CODEWITHSUNDEEP
pythonpandas
WindCheck
WindCheck

Reputation: 426

Dataframe row items to column

I have the following DataFrame:

       type             result
0      IP_1         ['a', 'b']
1      IP_2         ['c', 'd']

I want it reshaped to look like this:

         IP_1           IP_2         
0  ['a', 'b']     ['c', 'd']

but when I try df.pivot(columns='type', values='result') I got:

             IP_1              IP_2
0             Nan        ['a', 'b']       
1      ['c', 'd']               NaN

How can I solve that?

Upvotes: 0

Views: 89

Answers (5)

Umar.H
Umar.H

Reputation: 23099

If i understand your correctly you need to create create a new index then unstack based the position of each unique type value.

df.set_index([df.groupby("type").cumcount(), df["type"]])["result"].unstack()

type        IP_1        IP_2
0     ['a', 'b']  ['c', 'd']

Upvotes: 1

Mehdi Golzadeh
Mehdi Golzadeh

Reputation: 2583

You should have an the same index for both:

df = pd.DataFrame([1,["IP_1","['a', 'b']"],[2,"IP_2","['c', 'd']"]],columns=['ind','type','result'])
print(df.pivot(columns=['type'],index='ind'))

result            
type        IP_1        IP_2
ind                         
1     ['a', 'b']  ['c', 'd']

Upvotes: 1

Sreeram TP
Sreeram TP

Reputation: 11917

Try transpose after changing index to type,

df.set_index("type").transpose().reset_index(drop=True)


type    IP_1    IP_2
0      [a, b]   [c, d]

Upvotes: 1

AlexisG
AlexisG

Reputation: 2484

I've done it with a transpose

df = pd.DataFrame([
    {"type": "IP_1", "result": ['a', 'b']},
    {"type": "IP_2", "result": ['c', 'd']},
])
df = df.set_index('type').T

Output:

type      IP_1    IP_2
result  [a, b]  [c, d]

Upvotes: 3

MaPy
MaPy

Reputation: 505

Hi you can use unstack as below:

df.unstack().unstack(level=0)

Upvotes: 1

Related Questions