reorganising stats and preprocessing nodes. About to modify nodes so

they chain one to another rather than using pipeline.

nilmtk/elecmeter.py

@@ -1,5 +1,5 @@
 from __future__ import print_function, division
-from .pipeline import Pipeline, ClipNode, EnergyNode, LocateGoodSectionsNode
+from .pipeline import Pipeline, Clip, TotalEnergy, GoodSections
 from .hashable import Hashable
 from .appliance import Appliance
 from .datastore import Key
@@ -229,7 +229,7 @@ def total_energy(self, **load_kwargs):
         -------
         nilmtk.pipeline.EnergyResults object
         """
-        nodes = [ClipNode(), EnergyNode()]
+        nodes = [Clip(), TotalEnergy()]
         results = self._run_pipeline(nodes, **load_kwargs)
         return results['energy']
 
@@ -243,7 +243,7 @@ def good_sections(self):
         -------
         sections: list of nilmtk.TimeFrame objects
         """
-        nodes = [LocateGoodSectionsNode()]
+        nodes = [GoodSections()]
         results = self._run_pipeline(nodes)
         return results['good_sections']
 

nilmtk/pipeline/clipnode.py → nilmtk/preprocessing/clip.py

@@ -3,7 +3,7 @@
 from warnings import warn
 from nilmtk.utils import index_of_column_name
 
-class ClipNode(Node):
+class Clip(Node):
 
     # Not very well specified.  Really want to specify that 
     # we need 'lower_limit' and 'upper_limit' to be specified in

nilmtk/pipeline/locategoodsectionsnode.py → nilmtk/stats/goodsections.py

@@ -1,13 +1,13 @@
 from __future__ import print_function, division
-from node import Node, UnsatisfiedRequirementsError
+from node import Node
 import numpy as np
 from numpy import diff, concatenate
 from nilmtk import TimeFrame
 from nilmtk.utils import timedelta64_to_secs
-from locategoodsectionsresults import LocateGoodSectionsResults
+from goodsectionsresults import GoodSectionsResults
 
 
-class LocateGoodSectionsNode(Node):
+class GoodSections(Node):
     """Locate sections of data where the sample period is <= max_sample_period.
 
     Attributes
@@ -105,54 +105,9 @@ def process(self, df, metadata):
         self.previous_chunk_ended_with_open_ended_good_section = (
             ends_with_open_ended_good_section)
 
-        good_section_results = LocateGoodSectionsResults(max_sample_period)
+        good_section_results = GoodSectionsResults(max_sample_period)
         good_section_results.append(df.timeframe, {'sections': [sections]})
         results = getattr(df, 'results', {})
         results[self.name] = good_section_results
         df.results = results
         return df
-
-
-# reframe_index is perhaps not needed any more.  Might be 
-# safe to remove it.  Leaving it here for now as it might
-# come in handy, and is well tested at the moment.
-def reframe_index(index, window_start=None, window_end=None):
-    """
-    Parameters
-    ----------
-    index : pd.DatetimeIndex
-
-    window_start, window_end : pd.Timestamp
-        The start and end of the window of interest. If this window
-        is larger than the duration of `data` then a single timestamp will be
-        inserted at `window_start` or `window_end` as necessary. If this window
-        is shorter than the duration of `data` then `data` will be cropped.
-
-    Returns
-    -------
-    index : pd.DatetimeIndex
-    """
-
-    tz = index.tzinfo
-    reset_tz = False
-
-    # Handle window...
-    if window_start is not None:
-        if window_start >= index[0]:
-            index = index[index >= window_start]
-        else:
-            index = index.insert(0, window_start)
-            reset_tz = True
-
-    if window_end is not None:
-        if window_end <= index[-1]:
-            index = index[index <= window_end]
-        else:
-            index = index.insert(len(index), window_end)
-            reset_tz = True
-
-    if reset_tz:
-        # index.insert breaks timezone.
-        index = index.tz_localize('UTC').tz_convert(tz)
-
-    return index

nilmtk/pipeline/locategoodsectionsresults.py → nilmtk/stats/goodsectionsresults.py

@@ -2,7 +2,7 @@
 import pandas as pd
 from datetime import timedelta
 
-class LocateGoodSectionsResults(Results):
+class GoodSectionsResults(Results):
     """
     Attributes
     ----------
@@ -15,7 +15,7 @@ class LocateGoodSectionsResults(Results):
 
     def __init__(self, max_sample_period):
         self.max_sample_period_td = timedelta(seconds=max_sample_period)
-        super(LocateGoodSectionsResults, self).__init__()
+        super(GoodSectionsResults, self).__init__()
 
     def append(self, timeframe, new_results):
         """Append a single result.
@@ -25,7 +25,7 @@ def append(self, timeframe, new_results):
         timeframe : nilmtk.TimeFrame
         new_results : {'sections': list of TimeFrame objects}
         """
-        super(LocateGoodSectionsResults, self).append(timeframe, new_results)
+        super(GoodSectionsResults, self).append(timeframe, new_results)
 
