Re: Performance tuning for a simple SLE solver

2009-11-23 Thread luskwater
I replaced the calls to get-in with a my-get-in macro that "compiled" them to nested calls on the map or vector: (my-get-in sle [:A 1 2]) => (((sle :A) 1) 2) The form (data key) seems to be about 10 to 15 times faster than the equivalent (get-in data [key]) on my laptop. It cut the runtime

Re: Performance tuning for a simple SLE solver

2009-10-11 Thread B Smith-Mannschott
On Fri, Oct 9, 2009 at 13:38, andi.xeno...@googlemail.com wrote: > > Hi, > > I'm new to Clojure, and let me first say that I love it! At least I > love the language, but I have some concerns regarding performance: > > My first try was to implement a Gauß elemination algorithm for solving > a syst

Performance tuning for a simple SLE solver

2009-10-09 Thread andi.xeno...@googlemail.com
Hi, I'm new to Clojure, and let me first say that I love it! At least I love the language, but I have some concerns regarding performance: My first try was to implement a Gauß elemination algorithm for solving a system of linear equations. Here is the code: http://www.xenoage.com/extern/zongblo