Pandas - Error when splitting row into multiple columns

Asked Apr 28 '20 at 12:13

Active Apr 28 '20 at 12:13

Viewed 14 times

I have a Dataframe in the below format:

issue_id, products
101, ['Apple','Orange']
102, ['Apple','Banana']
103, ['Grapes']

I am trying to have them split into rows based on ticket_id

issue_id, products
101, Apple
101, Orange
102, Apple
102, Banana
103, Grapes

I tried to explode as below:

df = df.reset_index()
df.set_index(['index', 'issue_id'])
                .stack()
                .str.split(',', expand=True)
                .stack()
                .unstack(-2)
                .reset_index(-1, drop=True)
                .reset_index()
                )

but this throws TypeError: unhashable type: 'Issue'

asked Apr 28 '20 at 12:13

scott martin

1,253
1
14
36

what is `print (df.columns.tolist())` after `df = df.reset_index()` ? – jezrael Apr 28 '20 at 12:17
@jezrael, it prints out `['index', 'issue_id', 'products'] ` – scott martin Apr 28 '20 at 12:18
Ok it is good. What is `print (df['issue_id'].tolist()[:5])` ? – jezrael Apr 28 '20 at 12:20
@jezrael it prints `[101, 102, 103]` – scott martin Apr 28 '20 at 12:21
It is good, what is `print (type(df.loc[0, 'products']))` ? – jezrael Apr 28 '20 at 12:23
@jezrael ,it returns `` – scott martin Apr 28 '20 at 12:34
So use `df = df.explode('products')` insted your solution – jezrael Apr 28 '20 at 12:34
@jezrael, is there an alternative to using explode.. I am using version `0.22 of pandas `and hence was performing this way of extracting data.. – scott martin Apr 28 '20 at 12:36
1

OK, added dupe, there is answers also for lower version of pandas – jezrael Apr 28 '20 at 12:39
1

@jezrael thank you.. – scott martin Apr 28 '20 at 13:19

Pandas - Error when splitting row into multiple columns

0 Answers0