Update events for pandas v3.0 compatibility

python/events/README.md

@@ -61,15 +61,15 @@ observations = observations.drop_duplicates(
 # Resample to hourly, keep first measurement in each 1-hour bin
 observations = observations.groupby([
     'usgs_site_code',
-    Grouper(key='value_date', freq='H')
+    Grouper(key='value_date', freq='h')
     ]).first().ffill()
 
 # Detect events
 events = observations['value'].groupby(
     level='usgs_site_code').apply(
         list_events_helper, 
         level='usgs_site_code', 
-        halflife='6H', 
+        halflife='6h', 
         window='7D'
     )
 

python/events/src/hydrotools/events/event_detection/_version.py

@@ -1 +1 @@
-__version__ = "1.1.5"
+__version__ = "1.1.6"

python/events/src/hydrotools/events/event_detection/decomposition.py

@@ -153,12 +153,12 @@ def event_boundaries(event_points: pd.Series):
 
         """
     # Identify event starts
-    forward_shift = event_points.shift(1).fillna(False)
+    forward_shift = event_points.shift(1).astype(bool).fillna(False)
     starts = (event_points & ~forward_shift)
     starts = starts[starts]
 
     # Identify event ends
-    backward_shift = event_points.shift(-1).fillna(False)
+    backward_shift = event_points.shift(-1).astype(bool).fillna(False)
     ends = (event_points & ~backward_shift)
     ends = ends[ends]
 
@@ -212,8 +212,8 @@ def mark_event_flows(
     series: pd.Series,
     halflife: Union[float, str, pd.Timedelta],
     window: Union[int, pd.tseries.offsets.DateOffset, pd.Index],
-    minimum_event_duration: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0H',
-    start_radius: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0H'
+    minimum_event_duration: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0h',
+    start_radius: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0h'
     ) -> pd.Series:
     """Model the trend in a streamflow time series by taking the max
         of two rolling minimum filters applied in a forward and 
@@ -236,10 +236,10 @@ def mark_event_flows(
         window: int, offset, or BaseIndexer subclass, required
             Size of the moving window for `pandas.Series.rolling.min`.
             This filter is used to model the trend in `series`.
-        minimum_event_duration: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0H'
+        minimum_event_duration: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0h'
             Enforce a minimum event duration. This should generally be set equal to 
             halflife to reduce the number of false positives flagged as events.
-        start_radius: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0H'
+        start_radius: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0h'
             Shift event starts to a local minimum. Phase shifts imparted on the 
             original signal may advance or delay event start times depending upon how 
             much smoothing is required to eliminate noise.
@@ -293,8 +293,8 @@ def list_events(
     series: pd.Series,
     halflife: Union[float, str, pd.Timedelta],
     window: Union[int, pd.tseries.offsets.DateOffset, pd.Index],
-    minimum_event_duration: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0H',
-    start_radius: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0H'
+    minimum_event_duration: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0h',
+    start_radius: Union[pd.Timedelta, datetime.timedelta, np.timedelta64, str, int] = '0h'
     ) -> pd.DataFrame:
     """Apply time series decomposition to mark event values in a streamflow
         time series. Discretize continuous event values into indiviual events.
@@ -312,10 +312,10 @@ def list_events(
         window: int, offset, or BaseIndexer subclass, required
             Size of the moving window for `pandas.Series.rolling.min`.
             This filter is used to model the trend in `series`.
-        minimum_event_duration: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0H'
+        minimum_event_duration: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0h'
             Enforce a minimum event duration. This should generally be set equal to 
             halflife to reduce the number of false positives flagged as events.
-        start_radius: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0H'
+        start_radius: pandas.Timedelta, datetime.timedelta, numpy.timedelta64, str, int, optional, default '0h'
             Shift event starts to a local minimum. Phase shifts imparted on the 
             original signal may advance or delay event start times depending upon how 
             much smoothing is required to eliminate noise.

python/events/tests/test_decomposition.py

@@ -23,12 +23,12 @@ def test_list_events():
         index=pd.date_range(
             start=pd.to_datetime('2018-01-01'),
             periods=len(t),
-            freq='H'),
+            freq='h'),
         name='streamflow'
     )
 
     # Detect event
-    events = ev.list_events(series, '6H', '7D')
+    events = ev.list_events(series, '6h', '7D')
 
     # Should detect a single event
     assert len(events.index) == 1
@@ -56,12 +56,12 @@ def test_list_events_noise():
         index=pd.date_range(
             start=pd.to_datetime('2018-01-01'),
             periods=len(t),
-            freq='H'),
+            freq='h'),
         name='streamflow'
     )
 
     # Detect event
-    events = ev.list_events(series, '6H', '7D', '6H', '7H')
+    events = ev.list_events(series, '6h', '7D', '6h', '7h')
 
     # Should detect a single event
     assert len(events.index) == 1
@@ -80,7 +80,7 @@ def test_local_minimum_datetime_exception():
     with pytest.raises(Exception):
         idx = ev.find_local_minimum(
             pd.Timestamp('2020-01-01 01:00'),
-            '3H',
+            '3h',
             series
         )
 
@@ -91,15 +91,15 @@ def test_origin_not_idx():
         index=pd.date_range(
             start=pd.to_datetime('2018-01-01'),
             periods=10,
-            freq='H'),
+            freq='h'),
         name='streamflow'
     )
 
     # Test local minimum
     with pytest.raises(Exception):
         idx = ev.find_local_minimum(
             pd.Timestamp('2020-01-01 01:00'),
-            '3H',
+            '3h',
             series
         )
 
@@ -110,7 +110,7 @@ def test_bad_time_series_idx():
         index=[i for i in range(5)]
     )
     with pytest.raises(Exception):
-        events = ev.list_events(series, '6H', '7D')
+        events = ev.list_events(series, '6h', '7D')
 
     # Not monotonic
     series = pd.Series(
@@ -125,7 +125,7 @@ def test_bad_time_series_idx():
         name='streamflow'
     )
     with pytest.raises(Exception):
-        events = ev.list_events(series, '6H', '7D')
+        events = ev.list_events(series, '6h', '7D')
 
     # Duplicated
     series = pd.Series(
@@ -140,7 +140,7 @@ def test_bad_time_series_idx():
         name='streamflow'
     )
     with pytest.raises(Exception):
-        events = ev.list_events(series, '6H', '7D')
+        events = ev.list_events(series, '6h', '7D')
 
 def test_null_warning():
     series = pd.Series(
@@ -155,4 +155,4 @@ def test_null_warning():
         name='streamflow'
     )
     with pytest.warns(UserWarning):
-        events = ev.list_events(series, '6H', '7D')
+        events = ev.list_events(series, '6h', '7D')