runtimecompare approaches one problem · three ways
Compare Approaches
The same problem solved different ways in Python, run on the same input, with what really differs and when to use each.
Memoising: dict vs lru_cache vs cache
Count ways to climb n stairs taking 1 or 2 steps, without recomputing.
what differs
All three store results by argument. The decorators also give cache_info() and cache_clear(), and are thread-safe. A module-level dict outlives every call, so it's easy to forget to clear.
at a glance
| approach | use it when |
|---|---|
| A · manual dict | You need control over the key or the stored value. |
| B · @lru_cache(maxsize=128) | Many distinct inputs: maxsize bounds memory by evicting the least recently used. |
| C · @cache | A pure function whose inputs are few enough to keep every result. |
Watch out: Arguments must be hashable (no lists). Caching a function with side effects or time-dependent results returns stale answers.
the outputs
Each approach runs on the same input.
concepts
memoizationfunctools.lru_cache
learn it
Caching: remember results instead of recomputing
related comparisons
Wrapper without wraps vs decorator with functools.wraps · Recursion vs iteration