[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Mata slow?
firstname.lastname@example.org (William Gould, Stata) writes:
> Brendan Halpin <email@example.com> explained that the Mata code we have
> been discussing is to be called _N*(_N-1)/2 times, so speed is important.
> To refresh your memories, Brendan wrote code that ran in .03906 seconds,
> which he noted was about 40 times slower than a plugin he wrote in C.
> I worked on the code on got the execution time down to .0228 seconds,
> or about 22 times slower than the plugin.
> Motivated by Brendan's latest comment, I have now improved the program so that
> it runs in .003 seconds. That's just 2 times slower than the plugin, and we
> have reduced the run time to just 7.5% of the original.
Many thanks for that! It really makes a difference.
Brendan Halpin, Department of Sociology, University of Limerick, Ireland
Tel: w +353-61-213147 f +353-61-202569 h +353-61-338562; Room F2-025 x 3147
* For searches and help try: