Declarative vs imperative — describing what, not how

Both produce the identical result. The imperative version spells out every mechanical step — initialize an index, check a bound, increment, check a condition, mutate an array. The declarative version states the goal ("the even numbers") and hands the mechanics to filter entirely — the reader doesn't have to trace through index bookkeeping to know what the code is for.

None of these are "functional programming" by name, but they all share the same underlying idea map/filter/reduce demonstrate at the smallest scale: say what you want, not the mechanical steps to produce it.

Imperative code's correctness depends on getting every mechanical detail right — the loop bound, the increment, the mutation order — any of which can individually be wrong in a way that's easy to miss on a quick read. Declarative code shifts that correctness burden onto a well-tested, already-verified building block (filter, the SQL engine, the rendering engine) — trusting numbers.filter(...) to correctly filter is a much smaller leap than manually re-verifying a hand-written loop's index arithmetic every single time you read it.

"Could this be written more declaratively?" is a genuinely useful question to ask when reviewing or writing a loop — not because imperative code is always wrong, but because a declarative version is very often shorter, easier to verify at a glance, and less prone to the small mechanical bugs (off-by-one errors, forgotten mutations) that hand-written loops are specifically prone to.