BUG,TST: Remove case where vectorization fails in pct_change groupby method, improve test. #21235

Author: simonariddell

pandas/core/groupby/groupby.py

@@ -3935,38 +3935,10 @@ def _apply_to_column_groupbys(self, func):
         return func(self)
 
     def pct_change(self, periods=1, fill_method='pad', limit=None, freq=None):
-        """Calculate pct_change of each value to previous entry in group"""
-        grouper = self.grouper
-        cache_exist = getattr(grouper, '_cache', False)
-        if cache_exist:
-            in_cache = True if 'is_monotonic' in cache_exist.keys() else False
-        else:
-            in_cache = False
-        m = grouper.is_monotonic if in_cache else False
-        if not m or fill_method is None:
-            return self.apply(lambda x: x.pct_change(periods=periods,
-                                                     fill_method=fill_method,
-                                                     limit=limit, freq=freq))
-
-        def get_invalid_index(x):
-            if periods == 0:
-                return x
-            elif periods > 0:
-                ax = Index(np.arange(min(x), min(x) + periods))
-                return ax
-            elif periods < 0:
-                ax = Index(np.arange(max(x), max(x) + periods, -1))
-                return ax
-
-        filled = getattr(self, fill_method)(limit=limit)
-        shifted = filled.shift(periods=periods, freq=freq)
-        pct_change = (filled / shifted) - 1
-
-        invalid_index = Index([])
-        for i in [get_invalid_index(v) for k, v in self.indices.items()]:
-            invalid_index = invalid_index.union(i)
-        pct_change.iloc[invalid_index] = np.nan
-        return pct_change
+        """Calcuate pct_change of each value to previous entry in group"""
+        return self.apply(lambda x: x.pct_change(periods=periods,
+                                                 fill_method=fill_method,
+                                                 limit=limit, freq=freq))
 
 
 class NDFrameGroupBy(GroupBy):

pandas/tests/groupby/test_transform.py

@@ -723,16 +723,13 @@ def interweave(list_obj):
 
 @pytest.mark.parametrize("test_series", [True, False])
 @pytest.mark.parametrize("shuffle", [True, False])
-@pytest.mark.parametrize("activate_cache", [True, False])
 @pytest.mark.parametrize("periods,fill_method,limit", [
     (1, 'ffill', None), (1, 'ffill', 1),
     (1, 'bfill', None), (1, 'bfill', 1),
     (-1, 'ffill', None), (-1, 'ffill', 1),
     (-1, 'bfill', None), (-1, 'bfill', 1),
-    (-1, None, None), (-1, None, 1),
-    (-1, None, None), (-1, None, 1)
 ])
-def test_pct_change(test_series, shuffle, activate_cache, periods, fill_method, limit):
+def test_pct_change(test_series, shuffle, periods, fill_method, limit):
     vals = [3, np.nan, 1, 2, 4, 10, np.nan, 9]
     keys = ['a', 'b']
     key_v = np.repeat(keys, len(vals))
@@ -751,9 +748,6 @@ def test_pct_change(test_series, shuffle, activate_cache, periods, fill_method,
     exp = pd.DataFrame(exp_vals.values, columns=['A'])
     grp = df.groupby('key')
 
-    if activate_cache:
-        grp.grouper.is_monotonic
-
     def get_result(grp_obj):
         return grp_obj.pct_change(periods=periods,
                                   fill_method=fill_method,