gym-csle-stopping-game 0.6.3__tar.gz → 0.6.5__tar.gz

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym_csle_stopping_game
-Version: 0.6.3
+Version: 0.6.5
 Summary: OpenAI gym reinforcement learning environment of a Dynkin (Optimal stopping) game in CSLE
 Author: Kim Hammar
 Author-email: hammar.kim@gmail.com

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/setup.cfg

@@ -20,11 +20,11 @@ classifiers =
 [options]
 install_requires = 
 	gymnasium>=0.27.1
-	csle-base>=0.6.3
-	csle-common>=0.6.3
-	csle-attacker>=0.6.3
-	csle-defender>=0.6.3
-	csle-collector>=0.6.3
+	csle-base>=0.6.5
+	csle-common>=0.6.5
+	csle-attacker>=0.6.5
+	csle-defender>=0.6.5
+	csle-collector>=0.6.5
 python_requires = >=3.8
 package_dir = 
 	=src

gym_csle_stopping_game-0.6.5/src/gym_csle_stopping_game/__version__.py

@@ -0,0 +1 @@
+__version__ = '0.6.5'

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game/dao/stopping_game_config.py

@@ -14,7 +14,7 @@ class StoppingGameConfig(SimulationEnvInputConfig):
                  T: npt.NDArray[Any], O: npt.NDArray[np.int_], Z: npt.NDArray[Any],
                  R: npt.NDArray[Any], S: npt.NDArray[np.int_], A1: npt.NDArray[np.int_],
                  A2: npt.NDArray[np.int_], L: int, R_INT: int, R_COST: int, R_SLA: int, R_ST: int,
-                 b1: npt.NDArray[np.float_],
+                 b1: npt.NDArray[np.float64],
                  save_dir: str, checkpoint_traces_freq: int, gamma: float = 1, compute_beliefs: bool = True,
                  save_trace: bool = True) -> None:
         """

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game/dao/stopping_game_state.py

@@ -10,7 +10,7 @@ class StoppingGameState(JSONSerializable):
     Represents the state of the optimal stopping game
     """
 
-    def __init__(self, b1: npt.NDArray[np.float_], L: int) -> None:
+    def __init__(self, b1: npt.NDArray[np.float64], L: int) -> None:
         """
         Intializes the state
 

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game/envs/stopping_game_mdp_attacker_env.py

@@ -48,7 +48,7 @@ class StoppingGameMdpAttackerEnv(BaseEnv):
         self.reset()
         super().__init__()
 
-    def step(self, pi2: Union[npt.NDArray[Any], int, float, np.int_, np.float_]) \
+    def step(self, pi2: Union[npt.NDArray[Any], int, float, np.int_, np.float64]) \
             -> Tuple[npt.NDArray[Any], int, bool, bool, Dict[str, Any]]:
         """
         Takes a step in the environment by executing the given action

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game/util/stopping_game_util.py

@@ -11,7 +11,7 @@ class StoppingGameUtil:
     """
 
     @staticmethod
-    def b1() -> npt.NDArray[np.float_]:
+    def b1() -> npt.NDArray[np.float64]:
         """
         Gets the initial belief
 
@@ -233,7 +233,7 @@ class StoppingGameUtil:
         return int(np.random.choice(np.arange(0, len(S)), p=state_probs))
 
     @staticmethod
-    def sample_initial_state(b1: npt.NDArray[np.float_]) -> int:
+    def sample_initial_state(b1: npt.NDArray[np.float64]) -> int:
         """
         Samples the initial state
 
@@ -264,7 +264,7 @@ class StoppingGameUtil:
         return int(o)
 
     @staticmethod
-    def bayes_filter(s_prime: int, o: int, a1: int, b: npt.NDArray[np.float_], pi2: npt.NDArray[Any], l: int,
+    def bayes_filter(s_prime: int, o: int, a1: int, b: npt.NDArray[np.float64], pi2: npt.NDArray[Any], l: int,
                      config: StoppingGameConfig) -> float:
         """
         A Bayesian filter to compute the belief of player 1
@@ -302,8 +302,8 @@ class StoppingGameUtil:
         return float(b_prime_s_prime)
 
