1
0
Fork 0
mirror of https://github.com/sockspls/badfish synced 2025-05-01 17:19:36 +00:00

Merge Joona's new aspiration window search

It seems very positive.

After 999 games at 1'+0" result is: +249 -216 =534 +11 ELO

And after another 456 games we still have: +122 -113 =221
This commit is contained in:
Marco Costalba 2009-04-16 08:35:26 +02:00
commit 4634be8ba6

View file

@ -47,6 +47,26 @@ namespace {
/// Types /// Types
// IterationInfoType stores search results for each iteration
//
// Because we use relatively small (dynamic) aspiration window,
// there happens many fail highs and fail lows in root. And
// because we don't do researches in those cases, "value" stored
// here is not necessarily exact. Instead in case of fail high/low
// we guess what the right value might be and store our guess
// as a "speculated value" and then move on. Speculated values are
// used just to calculate aspiration window width, so also if are
// not exact is not big a problem.
struct IterationInfoType {
IterationInfoType(Value v = Value(0), Value sv = Value(0))
: value(v), speculatedValue(sv) {}
Value value, speculatedValue;
};
// The BetaCounterType class is used to order moves at ply one. // The BetaCounterType class is used to order moves at ply one.
// Apart for the first one that has its score, following moves // Apart for the first one that has its score, following moves
// normally have score -VALUE_INFINITE, so are ordered according // normally have score -VALUE_INFINITE, so are ordered according
@ -196,7 +216,7 @@ namespace {
BetaCounterType BetaCounter; BetaCounterType BetaCounter;
// Scores and number of times the best move changed for each iteration: // Scores and number of times the best move changed for each iteration:
Value ValueByIteration[PLY_MAX_PLUS_2]; IterationInfoType IterationInfo[PLY_MAX_PLUS_2];
int BestMoveChangesByIteration[PLY_MAX_PLUS_2]; int BestMoveChangesByIteration[PLY_MAX_PLUS_2];
// MultiPV mode // MultiPV mode
@ -214,6 +234,7 @@ namespace {
bool AbortSearch; bool AbortSearch;
bool Quit; bool Quit;
bool FailHigh; bool FailHigh;
bool FailLow;
bool Problem; bool Problem;
bool PonderingEnabled; bool PonderingEnabled;
int ExactMaxTime; int ExactMaxTime;
@ -246,13 +267,10 @@ namespace {
/// Functions /// Functions
Value id_loop(const Position &pos, Move searchMoves[]); Value id_loop(const Position &pos, Move searchMoves[]);
Value root_search(Position &pos, SearchStack ss[], RootMoveList &rml); Value root_search(Position &pos, SearchStack ss[], RootMoveList &rml, Value alpha, Value beta);
Value search_pv(Position &pos, SearchStack ss[], Value alpha, Value beta, Value search_pv(Position &pos, SearchStack ss[], Value alpha, Value beta, Depth depth, int ply, int threadID);
Depth depth, int ply, int threadID); Value search(Position &pos, SearchStack ss[], Value beta, Depth depth, int ply, bool allowNullmove, int threadID);
Value search(Position &pos, SearchStack ss[], Value beta, Value qsearch(Position &pos, SearchStack ss[], Value alpha, Value beta, Depth depth, int ply, int threadID);
Depth depth, int ply, bool allowNullmove, int threadID);
Value qsearch(Position &pos, SearchStack ss[], Value alpha, Value beta,
Depth depth, int ply, int threadID);
void sp_search(SplitPoint *sp, int threadID); void sp_search(SplitPoint *sp, int threadID);
void sp_search_pv(SplitPoint *sp, int threadID); void sp_search_pv(SplitPoint *sp, int threadID);
void init_node(SearchStack ss[], int ply, int threadID); void init_node(SearchStack ss[], int ply, int threadID);
@ -380,6 +398,7 @@ void think(const Position &pos, bool infinite, bool ponder, int side_to_move,
