# flatten the groupby
flat_week_groups = week_groups.reset_index()
# create an index of all the weeks from start_date to now_date
# to pick up the missing weeks
dummy_index = []
for dtime in rrule.rrule(rrule.WEEKLY, dtstart=start_date, until=now_date):
    dt_week = dtime.isocalendar()[1]
    dt_year = dtime.isocalendar()[0]
    dummy_index.append(tuple([dt_year, dt_week]))
# add the missing weeks to flat_week_groups
for i in dummy_index:
    if i not in week_groups.index:
        flat_week_groups = flat_week_groups.append( \\
                 {'date_year': i[0], \\
                  'date_week': i[1], \\
                  'value': 0},ignore_index=True)
# sort by year and week
# re-stack the DataFrame with the added values for empty weeks
week_groups = flat_week_groups.sort_values( \\
     ['date_year','date_week']).set_index(['date_year','date_week'])
week_groups.plot(kind='bar',figsize=(10,5),legend=None)