如何在条件下对熊猫数据框进行子集

Question

I have a pandas dataframe like this:

 buyer_id meal_type
    139       Veg
    140    NonVeg
    140       Veg
     36    NonVeg
     79       Veg
     79    NonVeg
     79    NonVeg
     72    NonVeg
     72    NonVeg
     65    NonVeg
     65       Veg

Now I want all the buyer_id which has only Veg as a meal type,all the buyer_id which has Veg and NonVeg as meal type and all the buyer_id which has only NonVeg as a meal type.

so,

139 Veg
140 Veg and NonVeg
36  NonVeg
79  Veg and NonVeg

and so on. I am doing the following in Python to get all the buyer ids that have Veg , NonVeg or both.

segments_data.buyer_id[segments_data['meal_type']=='Veg' &      
segments_data['meal_type']=='NonVeg']

But It doesn't work. Please help.

Answer 1

Grouping by buyer_id and converting meal_type into a set works:

>>> df.groupby('buyer_id')['meal_type'].apply(set).str.join(' and ')

    buyer_id
36             NonVeg
65     Veg and NonVeg
72             NonVeg
79     Veg and NonVeg
139               Veg
140    Veg and NonVeg
Name: meal_type, dtype: object