helper.py

class anav:  # The class encapsulating the environment
    '''
    Actions [0 : Stand, 1 : Up, 2 : Right, 3 : Down, 4 : Left]
    These are the representing no.s for the mentioned actions
    '''

    def __init__(self, length=30, width=30, goalPositions=[15,29]):
        
        # The player start at random locations
        
        self.pA=[np.random.randint(30),np.random.randint(30)] 
        self.pO=[6,8]
            
        
        self.h = length   # Length of the Football Pitch    
        self.w = width    # Width of the Football Pitch
        
        self.goalPositions = np.array(goalPositions)   # This means that the middle 4 positions at the right and left are the goals
        
     
        
        self.reward = 0                            # Initially the reward is 0
        
        self.observation=np.random.rand(6,)
        self.done = bool(0)                          # This stores whether the game needs to be restart with new position (in the case of a goal)

    def reset(self):
        self.done = bool(0)
        self.reward = 0
        
        self.pA = np.array([np.random.randint(self.h), np.random.randint(self.h)])
        self.pA = np.array([15,0])
        #self.pO=[6,8]
        self.x_traj,self.y_traj=[],[]
        self.xo_traj,self.yo_traj=[],[]
        return np.array((*self.pA,(15-self.pA[0]),(15-self.pA[1]),self.theta,*self.pO)).astype(np.float32)

    def step(self, action):
        if self.done == bool(1):
          self.reset()
        self.move(first, action)                   # We chose the first player at random
        if self.done == bool(1):
          return self.observation, self.reward, self.done
        if not done:
            self.current_player_num = (self.current_player_num + 1) % 2   
        return self.observation,self.reward, self.done,{}

    def move(self, player, action):
        opponent = 1 - player
        
        newPosition = self.pA + self.actionToMove(action)
        
        if self.ballOwner is player and self.isInGoal(*newPosition) >= 0:
            self.done = bool(1)
            return 1 - self.isInGoal(*newPosition)
        # If it's in the board
        elif self.isInBoard(*newPosition):
            self.positions[player] = newPosition
        if(self.ballOwner!=0):
          self.reward=-1
        return -1



    def isInBoard(self, x, y):
        if(x<0 or x>(30)):
          return 0
        if(y<0 or y>(30)):
          return 0 
        return 1
        

    #def choosePlayer(self):
    #    return np.random.randint(0, 2)
    def render(self,mode="human"):
        

        plt.cla()
        plt.arrow(x_start, y_start, np.cos(theta_start),
                      np.sin(theta_start), color='r', width=0.1)
        plt.arrow(x_goal, y_goal, np.cos(theta_goal),
                      np.sin(theta_goal), color='g', width=0.1)
        plot_vehicle(x, y, theta, x_traj, y_traj)