It wasn't very complicated, I just started from zero experience worrying about
this stuff.
- switch map to loop+recur
- add primitive type hints to args and return values
- make sure you get the types right (e.g. make sure your longs are longs, not
ints)
Cheers, Jay
On Mar 28, 2012, at 10:17
On Tuesday, March 27, 2012 2:51:41 PM UTC-4, Jay Fields wrote:
I recently ran into some code** that was in Java, and ran in single
digit microseconds (not millis). I converted it to clojure, and got it
running at about the same speed... though it did take me a day to
figure out all the
On Mon, Mar 26, 2012 at 4:03 PM, Daniel Solano Gomez
cloj...@sattvik.com wrote:
On Mon Mar 26 15:15 2012, Cedric Greevey wrote:
What you're running into is the overhead of get the root value of vars
like swap! and move-up!. The only way to avoid this is use something
like definline.
I
An alternative to the complexities involved may be to write
number-crunching hot spots in Java. Thus I'd add to Tip 10 that it ain't so
bad if you exploit some JVM-platform-goodness *you* just wrote; by
definition it's not much code, it's much simpler and can be done in a
functional style.
On
On Mon, Mar 26, 2012 at 3:37 AM, Sergey Didenko sergey.dide...@gmail.comwrote:
Hi,
I believe I've heard claims that nothing stops Clojure 1.3 code to be
made very close to Java in terms of execution speed.
My personal experience is that it is very difficult to get Java execution
speed.
On
On Tue, Mar 27, 2012 at 2:18 PM, Mark Engelberg mark.engelb...@gmail.comwrote:
On Mon, Mar 26, 2012 at 3:37 AM, Sergey Didenko
sergey.dide...@gmail.comwrote:
Hi,
I believe I've heard claims that nothing stops Clojure 1.3 code to be
made very close to Java in terms of execution speed.
I recently ran into some code** that was in Java, and ran in single
digit microseconds (not millis). I converted it to clojure, and got it
running at about the same speed... though it did take me a day to
figure out all the tweaks.
It can be done, if you're willing to invest the time and learn
You can spot a weak argument when euphemisms and stuff we are
yet-to-encounter start to pop up.
I find puzzler's experience exactly to mine, point by point, and given the
little talk of multi-threading coding in this board I'd expect the same to
apply to many others. If anything, the
On Tue, Mar 27, 2012 at 3:12 PM, Armando Blancas abm221...@gmail.comwrote:
If anything, the java-in-parens gvec.clj *proves* how difficult it is to
get Java perf in Clojure. The good real world perf looks a lot like Java,
regardless of where you place the parens. So I agree with puzzler, Java
That's a distinction without a difference. I'm saying the host interop
overwhelms the cool stuff. That's the kind of code I find foolish to write
in product development. I don't advocate going back and forth to Java, but
when the host takes over, that's sign that *it* can be the simpler, better
On Tue, Mar 27, 2012 at 5:28 PM, Armando Blancas abm221...@gmail.comwrote:
That's a distinction without a difference. I'm saying the host interop
overwhelms the cool stuff. That's the kind of code I find foolish to write
in product development. I don't advocate going back and forth to Java,
On Tue, Mar 27, 2012 at 5:28 PM, Armando Blancas abm221...@gmail.com wrote:
That's a distinction without a difference. I'm saying the host interop
overwhelms the cool stuff. That's the kind of code I find foolish to write
in product development. I don't advocate going back and forth to Java,
Hi,
I believe I've heard claims that nothing stops Clojure 1.3 code to be
made very close to Java in terms of execution speed.
However when trying to match the speed of ad-hoc Heapsort
implementation I have faced essential obstacles.
Setting type hints and coercions was quite easy. A bit harder
I suggest examining the examples in test.benchmark to see how Java numeric
performance can be attained, http://github.com/clojure/test.benchmark
On Monday, March 26, 2012, Sergey Didenko sergey.dide...@gmail.com wrote:
Hi,
I believe I've heard claims that nothing stops Clojure 1.3 code to be
Hello,
Having given a talk on precisely this topic at Clojure/West, I may be
able to give you a couple of hints and explanations. You can find the
slides at
https://github.com/strangeloop/clojurewest2012-slides/raw/master/Solano-G%C3%B3mez-Crunching-Numbers-with-Clojure.pdf,
and I am planning on
What you're running into is the overhead of get the root value of vars
like swap! and move-up!. The only way to avoid this is use something
like definline.
I thought that in 1.3 those had been made much faster to access if
they didn't have ^{:dynamic true} applied to them?
--
You received
On Mon Mar 26 15:15 2012, Cedric Greevey wrote:
What you're running into is the overhead of get the root value of vars
like swap! and move-up!. The only way to avoid this is use something
like definline.
I thought that in 1.3 those had been made much faster to access if
they didn't have
17 matches
Mail list logo
