Re: [gnugo-devel] attack_either patches analysis

[Evan Berggren Daniel]

On Sun, 10 Nov 2002 address@hidden wrote:

>
> Evan wrote:
>
> > I did that, and the results look good.  Reading nodes in owl.tst go from
> > 25347652 (cvs) to 25278457 (with new patch), or about a 0.2% increase.
> > I'm not entirely sure how this translates to total time, or how comparable
> > it is to all of first_batch.
>
> Are these numbers switched?

They were.  Oops.

>
> I ran the regressions twice, with and without the patch, while
> I was not logged into this machine. I got 9001 seconds without
> the patch, and 9269. This is about a 3% increase in time. There
> are two new passes trevora:220 and 13x13:42.

Those are the same as before, which is good.  Thanks.

> The timing test needs to be repeated.

Probably so.  I'll try some depth restrictions on them too.

Thanks

Evan Daniel