map()。因此，處理Series 元素映射用map()，處理DataFrame 元素級變換用df.map()，而需沿軸應(yīng)用複雜函數(shù)時用apply()，總結(jié)：apply() 最靈活，map() 專用於Series 元素映射，applymap() 已過時應(yīng)使用df.map() 替代。

Great question — apply() , map() , and applymap() are all used to transform data in pandas, but they serve different purposes and work on different parts of a DataFrame or Series. Here's a clear breakdown of their differences:

1. apply() – Apply a function along an axis of a DataFrame or on a Series

• Works on : DataFrame (rows or columns) or Series
• Purpose : Apply a custom function to either:
  • Each column or row of a DataFrame (axis=0 or axis=1)
  • Each element of a Series
• Flexible : Can return a scalar, a Series, or a DataFrame

Example with DataFrame:

 import pandas as pd

df = pd.DataFrame({
    'A': [1, 2, 3],
    'B': [4, 5, 6]
})

# Apply sum to each column (default axis=0)
df.apply(sum)
# Output: A 6
# B 15

# Apply sum to each row
df.apply(sum, axis=1)
# Output: 0 5
# 1 7
# 2 9

Example with Series:

 s = pd.Series([1, 2, 3])
s.apply(lambda x: x**2)
# Output: 0 1
# 1 4
# 2 9

? Use apply() when you need to apply a function across rows, columns, or elements, especially when the function isn't vectorized.

2. map() – Element-wise transformation for a Series only

• Works on : Only Series
• Purpose : Map values using a dictionary, Series, or function
• Common use : Replace values or perform element-wise operations with lookup tables

Example using dictionary:

 s = pd.Series(['apple', 'banana', 'cherry'])
mapping = {'apple': 'A', 'banana': 'B'}
s.map(mapping)
# Output: 0 A
# 1 B
# 2 NaN

Example using function:

 s.map(lambda x: x.upper())
# Output: 0 APPLE
# 1 BANANA
# 2 CHERRY

?? map() is only for Series , and it's great for value replacement or simple element-wise mapping.

3. applymap() – Apply a function to every individual element in a DataFrame

• Works on : Entire DataFrame (element by element)
• Purpose : Transform each element individually
• Note : Deprecated since pandas 2.1.0 — use map() instead on DataFrames

Example (before deprecation):

 df = pd.DataFrame({
    'A': [1.1, 2.2, 3.3],
    'B': [4.4, 5.5, 6.6]
})

df.applymap(lambda x: f"{x:.1f}")
# Output: AB
# 0 1.1 4.4
# 1 2.2 5.5
# 2 3.3 6.6

Modern replacement (use map() on DataFrame):

 df.map(lambda x: f"{x:.1f}")

? Use df.map() (replaces applymap() ) when you want to apply a function to every element in a DataFrame.

Summary: Key Differences

Quick Rule of Thumb:

• Need to transform each element in a Series ? → Use map()
• Need to apply a function to each element in a DataFrame ? → Use df.map()
• Need to apply a function to rows/columns or more complex logic? → Use apply()

Basically, map() is for simple element-wise mapping (especially Series), apply() is for more flexible operations, and applymap() is outdated — just use df.map() now.

以上是python的pandas中的apply（），map（）和applymap（）有什麼區(qū)別？的詳細(xì)內(nèi)容。更多資訊請關(guān)注PHP中文網(wǎng)其他相關(guān)文章！