Vectorized part of on_off_status and updated test pickle files

Author: nickpowersys

caar/timeseries.py

@@ -103,13 +103,21 @@ def on_off_status(df, id, start, end, freq='1min'):
                            .snap(freq=freq)
                            .tolist())
     # Populate array
+    starts = _integer_index_based_on_freq(starts_by_freq, start, freq)
+    ends = _integer_index_based_on_freq(record_ends_by_freq, start, freq)
+
     for i in range(len(records)):
-        start_index = _index_of_timestamp(start, starts_by_freq[i], freq)
-        end_index = _index_of_timestamp(start, record_ends_by_freq[i], freq)
-        status_in_intervals[start_index:end_index + 1]['on'] = 1
+        status_in_intervals[starts[i]:ends[i] + 1]['on'] = 1
     return status_in_intervals
 
 
+def _integer_index_based_on_freq(datetimes, reference, frequency):
+    time_deltas = pd.Series(datetimes) - reference
+    freq = _timedelta_from_string(frequency)
+    indexes = np.array(time_deltas/freq).astype(np.int)
+    return indexes
+
+
 def temps_arr_by_freq(df, id, start, end, cols=None, freq='1min', actuals_only=False):
     """Returns NumPy array containing timestamps ('times') and temperatures at the specified frequency. If *actuals_only* is True, only the observed temperatures will be returned in an array. Otherwise, by default, intervals without observations are filled with zeros.
 

tests/data/test_TX_cycles.pickle

@@ -228,20 +228,20 @@ def test_df_creation_after_fixed_dict(data_file, states, thermostats, postal, cy
     df = df_creation_func(clean_dict, **kwargs)
     assert isinstance(df, pd.DataFrame)
 
-# @slow
-# @pytest.mark.parametrize("df_fixture, id, start, end, freq",
-#                          [(cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
-#                            dt.datetime(2012, 6, 18, 23, 0, 0), '1min30s'),
-#                           (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
-#                            dt.datetime(2012, 6, 18, 23, 0, 0), 'min30s'),
-#                           (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
-#                            dt.datetime(2012, 6, 18, 23, 0, 0), '2min'),
-#                           (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
-#                            dt.datetime(2012, 6, 18, 23, 0, 0), 'min')])
-# def test_on_off_status_by_interval(df_fixture, id, start, end, freq):
-#     kwargs = {'freq': freq}
-#     on_off = ts.on_off_status(df_fixture, id, start, end, **kwargs)
-#     assert len(on_off['times']) > 0
+
+@pytest.mark.parametrize("df_fixture, id, start, end, freq",
+                         [(cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
+                           dt.datetime(2012, 6, 18, 23, 0, 0), '1min30s'),
+                          (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
+                           dt.datetime(2012, 6, 18, 23, 0, 0), 'min30s'),
+                          (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
+                           dt.datetime(2012, 6, 18, 23, 0, 0), '2min'),
+                          (cycle_df_fixture(), THERMO_ID1, dt.datetime(2012, 6, 18, 21, 0, 0),
+                           dt.datetime(2012, 6, 18, 23, 0, 0), 'min')])
+def test_on_off_status_by_interval(df_fixture, id, start, end, freq):
+    kwargs = {'freq': freq}
+    on_off = ts.on_off_status(df_fixture, id, start, end, **kwargs)
+    assert len(on_off['times']) > 0
 #
 # @slow
 # @pytest.mark.parametrize("df_fixture, id, start, end, freq",