将数据框中的列追加到字典python

Question

I want to set two columns from my dataframe (EisDat_cvs) as dictionary key and value in python, and everything works fine using this code:

dict_EisukeID_HMDB = EisDat_cvs.set_index('Unnamed: 28')['Unnamed: 26'].to_dict()

However now I want to append two columns as values and one as key in the dictionary, and I tried to modify the previous one as:

dict_EisukeID_HMDB = EisDat_cvs.set_index('Unnamed: 28')['Unnamed: 26', 'Unnamed: 1'].to_dict()

But python throws me a key error...

EDIT

I have a dataframe EisDat_cvs, of which I am considering 3 columns (Unnamed: 28, Unnamed: 26, Unnamed: 1). And I want to get a dictionary that takes the values in column 28 as my keys, and values in column 26 and 1 as values in the dictionary. Something like this:

Unnamed: 28, Unnamed: 26, Unnamed: 1
bla.          1           90
cra           2           12
ta            3           12

and my output should look like

dict_EisukeID_HMDB = { 'bla': '1', '90'
                       'cra': '2', '12'
                       'ta':  '3', '12'}

Answer 1

You can try this:

dict(zip(
    EisDat_cvs['Unnamed: 28'],
    zip(EisDat_cvs['Unnamed: 26'], EisDat_cvs['Unnamed: 1'])
))

The issue you had was not with to_dict() but instead by passing the columns as a tuple, you are trying to access the dataframe through a hierarchal column structure , which if your dataframe does not have, will throw a KeyError .

---

Hierarchal columns example:

df = pd.DataFrame({
    ("A", "A0"): [10, 30, 50, 70],
    ("A", "A1"): [11, 31, 51, 71],
    ("B", "B0"): [20, 40, 60, 80],
    ("B", "B1"): [21, 41, 61, 81]
})

print(df["A", "A0"])

Output:

0    10
1    30
2    50
3    70
Name: (A, A0), dtype: int64