CODEWITHSUNDEEP
pythonpandasdataframeseries
lmac
lmac

Reputation: 127

How to reverse a pandas series

I have a pandas series that must be flipped upside-down before I concatenate it to the main DataFrame.

I can easily flip it with myseries = myseries.iloc[::-1]

But when I attach it to the main DataFrame, it attaches the default series and not the flipped version. Why doesn't the flipped series stay in place?

myseries = myseries.iloc[::-1] 
newdf = pd.concat([newdf, myseries], axis=1)

EDIT: So my guess is that the index is being flipped as well, and when I concatenate its probably using the index to attach the series. Is there a way to flip the values in the series but leave the index untouched? My index starts at 2.

Upvotes: 3

Views: 4143

Answers (3)

Derek O
Derek O

Reputation: 19565

This is occurring because concatenating is based on the index. You can save the original index, and then set the index of the reversed series equal to the original index:

myseries_index = myseries.index
myseries = myseries.iloc[::-1]
myseries.index = myseries_index

Upvotes: 3

Ravindu Hirimuthugoda
Ravindu Hirimuthugoda

Reputation: 109

try this.works for me:)

myseries = myseries.iloc[:,::-1]

example code:

import numpy as np
import pandas as pd

 
 
dataframe = pd.DataFrame([[1, 'A', "Student"],
                          [2, 'B', "Tutor"],
                          [3, 'C', "Instructor"]])
                    
dataframe1 = pd.DataFrame([[1, 'A', "Student"],
                          [2, 'B', "Tutor"],
                          [3, 'C', "Instructor"]])
 
# reversing the dataframe
print("Reversed DataFrame")
dataframe = dataframe.iloc[:,::-1]

dataframe1 = pd.concat([dataframe1,dataframe],axis=1);
print(dataframe1);

Upvotes: 1

Carlos Horn
Carlos Horn

Reputation: 1293

The concatenation looks at the index. So you just need to reverse the index of your series before concatenation. See the following example:

s = pd.Series([1,2,3], name='b')
s.index = s.index[::-1]
df = pd.DataFrame({'a': list('xyz')})
pd.concat([df, s], axis=1)

Upvotes: 1

Related Questions