synth-ai 0.2.14__py3-none-any.whl → 0.4.4__py3-none-any.whl

synth_ai/environments/examples/red/engine_helpers/reward_library/pallet_town_rewards.py

@@ -1,559 +0,0 @@
-"""
-Pallet Town Early Game Reward Components
-
-Rewards specifically designed for the first 50 steps of Pokemon Red,
-focusing on house exploration, town discovery, and story triggers.
-"""
-
-from typing import Any, Dict, Set
-
-from synth_ai.environments.environment.rewards.core import RewardComponent
-
-
-class LeaveStartingRoomReward(RewardComponent):
-    """Reward for going downstairs from bedroom - +15 points"""
-
-    def __init__(self):
-        self.triggered = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.triggered:
-            return 0.0
-
-        prev_map = action.get("prev_map_id", -1)
-        current_map = state["map_id"]
-        prev_y = action.get("prev_player_y", -1)
-        current_y = state["player_y"]
-
-        # Detect going downstairs from bedroom (map change + y coordinate change)
-        if prev_map != current_map and prev_y > current_y:
-            self.triggered = True
-            return 15.0
-        return 0.0
-
-
-class TalkToMomReward(RewardComponent):
-    """Reward for first conversation with mom - +10 points"""
-
-    def __init__(self):
-        self.mom_talked_to = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.mom_talked_to:
-            return 0.0
-
-        # Check if we're in mom's house and had a text interaction
-        if state["map_id"] in [1, 2] and state["text_box_active"]:  # Assuming house maps
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                self.mom_talked_to = True
-                return 10.0
-        return 0.0
-
-
-class InteractWithTVReward(RewardComponent):
-    """Reward for checking the TV downstairs - +5 points"""
-
-    def __init__(self):
-        self.tv_checked = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.tv_checked:
-            return 0.0
-
-        # Detect TV interaction in house
-        if state["map_id"] in [1, 2] and state["text_box_active"]:
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                # Simple heuristic: TV is usually in certain positions
-                player_x, player_y = state["player_x"], state["player_y"]
-                if (player_x, player_y) in [
-                    (3, 4),
-                    (4, 4),
-                    (5, 4),
-                ]:  # Common TV positions
-                    self.tv_checked = True
-                    return 5.0
-        return 0.0
-
-
-class CheckComputerReward(RewardComponent):
-    """Reward for interacting with PC in room - +5 points"""
-
-    def __init__(self):
-        self.pc_checked = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.pc_checked:
-            return 0.0
-
-        # Detect PC interaction in bedroom
-        if state["map_id"] == 1 and state["text_box_active"]:  # Bedroom
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                # PC is usually in upper right of bedroom
-                player_x, player_y = state["player_x"], state["player_y"]
-                if player_x >= 6 and player_y <= 3:
-                    self.pc_checked = True
-                    return 5.0
-        return 0.0
-
-
-class HouseFullyExploredReward(RewardComponent):
-    """Reward for checking all interactive objects in starting house - +20 points"""
-
-    def __init__(self):
-        self.interactions: Set[str] = set()
-        self.required_interactions = {"tv", "pc", "mom", "bookshelf", "poster"}
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if len(self.interactions) >= len(self.required_interactions):
-            return 0.0
-
-        # Track interactions in house
-        if state["map_id"] in [1, 2] and state["text_box_active"]:
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                player_x, player_y = state["player_x"], state["player_y"]
-
-                # Map positions to interaction types
-                if (player_x, player_y) in [(3, 4), (4, 4), (5, 4)]:
-                    self.interactions.add("tv")
-                elif player_x >= 6 and player_y <= 3:
-                    self.interactions.add("pc")
-                elif (player_x, player_y) in [(1, 4), (2, 4)]:
-                    self.interactions.add("mom")
-                # Add more position mappings as needed
-
-                if len(self.interactions) >= len(self.required_interactions):
-                    return 20.0
-        return 0.0
-
-
-class ExitHouseReward(RewardComponent):
-    """Reward for first time leaving the starting house - +20 points"""
-
-    def __init__(self):
