IEvaluator Class Reference

Abstract interface for evaluators. More...

#include <IEvaluator.h>

Inheritance diagram for IEvaluator:

Public Member Functions
virtual	~IEvaluator ()=default
virtual float	evaluate (const Board &board, int ply=0)=0
	Evaluates the board from the perspective of the side to move.
virtual WDLConverter::WDL	evaluateWDL (const Board &board, int ply=0)
	Evaluates the board and returns Win/Draw/Loss probabilities.
virtual float	evaluate (const ChessInput &input)
	Evaluates the position using pre-computed features.
virtual void	setAggression (float aggression)
	Sets the contempt or aggression factor for the evaluator.
virtual void	setEvalScale (int base, int weight)
	Sets the scaling parameters for the evaluation score.
virtual void	setEvalNormalization (bool enable)
	Enables or disables dynamic evaluation normalization.
virtual void	setIncrementalRebuildInterval (int interval)
	Sets the interval for forcing full feature rebuilds (to correct accumulation errors).
virtual uint64_t	getFullRebuilds () const
	Retrieves the number of full feature rebuilds performed (for profiling).

Detailed Description

Abstract interface for evaluators.

Implementations of this interface are responsible for providing static evaluations of chess positions, translating them into centipawn scores or Win/Draw/Loss probabilities.

Constructor & Destructor Documentation

~IEvaluator()

virtual IEvaluator::~IEvaluator ( )

virtualdefault

Member Function Documentation

evaluate() [1/2]

virtual float IEvaluator::evaluate ( const Board & board, int ply = 0 )

pure virtual

Evaluates the board from the perspective of the side to move.

Parameters

board	The current chess board state.
ply	The current depth from the root of the search (used for scaling).

Returns

Evaluation score in centipawns (e.g., 150 = +1.5 pawns advantage).

Implemented in BatchEvalContext, EvalContextMoECache, and SimpleEvalContext.

Here is the caller graph for this function:

evaluate() [2/2]

virtual float IEvaluator::evaluate ( const ChessInput & input )

inlinevirtual

Evaluates the position using pre-computed features.

This is an optional optimization path for evaluators that utilize incremental feature extraction (like MoE neural networks).

Parameters

input

The pre-extracted neural network features.

Returns

Evaluation score in centipawns.

Reimplemented in BatchEvalContext.

evaluateWDL()

virtual WDLConverter::WDL IEvaluator::evaluateWDL ( const Board & board, int ply = 0 )

inlinevirtual

Evaluates the board and returns Win/Draw/Loss probabilities.

Parameters

board	The current chess board state.
ply	The current depth from the root of the search.

Returns

WDL probability distribution.

Reimplemented in EvalContextMoECache.

getFullRebuilds()

virtual uint64_t IEvaluator::getFullRebuilds ( ) const

inlinevirtual

Retrieves the number of full feature rebuilds performed (for profiling).

Returns

The number of full rebuilds.

Reimplemented in EvalContextMoECache.

setAggression()

virtual void IEvaluator::setAggression ( float aggression )

inlinevirtual

Sets the contempt or aggression factor for the evaluator.

Parameters

aggression

The aggression factor (positive encourages risk, negative encourages draws).

Reimplemented in EvalContextMoECache.

setEvalNormalization()

virtual void IEvaluator::setEvalNormalization ( bool enable )

inlinevirtual

Enables or disables dynamic evaluation normalization.

Parameters

enable

True to enable normalization, false to use raw scores.

Reimplemented in EvalContextMoECache.

setEvalScale()

virtual void IEvaluator::setEvalScale ( int base, int weight )

inlinevirtual

Sets the scaling parameters for the evaluation score.

Parameters

base	The base scaling divisor.
weight	The material weight scaling factor.

Reimplemented in EvalContextMoECache.

setIncrementalRebuildInterval()

virtual void IEvaluator::setIncrementalRebuildInterval ( int interval )

inlinevirtual

Sets the interval for forcing full feature rebuilds (to correct accumulation errors).

Parameters

interval

The number of incremental updates before forcing a rebuild.

Reimplemented in EvalContextMoECache.

---

The documentation for this class was generated from the following file:

• engine/src/eval/IEvaluator.h