Group answers & index based on Questions in python

Question

I have a dataframe as such for analysis purpose, I need to create a list of dictionaries as:

TARGET OUTPUT

[
{ 'is my anti hiv test conclusive or--Bla bla': [0, 1, 2] }, 
{'I have some hip pain 9 weeks--bla bla': [3, 4, 5, 6]} 
]

Here the list is indices of answers and not the actual answers

Well yes, the obvious method is to use groupby but facing some errors

I tried printing before converting to list. And it seems fine actually,

Can y'all please help me figure out it's correct syntax so I could to my targeted output.

Dataset link If somebody needs the shared notebook link, let me know in the comments.

Answer 1

You need to actually select the column ("index") whose values you want to appear in the list:

df_ans = data.groupby(["question_text"])["index"].apply(list).to_dict()

instead of

df_ans = data.groupby(["question_text"]).apply(list).to_dict()

Otherwise you get a list of the columns, as in your example. That's what happens when you convert a DataFrame to a list, i.e. list(data) gives you the same as list(data.columns) .