Search found 132 matches

by Fabio Gobbato
Sat Oct 11, 2014 5:37 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A little improvement to the Texel's tuning method
Replies: 9
Views: 2950

A little improvement to the Texel's tuning method

I have used it to tune my engine Pedone with very good results. I have downloaded about 200000 games from CCRL played by the top engine at 40/40 time control. From each game I have extracted 7-8 random positions and I have put together in a file (fen.csv) the FEN string of every position, the final ...
by Fabio Gobbato
Tue Oct 07, 2014 2:35 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Your experience with PVS + Aspiration window
Replies: 3
Views: 1819

Your experience with PVS + Aspiration window

What is your experience with PVS + Aspiration window?
Is there a big difference between use aspiration window or not use it?
by Fabio Gobbato
Thu Oct 02, 2014 4:37 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Parallel search
Replies: 7
Views: 1466

Re: Parallel search

That solutions solved some error related to the compilation. Then there was some problems with uninitialized variables caught with valgrind. Now I solved the problem, I have used similar names for the mutex so I lock with one and unlock with the other. That was the problem and because I used really ...
by Fabio Gobbato
Wed Oct 01, 2014 12:20 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Parallel search
Replies: 7
Views: 1466

Re: Parallel search

I have solved it starting a new codeblocks project.
I'm sorry but I wasn't able to understand the problem.
by Fabio Gobbato
Tue Sep 30, 2014 7:42 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Parallel search
Replies: 7
Views: 1466

Parallel search

When I launch a similar code I always get threads blocked for waiting a lock. Maybe it's a stupid thing but I can't understand where the code could leave a lock locked. Could someone help me? idle_loop { mutex_lock(SMPLock); ... cond_wait(Work,SMPLock); ... mutex_unlock(SMPL...
by Fabio Gobbato
Thu Sep 11, 2014 11:02 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Tune cut margins with Texel/gaviota tuning method
Replies: 9
Views: 1515

Tune cut margins with Texel/gaviota tuning method

In my engine I used a similar approach to Texel/Gaviota tuning method with good results. I was wondering about tune with this method futility-delta pruning and razoring margin. The advantage is that you can tune this values taking account the evaluation error of the search function of millions posit...
by Fabio Gobbato
Sun Aug 17, 2014 7:11 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killer and move encoding
Replies: 25
Views: 4778

Re: Killer and move encoding

I use killer moves from the same ply but not necessary have the same parent node.
And I also reuse killers from the last search shifted by two ply.
by Fabio Gobbato
Fri Aug 15, 2014 7:14 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killer and move encoding
Replies: 25
Views: 4778

Re: Killer and move encoding

It could swap a bishop move with a queen move but also a rook move with a pawn move.
I haven't checked every time the search use such a killer, I have only made a test with the two different solutions.
by Fabio Gobbato
Thu Aug 14, 2014 5:51 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killer and move encoding
Replies: 25
Views: 4778

Re: Killer and move encoding

Validate a killer if it's legal of course. For instance if you save a killer with only "from" and "to" you could store a move of a bishop that goes from c1 to f4 and when you have to use this killer it could match with a queen that goes from c1 to f4 because you don't check the piece type but only t...
by Fabio Gobbato
Thu Aug 14, 2014 4:38 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killer and move encoding
Replies: 25
Views: 4778

Re: Killer and move encoding

6 bit for origin square 6 bit for destination square 2 bit promotion or flag 2 bit flag I would like to use the same encoding for hash, killer and move lists but it seems that validate killers with only "from" and "to" squares behave differently from validate them with "piece type", "from", and "to"...