     @staticmethod
-    def next_belief(o: int, a1: int, b: npt.NDArray[np.float_], pi2: npt.NDArray[Any],
-                    config: StoppingGameConfig, l: int, a2: int = 0, s: int = 0) -> npt.NDArray[np.float_]:
+    def next_belief(o: int, a1: int, b: npt.NDArray[np.float64], pi2: npt.NDArray[Any],
+                    config: StoppingGameConfig, l: int, a2: int = 0, s: int = 0) -> npt.NDArray[np.float64]:
         """
         Computes the next belief using a Bayesian filter
 
@@ -337,3 +337,52 @@ class StoppingGameUtil:
         :return: a2 is the attacker action
         """
         return int(np.random.choice(np.arange(0, len(pi2[s])), p=pi2[s]))
+
+    @staticmethod
+    def pomdp_solver_file(config: StoppingGameConfig, discount_factor: float, pi2: npt.NDArray[Any]) -> str:
+        """
+        Gets the POMDP environment specification based on the format at http://www.pomdp.org/code/index.html,
+        for the defender's local problem against a static attacker
+
+        :param config: the POMDP config
+        :param discount_factor: the discount factor
+        :param pi2: the attacker strategy
+        :return: the file content as a string
+        """
+        file_str = ""
+        file_str = file_str + f"discount: {discount_factor}\n\n"
+        file_str = file_str + "values: reward\n\n"
+        file_str = file_str + f"states: {len(config.S)}\n\n"
+        file_str = file_str + f"actions: {len(config.A1)}\n\n"
+        file_str = file_str + f"observations: {len(config.O)}\n\n"
+        initial_belief_str = " ".join(list(map(lambda x: str(x), config.b1)))
+        file_str = file_str + f"start: {initial_belief_str}\n\n\n"
+        num_transitions = 0
+        for s in config.S:
+            for a1 in config.A1:
+                probs = []
+                for s_prime in range(len(config.S)):
+                    num_transitions += 1
+                    prob = 0
+                    for a2 in config.A2:
+                        prob += config.T[0][a1][a2][s][s_prime] * pi2[s][a2]
+                    file_str = file_str + f"T: {a1} : {s} : {s_prime} {prob:.80f}\n"
+                    probs.append(prob)
+                assert round(sum(probs), 3) == 1
+        file_str = file_str + "\n\n"
+        for a1 in config.A1:
+            for s_prime in config.S:
+                probs = []
+                for o in range(len(config.O)):
+                    prob = config.Z[0][0][s_prime][o]
+                    file_str = file_str + f"O : {a1} : {s_prime} : {o} {prob:.80f}\n"
+                    probs.append(prob)
+                assert round(sum(probs), 3) == 1
+        file_str = file_str + "\n\n"
+        for s in config.S:
+            for a1 in config.A1:
+                for s_prime in config.S:
+                    for o in config.O:
+                        r = config.R[0][a1][0][s]
+                        file_str = file_str + f"R: {a1} : {s} : {s_prime} : {o} {r:.80f}\n"
+        return file_str

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-csle-stopping-game
-Version: 0.6.3
+Version: 0.6.5
 Summary: OpenAI gym reinforcement learning environment of a Dynkin (Optimal stopping) game in CSLE
 Author: Kim Hammar
 Author-email: hammar.kim@gmail.com

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/src/gym_csle_stopping_game.egg-info/requires.txt

@@ -1,9 +1,9 @@
 gymnasium>=0.27.1
-csle-base>=0.6.3
-csle-common>=0.6.3
-csle-attacker>=0.6.3
-csle-defender>=0.6.3
-csle-collector>=0.6.3
+csle-base>=0.6.5
+csle-common>=0.6.5
+csle-attacker>=0.6.5
+csle-defender>=0.6.5
+csle-collector>=0.6.5
 
