Re: Set performance

2010-04-05 Thread Krukow
On Apr 5, 3:06 pm, ineol wrote: > It has nothing to do with Clojure but your colleague can look at the > Google collection library that contains an > ImmutableSet.http://guava-libraries.googlecode.com/svn/trunk/javadoc/com/google/co... Yes, there is also Scalas data structures (he was using th

Re: Set performance

2010-04-05 Thread Krukow
On Apr 2, 3:28 pm, Chas Emerick wrote: > As Clojure moves towards being self-hosted, fewer and fewer of the   > data structures will be implemented in Java, thereby ensuring   > dependence on the Clojure runtime.  Just FYI. Yes, I've realized that as well :-) I would have to maintain a Java "ba

Re: Set performance

2010-04-05 Thread Per Vognsen
I looked at the source code. ImmutableSet doesn't support adding adding or removing elements incrementally. You can make an ImmutableSet.Builder from an existing ImmutableSet of n elements and then make a batch of m updates from which it will construct a new ImmutableSet in O(n+m) time. It simply i

Re: Set performance

2010-04-05 Thread ineol
It has nothing to do with Clojure but your colleague can look at the Google collection library that contains an ImmutableSet. http://guava-libraries.googlecode.com/svn/trunk/javadoc/com/google/common/collect/ImmutableSet.html -- You received this message because you are subscribed to the Google G

Re: Set performance

2010-04-02 Thread Chas Emerick
As Clojure moves towards being self-hosted, fewer and fewer of the data structures will be implemented in Java, thereby ensuring dependence on the Clojure runtime. Just FYI. - Chas On Mar 31, 2010, at 12:56 PM, Krukow wrote: It would be nice to have a version of the clojure data structure

Re: Set performance

2010-04-01 Thread Julien
Did you run the test with the -server jvm option? This command line argument is usually recommended when measuring performance. -Julien On Mar 31, 10:56 am, Krukow wrote: > On Mar 29, 10:21 pm, Krukow wrote:> Hello, > [snip..] > > What was surprising to me wasn't that "inserts" are slower - tha

Re: Set performance

2010-03-31 Thread Krukow
On Mar 29, 10:21 pm, Krukow wrote: > Hello, [snip..] > What was surprising to me wasn't that "inserts" are slower - that is > ok and it could be improved with transients. The surprising thing was > that iterating through the entire set was significantly slower. [snip..] I've dug through the sou

Set performance

2010-03-29 Thread Krukow
Hello, a colleague of mine asked me a question about clojures set implementation performance. He needed an immutable set on the jvm platform for a Java/Scala project, and since I've been saying that Clojures data structures have good performance he tried clojures set. He was dissapointed though :-(