mdpsolver 0.9.8__tar.gz → 0.9.9__tar.gz

{mdpsolver-0.9.8/src/mdpsolver.egg-info → mdpsolver-0.9.9}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: mdpsolver
-Version: 0.9.8
+Version: 0.9.9
 Summary: A fast solver for Markov Decision Processes
 Author-email: Anders Reenberg Andersen <andersra@live.dk>, Jesper Fink Andersen <jesperfinkandersen@gmail.com>
 Project-URL: Homepage, https://github.com/areenberg/MDPSolver
@@ -8,10 +8,12 @@ Project-URL: Issues, https://github.com/areenberg/MDPSolver/issues
 Classifier: Development Status :: 3 - Alpha
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
-Classifier: Operating System :: OS Independent
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Operating System :: POSIX :: Linux
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Dynamic: license-file
 
 # MDPSolver
 

{mdpsolver-0.9.8 → mdpsolver-0.9.9}/pyproject.toml

@@ -10,7 +10,7 @@ where = ["src"]
 
 [project]
 name = "mdpsolver"
-version = "0.9.8"
+version = "0.9.9"
 authors = [
   { name="Anders Reenberg Andersen", email="andersra@live.dk" },
   { name="Jesper Fink Andersen", email="jesperfinkandersen@gmail.com" },
@@ -22,7 +22,8 @@ classifiers = [
     "Development Status :: 3 - Alpha",
     "Programming Language :: Python :: 3",
     "License :: OSI Approved :: MIT License",
-    "Operating System :: OS Independent",
+    "Operating System :: Microsoft :: Windows",
+    "Operating System :: POSIX :: Linux"
 ]
 
 [project.urls]