     @property
     def last_results(self):
@@ -64,4 +64,4 @@ def combined(self):
 
     def unify(self, other):
         # TODO!
-        super(LocateGoodSectionsResults, self).unify(other)
+        super(GoodSectionsResults, self).unify(other)

nilmtk/pipeline/tests/test_locategoodsections.py → nilmtk/stats/tests/test_locategoodsections.py

@@ -1,9 +1,9 @@
 #!/usr/bin/python
 from __future__ import print_function, division
 import unittest
-from nilmtk.pipeline import Pipeline, EnergyNode, LocateGoodSectionsNode
+from nilmtk.pipeline import Pipeline, TotalEnergy, GoodSections
 from nilmtk.pipeline.locategoodsectionsnode import reframe_index
-from nilmtk.pipeline.locategoodsectionsresults import LocateGoodSectionsResults
+from nilmtk.pipeline.locategoodsectionsresults import GoodSectionsResults
 from nilmtk.pipeline.energynode import _energy_for_power_series
 from nilmtk import TimeFrame, ElecMeter, HDFDataStore
 from nilmtk.elecmeter import ElecMeterID
@@ -28,7 +28,7 @@ def setUpClass(cls):
     def test_pipeline(self):
         meter = ElecMeter(store=self.datastore, metadata=self.meter_meta, 
                           meter_id=METER_ID)
-        nodes = [LocateGoodSectionsNode()]
+        nodes = [GoodSections()]
         pipeline = Pipeline(nodes)
         pipeline.run(meter)
 
@@ -97,7 +97,7 @@ def test_process(self):
         df.timeframe = TimeFrame(index[0], index[-1])
         df.look_ahead = pd.DataFrame()
 
-        locate = LocateGoodSectionsNode()
+        locate = GoodSections()
         locate.process(df, metadata)
         results = df.results['good_sections'].combined
         self.assertEqual(len(results), 4)
@@ -114,10 +114,10 @@ def test_process(self):
                       pd.Timestamp("2011-01-01 00:06:20")
         ]
         for split_point in [[4,6,9,17], [4,10,12,17]]:
-            locate = LocateGoodSectionsNode()
+            locate = GoodSections()
             df.results = {}
             prev_i = 0
-            aggregate_results = LocateGoodSectionsResults(MAX_SAMPLE_PERIOD)
+            aggregate_results = GoodSectionsResults(MAX_SAMPLE_PERIOD)
             for j,i in enumerate(split_point):
                 cropped_df = df.iloc[prev_i:i]
                 cropped_df.timeframe = TimeFrame(timestamps[j], timestamps[j+1])

nilmtk/pipeline/tests/test_energy.py → nilmtk/stats/tests/test_totalenergy.py

@@ -1,7 +1,7 @@
 #!/usr/bin/python
 from __future__ import print_function, division
 import unittest
-from nilmtk.pipeline import Pipeline, EnergyNode, ClipNode
+from nilmtk.pipeline import Pipeline, TotalEnergy, Clip
 from nilmtk.pipeline.energynode import _energy_for_power_series
 from nilmtk import TimeFrame, ElecMeter, HDFDataStore
 from nilmtk.elecmeter import ElecMeterID
@@ -45,7 +45,7 @@ def test_pipeline(self):
         meter = ElecMeter(store=self.datastore, 
                           metadata=self.meter_meta, 
                           meter_id=METER_ID)
-        nodes = [ClipNode(), EnergyNode()]
+        nodes = [Clip(), TotalEnergy()]
         pipeline = Pipeline(nodes)
         pipeline.run(meter)
         energy = deepcopy(pipeline.results['energy'])

nilmtk/pipeline/tests/test_energyresults.py → nilmtk/stats/tests/test_totalenergyresults.py

@@ -1,21 +1,21 @@
 #!/usr/bin/python
 from __future__ import print_function, division
 import unittest
-from nilmtk.pipeline.energyresults import EnergyResults
+from nilmtk.pipeline.energyresults import TotalEnergyResults
 from nilmtk import TimeFrame
 
 class TestEnergyResults(unittest.TestCase):
 
     def test_append(self):
-        er = EnergyResults()
+        er = TotalEnergyResults()
         tf = TimeFrame('2012-01-01','2012-01-02')
         er.append(tf, {'apparent':40, 'reactive':30, 'active':20})
         self.assertEqual(er._data.index.size, 1)
         self.assertEqual(er._data.index[0], tf.start)
         self.assertEqual(er._data['end'][tf.start], tf.end)
 
     def test_combined(self):
-        er = EnergyResults()
+        er = TotalEnergyResults()
         tf = TimeFrame('2012-01-01','2012-01-02')
         er.append(tf, {'apparent':40})
         tf2 = TimeFrame('2012-01-02','2012-01-03')