Comments on: cellfun – undocumented performance boost

By: Jordi Gutiérrez Hermoso

Jordi Gutiérrez Hermoso — Tue, 29 Nov 2011 16:30:38 +0000

I looked at your example, and I noticed that in Octave we hadn’t quite optimised this as much as possible. I went ahead and committed a change to fix this:

http://hg.savannah.gnu.org/hgweb/octave/rev/cf8cd43cdeb3

On my laptop with Intel Core 2 Duo @ 2.20G, I see the following:

octave:1> c = mat2cell(1:1e6,1,repmat(1,1,1e6));
octave:2> tic, d=cellfun('isempty',c); toc
Elapsed time is 0.0171831 seconds.
octave:3> tic, d=cellfun('isempty',c); toc
Elapsed time is 0.0182698 seconds.
octave:4> tic, d=cellfun('isempty',c); toc
Elapsed time is 0.0223808 seconds.
 
octave:5> tic, d=cellfun(@isempty,c); toc
Elapsed time is 0.0193319 seconds.
octave:6> tic, d=cellfun(@isempty,c); toc
Elapsed time is 0.01612 seconds.
octave:7> tic, d=cellfun(@isempty,c); toc
Elapsed time is 0.019449 seconds.

This should be part of our 3.6 release that should happen very soon!

Sadly, your preferred method of calling cellfun cannot be easily optimised:

octave:8> tic, d=cellfun(@(x) isempty(x),c); toc
Elapsed time is 0.924903 seconds.
octave:9> tic, d=cellfun(@(x) isempty(x),c); toc
Elapsed time is 0.873197 seconds.
octave:10> tic, d=cellfun(@(x) isempty(x),c); toc
Elapsed time is 0.875425 seconds.

Note that Octave still is single-threaded, so this does not benefit from any parallelisation right now. There’s work to build parallelisation into Octave, so perhaps we can see more dramatic speedups in the future.

By: Yair Altman

Yair Altman — Tue, 29 Nov 2011 15:05:23 +0000

@Jordi – thanks. If you have any other comparisons to Octave for any of the other articles here, please do post a comment.

By: Jordi Gutiérrez Hermoso

Jordi Gutiérrez Hermoso — Tue, 29 Nov 2011 14:59:38 +0000

This is very interesting. Our independent implementation of cellfun in Octave actually behaves very similarly! However, I did optimise it to check function handles for built-in string cases.

We have a thread about it:

http://octave.1599824.n4.nabble.com/More-cellfun-related-benchmarks-td3724314.html

By: tdc

tdc — Thu, 24 Nov 2011 11:48:11 +0000

I know this is old, but I just noticed it gets even worse if you use the other way of calling cellfun (which I’ve been using a lot):

>> c = mat2cell(1:1e6,1,repmat(1,1,1e6));
>> tic, d=cellfun('isempty',c); toc
Elapsed time is 0.034638 seconds.
 
>> tic, d=cellfun(@isempty,c); toc
Elapsed time is 0.859156 seconds.
 
>> tic, d=cellfun(@(x) isempty(x),c); toc
Elapsed time is 7.961039 seconds.

By: datestr performance | Undocumented Matlab

datestr performance | Undocumented Matlab — Wed, 05 Oct 2011 20:17:50 +0000

[...] Different performance hotspots can have different solutions: caching, vectorization, internal library functions, undocumented graphics properties, smart property selection, smart function selection, smart indexing, smart parameter selection etc. [...]

By: Ashish Sadanadnan

Ashish Sadanadnan — Wed, 13 May 2009 03:54:16 +0000

Yair,
I wasn’t disputing your results. Just wanted to show that the factor has improved significantly in newer version (65 to 17). Of course, 17 times faster is still very significant as you pointed out.

- Ashish.

By: Loren S

Loren S — Tue, 12 May 2009 11:10:39 +0000

Yair-

As I noted to you on my blog, MATLAB doesn’t convert from FH to string method because the user might have overridden whatever the method, e.g., isempty. MATLAB could, at runtime, see if it’s overridden, and if not, call the optimized version. But it can’t do that blindly without risk of wrong answers.

–Loren

By: Naor

Naor — Tue, 12 May 2009 01:12:26 +0000

wow. that’s a pretty significant unnecessary slowdown. at least this would be easy to catch with the profiler.

By: Yair Altman

Yair Altman — Mon, 11 May 2009 23:32:27 +0000

Ashish – actually your results show a factor of 17 between the slower @isempty and the faster ‘isempty’, consistent with the results I posted above (my reported factor of 65 is almost the same order of magnitude as 17, and may be due to external platform-dependent factors).

The absolute values of the results of course depend on the platform: my results were for a run-down heavily-loaded laptop… The important thing here is the factor between @isempty and ‘isempty’ – not the absolute values. And a factor of 17 is still high enough to be taken into consideration in a performance-critical application.

By: Ashish Sadanadnan

Ashish Sadanadnan — Mon, 11 May 2009 23:26:33 +0000

They seem to already have improved it quite a bit in R2009a; here are my results from running your code:

>> c = mat2cell(1:1e6,1,repmat(1,1,1e6));
>> tic, d=cellfun(‘isempty’,c); toc
Elapsed time is 0.032880 seconds.
>> tic, d=cellfun(@isempty,c); toc
Elapsed time is 0.563284 seconds.