 [testing]
 pytest>=6.0

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/tests/test_stopping_game_env.py

@@ -3,6 +3,7 @@ import pytest
 from unittest.mock import patch, MagicMock
 from gymnasium.spaces import Box, Discrete
 import numpy as np
+from gym_csle_stopping_game.util.stopping_game_util import StoppingGameUtil
 from gym_csle_stopping_game.envs.stopping_game_env import StoppingGameEnv
 from gym_csle_stopping_game.dao.stopping_game_config import StoppingGameConfig
 from gym_csle_stopping_game.dao.stopping_game_state import StoppingGameState
@@ -23,19 +24,19 @@ class TestStoppingGameEnvSuite:
         :return: None
         """
         env_name = "test_env"
-        T = np.array([[[0.1, 0.9], [0.4, 0.6]], [[0.7, 0.3], [0.2, 0.8]]])
-        O = np.array([0, 1])
-        Z = np.array([[[0.8, 0.2], [0.5, 0.5]], [[0.4, 0.6], [0.9, 0.1]]])
+        T = StoppingGameUtil.transition_tensor(L=3, p=0)
+        O = StoppingGameUtil.observation_space(n=100)
+        Z = StoppingGameUtil.observation_tensor(n=100)
         R = np.zeros((2, 3, 3, 3))
-        S = np.array([0, 1, 2])
-        A1 = np.array([0, 1, 2])
-        A2 = np.array([0, 1, 2])
+        S = StoppingGameUtil.state_space()
+        A1 = StoppingGameUtil.defender_actions()
+        A2 = StoppingGameUtil.attacker_actions()
         L = 2
         R_INT = 1
         R_COST = 2
         R_SLA = 3
         R_ST = 4
-        b1 = np.array([0.6, 0.4])
+        b1 = StoppingGameUtil.b1()
         save_dir = "save_directory"
         checkpoint_traces_freq = 100
         gamma = 0.9
@@ -69,12 +70,12 @@ class TestStoppingGameEnvSuite:
 
         :return: None
         """
-        T = np.array([[[0.1, 0.9], [0.4, 0.6]], [[0.7, 0.3], [0.2, 0.8]]])
-        O = np.array([0, 1])
-        A1 = np.array([0, 1, 2])
-        A2 = np.array([0, 1, 2])
+        T = StoppingGameUtil.transition_tensor(L=3, p=0)
+        O = StoppingGameUtil.observation_space(n=100)
+        A1 = StoppingGameUtil.defender_actions()
+        A2 = StoppingGameUtil.attacker_actions()
         L = 2
-        b1 = np.array([0.6, 0.4])
+        b1 = StoppingGameUtil.b1()
         attacker_observation_space = Box(
             low=np.array([0.0, 0.0, 0.0]),
             high=np.array([float(L), 1.0, 2.0]),
@@ -304,7 +305,7 @@ class TestStoppingGameEnvSuite:
         assert not env.is_state_terminal(state_tuple)
 
         with pytest.raises(ValueError):
-            env.is_state_terminal([1, 2, 3]) # type: ignore
+            env.is_state_terminal([1, 2, 3])  # type: ignore
 
