I currently have a pandas Series with dtype Timestamp, and I want to group it by date (and have many rows with different times in each group).
Series
Timestamp
To add another suggestion, I often use the following as it uses simple logic:
pd.Series(index=s.values).groupby(level=0)