How to get One-Hot encoded matrix from a survey table and vector of answers

Question

I have a surveys' answers from the participants in a pandas dataframe:

 ['A', 'B', 'C', 'A' ...],
 ['D', 'B', 'B', 'A' ...],
 ......................

 ['D', 'C', 'C', 'A' ...]]

and I have a vector of keys to the survey:

['D', 'B', 'B', 'A' ...]

I need to get a dataframe which displays the boolean results of survey like:

 [0, 1, 0, 1 ...],
 [1, 1, 1, 1 ...],
 ......................

 [1, 0, 0, 1 ...]]

I've tried to use pd.get_dummies(users_answ, keys) but that seems wrong

It is unclear from your question and data sample how your data is structured in terms of if answers are by user or by question. Your question needs a minimal reproducible example consisting of sample input, expected output, actual output, and only the relevant code necessary to reproduce the problem. See [How to make good reproducible pandas examples](https://stackoverflow.com/questions/20109391/how-to-make-good-reproducible-pandas-examples) for best practices related to Pandas questions. — itprorh66, Oct 06 '22 at 13:12
@itprorh66 already found, thanks anyway https://stackoverflow.com/questions/42790968/compare-a-matrix-against-a-column-vector — Nikita Lvov, Oct 06 '22 at 13:18

score 0 · Answer 1 · answered Oct 06 '22 at 13:15

You should be able to simply check the equality between the DataFrame and the list. The list should get aligned to the DataFrame across the columns:

df = pd.DataFrame([[*'ABCA'],[*'DBBA'],[*'DCCA']])
keys = [*'DBBA']

print(df)
   0  1  2  3
0  A  B  C  A
1  D  B  B  A
2  D  C  C  A

print(keys)
['D', 'B', 'B', 'A']

print(df == keys)
       0      1      2     3
0  False   True  False  True
1   True   True   True  True
2   True  False  False  True

# If you want actual integers instead of booleans
print((df == keys).astype(int))
   0  1  2  3
0  0  1  0  1
1  1  1  1  1
2  1  0  0  1

Nikita Lvov · Accepted Answer · 2022-10-06T14:43:07.417

The easiest way seems to use pandas eq function https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.eq.html#pandas.DataFrame.eq

So the whole solution oneline:

users_answ.eq(keys, axis=0)

Alternative solution:

#new array
checked_answ = []
#taking each row of surveys answers df
for r in range(0, users_answ.shape[0]): 
    row = users_answ.iloc[r].tolist()
    #creating the array for this row
    p = []
    for i in range(0, len(keys)):
        if(keys[i] == row[i]):
            p.append(1)
        else:
            p.append(0)
    checked_answ.append(p)

How to get One-Hot encoded matrix from a survey table and vector of answers

2 Answers2