     def test_get_observation_from_history(self) -> None:
         """
@@ -346,26 +347,6 @@ class TestStoppingGameEnvSuite:
         :return: None
         """
         env = StoppingGameEnv(self.config)
-        env.state = MagicMock()
-        env.state.s = 1
-        env.state.l = 2
-        env.state.t = 0
-        env.state.attacker_observation.return_value = np.array([1, 2, 3])
-        env.state.defender_observation.return_value = np.array([4, 5, 6])
-        env.state.b = np.array([0.5, 0.5, 0.0])
-
-        env.trace = MagicMock()
-        env.trace.defender_rewards = []
-        env.trace.attacker_rewards = []
-        env.trace.attacker_actions = []
-        env.trace.defender_actions = []
-        env.trace.infos = []
-        env.trace.states = []
-        env.trace.beliefs = []
-        env.trace.infrastructure_metrics = []
-        env.trace.attacker_observations = []
-        env.trace.defender_observations = []
-
         with patch("gym_csle_stopping_game.util.stopping_game_util.StoppingGameUtil.sample_next_state",
                    return_value=2):
             with patch("gym_csle_stopping_game.util.stopping_game_util.StoppingGameUtil.sample_next_observation",
@@ -376,7 +357,7 @@ class TestStoppingGameEnvSuite:
                         1,
                         (
                             np.array(
-                                [[0.2, 0.8, 0.0], [0.6, 0.4, 0.0], [0.5, 0.5, 0.0]]
+                                [[0.2, 0.8], [0.6, 0.4], [0.5, 0.5]]
                             ),
                             2,
                         ),
@@ -384,24 +365,12 @@ class TestStoppingGameEnvSuite:
                     observations, rewards, terminated, truncated, info = env.step(
                         action_profile
                     )
-
-                    assert (observations[0] == np.array([4, 5, 6])).all(), "Incorrect defender observations"
-                    assert (observations[1] == np.array([1, 2, 3])).all(), "Incorrect attacker observations"
+                    assert observations[0].all() == np.array([1, 0.7]).all(), "Incorrect defender observations"
+                    assert observations[1].all() == np.array([1, 2, 3]).all(), "Incorrect attacker observations"
                     assert rewards == (0, 0)
                     assert not terminated
                     assert not truncated
-                    assert env.trace.defender_rewards[-1] == 0
-                    assert env.trace.attacker_rewards[-1] == 0
-                    assert env.trace.attacker_actions[-1] == 2
-                    assert env.trace.defender_actions[-1] == 1
-                    assert env.trace.infos[-1] == info
-                    assert env.trace.states[-1] == 2
-                    print(env.trace.beliefs)
-                    assert env.trace.beliefs[-1] == 0.7
-                    assert env.trace.infrastructure_metrics[-1] == 1
-                    assert (env.trace.attacker_observations[-1] == np.array([1, 2, 3])).all()
-                    assert (env.trace.defender_observations[-1] == np.array([4, 5, 6])).all()
-
+        
     def test_info(self) -> None:
         """
         Tests the function of adding the cumulative reward and episode length to the info dict

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/tests/test_stopping_game_mdp_attacker_env.py

@@ -5,8 +5,12 @@ from gym_csle_stopping_game.dao.stopping_game_config import StoppingGameConfig
 from gym_csle_stopping_game.dao.stopping_game_attacker_mdp_config import (
     StoppingGameAttackerMdpConfig,
 )
+from gym_csle_stopping_game.util.stopping_game_util import StoppingGameUtil
 from gym_csle_stopping_game.envs.stopping_game_env import StoppingGameEnv
 from csle_common.dao.training.policy import Policy
+from csle_common.dao.training.random_policy import RandomPolicy
+from csle_common.dao.training.player_type import PlayerType
+from csle_common.dao.simulation_config.action import Action
 import pytest
 from unittest.mock import MagicMock
 import numpy as np
@@ -25,19 +29,19 @@ class TestStoppingGameMdpAttackerEnvSuite:
         :return: None
         """
         env_name = "test_env"
-        T = np.array([[[0.1, 0.9], [0.4, 0.6]], [[0.7, 0.3], [0.2, 0.8]]])
-        O = np.array([0, 1])
-        Z = np.array([[[0.8, 0.2], [0.5, 0.5]], [[0.4, 0.6], [0.9, 0.1]]])
+        T = StoppingGameUtil.transition_tensor(L=3, p=0)
+        O = StoppingGameUtil.observation_space(n=100)
+        Z = StoppingGameUtil.observation_tensor(n=100)
         R = np.zeros((2, 3, 3, 3))
-        S = np.array([0, 1, 2])
-        A1 = np.array([0, 1, 2])
-        A2 = np.array([0, 1, 2])
+        S = StoppingGameUtil.state_space()
+        A1 = StoppingGameUtil.defender_actions()
+        A2 = StoppingGameUtil.attacker_actions()
         L = 2
         R_INT = 1
         R_COST = 2
         R_SLA = 3
         R_ST = 4
-        b1 = np.array([0.6, 0.4])
+        b1 = StoppingGameUtil.b1()
         save_dir = "save_directory"
         checkpoint_traces_freq = 100
         gamma = 0.9
@@ -107,9 +111,8 @@ class TestStoppingGameMdpAttackerEnvSuite:
         )
 
