I have a dataframe that has two columns and I want to create a list containing all the values in the second column for the same value in column one.

If I have a dataframe that looks like:

So I want to be either be able to divide this datafram into three separate df for Cars, Fruits, and Countries. Or I want to have a list for Cars, Fruits, and Countries that would like this:

Cars = ['Toyota', 'Honda', 'Tesla']
Fruits = ['Apple', 'Orange']
Countries = ['USA, 'Mexico']

This is just an example, my dataframe is huge so I want to have a function that does this without having to manually type in each Type. I tried looking up groupby function for pandas but don't think I was able to find how I can use it to do what I need to.

Any help is appreciated.

CodePudding user response：

You can try this :

dict_ = df.groupby('Type').agg(list).T.to_dict()

for key in dict_:
  li_ = dict_.get(key).get("Item")
  globals()[key] = li_

Also, you can use locals depends on your scope :

dict_ = df.groupby('Type').agg(list).T.to_dict()

for key in dict_:
  li_ = dict_.get(key).get("Item")
  locals()[key] = li_

You can get :

locals()["Cars"]
Out[1]: ['Toyota', 'Honda', 'Teska']

globals()["Cars"]
Out[2]: ['Toyota', 'Honda', 'Teska']

CodePudding user response：

My try to solve your question :)

import pandas as pd

df = pd.DataFrame({'Type': ['Cars', 'Cars', 'Cars', 'Fruits', 'Fruits', 'Countries', 'Countries'],
                   'Item': ['Toyota', 'Honda', 'Tesla', 'Apple', 'Orange', 'USA', 'Mexico']})

grouped = df.groupby('Type')['Item'].apply(lambda tags:', '.join(tags)).to_frame()

print(grouped)

Output:

                           Item
Type                           
Cars       Toyota, Honda, Tesla
Countries           USA, Mexico
Fruits            Apple, Orange