Merge pull request #44 from jgrguric96/feature/ABC_model

First version of illuminator model interface

Models/construct_model.py

@@ -0,0 +1,105 @@
+from abc import ABC, abstractmethod
+from dataclasses import dataclass
+from enum import Enum
+from mosaik_api import Simulator
+
+class SimulatorType(Enum):
+    TIME_BASED = 'time-based'
+    EVENT_BASED = 'event-based'
+    HYBRID = 'hybrid'
+
+# NOTE: I have also inherited from the mosaik_api.Simulator class to integrate this model
+# This seems naïve to me. Lots of redundancy with defining which methods are mandatory.
+# As for the: the models created seem to have a wide variety of attributes and parameters that are defined.
+# In addition, the naming convection of the methods created in them is also non-existant
+@dataclass
+class IlluminatorModel(Simulator, ABC):
+    """ 
+    A common interface for adding energy models to the Illuminator.
+    This class combines the creation of model and simulators for the Mosaik engine.
+
+    ...
+
+    Parameters
+    ----------
+    simulator_type: SimulatorType.TIME_BASED
+        Defines how the simulator is advanced through time and whether its attributes are persistent in time or transient
+    model_name: str
+        The name of the model
+    public: bool
+        Determines whether a model can be instantiated by a user 
+    parameters: list 
+        Variables which values can be set when instantiating the model (can this be always considered inputs?)
+    attrs: list 
+        Model properties that can be read from (can these be always considered outputs?)
+    trigger: list 
+        Attribute names that cause the simulator to be stepped when another simulator provides output which is connected to one of those.
+        Empty if no triggers.
+    inputs: list
+        ???
+    outputs: list
+        ???
+    states: list
+        ???
+    """
+
+    # TODO:the ideal solution provides a single interface for creating a model and its simulator
+
+    ### meta data ###
+    model_name: str
+    parameters: list # variables which values can be set when instantiating the model (can this be always considered inputs?)
+    attrs: list  # model properties that can be read from (can these be always considered outputs?)
+    trigger: list
+
+    inputs: list
+    outputs: list
+    states: list
+
+    simulator_type: SimulatorType = SimulatorType.TIME_BASED
+    public: bool = True
+    #################
+
+    @classmethod
+    def __test__(cls, subclass):
+        return hasattr(subclass, "__init__")
+
+    def __post_init__(self):
+        """
+        Post-init method which creates and initializes the metadata used by mosaik_api
+        """
+        META = {
+            'type': self.simulator_type,
+            'models': {
+                self.model_name: {
+                    'public': self.public,
+                    'params': self.parameters,
+                    'attrs': self.attrs, 
+                    'trigger': self.trigger
+                },
+            },
+        }
+        super().__init__(META)
+
+    @abstractmethod
+    def init(self, sid, time_resolution, *args):
+        """ Initialize the simulator with the ID `sid` and pass the `time_resolution` and additional parameters sent by mosaik. """
+        raise NotImplementedError
+
+    @abstractmethod
+    def create(self, num, model, *args):
+        """ Create `num` instances of `model` using the provided additional arguments """
+        raise NotImplementedError
+
+    @abstractmethod
+    def step(self, time, inputs, max_advance):
+        """ 
+        Perform the next simulation step from time `time` using input values
+        from `inputs` and return the new simulation time (the time at which
+        `step()` should be called again) 
+        """
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_data(self, outputs):
+        """ Return the data for the requested outputs """
+        raise NotImplementedError