使用每列中的所有字符串数据旋转 pandas dataframe

Question

Sample dataframe:

Name   Attribute     Response
Joe    A             Yes
Joe    B             smoking 
Joe    B             headache
Mary   A             Null
Mary   B             Never
Bob    C             Today
Mary   A             Tomorrow

I have tried for several hours and searching through all apparently similar SO questions to pivot this df to the below desired output. Note, Joe and Mary have more than one row in which the Attribute is the same, but the response is different.

Desired output

Name    A                    B                     C
Joe    Yes                   smoking, headache     Null
Mary   Null, tomorrow        Never                 Null
Bob    Null                  Null                  Today

Again, to reiterate, I have looked through every SO response regarding reshaping dataframes from long to wide and none of them involved this precise question. Furthermore, each of those responses involved answers which I implemented and all resulted in errors, either a Value Error or Data Error, especially an error stating the index contained duplicate values. So, your help is appreciated.

Answer 1

You can do .pivot_table() with aggfunc=list :

print(
    df.pivot_table(
        index="Name", columns="Attribute", aggfunc=list, fill_value="Null"
    ).droplevel(0, axis=1)
)

Prints:

Attribute                 A                    B        C
Name                                                     
Bob                    Null                 Null  [Today]
Joe                   [Yes]  [smoking, headache]     Null
Mary       [Null, Tomorrow]              [Never]     Null

---

Or if you don't want lists:

print(
    df.pivot_table(
        index="Name",
        columns="Attribute",
        aggfunc=",".join,
        fill_value="Null",
    ).droplevel(0, axis=1)
)

Prints:

Attribute              A                 B      C
Name                                             
Bob                 Null              Null  Today
Joe                  Yes  smoking,headache   Null
Mary       Null,Tomorrow             Never   Null

---

EDIT: To rename indices:

df = df.pivot_table(
    index="Name",
    columns="Attribute",
    aggfunc=",".join,
    fill_value="Null",
)

df.index.name = ""
df.columns.name = ""