[Shootout-list] Re: Python Reverse Complement

Alan Post apost@recalcitrant.org
Tue, 15 Mar 2005 21:27:41 +0000 (UTC)


In article <20050315211149.16920.qmail@web60508.mail.yahoo.com>, Isaac Gouy wrote:
>> What is the goal of the reverse-complement benchmark? 
> 
> Replacement for reverse-file.
> afaict Reverse-complement seems to be a common operation in sequence analysis.

By "replacement for reverse-file" do you mean "measure speed of
line-oriented I/O in standard library"?  I don't see what else is
being measured by reverse-complement.

The reverse-file benchmark doesn't actually require line-oriented I/O.
Instead, (if programs use the ocaml/C algorithm) it measures the speed
of:

 * building a linked list of large input buffers 
 * scanning through character buffers looking for line breaks
 * outputting many small buffers.

  Alan