-        self.house_exited = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.house_exited:
-            return 0.0
-
-        prev_map = action.get("prev_map_id", -1)
-        current_map = state["map_id"]
-
-        # Exit from house to town
-        if prev_map in [1, 2] and current_map == 0:  # House to Pallet Town
-            self.house_exited = True
-            return 20.0
-        return 0.0
-
-
-class ExploreTownReward(RewardComponent):
-    """Reward for each new building/house entered - +5 points"""
-
-    def __init__(self):
-        self.buildings_entered: Set[int] = set()
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        prev_map = action.get("prev_map_id", -1)
-        current_map = state["map_id"]
-
-        # Entering a new building from town
-        if (
-            prev_map == 0 and current_map > 0 and current_map not in [1, 2]
-        ):  # From town to new building
-            if current_map not in self.buildings_entered:
-                self.buildings_entered.add(current_map)
-                return 5.0
-        return 0.0
-
-
-class TalkToNPCsReward(RewardComponent):
-    """Reward for each unique NPC conversation in Pallet Town - +8 points"""
-
-    def __init__(self):
-        self.npcs_talked_to: Set[tuple] = set()
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        # Detect NPC conversations in Pallet Town
-        if state["map_id"] == 0 and state["text_box_active"]:  # Pallet Town
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                # Use position as NPC identifier
-                npc_key = (state["player_x"], state["player_y"], state["map_id"])
-                if npc_key not in self.npcs_talked_to:
-                    self.npcs_talked_to.add(npc_key)
-                    return 8.0
-        return 0.0
-
-
-class OakLabDiscoveryReward(RewardComponent):
-    """Reward for finding and entering Oak's lab - +25 points"""
-
-    def __init__(self):
-        self.lab_discovered = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.lab_discovered:
-            return 0.0
-
-        prev_map = action.get("prev_map_id", -1)
-        current_map = state["map_id"]
-
-        # Entering Oak's lab (map 3)
-        if prev_map == 0 and current_map == 3:
-            self.lab_discovered = True
-            return 25.0
-        return 0.0
-
-
-class AttemptRoute1Reward(RewardComponent):
-    """Reward for trying to leave town (triggers Oak encounter) - +30 points"""
-
-    def __init__(self):
-        self.route_attempted = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.route_attempted:
-            return 0.0
-
-        # Detect reaching the edge of Pallet Town (attempting to go north)
-        if state["map_id"] == 0:  # In Pallet Town
-            if state["player_y"] <= 1:  # At northern edge
-                self.route_attempted = True
-                return 30.0
-        return 0.0
-
-
-class OakEncounterReward(RewardComponent):
-    """Reward for triggering Professor Oak to stop you - +50 points"""
-
-    def __init__(self):
-        self.oak_encountered = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.oak_encountered:
-            return 0.0
-
-        # Detect Oak stopping you (usually involves specific dialogue)
-        if state["text_box_active"] and not action.get("prev_text_box_active", False):
-            # Check if we're in a situation where Oak would appear
-            if state["map_id"] == 0 and state["player_y"] <= 2:
-                self.oak_encountered = True
-                return 50.0
-        return 0.0
-
-
-class FollowOakToLabReward(RewardComponent):
-    """Reward for returning to lab with Oak - +40 points"""
-
-    def __init__(self):
-        self.followed_oak = False
-        self.oak_encounter_happened = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.followed_oak:
-            return 0.0
-
-        # Track Oak encounter first
-        if not self.oak_encounter_happened and state["map_id"] == 0 and state["player_y"] <= 2:
-            if state["text_box_active"]:
-                self.oak_encounter_happened = True
-
-        # Then reward entering lab after encounter
-        if self.oak_encounter_happened:
-            prev_map = action.get("prev_map_id", -1)
-            current_map = state["map_id"]
-            if prev_map == 0 and current_map == 3:  # Town to lab
-                self.followed_oak = True
-                return 40.0
-        return 0.0
-
-
-class ChooseStarterPokemonReward(RewardComponent):
-    """Reward for selecting first Pokemon - +100 points"""
-
-    def __init__(self):
-        self.starter_chosen = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.starter_chosen:
-            return 0.0
-
