FieldAnalyzer.md

This is probably the most important class to understand how Agent2D works. As the name says for it self, this class is responsible for analyzing the field using a Voronoi Diagram to decide which action should be taken by the agent.

A brief explanation of a Voronoi Diagram can be found in this video: https://www.youtube.com/watch?v=Q804hv73L6U

To understand how this diagram is implemented, see the class Voronoi Diagram.

In SamplePlayer's actionImpl(), which is called in every game cycle, an instance of FieldAnalyzer is called too. Before this call, methods related to game strategy are also used (see Strategy)

void
SamplePlayer::actionImpl()
{
    //
    // update strategy and analyzer
    //
    Strategy::instance().update( world() );
    FieldAnalyzer::instance().update( world() );

   //[...]

Before generating a chain of actions, the field is analyzed at every game cycle. Note that is very important that you've grasped the concept of a Voronoi Diagram. The SamplePlayer class has the following source:

void
FieldAnalyzer::update( const WorldModel & wm )
{
    static GameTime s_update_time( 0, 0 );

    if ( s_update_time == wm.time() )
    {
        return;
    }
    s_update_time = wm.time();

    if ( wm.gameMode().type() == GameMode::BeforeKickOff
         || wm.gameMode().type() == GameMode::AfterGoal_
         || wm.gameMode().isPenaltyKickMode() )
    {
        return;
    }
    
#ifdef DEBUG_PRINT
    Timer timer;
#endif

    //updateVoronoiDiagram( wm );

#ifdef DEBUG_PRINT
    dlog.addText( Logger::TEAM,
                  "FieldAnalyzer::update() elapsed %f [ms]",
                  timer.elapsedReal() );
    writeDebugLog();
#endif

You can see that the method responsible for updating the Voronoi Diagram is commented out. More tests need to be done to know exactly what this method does. This method excludes some game modes in which the diagram is not analyzed. They are BeforeKickOff, AfterGoal_ e isPenaltyKickMode.