mdpsolver-0.9.9/src/mdpsolver/__init__.py

@@ -0,0 +1 @@
+from mdpsolver.model import model

mdpsolver-0.9.9/src/mdpsolver/model.py

@@ -0,0 +1,189 @@
+"""
+MDPSolver
+
+This file defines a `model` class that wraps around the `solvermodule` to create
+and solve Markov Decision Processes (MDPs). It provides methods to configure the model,
+solve it, and extract the resulting policy and value vectors.
+"""
+
+from mdpsolver import solvermodule
+
+
+class model:
+    """
+    Creates an MDPSolver object.
+
+    This class provides methods to initialize, solve, and return results
+    from an MDP model.
+    """
+
+    def __init__(self):
+        """Initialize the model by creating a solver module object."""
+        self.initialize()
+
+    def initialize(self):
+        """Create and initialize the solver module object."""
+        self.mdl = solvermodule.Model()
+
+    def solve(
+        self,
+        algorithm="mpi",
+        tolerance=1e-3,
+        update="standard",
+        criterion="discounted",
+        parIterLim=100,
+        SORrelaxation=1.0,
+        initPolicy=list(),
+        initValueVector=list(),
+        verbose=False,
+        postProcessing=True,
+        makeFinalCheck=True,
+        parallel=True,
+    ):
+        """
+        Derive an epsilon-optimal policy for the selected MDP model.
+
+        Args:
+            algorithm (str): Algorithm to use.
+            tolerance (float): Convergence threshold for the algorithm.
+            update (str): The value-update method.
+            criterion (str): The optimality criterion.
+            parIterLim (int): The partial evaluation limit employed in the modified policy iteration algorithm.
+            SORrelaxation (float): Relaxation parameter for the Successive Over-Relaxation method.
+            initPolicy (list): A 1D-list defining the initial policy. This option can be used for "warm starting" the optimization procedure.
+            initValueVector (list): A 1D-list defining the initial value vector. This option can be used for "warm starting" the optimization procedure.
+            verbose (bool): If True, prints solver progress to console.
+            postProcessing (bool): If True, performs post-processing after solving.
+            makeFinalCheck (bool): If True, makes a final check of the value vector. This process checks if the resulting values are reasonable.
+            parallel (bool): If True, enables parallel computation for faster solving. This option is only available for the custom MDP model with standard updates.
+
+        Returns:
+            None
+        """
+        self.mdl.solve(
+            algorithm=algorithm,
+            tolerance=tolerance,
+            update=update,
+            criterion=criterion,
+            parIterLim=parIterLim,
+            SORrelaxation=SORrelaxation,
+            initPolicy=initPolicy,
+            initValueVector=initValueVector,
+            verbose=verbose,
+            postProcessing=postProcessing,
+            makeFinalCheck=makeFinalCheck,
+            parallel=parallel,
+        )
+
+    def getRuntime(self):
+        """
+        Get the runtime of the last solver execution.
+
+        Returns:
+            float: Runtime in milliseconds.
+        """
+        return self.mdl.getRuntime()
+
+    def printPolicy(self):
+        """Print the entire policy to the terminal."""
+        self.mdl.printPolicy()
+
+    def printValueVector(self):
+        """Print the entire value vector to the terminal."""
+        self.mdl.printValueVector()
+
+    def getAction(self, stateIndex=0):
+        """
+        Get the action from the optimized policy for a specific state.
+
+        Args:
+            stateIndex (int): Index of the state.
+
+        Returns:
+            int: Action for the given state.
+        """
+        return self.mdl.getAction(stateIndex=stateIndex)
+
+    def getValue(self, stateIndex=0):
+        """
+        Get the value from the optimized value vector for a specific state.
+
+        Args:
+            stateIndex (int): Index of the state.
+
+        Returns:
+            float: Value for the given state.
+        """
+        return self.mdl.getValue(stateIndex=stateIndex)
+
+    def getPolicy(self):
+        """
+        Get the entire optimized policy.
+
+        Returns:
+            list: Optimized policy.
+        """
+        return self.mdl.getPolicy()
+
+    def getValueVector(self):
+        """
+        Get the entire optimized value vector.
+
+        Returns:
+            list: Optimized value vector.
+        """
+        return self.mdl.getValueVector()
+
+    def saveToFile(self, fileName="result.csv", type="policy"):
+        """
+        Save the optimized policy or value vector to a file.
+
+        Args:
+            fileName (str): Name of the output file.
+            type (str): Type of result to save ('policy' or 'value').
+
+        Returns:
+            None
+        """
+        return self.mdl.saveToFile(fileName=fileName, type=type)
+
+    def mdp(
+        self,
+        discount=0.99,
+        rewards=list(),
+        rewardsElementwise=list(),
+        rewardsFromFile="rewards.csv",
+        tranMatWithZeros=list(),
+        tranMatElementwise=list(),
+        tranMatProbs=list(),
+        tranMatColumns=list(),
+        tranMatFromFile="transitions.csv",
+    ):
+        """
+        Define the generic MDP model.
+
+        Args:
+            discount (float): Discount factor.
+            rewards (list): A 2D-list containing the reward (float) of a particular action in a particular state.
+            rewardsElementwise (list): Alternative reward format. A 2D-list where each row corresponds to a combination of a state and an action.
+            rewardsFromFile (str): Load the rewards from a comma-separated (,) file.
+            tranMatWithZeros (list): A 3D-list containing the transition probabilities.
+            tranMatElementwise (list): Sparse transition probabilities (option 1). A 2D-list where each row corresponds to a combination of a current state, an action, and a next state.
+            tranMatProbs (list): Sparse transition probabilities (option 2, part 1). A 3D-list containing the non-zero transition probabilities.
+            tranMatColumns (list): Sparse transition probabilities (option 2, part 2). A 3D-list containing the columns of the non-zero transition probabilities.
+            tranMatFromFile (str): Load the transition probabilities from a comma-separated (,) file.
+
+        Returns:
+            None
+        """
+        self.mdl.mdp(
+            discount=discount,
+            rewards=rewards,
+            rewardsElementwise=rewardsElementwise,
+            rewardsFromFile=rewardsFromFile,
+            tranMatWithZeros=tranMatWithZeros,
+            tranMatElementwise=tranMatElementwise,
+            tranMatProbs=tranMatProbs,
+            tranMatColumns=tranMatColumns,
+            tranMatFromFile=tranMatFromFile,
+        )

