Comments on: Space vs Time http://daniel.gredler.net/2008/01/26/space-vs-time/ Tue, 29 Dec 2009 03:36:29 +0000 http://wordpress.com/ hourly 1 By: Dmitry Zhelezov http://daniel.gredler.net/2008/01/26/space-vs-time/#comment-625 Dmitry Zhelezov Wed, 17 Jun 2009 11:46:07 +0000 http://gredler.wordpress.com/?p=60#comment-625 @Daniel: Actually, performance for HashMap depends on hash function and how often collisions occur. And as I know put() and get() are constant-time only when there are no collisions at all @Daniel: Actually, performance for HashMap depends on hash function and how often collisions occur. And as I know put() and get() are constant-time only when there are no collisions at all

]]> By: Daniel Gredler http://daniel.gredler.net/2008/01/26/space-vs-time/#comment-446 Daniel Gredler Wed, 12 Mar 2008 03:41:56 +0000 http://gredler.wordpress.com/?p=60#comment-446 @Hank: That would probably work. The BOs would be declared to have Sets, which in actuality would be PersistentSets, which wrap HashSets, which themselves wrap HashMaps (phew!). But you lose order and bag semantics, which we actually want to keep. @Blackjack: How do you figure? You'd have to have a single bucket in the HashMap to get O(n log n) performance, no? HashMap guarantees constant-time performance for get() and put(), meaning constant-time performance for containsKey()... meaning O(n) overall performance on average. Am I missing something? @Hank: That would probably work. The BOs would be declared to have Sets, which in actuality would be PersistentSets, which wrap HashSets, which themselves wrap HashMaps (phew!). But you lose order and bag semantics, which we actually want to keep.

@Blackjack: How do you figure? You’d have to have a single bucket in the HashMap to get O(n log n) performance, no? HashMap guarantees constant-time performance for get() and put(), meaning constant-time performance for containsKey()… meaning O(n) overall performance on average. Am I missing something?

]]> By: Blackjack http://daniel.gredler.net/2008/01/26/space-vs-time/#comment-444 Blackjack Mon, 03 Mar 2008 08:09:55 +0000 http://gredler.wordpress.com/?p=60#comment-444 Actually, your algo has O(n log n) performance. But that's also great :) Actually, your algo has O(n log n) performance. But that’s also great :)

]]> By: Hank http://daniel.gredler.net/2008/01/26/space-vs-time/#comment-401 Hank Sat, 26 Jan 2008 22:03:06 +0000 http://gredler.wordpress.com/?p=60#comment-401 What about using a <code>HashSet</code> and leveraging <code>addAll()</code> and <code>retainAll()</code>? What about using a HashSet and leveraging addAll() and retainAll()?

]]>