Merge pull request #1140 from changhiskhan/loffset

Implement loffset parameter in resample, close #1127

Author: wesm

pandas/core/generic.py

@@ -1,10 +1,12 @@
 # pylint: disable=W0231
+from datetime import timedelta
 
 import numpy as np
 
 from pandas.core.common import save, load
 from pandas.core.index import MultiIndex
 from pandas.tseries.index import DatetimeIndex
+from pandas.tseries.offsets import DateOffset
 
 #-------------------------------------------------------------------------------
 # Picklable mixin
@@ -158,7 +160,8 @@ def asfreq(self, freq, method=None, how=None):
         return asfreq(self, freq, method=method, how=how)
 
     def resample(self, rule, how='mean', axis=0, as_index=True,
-                 fill_method=None, closed='right', label='right', kind=None):
+                 fill_method=None, closed='right', label='right', kind=None,
+                 loffset=None):
         """
         Convenience method for frequency conversion and resampling of regular
         time-series data.
@@ -174,6 +177,8 @@ def resample(self, rule, how='mean', axis=0, as_index=True,
         label : {'right', 'left'}, default 'right'
             Which bin edge label to label bucket with
         as_index : see synonymous argument of groupby
+        loffset : timedelta
+            Adjust the resampled time labels
         """
         from pandas.tseries.resample import TimeGrouper
 
@@ -193,6 +198,9 @@ def resample(self, rule, how='mean', axis=0, as_index=True,
             # upsampling
             result = self.reindex(grouper.binner[1:], method=fill_method)
 
+        if isinstance(loffset, (DateOffset, timedelta)):
+            if len(result.index) > 0:
+                result.index = result.index + loffset
         return result
 
     def first(self, offset):

pandas/tests/test_index.py

@@ -221,8 +221,13 @@ def test_add(self):
         tm.assert_contains_all(self.strIndex, secondCat)
         tm.assert_contains_all(self.dateIndex, firstCat)
 
+    def test_timedelta(self):
         # this is valid too
         shifted = self.dateIndex + timedelta(1)
+        back = shifted + timedelta(-1)
+        self.assert_(tm.equalContents(self.dateIndex, back))
+        self.assertEqual(shifted.freq, self.dateIndex.freq)
+        self.assertEqual(shifted.freq, back.freq)
 
     def test_append_multiple(self):
         index = Index(['a', 'b', 'c', 'd', 'e', 'f'])

pandas/tseries/index.py

@@ -150,7 +150,6 @@ class DatetimeIndex(Int64Index):
     __le__ = _dt_index_cmp('__le__')
     __ge__ = _dt_index_cmp('__ge__')
 
-    __add__ = _dt_index_op('__add__')
     __sub__ = _dt_index_op('__sub__')
 
     # structured array cache for datetime fields

pandas/tseries/tests/test_resample.py

@@ -1,11 +1,11 @@
-from datetime import datetime
+from datetime import datetime, timedelta
 
 import numpy as np
 
 from pandas import Series, DataFrame, isnull, notnull
 
 from pandas.tseries.index import date_range
-from pandas.tseries.offsets import Minute
+from pandas.tseries.offsets import Minute, bday
 from pandas.tseries.period import period_range
 from pandas.tseries.resample import DatetimeIndex, TimeGrouper
 import pandas.tseries.offsets as offsets
@@ -133,6 +133,26 @@ def test_resample_basic(self):
         self.assertEquals(result.irow(1), s['1/4/2005'])
         self.assertEquals(result.irow(5), s['1/10/2005'])
 
+    def test_resample_loffset(self):
+        rng = date_range('1/1/2000 00:00:00', '1/1/2000 00:13:00', freq='min')
+        s = Series(np.random.randn(14), index=rng)
+        result = s.resample('5min', how='mean', closed='right', label='right',
+                            loffset=timedelta(minutes=1))
+        idx = date_range('1/1/2000', periods=4, freq='5min')
+        expected = Series([s[0], s[1:6].mean(), s[6:11].mean(), s[11:].mean()],
+                          index=idx + timedelta(minutes=1))
+        assert_series_equal(result, expected)
+
+                # from daily
+        dti = DatetimeIndex(start=datetime(2005,1,1), end=datetime(2005,1,10),
+                            freq='D')
+        ser = Series(np.random.rand(len(dti)), dti)
+
+        # to weekly
+        result = ser.resample('w-sun', how='last')
+        expected = ser.resample('w-sun', how='last', loffset=-bday)
+        self.assertEqual(result.index[0] - bday, expected.index[0])
+
     def test_resample_upsample(self):
         # from daily
         dti = DatetimeIndex(start=datetime(2005,1,1), end=datetime(2005,1,10),

pandas/util/testing.py

@@ -163,10 +163,6 @@ def makeFloatIndex(k):
     values = sorted(np.random.random_sample(k)) - np.random.random_sample(1)
     return Index(values * (10 ** np.random.randint(0, 9)))
 
-def makeDateIndex(k):
-    dates = list(bdate_range(datetime(2000, 1, 1), periods=k))
-    return Index(dates)
-
 def makeFloatSeries():
     index = makeStringIndex(N)
     return Series(randn(N), index=index)