Applying Functions (apply, map, applymap)

First: prefer vectorized operations

Pandas is fastest when you use vectorized operations.

Vectorized example

import pandas as pd
 
df = pd.DataFrame({"amount": [100, 200, 150]})
df["amount_with_tax"] = df["amount"] * 1.18
print(df)

Use apply/map when you really need custom logic.

From the book

McKinney’s naming logic: applymap got its name because it applies a Series’ map element-wise across an entire DataFrame. So the three tools nest neatly — map works on one Series, apply works on a Series or across rows/columns of a DataFrame, and applymap is map broadcast over every cell of a DataFrame.

map() (Series only)

map transforms values in a single Series.

Example: map categories

map categories

import pandas as pd
 
s = pd.Series(["pune", "delhi", "mumbai"])
state_map = {"pune": "MH", "delhi": "DL", "mumbai": "MH"}
 
print(s.map(state_map))

If a value is missing from the map, it becomes NaN.

apply() (Series or DataFrame)

Apply on a Series

apply on Series

import pandas as pd
 
s = pd.Series([1, 2, 3, 4])
print(s.apply(lambda x: x * x))

Apply on rows (axis=1)

apply on rows

import pandas as pd
 
df = pd.DataFrame({
    "name": ["Asha", "Ravi"],
    "score": [88, 91],
    "bonus": [5, 2],
})
 
def final_score(row):
    return row["score"] + row["bonus"]
 
df["final"] = df.apply(final_score, axis=1)
print(df)

Row-wise apply is easy to write but can be slower for large data.

applymap() (DataFrame element-wise)

Element-wise transform for the whole DataFrame.

applymap

import pandas as pd
 
df = pd.DataFrame({"a": [1, 2], "b": [3, 4]})
print(df.applymap(lambda x: x * 10))

Quick guidance

• Use vectorized operations if possible
• Use map for simple value mapping (Series)
• Use apply for custom logic on a Series or across rows/columns
• Use applymap for element-wise transforms

Visualize it

diagram map, apply, applymap at a glance mermaid

  flowchart LR
  A["Series.map(fn)"] --> R1["New Series, same shape"]
  B["Series/DataFrame.apply(fn)"] --> R2["Scalar, Series, or DataFrame"]
  C["DataFrame.applymap(fn)"] --> R3["New DataFrame, every cell transformed"]

map transforms one Series; apply works on a Series or DataFrame rows/columns; applymap runs map over every cell.

🧪 Try It Yourself

Exercise 1 – Map Values With a Dictionary

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Apply a Custom Function to a Series

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Apply Row-Wise With axis=1

pch.exerciseTag pch.exerciseTitle

Next

Custom transforms often need to handle gaps in the data — continue to Handling Missing Data (isna, fillna, dropna).