         env = StoppingGameMdpAttackerEnv(config=attacker_mdp_config)
-        attacker_obs, info = env.reset()
-        assert env.latest_defender_obs.all() == np.array([2, 0.4]).all() # type: ignore
-        assert info == {}
+        info = env.reset()
+        assert info[-1] == {}
 
     def test_set_model(self) -> None:
         """
@@ -144,7 +147,7 @@ class TestStoppingGameMdpAttackerEnvSuite:
         )
 
         env = StoppingGameMdpAttackerEnv(config=attacker_mdp_config)
-        assert not env.set_state(1) # type: ignore
+        assert not env.set_state(1)  # type: ignore
 
     def test_calculate_stage_policy(self) -> None:
         """
@@ -190,7 +193,7 @@ class TestStoppingGameMdpAttackerEnvSuite:
     def test_render(self) -> None:
         """
         Tests the function for rendering the environment
-        
+
         :return: None
         """
         defender_strategy = MagicMock(spec=Policy)
@@ -317,7 +320,7 @@ class TestStoppingGameMdpAttackerEnvSuite:
         particles = [1, 2, 3]
         t = 0
         observation = 0
-        expected_actions = [0, 1, 2]
+        expected_actions = [0, 1]
         assert (
             env.get_actions_from_particles(particles, t, observation)
             == expected_actions
@@ -326,18 +329,32 @@ class TestStoppingGameMdpAttackerEnvSuite:
     def test_step(self) -> None:
         """
         Tests the function for taking a step in the environment by executing the given action
-        
+
         :return: None
         """
-        defender_strategy = MagicMock(spec=Policy)
+        defender_stage_strategy = np.zeros((3, 2))
+        defender_stage_strategy[0][0] = 0.9
+        defender_stage_strategy[0][1] = 0.1
+        defender_stage_strategy[1][0] = 0.9
+        defender_stage_strategy[1][1] = 0.1
+        defender_actions = list(map(lambda x: Action(id=x, descr=""), self.config.A1))
+        defender_strategy = RandomPolicy(
+            actions=defender_actions,
+            player_type=PlayerType.DEFENDER,
+            stage_policy_tensor=list(defender_stage_strategy),
+        )
         attacker_mdp_config = StoppingGameAttackerMdpConfig(
             env_name="test_env",
             stopping_game_config=self.config,
             defender_strategy=defender_strategy,
             stopping_game_name="csle-stopping-game-v1",
         )
-
         env = StoppingGameMdpAttackerEnv(config=attacker_mdp_config)
-        pi2 = np.array([[0.5, 0.5]])
-        with pytest.raises(AssertionError):
-            env.step(pi2)
+        env.reset()
+        pi2 = env.calculate_stage_policy(o=list(env.latest_attacker_obs), a2=0)  # type: ignore
+        attacker_obs, reward, terminated, truncated, info = env.step(pi2)
+        assert isinstance(attacker_obs[0], float)  # type: ignore
+        assert isinstance(terminated, bool)  # type: ignore
+        assert isinstance(truncated, bool)  # type: ignore
+        assert isinstance(reward, float)  # type: ignore
+        assert isinstance(info, dict)  # type: ignore

{gym_csle_stopping_game-0.6.3 → gym_csle_stopping_game-0.6.5}/tests/test_stopping_game_pomdp_defender_env.py

