How to Use Lambda Functions in Python

To use lambda functions in Python, write lambda, the parameters, a colon and one expression. Python returns that expression’s value automatically, so a lambda is a tiny unnamed function you can pass straight into sorted(), filter() or map():

square = lambda x: x * x
print(square(7))                                  # 49
print(sorted(["Ohio", "Iowa", "Texas"], key=lambda s: len(s)))

Quick setup note: my test machine had Python 3.12.5 on Windows 11, and each output block is a straight copy of what Command Prompt printed.

In this tutorial, we’ll cover the syntax, sorting keys, filter/map/reduce, if-else lambdas, closures and when you should pick def instead.

What is the syntax of a lambda function in Python?

The pattern is lambda parameters: expression, with zero or more parameters and defaults allowed. Here’s a sales-tax helper written both ways:

add_tax = lambda price, rate=0.0825: round(price * (1 + rate), 2)

def add_tax_def(price, rate=0.0825):
    return round(price * (1 + rate), 2)

print(add_tax(49.99))
print(add_tax_def(49.99))
print(add_tax(49.99, rate=0.06))
print(add_tax.__name__, "vs", add_tax_def.__name__)

Output:

54.11
54.11
52.99
<lambda> vs add_tax_def

Every lambda is named <lambda>. If you need a name, docs or several lines, use a regular function definition with a return statement.

How do you sort a list with a lambda key in Python?

This is where I use lambdas the most. The key function picks the value to compare, so a lambda sorts dicts by any field:

orders = [
    {"customer": "Megan Bowen", "total": 128.40, "date": "2026-03-14"},
    {"customer": "Alex Wilber", "total": 42.75, "date": "2026-01-09"},
    {"customer": "Adele Vance", "total": 310.00, "date": "2026-02-22"},
    {"customer": "Lee Gu", "total": 42.75, "date": "2026-02-01"},
]

for o in sorted(orders, key=lambda o: o["total"], reverse=True):
    print(f'{o["customer"]:<12} ${o["total"]:.2f}')

by_total_then_date = sorted(orders, key=lambda o: (o["total"], o["date"]))
print([o["customer"] for o in by_total_then_date])

biggest = max(orders, key=lambda o: o["total"])
print("Largest order:", biggest["customer"], f'${biggest["total"]:.2f}')

Output:

Adele Vance  $310.00
Megan Bowen  $128.40
Alex Wilber  $42.75
Lee Gu       $42.75
['Alex Wilber', 'Lee Gu', 'Megan Bowen', 'Adele Vance']
Largest order: Adele Vance $310.00

A tuple key sorts by the first item, then the next on ties; see sorting by the second element of a tuple. max() takes the same key, as in finding the maximum value with max().

Use lambda with filter(), map() and reduce() in Python

Each of these takes a function first. filter() keeps items where the lambda is true, map() transforms each one, and reduce() folds them into one value:

from functools import reduce

prices = [19.99, 5.49, 120.00, 64.50, 3.75]

over_20 = list(filter(lambda p: p > 20, prices))
with_tax = list(map(lambda p: round(p * 1.0825, 2), prices))
cart_total = reduce(lambda total, p: total + p, prices, 0)

print("filter:", over_20)
print("map:   ", with_tax)
print(f"reduce: ${cart_total:.2f}")
print("comprehension:", [round(p * 1.0825, 2) for p in prices if p > 20])

Output:

filter: [120.0, 64.5]
map:    [21.64, 5.94, 129.9, 69.82, 4.06]
reduce: $213.73
comprehension: [129.9, 69.82]

Both filter() and map() return lazy iterators, so wrap them in list(). See the map() function in Python and the reduce() function in Python.

The last line gets the same taxed prices with a Python list comprehension, often easier to read. More options are in ways to filter lists in Python.

Can a Python lambda use if-else or return another function?

Yes. You can’t put an if statement in a lambda, but you can use a if condition else b. A function can also return a lambda that remembers its arguments (a closure):

shipping = lambda total: 0.00 if total >= 50 else 5.99
print([shipping(t) for t in (19.99, 50.00, 75.25)])

def make_discount(pct):
    return lambda price: round(price * (1 - pct / 100), 2)

black_friday = make_discount(25)
print(black_friday(80.00), black_friday(19.99))

fees = [lambda amt: amt * r for r in (0.01, 0.02, 0.03)]
print("late binding:", [f(100) for f in fees])

fees = [lambda amt, r=r: amt * r for r in (0.01, 0.02, 0.03)]
print("fixed:       ", [f(100) for f in fees])

Output:

[5.99, 0.0, 0.0]
60.0 14.99
late binding: [3.0, 3.0, 3.0]
fixed:        [1.0, 2.0, 3.0]

Watch the late binding line: loop lambdas read r when they run, so all three used 0.03. A default, r=r, freezes each value. For more, see passing a function as a parameter in Python.

What are the limitations of lambda functions in Python?

You get exactly one expression. Statements like return, assignments or loops aren’t allowed, and tracebacks show <lambda> instead of a useful name:

import traceback

try:
    compile("double = lambda x: return x * 2", "demo.py", "exec")
except SyntaxError as err:
    print("SyntaxError:", err.msg)

to_cents = lambda dollars: int(dollars * 100)
try:
    to_cents(None)
except TypeError as err:
    print("TypeError:", err)
    print("Traceback names the function:", traceback.extract_tb(err.__traceback__)[-1].name)

Output:

SyntaxError: invalid syntax
TypeError: unsupported operand type(s) for *: 'NoneType' and 'int'
Traceback names the function: <lambda>

Lambdas also can’t carry a docstring or type hints, and PEP 8 recommends def over f = lambda .... My rule of thumb: if it fits inside the call that uses it, keep the lambda; otherwise write a function.

Using lambdas in data or GUI code? These pick up from here:

• Lambda functions on a pandas DataFrame
• Apply a function to each row in pandas
• Tkinter button command (with lambda arguments)
• Built-in functions in Python

Frequently asked questions