remove old Normalize and refs

src/pymgrid/modules/base/base_module.py

@@ -8,7 +8,6 @@
 
 from pymgrid.utils.logger import ModularLogger
 from pymgrid.utils.space import ModuleSpace
-from pymgrid.modules.utils.normalize import Normalize, IdentityNormalize
 from pymgrid.utils.serialize import add_numpy_pandas_representers, add_numpy_pandas_constructors, dump_data
 
 
@@ -38,59 +37,16 @@ class BaseMicrogridModule(yaml.YAMLObject):
     def __init__(self,
                  raise_errors,
                  provided_energy_name='provided_energy',
-                 absorbed_energy_name='absorbed_energy',
-                 normalize_pos=...
+                 absorbed_energy_name='absorbed_energy'
                  ):
 
         self.raise_errors = raise_errors
         self._current_step = 0
-        self._obs_normalizer = self._get_normalizer(normalize_pos, obs=True)
-        self._act_normalizer = self._get_normalizer(normalize_pos, act=True)
         self._action_space = self._get_action_spaces()
         self._observation_space = self._get_observation_spaces()
         self.provided_energy_name, self.absorbed_energy_name = provided_energy_name, absorbed_energy_name
         self._logger = ModularLogger()
-        self.name = (None, None) # set by ModularMicrogrid
-
-    def _get_normalizer(self, normalize_pos, obs=False, act=False):
-        assert obs+act == 1, 'Must be initiating normalizer for obs or act but not both or neither.'
-        _str = 'obs' if obs else 'act'
-        try:
-            if obs:
-                val_min, val_max = self.min_obs, self.max_obs
-            else:
-                val_min, val_max = self.min_act, self.max_act
-
-        except AttributeError:
-            script_logger.debug(f'min_{_str} and max_{_str} attributes not found for module {self.__class__.__name__}. '
-                                f'Returning identity normalizer')
-
-            return IdentityNormalize()
-
-        try:
-            normalize_position = normalize_pos['obs' if obs else 'act']
-            val_min, val_max = val_min[normalize_position], val_max[normalize_position]
-        except TypeError:
-            pass
-
-        if val_min is None or np.isnan(val_min).any() or val_max is None or np.isnan(val_max).any():
-            script_logger.debug(f'One of min_{_str} or max_{_str} attributes is None or NaN for module {self.__class__.__name__}. Returni'
-                          f'ng identity normalizer')
-            return IdentityNormalize()
-
-        elif np.isinf(np.array(val_min)).any() or np.isinf(np.array(val_max)).any():
-            script_logger.debug(f'One of min_{_str} or max_{_str} attributes is Infinity for module {self.__class__.__name__}. Returni'
-                  f'ng identity normalizer')
-            return IdentityNormalize()
-        elif isinstance(val_min, np.ndarray) and isinstance(val_max, np.ndarray) and len(val_min) == 0 and len(val_max) == 0:
-            return IdentityNormalize()
-
-        try:
-            assert val_max >= val_min, f'max_{_str} must be greater than min_{_str}'
-        except ValueError:
-            assert (val_max >= val_min).all(), f'max_{_str} must be greater than min_{_str}'
-
-        return Normalize(val_min, val_max)
+        self.name = (None, None)  # set by ModularMicrogrid
 
     def _get_action_spaces(self):
         unnormalized_low = self.min_act if isinstance(self.min_act, np.ndarray) else np.array([self.min_act])
@@ -537,30 +493,6 @@ def current_step(self):
         """
         return self._current_step
 
-    @property
-    def act_normalizer(self):
-        """
-        Module's action normalizer.
-
-        Returns
-        -------
-        Normalize
-
-        """
-        return self._act_normalizer
-
-    @property
-    def obs_normalizer(self):
-        """
-        Module's observation normalizer.
-
-        Returns
-        -------
-        Normalize
-
-        """
-        return self._obs_normalizer
-
     @property
     @abstractmethod
     def min_obs(self):

src/pymgrid/modules/base/timeseries/base_timeseries_module.py

@@ -44,10 +44,8 @@ def __init__(self,
                                  ]
                                  }
 
-        super().__init__(raise_errors,
-                         provided_energy_name=provided_energy_name,
-                         absorbed_energy_name=absorbed_energy_name,
-                         normalize_pos=normalize_pos)
+        super().__init__(raise_errors, provided_energy_name=provided_energy_name,
+                         absorbed_energy_name=absorbed_energy_name)
 
     def _set_time_series(self, time_series):
         _time_series = np.array(time_series)

src/pymgrid/modules/battery_module.py

@@ -87,9 +87,7 @@ def __init__(self,
         self.init_charge, self.init_soc = init_charge, init_soc
         self._current_charge, self._soc = self._init_battery(init_charge, init_soc)
         self.name = ('battery', None)
-        super().__init__(raise_errors,
-                         provided_energy_name='discharge_amount',
-                         absorbed_energy_name='charge_amount')
+        super().__init__(raise_errors, provided_energy_name='discharge_amount', absorbed_energy_name='charge_amount')
 
     def _init_battery(self, init_charge, init_soc):
         if init_charge is not None:

src/pymgrid/modules/genset_module.py

@@ -91,10 +91,7 @@ def __init__(self,
         self._steps_until_up, self._steps_until_down = self._reset_up_down_times()
         self.name = ('genset', None)
 
-        super().__init__(raise_errors,
-                         provided_energy_name=provided_energy_name,
-                         absorbed_energy_name=None,
-                         normalize_pos=dict(obs=..., act=-1))
+        super().__init__(raise_errors, provided_energy_name=provided_energy_name, absorbed_energy_name=None)
 
     def step(self, action, normalized=True):
         goal_status = action[0]

src/pymgrid/modules/unbalanced_energy_module.py

@@ -16,9 +16,7 @@ def __init__(self,
                  overgeneration_cost=2.0
                  ):
 
-        super().__init__(raise_errors,
-                         provided_energy_name='loss_load',
-                         absorbed_energy_name='overgeneration')
+        super().__init__(raise_errors, provided_energy_name='loss_load', absorbed_energy_name='overgeneration')
         self.loss_load_cost, self.overgeneration_cost = loss_load_cost, overgeneration_cost
         self.name = ('unbalanced_energy', None)