code sample for backport pandas-dev#37305

Author: simonjayhawkins

bisect/37305.py

@@ -0,0 +1,74 @@
+import numpy as np
+import pandas as pd
+
+print(pd.__version__)
+
+df = pd.DataFrame(
+    {"col": range(10)}, index=pd.date_range("2012-01-01", periods=10, freq="20min")
+)
+df
+
+df.index
+
+# res = df.resample("H").apply(lambda group: len(group["col"].unique()))
+# res
+
+# res.index
+
+res = df.resample("H")
+res
+
+res.axis
+
+res.binner
+
+res.grouper
+
+res._set_binner()
+res
+
+res.binner
+
+res.grouper
+
+func = lambda group: len(group["col"].unique())
+func
+
+# from pandas.core.aggregation import aggregate
+
+# result, how = aggregate(res, func)
+# result, how # (None, True)
+
+# res.binner
+# # DatetimeIndex(['2012-01-01 00:00:00', '2012-01-01 01:00:00',
+# #                '2012-01-01 02:00:00', '2012-01-01 03:00:00',
+# #                '2012-01-01 04:00:00'],
+# #               dtype='datetime64[ns]', freq='H')
+
+# res.grouper
+# # <pandas.core.groupby.ops.BinGrouper at 0x2768ebbc550>
+
+obj = res._selected_obj
+obj
+
+obj is df
+
+obj.index
+
+grouper = res.grouper
+grouper
+
+axis = res.axis
+axis
+
+from pandas.core.groupby.groupby import get_groupby
+
+grouped = get_groupby(obj, by=None, grouper=grouper, axis=axis)
+grouped
+
+# result = grouped._aggregate_item_by_item(func)
+
+result = grouped.apply(func)
+print(result)
+
+assert result.index.freq is None