Create columns based on rows

Question

I am new to pandas and I am looking for a nice way to transform this dataframe:

Date	Name	Value
01-01-2022	A	0
01-01-2022	B	1
01-01-2022	C	1
02-01-2022	A	1
02-01-2022	B	1
02-01-2022	C	0

To this dataframe:

Name	Value_before	Value_after
A	0	1
B	1	1
C	1	0

First table contains only data from two dates.

Asked By: kstajer

||

Source

Answer 1

Assuming:

that you have only 2 dates
that there are no duplicated Name per date

You can use a pivot taking advantage of the fact the pivot sorts the columns, then set_axis to use your custom names

out = (df
   .assign(Date=pd.to_datetime(df['Date']))  # ensure datetime for correct sorting
   .pivot('Name', 'Date', 'Value')
   .set_axis(['Value_before', 'Value_after'], axis=1)
   .reset_index()
)

output:

  Name  Value_before  Value_after
0    A             0            1
1    B             1            1
2    C             1            0

Answered By: mozway

Answer 2

What you are looking for are pivot tables

import pandas as pd
from datetime import date

df = pd.DataFrame({
    "Date": [date(2022,1,1), date(2022,1,1), date(2022,1,1), date(2022,1,2), date(2022,1,2), date(2022,1,2)],
    "Name": ["A", "B", "C", "A", "B", "C"],
    "Value": [0, 1, 1, 1, 1, 0]
})

print(pd.pivot_table(df, values="Value", index="Name", columns="Date"))

Answered By: Matteo Zanoni

Answer 3

def function1(dd:pd.DataFrame):
    dd1=dd.loc[:,'Value']
    dd1.index=['Value_before','Value_after']
    return dd1
df1.groupby("Name").apply(function1)

Value  Value_before  Value_after
Name                            
A                 0            1
B                 1            1
C                 1            0

Answered By: G.G

Create columns based on rows

Question:

Answers: