Pivot Tables, Melt, Stack, and Unstack

Pandas Basics

2 min read

Published Sep 29 2025, updated Oct 24 2025

PandasPython

There are various ways to reshape your data between wide, long, and hierarchical forms.

Pivot Table → reshapes data from long → wide, useful for summarisation (like Excel pivot tables).
Melt → reshapes data from wide → long, useful for tidy data & plotting.
Stack → Takes columns and “stacks” them into the row index.
Unstack → Moves an index level back into columns.

Pivot Table

A pivot table summarises and reshapes data by aggregating values across one or more categories.

Syntax:

pd.pivot_table(

data,

values=None,

index=None,

columns=None,

aggfunc="mean",

fill_value=None,

margins=False

)

Key Parameters:

data → DataFrame
values → column(s) to aggregate
index → row grouping
columns → column grouping
aggfunc → aggregation function ('mean', 'sum', 'count', custom funcs)
fill_value → replace missing values in result
margins=True → adds totals (like Excel “All”)

Example:

import pandas as pd

data = {

"Department": ["Sales", "Sales", "IT", "IT", "HR"],

"Employee": ["Alice", "Bob", "Charlie", "Dave", "Eve"],

"Salary": [50000, 60000, 70000, 80000, 45000],

}

df = pd.DataFrame(data)

pivot = pd.pivot_table(

df,

values="Salary",

index="Department",

aggfunc="mean"

)

print(pivot)

Output:

Salary

Department

HR 45000

IT 75000

Sales 55000

Melt

melt() is the reverse of pivot. It transforms a wide DataFrame into a long (tidy) format.

Syntax:

pd.melt(

frame,

id_vars=None,

value_vars=None,

var_name="variable",

value_name="value"

)

Key Parameters:

frame → DataFrame
id_vars → columns to keep fixed (like identifiers)
value_vars → columns to unpivot
var_name → new column name for variable names
value_name → new column name for values

Example:

df = pd.DataFrame({

"Employee": ["Alice", "Bob"],

"Sales_Q1": [200, 150],

"Sales_Q2": [210, 160]

})

melted = pd.melt(

df,

id_vars=["Employee"],

value_vars=["Sales_Q1", "Sales_Q2"],

var_name="Quarter",

value_name="Sales"

)

print(melted)

Output:

Employee Quarter Sales

0 Alice Sales_Q1 200

1 Bob Sales_Q1 150

2 Alice Sales_Q2 210

3 Bob Sales_Q2 160

Stack

Takes columns and “stacks” them into the row index. Useful for moving from a wide to a longer DataFrame, especially when you already have a MultiIndex.

Example:

import pandas as pd

df = pd.DataFrame({

"Name": ["Alice", "Bob"],

"Math": [90, 80],

"Science": [85, 95]

})

print("Original DF:\n", df)

stacked = df.set_index("Name").stack()

print("\nStacked:\n", stacked)

Output:

Original DF: