Die GNU autotools

[Shlomi Fish]

On Friday 14 Jan 2011 21:54:17 Shachar Shemesh wrote:
> On 14/01/11 21:51, Shlomi Fish wrote:
> > However it is a 138% improvement:
> > [shell]
> > shlomif:~$ perl -E 'say +(224/94-1)*100'
> > 138.297872340426
> > shlomif:~$ perl -E 'say 94*(138/100+1)'
> > 223.72
> > [/shell]
> 
> Your problem is with the percentages, not with the algebra. Your base
> speed is the speed before optimization. That is your 100%. You took your
> 100% to be the speed after optimizations, which is not the correct way
> to measure it.

Well, by that logic if my program initially took 100 seconds to run, and it 
ended up taking 1 second to run, then that would be a 99% improvement, instead 
of a 9,900% improvement, even though I've made my program 100 times faster. 
Similarly, cutting down the program to 50 seconds will make it a 50% 
improvement instead of a 100% speed improvement, which is only about half of 
the 99% improvement earlier.

One way to think about it is to measure speed not in seconds, but in 
frequency/Hz : i.e: how many programs can I run in a second. So the initial 
program has a freq of 0.01 Hz and the final program has a freq of 1 Hz which 
is a 9,900% improvement.

Regards,

	Shlomi Fish

-- 
-----------------------------------------------------------------
Shlomi Fish       http://www.shlomifish.org/
http://www.shlomifish.org/humour/ways_to_do_it.html

Chuck Norris can make the statement "This statement is false" a true one.

Please reply to list if it's a mailing list post - http://shlom.in/reply .