AbortSearch = false; AbortSearch = false;
Quit = false; Quit = false;
FailHigh = false; FailHigh = false;
FailLow = false;
Problem = false; Problem = false;
ExactMaxTime = maxTime; ExactMaxTime = maxTime;
@ -654,14 +673,13 @@ namespace {
ss[i].init(i); ss[i].init(i);
ss[i].initKillers(); ss[i].initKillers();
} }
ValueByIteration[0] = Value(0); IterationInfo[1] = IterationInfoType(rml.get_move_score(0), rml.get_move_score(0));
ValueByIteration[1] = rml.get_move_score(0);
Iteration = 1; Iteration = 1;
EasyMove = rml.scan_for_easy_move(); EasyMove = rml.scan_for_easy_move();
// Iterative deepening loop // Iterative deepening loop
while (!AbortSearch && Iteration < PLY_MAX) while (Iteration < PLY_MAX)
{ {
// Initialize iteration // Initialize iteration
rml.sort(); rml.sort();
@ -672,8 +690,62 @@ namespace {
std::cout << "info depth " << Iteration << std::endl; std::cout << "info depth " << Iteration << std::endl;
// Calculate dynamic search window based on previous iterations
Value alpha, beta;
if (MultiPV == 1 && Iteration >= 6)
{
int prevDelta1 = IterationInfo[Iteration - 1].speculatedValue - IterationInfo[Iteration - 2].speculatedValue;
int prevDelta2 = IterationInfo[Iteration - 2].speculatedValue - IterationInfo[Iteration - 3].speculatedValue;
int delta = Max(2 * abs(prevDelta1) + abs(prevDelta2), ProblemMargin);
alpha = Max(IterationInfo[Iteration - 1].value - delta, -VALUE_INFINITE);
beta = Min(IterationInfo[Iteration - 1].value + delta, VALUE_INFINITE);
}
else
{
alpha = - VALUE_INFINITE;
beta = VALUE_INFINITE;
}
// Search to the current depth // Search to the current depth
ValueByIteration[Iteration] = root_search(p, ss, rml); Value value = root_search(p, ss, rml, alpha, beta);
// Write PV to transposition table, in case the relevant entries have
// been overwritten during the search.
TT.insert_pv(p, ss[0].pv);
if (AbortSearch)
break; // Value cannot be trusted. Break out immediately!
//Save info about search result
Value speculatedValue;
bool fHigh = false;
bool fLow = false;
Value delta = value - IterationInfo[Iteration - 1].value;
if (value >= beta)
{
assert(delta > 0);
fHigh = true;
speculatedValue = value + delta;
BestMoveChangesByIteration[Iteration] += 2; // Allocate more time
}
else if (value <= alpha)
{
assert(value == alpha);
assert(delta < 0);
fLow = true;
speculatedValue = value + delta;
BestMoveChangesByIteration[Iteration] += 3; // Allocate more time
} else
speculatedValue = value;
speculatedValue = Min(Max(speculatedValue, -VALUE_INFINITE), VALUE_INFINITE);
IterationInfo[Iteration] = IterationInfoType(value, speculatedValue);
// Erase the easy move if it differs from the new best move // Erase the easy move if it differs from the new best move
if (ss[0].pv[0] != EasyMove) if (ss[0].pv[0] != EasyMove)
@ -692,13 +764,15 @@ namespace {
// Stop search early when the last two iterations returned a mate score // Stop search early when the last two iterations returned a mate score
if ( Iteration >= 6 if ( Iteration >= 6
&& abs(ValueByIteration[Iteration]) >= abs(VALUE_MATE) - 100 && abs(IterationInfo[Iteration].value) >= abs(VALUE_MATE) - 100
&& abs(ValueByIteration[Iteration-1]) >= abs(VALUE_MATE) - 100) && abs(IterationInfo[Iteration-1].value) >= abs(VALUE_MATE) - 100)
stopSearch = true; stopSearch = true;
// Stop search early if one move seems to be much better than the rest // Stop search early if one move seems to be much better than the rest
int64_t nodes = nodes_searched(); int64_t nodes = nodes_searched();
if ( Iteration >= 8 if ( Iteration >= 8
&& !fLow
&& !fHigh
&& EasyMove == ss[0].pv[0] && EasyMove == ss[0].pv[0]
