---

The hash compaction algorithm stores compressed versions of states in a hash table. This means the memory consumption of the algorithm is very low and independent of the size of the individual states (the compressed states are always the same size). However, if two distinct states have the same compressed representation, the algorithm will produce an incorrect state graph. Fortunately, this probability is very low (less than 0.1%), and can made arbitrarily small.