BUG: fix BH resampling (GH12351)

close pandas-dev#12351 xref pandas-dev#13364

Author: scari

pandas/core/resample.py

@@ -10,10 +10,10 @@
                                  SeriesGroupBy, groupby, PanelGroupBy)
 
 from pandas.tseries.frequencies import to_offset, is_subperiod, is_superperiod
-from pandas.core.indexes.datetimes import DatetimeIndex, date_range
-from pandas.core.indexes.timedeltas import TimedeltaIndex
-from pandas.tseries.offsets import DateOffset, Tick, Day, _delta_to_nanoseconds
-from pandas.core.indexes.period import PeriodIndex, period_range
+from pandas.tseries.index import DatetimeIndex, date_range
+from pandas.tseries.tdi import TimedeltaIndex
+from pandas.tseries.offsets import DateOffset, Tick, Day, BusinessHour, _delta_to_nanoseconds
+from pandas.tseries.period import PeriodIndex, period_range
 import pandas.core.common as com
 import pandas.core.algorithms as algos
 
@@ -1016,7 +1016,7 @@ def __init__(self, freq='Min', closed=None, label=None, how='mean',
                  convention=None, base=0, **kwargs):
         freq = to_offset(freq)
 
-        end_types = set(['M', 'A', 'Q', 'BM', 'BA', 'BQ', 'W'])
+        end_types = set(['M', 'A', 'Q', 'BH', 'B', 'BM', 'BA', 'BQ', 'W'])
         rule = freq.rule_code
         if (rule in end_types or
                 ('-' in rule and rule[:rule.find('-')] in end_types)):
@@ -1298,7 +1298,8 @@ def _get_range_edges(first, last, offset, closed='left', base=0):
             return _adjust_dates_anchored(first, last, offset,
                                           closed=closed, base=base)
 
-    if not isinstance(offset, Tick):  # and first.time() != last.time():
+    # GH12351
+    if not isinstance(offset, Tick) and not isinstance(offset, BusinessHour):
         # hack!
         first = first.normalize()
         last = last.normalize()

pandas/tests/test_resample.py

@@ -2572,6 +2572,26 @@ def test_upsample_daily_business_daily(self):
         expected = ts.asfreq('H', how='s').reindex(exp_rng)
         assert_series_equal(result, expected)
 
+    '''
+    def test_resample_business_hourly(self):
+        rng = pd.date_range(start='2016-02-11 00:00:00',
+                            end='2016-02-12 23:00:00',
+                            freq='H')
+        expected_rng = pd.date_range(start='2016-02-11 00:00:00',
+                                     end='2016-02-12 23:00:00',
+                                     freq='H')
+        df = pd.DataFrame(index=rng)
+
+        df.resample('BH').count()
+    '''
+
+    def test_resample_empty(self):
+        ts = _simple_pts('1/1/2000', '2/1/2000')[:0]
+
+        result = ts.resample('A').asfreq()
+        self.assertEqual(len(result), 0)
+
+
     def test_resample_irregular_sparse(self):
         dr = date_range(start='1/1/2012', freq='5min', periods=1000)
         s = Series(np.array(100), index=dr)

pandas/tseries/frequencies.py

@@ -1114,7 +1114,7 @@ def is_subperiod(source, target):
         return source in ['C', 'H', 'T', 'S', 'L', 'U', 'N']
     elif target == 'D':
         return source in ['D', 'H', 'T', 'S', 'L', 'U', 'N']
-    elif target == 'H':
+    elif _is_hourly(source):
         return source in ['H', 'T', 'S', 'L', 'U', 'N']
     elif target == 'T':
         return source in ['T', 'S', 'L', 'U', 'N']
@@ -1170,7 +1170,7 @@ def is_superperiod(source, target):
         return target in ['D', 'C', 'B', 'H', 'T', 'S', 'L', 'U', 'N']
     elif source == 'D':
         return target in ['D', 'C', 'B', 'H', 'T', 'S', 'L', 'U', 'N']
-    elif source == 'H':
+    elif _is_hourly(source):
         return target in ['H', 'T', 'S', 'L', 'U', 'N']
     elif source == 'T':
         return target in ['T', 'S', 'L', 'U', 'N']
@@ -1213,6 +1213,11 @@ def _is_weekly(rule):
     return rule == 'W' or rule.startswith('W-')
 
 
+def _is_hourly(rule):
+    rule = rule.upper()
+    return rule == 'H' or rule.startswith('BH')
+
+
 DAYS = ['MON', 'TUE', 'WED', 'THU', 'FRI', 'SAT', 'SUN']
 
 MONTHS = tslib._MONTHS