CODEWITHSUNDEEP
pythonpandas
goerge
goerge

Reputation: 25

concatenate column values pandas dataframe

Hello I have a dateframe1 of values and i want to transform it into a new dataframe2 by concatenating values of columns in the original dataframe1 i.e

dataframe1
ProductName  Value otherValue
Product1      2     5
Product2      3     2
Product1      1     5
Product3      4     7
Product3      5     7
Product1      5     5
Product2      9     2

dataframe2
ProductName  Value     otherValue
Product1      2 1 5       5
Product2      3 9         2
Product3      4 5         7

Upvotes: 0

Views: 110

Answers (2)

Dani Mesejo
Dani Mesejo

Reputation: 61900

You could groupby ProductName and aggregate using ' '.join on Value and first on otherValue:

result = df.assign().groupby('ProductName', as_index=False).agg({ 'Value' : lambda x : ' '.join(map(str, x)), 'otherValue' : 'first' } )

print(result)

Output

  ProductName  Value  otherValue
0    Product1  2 1 5           5
1    Product2    3 9           2
2    Product3    4 5           7

Note that this solution assumes the column Value is not a string, otherwise you can use ' '.join directly

Upvotes: 2

Celius Stingher
Celius Stingher

Reputation: 18367

You can try with this in two lines. first we need to turn the column Value into strings so we can perform join and operations and the second are all the operations to return the desired output:

import pandas as pd
import numpy as np 
df = pd.DataFrame(data={'ProductName':['Product1','Product2','Product1','Product3','Product3','Product1','Product2'],'Value':[2,3,1,4,5,5,9],'otherValue':[5,2,5,7,7,5,2]})
df['Value'] = df['Value'].astype(str)
df = df.merge(df.groupby('ProductName',as_index=True)['Value'].apply(' '.join).reset_index(),how='left',left_on='ProductName',right_on='ProductName').drop('Value_x',axis=1).drop_duplicates().rename(columns={'Value_y':'Value'})

print(df) Output:

  ProductName  otherValue   Value
0    Product1           5   2 1 5
1    Product2           2     3 9
3    Product3           7     4 5

Upvotes: 0

Related Questions