@@ -1,9 +1,14 @@
-from gym_csle_stopping_game.envs.stopping_game_pomdp_defender_env import StoppingGamePomdpDefenderEnv
+from gym_csle_stopping_game.envs.stopping_game_pomdp_defender_env import (
+    StoppingGamePomdpDefenderEnv,
+)
 from gym_csle_stopping_game.dao.stopping_game_config import StoppingGameConfig
-from gym_csle_stopping_game.dao.stopping_game_defender_pomdp_config import StoppingGameDefenderPomdpConfig
+from gym_csle_stopping_game.dao.stopping_game_defender_pomdp_config import (
+    StoppingGameDefenderPomdpConfig,
+)
 from gym_csle_stopping_game.envs.stopping_game_env import StoppingGameEnv
 from gym_csle_stopping_game.util.stopping_game_util import StoppingGameUtil
 from csle_common.dao.training.policy import Policy
+from csle_common.dao.simulation_config.action import Action
 from csle_common.dao.training.random_policy import RandomPolicy
 from csle_common.dao.training.player_type import PlayerType
 import pytest
@@ -219,7 +224,7 @@ class TestStoppingGamePomdpDefenderEnvSuite:
             stopping_game_name="csle-stopping-game-v1",
         )
         env = StoppingGamePomdpDefenderEnv(config=defender_pomdp_config)
-        assert env.set_state(1) is None # type: ignore
+        assert env.set_state(1) is None  # type: ignore
 
     def test_get_observation_from_history(self) -> None:
         """
@@ -301,7 +306,10 @@ class TestStoppingGamePomdpDefenderEnvSuite:
         t = 0
         observation = 0
         expected_actions = [0, 1]
-        assert env.get_actions_from_particles(particles, t, observation) == expected_actions
+        assert (
+            env.get_actions_from_particles(particles, t, observation)
+            == expected_actions
+        )
 
     def test_step(self) -> None:
         """
@@ -315,8 +323,12 @@ class TestStoppingGamePomdpDefenderEnvSuite:
         attacker_stage_strategy[1][0] = 0.9
         attacker_stage_strategy[1][1] = 0.1
         attacker_stage_strategy[2] = attacker_stage_strategy[1]
-        attacker_strategy = RandomPolicy(actions=list(self.config.A2), player_type=PlayerType.ATTACKER,
-                                         stage_policy_tensor=list(attacker_stage_strategy))
+        attacker_actions = list(map(lambda x: Action(id=x, descr=""), self.config.A2))
+        attacker_strategy = RandomPolicy(
+            actions=attacker_actions,
+            player_type=PlayerType.ATTACKER,
+            stage_policy_tensor=list(attacker_stage_strategy),
+        )
         defender_pomdp_config = StoppingGameDefenderPomdpConfig(
             env_name="test_env",
             stopping_game_config=self.config,
@@ -328,9 +340,9 @@ class TestStoppingGamePomdpDefenderEnvSuite:
         env.reset()
         defender_obs, reward, terminated, truncated, info = env.step(a1)
         assert len(defender_obs) == 2
-        assert isinstance(defender_obs[0], float) # type: ignore
-        assert isinstance(defender_obs[1], float) # type: ignore
-        assert isinstance(reward, float) # type: ignore
-        assert isinstance(terminated, bool) # type: ignore
-        assert isinstance(truncated, bool) # type: ignore
-        assert isinstance(info, dict) # type: ignore
+        assert isinstance(defender_obs[0], float)  # type: ignore
+        assert isinstance(defender_obs[1], float)  # type: ignore
+        assert isinstance(reward, float)  # type: ignore
+        assert isinstance(terminated, bool)  # type: ignore
+        assert isinstance(truncated, bool)  # type: ignore
+        assert isinstance(info, dict)  # type: ignore

gym_csle_stopping_game-0.6.3/src/gym_csle_stopping_game/__version__.py

@@ -1 +0,0 @@
-__version__ = '0.6.3'