pacai.gridworld.board

  1import typing
  2
  3import pacai.core.board
  4
  5MARKER_TERMINAL: pacai.core.board.Marker = pacai.core.board.Marker('T')
  6MARKER_DISPLAY_VALUE: pacai.core.board.Marker = pacai.core.board.Marker('V')
  7MARKER_DISPLAY_QVALUE: pacai.core.board.Marker = pacai.core.board.Marker('Q')
  8MARKER_SEPARATOR: pacai.core.board.Marker = pacai.core.board.Marker('X')
  9
 10AGENT_MARKER: pacai.core.board.Marker = pacai.core.board.MARKER_AGENT_0
 11""" The fixed marker of the only agent. """
 12
 13BOARD_COL_DELIM: str = ','
 14
 15class Board(pacai.core.board.Board):
 16    """
 17    A board for GridWorld.
 18
 19    In addition to walls, grid worlds also have terminal states.
 20    These will be represented as `T-?\\d`, so a 'T' followed by the value of the terminal state,
 21    e.g., `T1`, `T100`, `T-5`.
 22    """
 23
 24    def __init__(self, *args: typing.Any,
 25            additional_markers: list[str] | None = None,
 26            qdisplay: bool = False,
 27            **kwargs: typing.Any) -> None:
 28        """
 29        Construct a GridWorld board.
 30        If qdisplay is true, then the existing board will be extended to display Q-Values and policies.
 31        """
 32
 33        if (additional_markers is None):
 34            additional_markers = []
 35
 36        additional_markers += [
 37            MARKER_TERMINAL,
 38        ]
 39
 40        kwargs['strip'] = False
 41
 42        self._terminal_values: dict[pacai.core.board.Position, int] = {}
 43        """ Values for each terminal position. """
 44
 45        # Ensure that super's init() is called after self._terminal_values exists.
 46        super().__init__(*args, additional_markers = additional_markers, **kwargs)  # type: ignore
 47
 48        self._original_height: int = self.height
 49        """ The original height for this board. """
 50
 51        self._original_width: int = self.width
 52        """ The original width for this board. """
 53
 54        if (qdisplay):
 55            self._add_qvalue_display()
 56
 57    def _split_line(self, line: str) -> list[str]:
 58        # Skip empty lines.
 59        if (len(line.strip()) == 0):
 60            return []
 61
 62        return line.split(BOARD_COL_DELIM)
 63
 64    def _translate_marker(self, text: str, position: pacai.core.board.Position) -> pacai.core.board.Marker | None:
 65        # GridWorld markers may have additional whitespace or repeated characters.
 66        # Only simple markers (not terminals) can have repeated characters.
 67
 68        text = text.strip()
 69        if (len(text) == 0):
 70            text = ' '
 71
 72        if (not text.startswith(MARKER_TERMINAL)):
 73            return super()._translate_marker(text[0], position)
 74
 75        value = int(text[1:])
 76        self._terminal_values[position] = value
 77
 78        return MARKER_TERMINAL
 79
 80    def is_terminal_position(self, position: pacai.core.board.Position) -> bool:
 81        """ Check if the given position is a terminal. """
 82
 83        return (position in self._terminal_values)
 84
 85    def get_terminal_value(self, position: pacai.core.board.Position) -> int:
 86        """ Get the value of this terminal position (or raise an exception if the position is not a terminal). """
 87
 88        return self._terminal_values[position]
 89
 90    def display_qvalues(self) -> bool:
 91        """ Check if this board is displaying Q-Values. """
 92
 93        return len(self.get_marker_positions(MARKER_SEPARATOR)) > 0
 94
 95    def _add_qvalue_display(self) -> None:
 96        """
 97        Add a Q-Value display, which includes a section for values and q-values.
 98        The board will be doubled in size (along each dimension).
 99        The top-left will have the original board (with agent),
100        the top-right will have the values and policies,
101        and the bottom-left will have the q-values.
102        """
103
104        if (len(self.get_marker_positions(MARKER_SEPARATOR)) > 0):
105            raise ValueError("Already added Q-Value display.")
106
107        self._original_height = self.height
108        self._original_width = self.width
109        base_walls = self._walls.copy()
110
111        # Grow the board.
112        self.height = (self._original_height * 2) + 1
113        self.width = (self._original_width * 2) + 1
114
115        offset_values = pacai.core.board.Position(0, self._original_width + 1)  # Values (Top-Right)
116        offset_qvalues = pacai.core.board.Position(self._original_height + 1, 0)  # Q-Values (Bottom-Left)
117
118        # Add separators to quarter off sections of the new board.
119
120        for row in range(self.height):
121            self.place_marker(MARKER_SEPARATOR, pacai.core.board.Position(row, self._original_width))
122
123        for col in range(self.width):
124            self.place_marker(MARKER_SEPARATOR, pacai.core.board.Position(self._original_height, col))
125
126        # Empty out the blank section.
127        for row in range(self._original_height):
128            for col in range(self._original_width):
129                position = pacai.core.board.Position(self._original_height + 1 + row, self._original_width + 1 + col)
130                self.place_marker(MARKER_SEPARATOR, position)
131
132        # Duplicate the walls on the new sections.
133
134        for offset in [offset_values, offset_qvalues]:
135            for base_wall in base_walls:
136                self._walls.add(base_wall.add(offset))
137
138        # Place the markers to display values/q-values.
139
140        for base_row in range(self._original_height):
141            for base_col in range(self._original_width):
142                base_position = pacai.core.board.Position(base_row, base_col)
143                if (self.is_wall(base_position) or self.is_marker(MARKER_TERMINAL, base_position)):
144                    continue
145
146                self.place_marker(MARKER_DISPLAY_VALUE, base_position.add(offset_values))
147                self.place_marker(MARKER_DISPLAY_QVALUE, base_position.add(offset_qvalues))
148
149        # Copy terminal markers.
150
151        for (base_position, value) in list(self._terminal_values.items()):
152            for offset in [offset_values, offset_qvalues]:
153                position = base_position.add(offset)
154                self._terminal_values[position] = value
155                self.place_marker(MARKER_TERMINAL, position)
156
157    def to_dict(self) -> dict[str, typing.Any]:
158        data = super().to_dict()
159        data['_terminal_values'] = [(position.to_dict(), value) for (position, value) in self._terminal_values.items()]
160        return data
161
162    @classmethod
163    def from_dict(cls, data: dict[str, typing.Any]) -> typing.Any:
164        board = super().from_dict(data)
165        board._terminal_values = {pacai.core.board.Position.from_dict(raw): value for (raw, value) in data['_terminal_values']}
166        return board