all is explained in the README on github :-) and here is an example Splicer:
https://github.com/JulesGosnell/seqspert/blob/master/src/java/clojure/lang/ArrayNodeAndArrayNodeSplicer.java Jules On Monday, 17 November 2014 23:01:44 UTC, Fluid Dynamics wrote: > > On Monday, November 17, 2014 5:00:23 PM UTC-5, Jules wrote: >> >> Back in February, I posted about some prototype code I had put together >> that attempted an alternative approach to map-merging that promised to be >> faster and generate less churn than the standard approach. >> >> Glen Mailer suggested that I put it in a standalone library. At the time, >> I thought that this would not be possible because of its necessarily tight >> coupling to clojure.lang.PersistentHashMap, but then I found some more >> time, went away and had a go. >> >> Here it is: >> >> https://github.com/JulesGosnell/seqspert >> >> along with some other hash-set, vector and array related stuff that I >> have been playing with, instructions to set it up in Lein and example code >> to paste straight into the repl to see how well it performs in your test >> environment. >> >> Some simple testing indicates that the larger the map and the greater the >> number of threads that you can run in parallel, the better results you will >> see. I have seen 30x speedups on a 16-way box merging two hash-maps of 10M >> entries each. >> >> Please take a look at it and let me know about any issues, improvements, >> results etc - I'd be very interested to hear from you. >> >> I'll put out a stable release in a while, once interested parties have >> kicked the tyres a bit. >> >> regards, >> >> >> Jules >> > > Looks like you're seeing a general 2x speedup, and an orthogonal > ${num_cores} speedup from parallelization. What's the source of the general > 2x speedup? Algorithmic improvement? Use of transients? > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
