We use cookies to ensure you have the best browsing experience on our website. Please read our cookie policy for more information about how we use cookies.
Imagine all strings in test have same hashCodes. String length <= 100. It's very easy to achieve. If author wanted, he could've made such a test. Every hash algorithm is vulnerable to that kind of test. And your O(N+Q) hashtable solution will degrade to O(N*Q).
So when talking about "big O", which is upper bound, "hashtable" solution has same complexity, which is O(N*Q).
Better and stable algorithm would be to use TreeMap with lexicographic comparator. Which would give you O(Q*logN).
By the way, starging from Java8 they combined the "best of two worlds". HashMap class uses binary tree (if keys are "comparable") instead of linked list in case there are a lots of keys with same hashCode.
So in java8 it would give you O(Q*logN) if using HashMap solution.
Sparse Arrays
You are viewing a single comment's thread. Return to all comments →
Imagine all strings in test have same hashCodes. String length <= 100. It's very easy to achieve. If author wanted, he could've made such a test. Every hash algorithm is vulnerable to that kind of test. And your O(N+Q) hashtable solution will degrade to O(N*Q).
So when talking about "big O", which is upper bound, "hashtable" solution has same complexity, which is O(N*Q).
Better and stable algorithm would be to use TreeMap with lexicographic comparator. Which would give you O(Q*logN).
By the way, starging from Java8 they combined the "best of two worlds". HashMap class uses binary tree (if keys are "comparable") instead of linked list in case there are a lots of keys with same hashCode. So in java8 it would give you O(Q*logN) if using HashMap solution.