Lambda Expressions

Lambda Expressions are ideally used when we need to do something simple and are more interested in getting the job done quickly rather than formally naming the function. Lambda expressions are also known as anonymous functions.

Lambda Expressions in Python are a short way to declare small and anonymous functions (it is not necessary to provide a name for lambda functions).

Lambda functions behave just like regular functions declared with the def keyword. They come in handy when you want to define a small function in a concise way. They can contain only one expression, so they are not best suited for functions with control-flow statements.

Syntax of a Lambda Function

lambda arguments: expression

Lambda functions can have any number of arguments but only one expression.

Example code

# Lambda function to calculate square of a number
square = lambda x: x ** 2
print(square(3)) # Output: 9

# Traditional function to calculate square of a number
def square1(num):
  return num ** 2
print(square(5)) # Output: 25

In the above lambda example, lambda x: x ** 2 yields an anonymous function object which can be associated with any name. So, we associated the function object with square. So from now on we can call the square object like any traditional function, for example square(10)

Examples of lambda functions

Beginner

lambda_func = lambda x: x**2 # Function that takes an integer and returns its square
lambda_func(3) # Returns 9

Intermediate

lambda_func = lambda x: True if x**2 >= 10 else False
lambda_func(3) # Returns False
lambda_func(4) # Returns True

Complex

my_dict = {"A": 1, "B": 2, "C": 3}
sorted(my_dict, key=lambda x: my_dict[x]%3) # Returns ['C', 'A', 'B']

Use-case

Let’s say you want to filter out odd numbers from a list. You could use a for loop:

my_list = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
filtered = []

for num in my_list:
     if num % 2 != 0:
         filtered.append(num)

print(filtered)      # Python 2: print filtered
# [1, 3, 5, 7, 9]

Or you could write this as a one liner with list-comprehensions:

filtered = [x for x in [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] if x % 2 != 0]

But you might be tempted to use the built-in filter function. Why? The first example is a bit too verbose and the one-liner can be harder to understand. But filter offers the best of both words. What is more, the built-in functions are usually faster.

my_list = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

filtered = filter(lambda x: x % 2 != 0, my_list)

list(filtered)
# [1, 3, 5, 7, 9]

NOTE: in Python 3 built in functions return generator objects, so you have to call list. In Python 2, on the other hand, they return a list, tupleor string.

So what happened? You told filter to take each element in my_list and apply the lambda expressions. The values that return False are filtered out.

More Information: