Memoization trades memory for speed by caching inputs to outputs. It only works correctly for pure functions with stable argument comparison.
Common Interview Questions
Q: Implement basic memoize for single-argument functions
Answer: Map keys use SameValueZero equality. For multiple arguments, serialize with JSON or a custom key function — mind object key identity.
function memoize(fn) {
const cache = new Map();
return (arg) => {
if (cache.has(arg)) return cache.get(arg);
const result = fn(arg);
cache.set(arg, result);
return result;
};
}
Q: When should you NOT memoize?
Answer: Impure functions, functions with side effects, rapidly changing inputs, or when memory cost exceeds compute savings. Also avoid unbounded caches in long-lived processes.
Q: How does React's useMemo relate?
Answer: useMemo memoizes a computed value between renders when dependencies are unchanged. It is a React optimization, not a general-purpose cache across calls outside the component.
Q: What is an LRU cache?
Answer: Least Recently Used eviction removes oldest entries when capacity is exceeded. Useful for bounded memoization in interview system design follow-ups.
Key Points for Interviews
-
Requires pure functions for correctness
-
Argument serialization is the hard part for multi-arg memoize
-
Fibonacci recursive demo is the classic teaching example
-
Bound cache size in production to prevent memory growth
-
Distinguish memoization from debouncing/throttling