Is there a way to group id's with corresponding alternate id in python

Question

Currently, I am working on grouping transactions based on a product ID. The problem is that a product can contain various alternate IDs that can be located in two columns. For example, a product can look like this in the dataset;

requested ID	Item ID
123t-34	4s4532t
:------------	-------:
4s4532t.	123t-34
:------------	-------:
123t-34.	4s4532t

This is an example of a much bigger dataset. Any input on how I could group them to look something similar to this;

Id.	Alternate Ids
123t-34	(4s4532t,123t-34)

This is my attempt to solve the problem. data_dict = data.groupby('requested_id')['Item ID'].agg('unique')

But this only solved half of the grouping based on one column.

score 0 · Answer 1 · answered Sep 26 '22 at 23:20

You can use pandas.DataFrame.itertuples to build a tuple from two or more columns.

Try this :

(
    data.assign(AlternateIDs= list(df[df.columns[::-1]].itertuples(index=False, name=None)))
        .rename(columns={'requested ID': 'Id.'})
        .drop(columns='Item ID')
)

# Output :

        Id.         AlternateIDs
0   123t-34   (4s4532t, 123t-34)
1  4s4532t.  (123t-34, 4s4532t.)
2  123t-34.  (4s4532t, 123t-34.)

Is there a way to group id's with corresponding alternate id in python

1 Answers1

# Output :