Re:Robin Hood Hashing should be the default hash table imple...
by pkhuong on Wednesday August 31, 2016 @02:57PM
(#12400578)
You can have preallocated buckets in an array. What's the difference you ask? When you have more elements per buckets, the variance becomes smaller as a fraction of the bucket's size. Imagine you had a hash table with buckets of 1000 elements. You'd probably feel comfortable with a 90% load even without any strategy to handle bucket overflow except growing and rehashing. Obviously, the downside is that we have to do more work to search in buckets.
The difference between linear probing and buckets is a big one: buckets are linked lists. That means that hitting a bucket doesn't pull everything else inside the bucket into the CPU cache. So iterating through the bucket may incur a cache miss for every element. On the other hand, linear probing in open addressing does away with buckets and instead searches the neighbors of the collision. Because they're nearby in memory, it's a lot quicker to search.