FilterHash | An open addressing based hash table implementation

FilterHash is a Java library. FilterHash has no bugs, it has no vulnerabilities, it has build file available and it has low support. You can download it from GitHub.

An open addressing based hash table implementation. Terminology We are hashing n elements into k locations. The load factor is α=n/k. Description of the algorithm The idea is to partition the set of indexes (i.e. the underlying array) into i subsets. Notice that the partition is purely conceptual: In practice there is only one array, but our algorithm will operate on the array as if it was split into i subarrays. Order the subsets by descending size. Each of these subsets will be mapped by a hash table with the following closed addressing collision strategy: Rehash the colliding item into the next hash table in the ordered sequence of hash tables. Thus, the first (and biggest) hash table will receive all of the n items we are hashing. And the expected number of collisions in one hash table will be the input for the next hash table. We can achieve any desired load factor α=n/k in a given hash table, by choosing k small enough relative to n. And if all i hash tables have an expected load factor of α, the whole data structure as such will have a load factor of α.