timeit and (micro)benchmarks (was Re: [PATCH] performance: disable workaround for an old bug of Python)

Bryan O'Sullivan bos at serpentine.com
Tue Aug 23 12:13:09 EDT 2016


On Mon, Aug 22, 2016 at 7:13 AM, Augie Fackler <raf at durin42.com> wrote:

> You've got a lot more experience with Python performance than I - for
> little microbenchmarks is timeit still not good? If so, what would you
> recommend for (semi-disposable) microbenchmarks?
>

I don't believe there are any trustworthy Python benchmarking libraries.

The state of the art is advancing fairly quickly in this area in general.
Here's a good paper that was put up on arxiv just last week:
http://arxiv.org/pdf/1608.04295.pdf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mercurial-scm.org/pipermail/mercurial-devel/attachments/20160823/869e44fc/attachment.html>


More information about the Mercurial-devel mailing list