-        # Detect getting first Pokemon
-        prev_party_count = len(action.get("prev_party", []))
-        current_party_count = len(state.get("party", []))
-
-        if prev_party_count == 0 and current_party_count == 1:
-            if state["map_id"] == 3:  # In Oak's lab
-                self.starter_chosen = True
-                return 100.0
-        return 0.0
-
-
-class RivalEncounterReward(RewardComponent):
-    """Reward for meeting and naming rival - +30 points"""
-
-    def __init__(self):
-        self.rival_met = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.rival_met:
-            return 0.0
-
-        # Detect rival encounter (specific dialogue patterns)
-        if state["map_id"] == 3 and state["text_box_active"]:  # In Oak's lab
-            # This is a simplified check - in reality you'd analyze dialogue content
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active and state["text_box_active"]:
-                # Check if we have at least one Pokemon (starter chosen)
-                if len(state.get("party", [])) >= 1:
-                    self.rival_met = True
-                    return 30.0
-        return 0.0
-
-
-class FirstPokemonBattleReward(RewardComponent):
-    """Reward for the first battle with rival - +75 points"""
-
-    def __init__(self):
-        self.first_battle_done = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.first_battle_done:
-            return 0.0
-
-        # Detect entering battle for first time
-        prev_in_battle = action.get("prev_in_battle", False)
-        current_in_battle = state["in_battle"]
-
-        if not prev_in_battle and current_in_battle:
-            if state["map_id"] == 3:  # In Oak's lab
-                self.first_battle_done = True
-                return 75.0
-        return 0.0
-
-
-class MenuDiscoveryReward(RewardComponent):
-    """Reward for opening START menu for first time - +10 points"""
-
-    def __init__(self):
-        self.menu_discovered = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.menu_discovered:
-            return 0.0
-
-        # This would need menu state tracking - simplified implementation
-        # In real implementation, you'd track when START is pressed
-        buttons_pressed = action.get("buttons_pressed", [])
-        if "START" in buttons_pressed and not self.menu_discovered:
-            self.menu_discovered = True
-            return 10.0
-        return 0.0
-
-
-class PokemonMenuReward(RewardComponent):
-    """Reward for checking Pokemon party status - +15 points"""
-
-    def __init__(self):
-        self.pokemon_menu_checked = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.pokemon_menu_checked:
-            return 0.0
-
-        # Simplified - would need menu navigation tracking
-        # This is a placeholder for actual menu state detection
-        if len(state.get("party", [])) > 0:  # Has Pokemon
-            # Assume menu was checked if we have Pokemon and certain conditions
-            self.pokemon_menu_checked = True
-            return 15.0
-        return 0.0
-
-
-class BagDiscoveryReward(RewardComponent):
-    """Reward for opening bag/items menu - +10 points"""
-
-    def __init__(self):
-        self.bag_discovered = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.bag_discovered:
-            return 0.0
-
-        # Simplified implementation
-        if len(state.get("inventory", [])) > 0:
-            self.bag_discovered = True
-            return 10.0
-        return 0.0
-
-
-class SaveGameReward(RewardComponent):
-    """Reward for saving the game for first time - +20 points"""
-
-    def __init__(self):
-        self.game_saved = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.game_saved:
-            return 0.0
-
-        # This would need save state detection
-        # Placeholder implementation
-        if state.get("game_saved", False):
-            self.game_saved = True
-            return 20.0
-        return 0.0
-
-
-class TryAllDirectionsReward(RewardComponent):
-    """Reward for attempting movement in all 4 directions - +5 points"""
-
-    def __init__(self):
-        self.directions_tried: Set[str] = set()
-        self.reward_given = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.reward_given:
-            return 0.0
-
-        # Track movement directions
-        buttons_pressed = action.get("buttons_pressed", [])
-        for button in buttons_pressed:
-            if button in ["UP", "DOWN", "LEFT", "RIGHT"]:
-                self.directions_tried.add(button)
-
-        if len(self.directions_tried) >= 4:
-            self.reward_given = True
-            return 5.0
-        return 0.0
-
-
