Filter groupby object: select all rows associated with customers satisfying certain criteria

In the fictitious data here, I want to identify any customers who did not purchase alcoholic beverages (specifically "Beer" or "Wine" in example) across all visits combined.

df = pd.DataFrame({"customer": ['A55', 'A55', 'A55', 
                       'B080', 'B080', 
                       'D900', 'D900', 
                       'Z900', 'Z900', 'Z900', 'Z900'],
               "date":['01/11/2016', '01/21/2016', '02/11/2016', 
                       '08/17/2016', '6/17/2016',
                       '03/01/2016','04/30/2016', 
                       '05/16/2016','09/27/2016', '10/05/2016', '11/11/2016'],
               "item": ['Beer', 'Beer', 'Wine', 
                        'Coffee', 'Gatorade', 
                        'Wine', 'Coffee', 
                        'Beer', 'Wine', 'Beer', 'Coffee']
})

    ##Tag visit number for each customer
df["customer_visit_number"] = df.groupby("customer").cumcount().add(1)

In the dataframe example below, only customer B080 never bought any Beer or Wine. By contrast the others bought Beer or Wine on occasion or all visits. To tag customer visits I added a cumcount() operation but not all customers visit the store same number of times.

Is there a simple way in pandas to filter all rows associated with this customer only after doing the groupby operation at customer level?

See dataframe below

>Solution :

Identify customers that purchased alcoholic beverages:

drop = (df['item'].isin(['Beer', 'Wine'])
        .groupby(df['customer']).any()
       )

Filter with boolean indexing:

out = df[~df['customer'].isin(drop[drop].index)]

Or with groupby.transform to directly obtain a Series of booleans for boolean indexing:

m = (df['item'].isin(['Beer', 'Wine'])
     .groupby(df['customer'])
     .transform('any')
    )
out = df[~m]

Output:

  customer        date      item
3     B080  08/17/2016    Coffee
4     B080   6/17/2016  Gatorade

Intermediates:

# drop
customer
A55      True
B080    False
D900     True
Z900     True
Name: item, dtype: bool

# m
0      True
1      True
2      True
3     False
4     False
5      True
6      True
7      True
8      True
9      True
10     True
Name: item, dtype: bool