On Wed, 9 Jan 2008 15:45:41 -0800 (PST) "[EMAIL PROTECTED]" <[EMAIL PROTECTED]> wrote:
> Okay I profiled the code and here is the output: > > http://heightened.files.wordpress.com/2008/01/output.txt > > It seems that the function it spends the longest on is the red_points > function that he uses to find the points. Ok, so what's that look like? Python - used properly - can be quite fast. We process 1.5 billion rows a day through our python-based ETL system. The trick is to arrange things so that the cpu-intensive work gets done by C code. Preferably by tightly coded C written by very sharp people and extensively tweaked to make it go fast. In our case, the T part of our ETL system is handled by a custom C library that's been around - and being debugged and tweaked - for quite some time. Other examples include much of the code for python's builtins, and things like the Numpy extension library. <mike -- Mike Meyer <[EMAIL PROTECTED]> http://www.mired.org/consulting.html Independent Network/Unix/Perforce consultant, email for more information. -- http://mail.python.org/mailman/listinfo/python-list
