Memoized

Memoized is an adjective used in computing to describe a function, value, or computation whose results are cached after the first evaluation, so that subsequent calls with the same inputs can return the cached result quickly without re-execution.

Memoization is the technique of caching results of expensive function calls based on their inputs. It is a form of caching tailored to single-function reuse rather than general data caching. A memoized function maintains a mapping from input argument(s) to output value; on invocation, it first checks the cache, and returns the cached value if present; otherwise it computes the result, stores it in the cache, and returns it.

Caches are typically stored in memory, and keys are derived from the function’s arguments. Implementations vary: some use a simple dictionary or map; others employ eviction strategies such as least-recently-used (LRU) to bound memory use. Memoization can be applied automatically by language runtimes (as in Python’s functools.lru_cache, or JavaScript closures) or manually by developers.

Considerations include correctness and purity: memoized functions should be side-effect-free with respect to the input, or those effects must be considered in invalidation. In languages with immutable data, memoization is often straightforward; in others, cache invalidation, coherence, and memory pressure must be managed. Common applications include dynamic programming, repeated expensive computations, and memoizing API calls.