Memoize function results to avoid expensive recomputation.

Indicates that the decorated function's results should be cached for performance optimization.

When a pure function is called repeatedly with the same arguments When the function computation is expensive relative to cache lookup When the function's output depends only on its inputs (pure function)

Performance optimization in recursive algorithms (e.g., Fibonacci, factorial) Caching expensive I/O or computation results in web services Memoizing pure functions in functional-style Python code

@functools.lru_cache(maxsize=<N>)

@functools.lru_cache

{"maxsize":"integer >= 1, or None for unlimited cache"}

• def def return @ functools

• @lru_cache @functools.lru_cache() @functools.lru_cache(maxsize=None)

Using lru_cache on functions with mutable arguments (e.g., lists, dicts) leading to incorrect caching Forgetting to import functools Setting maxsize too low causing excessive evictions, or too high causing excessive memory use

@functools.lru_cache vs @functools.cache (Python 3.9+): cache is unbounded lru_cache wrapper @functools.lru_cache vs @functools.lru_cache(maxsize=None): unbounded LRU cache

Applying lru_cache to recursive functions without proper base case can cause infinite recursion due to caching intermediate results Using lru_cache on methods without considering self argument can cause memory leaks if not careful

• functools decorators
• memoization techniques
• functional programming patterns

The cache is per-function instance; each decorated function gets its own cache. The maxsize argument controls the maximum number of entries before LRU eviction begins.

Signals to readers that the function is pure and performance-sensitive, encouraging reuse of results.

@functools.lru_cache(maxsize=128)

Python 3.9+ includes functools.cache as an alias for lru_cache(maxsize=None).

Consider using @functools.cache (unbounded) or @functools.lru_cache(maxsize=None) for unbounded caching, or @functools.lru_cache(maxsize=2**30) for very large caches.