Better handling for pandas stacked column dataframe

[LeeTZ]

For a simple file beakerxdata.csv:

time,downloads,package
20180906 20:00:00 -0400,21,beakerx
20180906 20:00:00 -0400,2,jlab
20180907 20:00:00 -0400,23,beakerx
20180907 20:00:00 -0400,1,jlab
20180908 20:00:00 -0400,3,beakerx
20180908 20:00:00 -0400,123,jlab
20180909 20:00:00 -0400,32,beakerx
20180909 20:00:00 -0400,31,jlab
20181006 20:00:00 -0400,21,beakerx
20181006 20:00:00 -0400,2,jlab
20181007 20:00:00 -0400,23,beakerx
20181007 20:00:00 -0400,1,jlab
20181008 20:00:00 -0400,3,beakerx
20181008 20:00:00 -0400,123,jlab
20181009 20:00:00 -0400,32,beakerx
20181009 20:00:00 -0400,31,jlab

If we do

mini_df = pd.read_csv('beakerxdata.csv')
mini_df['downloads'] = pd.to_numeric(mini_df['downloads'])
mini_df['time'] = pd.DatetimeIndex(mini_df['time']).to_period('M').to_timestamp(how='end')
s = mini_df.groupby([pd.Grouper(key='time',freq='M'), 'package']).sum().reset_index()
n = s[['time','package','downloads']].set_index(['time','package']).unstack('time')
n

This should be the expected behavior:

However in our table the columns are shown in timestamps, not in readable format