mdpsolver-0.9.9/src/mdpsolver/problems/cbmaux.py

@@ -0,0 +1,40 @@
+import math
+
+# Condition-based maintenance problem
+
+# Note: The class is only used for retrieving/converting information
+# (i.e. returning an action of the policy). The policy is optimized
+# in the 'model' (model.py) class.
+
+
+class cbmaux:
+
+    def __init__(self, components, stages):
+        self.components = components
+        self.stages = stages
+
+    def compToState(self, compStates):
+        # input: list containing the stage of each component
+        # output: the associated state index
+        prod = self.components * [1]
+        if self.components > 1:
+            for i in range(self.components - 2, -1, -1):
+                prod[i] = prod[i + 1] * self.stages
+        stateIndex = 0
+        for i in range(self.components):
+            stateIndex += compStates[i] * prod[i]
+        return stateIndex
+
+    def actionToReplace(self, actionIndex):
+        # input: action index from the optimized policy
+        # output: list of bools indicating which comps to replace
+        prod = self.components * [1]
+        if self.components > 1:
+            for i in range(self.components - 2, -1, -1):
+                prod[i] = prod[i + 1] * 2
+        replace = self.components * [False]
+        for i in range(self.components):
+            if math.floor(actionIndex / prod[i]) > 0:
+                replace[i] = True
+                actionIndex -= prod[i]
+        return replace

mdpsolver-0.9.9/src/mdpsolver/problems/tbmaux.py

@@ -0,0 +1,40 @@
+import math
+
+# Time-based maintenance problem
+
+# Note: The class is only used for retrieving/converting information
+# (i.e. returning an action of the policy). The policy is optimized
+# in the 'model' (model.py) class.
+
+
+class tbmaux:
+
+    def __init__(self, components, stages):
+        self.components = components
+        self.stages = stages
+
+    def compToState(self, compStates):
+        # input: list containing the stage of each component
+        # output: the associated state index
+        prod = self.components * [1]
+        if self.components > 1:
+            for i in range(self.components - 2, -1, -1):
+                prod[i] = prod[i + 1] * self.stages
+        stateIndex = 0
+        for i in range(self.components):
+            stateIndex += compStates[i] * prod[i]
+        return stateIndex
+
+    def actionToReplace(self, actionIndex):
+        # input: action index from the optimized policy
+        # output: list of bools indicating which comps to replace
+        prod = self.components * [1]
+        if self.components > 1:
+            for i in range(self.components - 2, -1, -1):
+                prod[i] = prod[i + 1] * 2
+        replace = self.components * [False]
+        for i in range(self.components):
+            if math.floor(actionIndex / prod[i]) > 0:
+                replace[i] = True
+                actionIndex -= prod[i]
+        return replace

{mdpsolver-0.9.8 → mdpsolver-0.9.9/src/mdpsolver.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: mdpsolver
-Version: 0.9.8
+Version: 0.9.9
 Summary: A fast solver for Markov Decision Processes
 Author-email: Anders Reenberg Andersen <andersra@live.dk>, Jesper Fink Andersen <jesperfinkandersen@gmail.com>
 Project-URL: Homepage, https://github.com/areenberg/MDPSolver
@@ -8,10 +8,12 @@ Project-URL: Issues, https://github.com/areenberg/MDPSolver/issues
 Classifier: Development Status :: 3 - Alpha
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
-Classifier: Operating System :: OS Independent
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Operating System :: POSIX :: Linux
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Dynamic: license-file
 
 # MDPSolver