&& ( ( rml.get_move_cumulative_nodes(0) > (nodes * 85) / 100 && ( ( rml.get_move_cumulative_nodes(0) > (nodes * 85) / 100
&& current_search_time() > MaxSearchTime / 16) && current_search_time() > MaxSearchTime / 16)
@ -719,15 +793,13 @@ namespace {
if (stopSearch) if (stopSearch)
{ {
//FIXME: Implement fail-low emergency measures
if (!PonderSearch) if (!PonderSearch)
break; break;
else else
StopOnPonderhit = true; StopOnPonderhit = true;
} }
} }
// Write PV to transposition table, in case the relevant entries have
// been overwritten during the search:
TT.insert_pv(p, ss[0].pv);
if (MaxDepth && Iteration >= MaxDepth) if (MaxDepth && Iteration >= MaxDepth)
break; break;
@ -784,15 +856,23 @@ namespace {
// scheme (perhaps we should try to use this at internal PV nodes, too?) // scheme (perhaps we should try to use this at internal PV nodes, too?)
// and prints some information to the standard output. // and prints some information to the standard output.
Value root_search(Position &pos, SearchStack ss[], RootMoveList &rml) { Value root_search(Position &pos, SearchStack ss[], RootMoveList &rml, Value alpha, Value beta) {
Value alpha = -VALUE_INFINITE; Value oldAlpha = alpha;
Value beta = VALUE_INFINITE, value; Value value;
Bitboard dcCandidates = pos.discovered_check_candidates(pos.side_to_move()); Bitboard dcCandidates = pos.discovered_check_candidates(pos.side_to_move());
// Loop through all the moves in the root move list // Loop through all the moves in the root move list
for (int i = 0; i < rml.move_count() && !AbortSearch; i++) for (int i = 0; i < rml.move_count() && !AbortSearch; i++)
{ {
if (alpha >= beta)
{
// We failed high, invalidate and skip next moves, leave node-counters
// and beta-counters as they are and quickly return, we will try to do
// a research at the next iteration with a bigger aspiration window.
rml.set_move_score(i, -VALUE_INFINITE);
continue;
}
int64_t nodes; int64_t nodes;
Move move; Move move;
StateInfo st; StateInfo st;
@ -825,12 +905,12 @@ namespace {
if (i < MultiPV) if (i < MultiPV)
{ {
value = -search_pv(pos, ss, -beta, VALUE_INFINITE, newDepth, 1, 0); value = -search_pv(pos, ss, -beta, -alpha, newDepth, 1, 0);
// If the value has dropped a lot compared to the last iteration, // If the value has dropped a lot compared to the last iteration,
// set the boolean variable Problem to true. This variable is used // set the boolean variable Problem to true. This variable is used
// for time managment: When Problem is true, we try to complete the // for time managment: When Problem is true, we try to complete the
// current iteration before playing a move. // current iteration before playing a move.
Problem = (Iteration >= 2 && value <= ValueByIteration[Iteration-1] - ProblemMargin); Problem = (Iteration >= 2 && value <= IterationInfo[Iteration-1].value - ProblemMargin);
if (Problem && StopOnPonderhit) if (Problem && StopOnPonderhit)
StopOnPonderhit = false; StopOnPonderhit = false;
@ -855,7 +935,7 @@ namespace {
// was aborted because the user interrupted the search or because we // was aborted because the user interrupted the search or because we
// ran out of time. In this case, the return value of the search cannot // ran out of time. In this case, the return value of the search cannot
// be trusted, and we break out of the loop without updating the best // be trusted, and we break out of the loop without updating the best
// move and/or PV: // move and/or PV.
if (AbortSearch) if (AbortSearch)
break; break;
@ -874,7 +954,7 @@ namespace {
rml.set_move_score(i, -VALUE_INFINITE); rml.set_move_score(i, -VALUE_INFINITE);
else else
{ {
// New best move! // PV move or new best move!
