How to use different combinations of group by while trying to get the top most viewed

Question

Am trying to get the top most rating using groupby of multiple columns and if there is no combination of that particular groupby, its throwing me an error . how to do multiple c

Answer 1

You can use *args for dynamic input, (ordering of values cannot be changed) with query for filtering:

def get_top(*args):     
    c = ['maritalstatus', 'gender', 'age_range', 'occ']
    m = (df.groupby(c)['rating'].apply(lambda x: x.value_counts().index[0])
           .reset_index())
    args = list(args)
    while True:
        d = dict(zip(c, args))
        #https://stackoverflow.com/a/48371587/2901002
        q = ' & '.join((('({} ==  "{}")').format(i, j)) for i, j in d.items())
        m1 = m.query(q)['rating']  
        if m1.empty and len(args) > 1:
            args.pop()
        else:
            return m1

print(get_top('ma', 'M', 'young','teacher'))
1    PG
Name: rating, dtype: object