Author: Robert Hyatt
Date: 07:27:27 09/07/02
Go up one level in this thread
On September 07, 2002 at 09:25:57, Ralf Elvsén wrote: >On September 06, 2002 at 11:53:13, Robert Hyatt wrote: > >>I have posted the raw data logs, the "cooked data" that I extracted from the >>logs, and the speedup tables (those for Martin last nite). It might be >>interesting to take the cb.c program I also posted and change the speedup >>format to show 3 decimel places (I used 2 as Martin had suggested that would >>be better.) >> >>It would be interesting to run the program with 1, 2 and 3 decimel place >>accuracy, and let everyone look at the three tables and decide which one >>_really_ provides the most useful information. I'll bet everyone likes >>.1 better than .11 because is .01 really significant? Or is it just random >>noise? >> >>I will let someone else run this as I have supplied the raw data and program >>on my ftp machine. that way I can't be accused of biasing the results in any >>way. :) > >Keep as many digits as your machine allows you until the bitter end. >Then you make an estimation of the uncertainty, and present a properly >rounded value together with the uncertainty. Rounding before this >stage is a no-no (although I'm not sure that was what you intended >to do). > >Ralf It depends on circumstances. IE in my dissertation, I had the freedom to run each test hundreds or thousands of times, which I did. For the DTS paper that was not an option. With a single position, precision to 2 decimel places is pointless as the test logs on my ftp machine shows...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.