What does the "yield" keyword do in Python?

python

yield

generators

itertools

byAlex Kataev·Aug 4, 2024

The yield keyword essentially bends a function into a generator, a magical memory-efficient construct that doles out values one at a time. Rather than eagerly generating all values at once and causing a potential memory avalanche, each yield forms a pause point, saving the function's state.

# Quick generator
def countdown(n):
    while n > 0:
        yield n   # Bam! A yield in the wild.
        n -= 1 

# Efficient run: prints 5, 4, 3, 2, 1
for number in countdown(5):
    print(number)

With the countdown generator, you're printing descending numbers, one at a time, using a memory-friendly iterator.

Deep Dive into yield

Preservation of state

yield provides a neat stateful iteration feature whereby the generator preserves its status quo between calls, enabling smooth resume of operations. It's like binge-watching Netflix: you can pause in between and pick up where you left off!

Memory friendliness: Generators vs lists

Compared to lists that eagerly munch your memory, generators enter the field one-by-one, serving items on demand thus minimizing memory usage. Consequently, methods like extend() become more efficient when operating on dynamic lists.

Behind the scenes magic: Iteration Protocol

Understand the magic behind yield by mastering Python's iteration protocol—the combo of __iter__() and __next__(). These methods allow the transformation of an iterable into an iterator, ensuring an ordered party-line procession until the bouncer (StopIteration exception) calls it a night.

Coroutine capabilities & advanced interaction

In addition to basic iterations, generators can also moonlight as coroutines. They have the wax on, wax off talent of pausing and resuming execution, enabling an interactive tango with send(), close(), and throw(), giving you power over control flows and data exchange.

Advanced yield techniques

One way trip: Reusability and exhaustion

A generator is a one-time ride; once exhausted, it needs a fresh token to start again. Think of it as a ticket for a single joyride: once you complete the circuit, you need to get a new ticket for another spin.

Toolbox expansion: Integrating with `itertools`

Think of the itertools library as a Nürburgring for iterators. With its various techniques, you can manipulate generator results, allowing for high-speed maneuvers like chaining multiple generators or applying islice() for precision slicing.

Cleaner delegation: `yield from`

yield from is the ** Middleman Terminator**. It simplifies the process of delegating to sub-generators. This comes in handy when your generators are yielding from other generators, thus enabling nested iterations with less fuss and more elegance.

explain-codes / Python / What does the "yield" keyword do in Python?

Linked

Difference Between Python's Generators and Iterators



What is memoization and how can I use it in Python?



Call int() function on every list element?



How do I split a list into equally-sized chunks?



Create list of single item repeated N times

