[Kst] New equation parser and interpreter

[Barth Netterfield]

Cool...

What did you do?

cbn

On February 12, 2004 12:44 pm, George Staikos wrote:
> I rewrote the equation parser and interpreter today and I think the
> preliminary results speak for themself:
>
> kst: Trying my parser on [2*sin(x)*cos(x) + 0.95e-1-10*log(2)].
> kst: Parsing time: 0
> kst: Execution time: 890000         <--------- new interpreter
> kst: FillY <<<<<<<
> kst: Execution time: 6840000        <--------- old interpreter
> kst: FillY >>>>>>>
>
>    That's about 13% of the previous time required to interpret an equation
> across one million data points - it's over 7 times faster!  Those units are
> in clock_t.  I expect this number to change as I finish implementing it
> because this was not calling sin(), cos(), or log() - that's not
> implemented yet.  Those just return 0.  However I think this is very
> promising.  Actually it's possible to make it faster than my implementation
> too, but I'm not sure it's worth the extra effort.