CODEWITHSUNDEEP
python-3.xpandas
Naga
Naga

Reputation: 452

Python Pandas groupby data on a "list" column

Hi I have a list column in my data-frame and would like group by values in the list. Below is the piece of code and expected output

import pandas as pd
df=pd.DataFrame([ { "id" : "1", "fix": ["fix1", "fix2"] }, { "id" : "2", "fix": ["fix2", "fix3"] } ])
res=df.groupby(["fix"]).count() /* this does not work */

Output 

fix1 - 1
fix2- 2  
fix3 -1

Upvotes: 0

Views: 18

Answers (1)

jezrael
jezrael

Reputation: 862661

Use DataFrame.explode first:

res=df.explode('fix').groupby(["fix"]).count() 
print (res)
fix     
fix1   1
fix2   2
fix3   1

Similar solution with Series.explode and Series.value_counts:

res=df['fix'].explode().value_counts()
print (res)
fix2    2
fix3    1
fix1    1
Name: fix, dtype: int64

Upvotes: 2

Related Questions