// Update PV // Update PV
rml.set_move_score(i, value); rml.set_move_score(i, value);
@ -906,11 +986,12 @@ namespace {
LogFile << pretty_pv(pos, current_search_time(), Iteration, nodes_searched(), value, ss[0].pv) LogFile << pretty_pv(pos, current_search_time(), Iteration, nodes_searched(), value, ss[0].pv)
<< std::endl; << std::endl;
alpha = value; if (value > alpha)
alpha = value;
// Reset the global variable Problem to false if the value isn't too // Reset the global variable Problem to false if the value isn't too
// far below the final value from the last iteration. // far below the final value from the last iteration.
if (value > ValueByIteration[Iteration - 1] - NoProblemMargin) if (value > IterationInfo[Iteration - 1].value - NoProblemMargin)
Problem = false; Problem = false;
} }
else // MultiPV > 1 else // MultiPV > 1
@ -934,7 +1015,11 @@ namespace {
} }
alpha = rml.get_move_score(Min(i, MultiPV-1)); alpha = rml.get_move_score(Min(i, MultiPV-1));
} }
} } // New best move case
assert(alpha >= oldAlpha);
FailLow = (alpha == oldAlpha);
} }
return alpha; return alpha;
} }
@ -1083,7 +1168,7 @@ namespace {
// (from the computer's point of view) since the previous iteration: // (from the computer's point of view) since the previous iteration:
if ( ply == 1 if ( ply == 1
&& Iteration >= 2 && Iteration >= 2
&& -value <= ValueByIteration[Iteration-1] - ProblemMargin) && -value <= IterationInfo[Iteration-1].value - ProblemMargin)
Problem = true; Problem = true;
} }
@ -1796,7 +1881,7 @@ namespace {
// (from the computer's point of view) since the previous iteration. // (from the computer's point of view) since the previous iteration.
if ( sp->ply == 1 if ( sp->ply == 1
&& Iteration >= 2 && Iteration >= 2
&& -value <= ValueByIteration[Iteration-1] - ProblemMargin) && -value <= IterationInfo[Iteration-1].value - ProblemMargin)
Problem = true; Problem = true;
} }
lock_release(&(sp->lock)); lock_release(&(sp->lock));
@ -2425,8 +2510,8 @@ namespace {
return; return;
bool overTime = t > AbsoluteMaxSearchTime bool overTime = t > AbsoluteMaxSearchTime
|| (RootMoveNumber == 1 && t > MaxSearchTime + ExtraSearchTime) || (RootMoveNumber == 1 && t > MaxSearchTime + ExtraSearchTime && !FailLow) //FIXME: We are not checking any problem flags, BUG?
|| ( !FailHigh && !fail_high_ply_1() && !Problem || ( !FailHigh && !FailLow && !fail_high_ply_1() && !Problem
&& t > 6*(MaxSearchTime + ExtraSearchTime)); && t > 6*(MaxSearchTime + ExtraSearchTime));
if ( (Iteration >= 3 && (!InfiniteSearch && overTime)) if ( (Iteration >= 3 && (!InfiniteSearch && overTime))
@ -2447,8 +2532,8 @@ namespace {
(!InfiniteSearch && (StopOnPonderhit || (!InfiniteSearch && (StopOnPonderhit ||
t > AbsoluteMaxSearchTime || t > AbsoluteMaxSearchTime ||
(RootMoveNumber == 1 && (RootMoveNumber == 1 &&
t > MaxSearchTime + ExtraSearchTime) || t > MaxSearchTime + ExtraSearchTime && !FailLow) ||
(!FailHigh && !fail_high_ply_1() && !Problem && (!FailHigh && !FailLow && !fail_high_ply_1() && !Problem &&
t > 6*(MaxSearchTime + ExtraSearchTime))))) t > 6*(MaxSearchTime + ExtraSearchTime)))))
AbortSearch = true; AbortSearch = true;
} }