Author: James Swafford
Date: 14:34:30 12/16/01
Go up one level in this thread
On December 16, 2001 at 14:53:43, Matthias Gemuh wrote: If you are probing the table for exact scores, and storing that move to pv (if appropriate), then what you're seeing is normal. It'd be pretty tough to get a pv "more complete" than that. -- James > >Hi Experts, > >I hate my program! >It outputs some solutions correctly (see last one), but some are thus truncated: > >2rq1rk1/1b1nbpp1/1p5p/p2pNB2/3p1N2/4P3/PPQ2PPP/2RR2K1 w - - bm Bh7+; id >"BWTC.462"; >[4] (+Mate in 6) 1.Bh7 Kh8 2.Nxf7 (00h:00m:00s) > >r4rk1/p1pb1qp1/3n1p1R/2pPpN2/2P5/P1PB4/4Q1PP/5RK1 w - - bm Ne7+; id "BWTC.510"; >[4] (+Mate in 5) 1.Ne7 Qxe7 2.Rh8 (00h:00m:00s) > >2k4r/ppp3p1/3p2q1/3P4/2P1p3/1P6/P3QbNB/R4R1K b - - bm Rxh2+; id "BWTC.584"; >[4] (-Mate in 6) 1 ...Rxh2 2.Kxh2 Qh6 (00h:00m:00s) > >5rk1/pR4pp/2p5/4p3/2P1P1q1/6P1/PP1Qn1BP/7K b - - bm Rf1+; id "BWTC.614"; >[7] (-Mate in 5) 1 ...Rf1 2.Bxf1 Qxe4 (00h:00m:00s) > >r2q1rk1/1b1nbppp/p1p1p3/1pPpN3/3P1N2/P2BP3/1PQ2PPP/R4RK1 w - - bm Bxh7+; id >"BWTC.846"; >[4] (+Mate in 6) 1.Bxh7 Kh8 2.Nxf7 (00h:00m:00s) > >5r1k/pR2Q1pp/8/q2pP3/3Pn3/4B1P1/7P/7K b - - bm Qe1+; id "BWTC.626"; >[4] (-Mate in 7) 1 ...Qe1 2.Bg1 Nf2 3.Kg2 Qe4 4.Kf1 Qd3 5.Kg2 Qf3 6.Kf1 Ne4 >7.Bf2 Qxf2 (00h:00m:00s) > > >I use Aspiration/PV Search and have implemented main hash table, null move, >internal iterative deepening, extensions, quiescence. >I collect the PV (also after retrieving exact hash score/move) as follows: > >PV_Length[nPly] = nPly; (at right place) > >Then later, >PV[nPly][nPly] = BigMoveListStack[i] (or HashMove); >for (int j = nPly + 1; j < PV_Length[nPly + 1]; j++) PV[nPly][j] = PV[nPly + >1][j]; >PV_Length[nPly] = PV_Length[nPly + 1]; > >How can I collect the PV more completely? I want my prog to spit out the full >string. > > >Thanks, >Matthias.
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.