Experimenting with slight change of design. EMeters now contain list

of appliances they measure.  Tests pass.

nilmtk/__init__.py

@@ -3,6 +3,6 @@
 from nilmtk.timeframe import TimeFrame
 from nilmtk.emeter import EMeter
 from nilmtk.datastore import DataStore, HDFDataStore
-from nilmtk.appliancegroup import ApplianceGroup
+from nilmtk.metergroup import MeterGroup
 from nilmtk.appliance import Appliance
 from nilmtk.mains import Mains

nilmtk/appliance.py

@@ -1,6 +1,4 @@
-from mains import Mains
-
-class Appliance(Mains):
+class Appliance(object):
     """
     Attributes
     ----------
@@ -17,13 +15,7 @@ class Appliance(Mains):
        
        on_power_threshold : float, watts
        minimum_off_duration : timedelta
-       minimum_on_duration : timedelta
-       
-    mains : Mains (used so appliance methods can default to use
-      the same measured parameter (active / apparent / reactive) 
-      as Mains; and also for use in proportion of energy submetered
-      and for voltage normalisation.)
-       
+       minimum_on_duration : timedelta       
     """
 
     # TODO: appliance_types will be loaded from disk
@@ -88,38 +80,4 @@ def __repr__(self):
                         md.get('type'), md.get('instance'), 
                         md.get('dataset'), md.get('building')))
 
-    def total_on_duration(self):
-        """Return timedelta"""
-        raise NotImplementedError
-
-    def on_durations(self):
-        raise NotImplementedError
-
-    def activity_distribution(self, bin_size, timespan):
-        raise NotImplementedError
-
-    def when_on(self):
-        """Return Series of bools"""
-        raise NotImplementedError    
-
-    def on_off_events(self):
-        # use self.metadata.minimum_[off|on]_duration
-        raise NotImplementedError
-
-    def discrete_appliance_activations(self):
-        """
-        Return a Mask defining the start and end times of each appliance
-        activation.
-        """
-        raise NotImplementedError
-
-    def proportion_of_energy(self):
-        # get the old mask so we can put it back at the end
-        old_mask = self.get_loader_attribute('mask')
-
-        # Mask out gaps from mains
-        self.set_loader_attributes(mask = self.mains.gaps())
-        proportion_of_energy = self.total_energy() / self. mains.total_energy()
-        self.set_loader_attributes(mask = old_mask)
-        return proportion_of_energy 
 

nilmtk/electricity.py

@@ -1,7 +1,7 @@
 from __future__ import print_function, division
-from .appliancegroup import ApplianceGroup
+from .appliancegroup import MeterGroup
 
-class Electricity(ApplianceGroup):
+class Electricity(MeterGroup):
     """Represents mains circuit in a single building.
     
     Attributes
@@ -21,7 +21,7 @@ def meters_directly_downstream_of_mains(self):
         """
         raise NotImplementedError
         wiring_graph = self.wiring_graph()
-        # return meters (not appliances) one hop from root (mains)
+        # return meters (not meters) one hop from root (mains)
 
     def proportion_energy_submetered(self):
         good_mains_timeframes = self.mains.good_timeframes()

nilmtk/emeter.py

@@ -7,6 +7,18 @@ class EMeter(object):
     
     Attributes
     ----------
+    appliances : list of Appliance objects connected immediately downstream
+      of this meter.  Will be [] if no appliances are connected directly
+      to this meter.
+
+    dominant_appliance : reference to Appliance which is responsibly for 
+      most of the power demand on this channel.
+
+    mains : Mains (used so appliance methods can default to use
+      the same measured parameter (active / apparent / reactive) 
+      as Mains; and also for use in proportion of energy submetered
+      and for voltage normalisation.)
+
     store : nilmtk.DataStore
 
     key : key into nilmtk.DataStore to access data
@@ -45,6 +57,9 @@ def __init__(self):
         self.metadata = {}
         self.store = None
         self.key = None
+        self.appliances = []
+        self.mains = None
+        self.dominant_appliance = None
 
     @classmethod
     def _load_meter_devices(cls, store):
@@ -68,7 +83,40 @@ def save(self, destination, key):
         """
         # destination.write_metadata(key, self.metadata)
         raise NotImplementedError
-
+
+    def matches(self, key):
+        """
+        Parameters
+        ----------
+        key : dict
+
+        Returns
+        -------
+        True if all key:value pairs in `key` match any appliance
+        in `self.appliances`.
+        """
+        for appliance in self.appliances:
+            if appliance.matches(key):
+                return True
+        return False
+
+    @property
+    def id(self):
+        key_attrs = ['id', 'dataset', 'building']
+        id_dict = {}
+        for key in key_attrs:
+            id_dict[key] = self.metadata.get(key)
+        return id_dict
+
+    def __eq__(self, other):
+        if isinstance(other, EMeter):
+            return self.id == other.id
+        else:
+            return False
+
+    def __hash__(self):
+        return hash(((k,v) for k,v in self.id.iteritems()))
+
     def power_series(self, measurement_preferences=None, 
                      required_measurement=None,
                      normalise=False, voltage_series=None, 
@@ -112,12 +160,6 @@ def total_energy(self, **load_kwargs):
         nodes = [ClipNode(), EnergyNode()]
         results = self._run_pipeline(nodes, **load_kwargs)
         return results['energy']
-
-    def _sanity_check_before_processing(self):
-        if self.store is None:
-            msg = ("'meter.loader' is not set!"
-                   " Cannot process data without a loader!")
-            raise RuntimeError(msg)
 
     def dropout_rate(self):
         """returns a DropoutRateResults object."""
@@ -132,13 +174,39 @@ def good_sections(self):
         nodes = [LocateGoodSectionsNode()]
         results = self._run_pipeline(nodes)
         return results['good_sections']
-
-    def _run_pipeline(self, nodes, **load_kwargs): 
-        self._sanity_check_before_processing()
-        pipeline = Pipeline(nodes)
-        pipeline.run(meter=self, **load_kwargs)
-        return pipeline.results        
 
+    def total_on_duration(self):
+        """Return timedelta"""
+        raise NotImplementedError
+
+    def on_durations(self):
+        raise NotImplementedError
+
+    def activity_distribution(self, bin_size, timespan):
+        raise NotImplementedError
+
+    def when_on(self):
+        """Return Series of bools"""
+        raise NotImplementedError    
+
+    def on_off_events(self):
+        # use self.metadata.minimum_[off|on]_duration
+        raise NotImplementedError
+
+    def discrete_appliance_activations(self):
+        """
+        Return a Mask defining the start and end times of each appliance
+        activation.
+        """
+        raise NotImplementedError
+
+    def proportion_of_energy(self):
+        # Mask out gaps from mains
+        good_mains_timeframes = self.mains.good_timeframes()
+        proportion_of_energy = (self.total_energy(timeframes=good_mains_timeframes) / 
+                                self. mains.total_energy())
+        return proportion_of_energy 
+
     def contiguous_sections(self):
         """retuns Mask object"""
         raise NotImplementedError
@@ -150,3 +218,11 @@ def clean_and_export(self, destination_datastore):
         implausible values removed)"""
         raise NotImplementedError
 
+    def _run_pipeline(self, nodes, **load_kwargs): 
+        if self.store is None:
+            msg = ("'meter.loader' is not set!"
+                   " Cannot process data without a loader!")
+            raise RuntimeError(msg)
+        pipeline = Pipeline(nodes)
+        pipeline.run(meter=self, **load_kwargs)
+        return pipeline.results