Insert keys and watch the hash function scatter them across buckets.
The hash decides the bucket; collisions are resolved by chaining, so lookups stay O(1) on average.
Insert a key to fill the table.