-class DoorInteractionReward(RewardComponent):
-    """Reward for trying to enter each door/building - +3 points per door"""
-
-    def __init__(self):
-        self.doors_tried: Set[tuple] = set()
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        # Detect door interaction attempts
-        prev_map = action.get("prev_map_id", -1)
-        current_map = state["map_id"]
-
-        if prev_map != current_map and prev_map == 0:  # From town to building
-            door_key = (state["player_x"], state["player_y"], current_map)
-            if door_key not in self.doors_tried:
-                self.doors_tried.add(door_key)
-                return 3.0
-        return 0.0
-
-
-class ObjectInteractionReward(RewardComponent):
-    """Reward for pressing A on various objects - +3 points per object"""
-
-    def __init__(self):
-        self.objects_interacted: Set[tuple] = set()
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        # Detect A button interactions
-        buttons_pressed = action.get("buttons_pressed", [])
-        if "A" in buttons_pressed and state["text_box_active"]:
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active:
-                object_key = (state["player_x"], state["player_y"], state["map_id"])
-                if object_key not in self.objects_interacted:
-                    self.objects_interacted.add(object_key)
-                    return 3.0
-        return 0.0
-
-
-class SignReadingReward(RewardComponent):
-    """Reward for reading town sign and other informational signs - +5 points"""
-
-    def __init__(self):
-        self.signs_read: Set[tuple] = set()
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        # Detect sign reading (specific positions in town)
-        if state["map_id"] == 0 and state["text_box_active"]:  # Pallet Town
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active:
-                # Town sign is usually at specific coordinates
-                sign_positions = [(5, 8), (6, 8), (7, 8)]  # Common sign positions
-                player_pos = (state["player_x"], state["player_y"])
-                if player_pos in sign_positions:
-                    sign_key = (state["player_x"], state["player_y"])
-                    if sign_key not in self.signs_read:
-                        self.signs_read.add(sign_key)
-                        return 5.0
-        return 0.0
-
-
-class CompleteTownExplorationReward(RewardComponent):
-    """Reward for visiting every accessible location - +50 points"""
-
-    def __init__(self):
-        self.locations_visited: Set[tuple] = set()
-        self.required_locations = 20  # Estimated accessible tiles in Pallet Town
-        self.reward_given = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.reward_given:
-            return 0.0
-
-        if state["map_id"] == 0:  # In Pallet Town
-            location_key = (state["player_x"], state["player_y"])
-            self.locations_visited.add(location_key)
-
-            if len(self.locations_visited) >= self.required_locations:
-                self.reward_given = True
-                return 50.0
-        return 0.0
-
-
-class AllNPCsTalkedToReward(RewardComponent):
-    """Reward for speaking with every NPC in town - +30 points"""
-
-    def __init__(self):
-        self.npcs_talked_to: Set[tuple] = set()
-        self.required_npcs = 5  # Estimated NPCs in Pallet Town
-        self.reward_given = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.reward_given:
-            return 0.0
-
-        # Track NPC conversations
-        if state["map_id"] == 0 and state["text_box_active"]:
-            prev_text_active = action.get("prev_text_box_active", False)
-            if not prev_text_active:
-                npc_key = (state["player_x"], state["player_y"])
-                self.npcs_talked_to.add(npc_key)
-
-                if len(self.npcs_talked_to) >= self.required_npcs:
-                    self.reward_given = True
-                    return 30.0
-        return 0.0
-
-
-class ReadyForAdventureReward(RewardComponent):
-    """Reward for having starter Pokemon and being ready to leave town - +60 points"""
-
-    def __init__(self):
-        self.ready_reward_given = False
-
-    async def score(self, state: Dict[str, Any], action: Dict[str, Any]) -> float:
-        if self.ready_reward_given:
-            return 0.0
-
-        # Check if player has starter and is at town exit
-        has_pokemon = len(state.get("party", [])) > 0
-        at_town_exit = state["map_id"] == 0 and state["player_y"] <= 2
-
-        if has_pokemon and at_town_exit:
-            self.ready_reward_given = True
-            return 60.0
-        return 0.0