Repository navigation
Microsoft.Data.Analysis.DataFrame Join should produce a result with a single joined column #6133
Description
Given two DataFame objects it should be easy to perform a simple outer join to achieve these results:
DataFrame1:
| Date | ClosePrice |
|---|---|
| 3/1/2022 12:00:00 AM | 10.5 |
| 3/2/2022 12:00:00 AM | 12.4 |
| 3/3/2022 12:00:00 AM | 11.3 |
DataFrame2:
| Date | ShortPercent |
|---|---|
| 3/1/2022 12:00:00 AM | 2.34 |
| 3/2/2022 12:00:00 AM | 2.36 |
| 3/3/2022 12:00:00 AM | 3.01 |
| 3/4/2022 12:00:00 AM | 3.04 |
Resulting DataFrame:
| Date | ClosePrice | ShortPercent |
|---|---|---|
| 3/1/2022 12:00:00 AM | 10.5 | 2.34 |
| 3/2/2022 12:00:00 AM | 12.4 | 2.36 |
| 3/3/2022 12:00:00 AM | 11.3 | 3.01 |
| 3/4/2022 12:00:00 AM | null | 3.04 |
This does not seem to be possible with the current Merge or Join methods, since they end up with two columns for Date (Date_left and Date_right) in the resulting DataFrame and combining those to a single column seems complex and error prone.
Reacted by Simon Felix, Nils Lande, superfamousdog and Alexey Shokov
Activity
michaelgsharp commentedon Mar 18, 2022
Thanks for pointing this out @olavt. This is the same issue that you and I were talking about right @luisquintanilla?
luisquintanilla commentedon Mar 21, 2022
Yup. That's right. Related to #6128.
Nils12345678901234567 commentedon Sep 5, 2024
I would regard this as a bug, not an enhanment. The Merge and Join methods deos nor work as expected.
superfamousdog commentedon Dec 18, 2024
I am surprised that this bug is still open-- this is basic functionality. Implementing this would go a long way in enabling developers to use DataFrames in C#!
rosebyte commentedon Sep 30, 2026
Triage: it feels a meaningful request, any contribution welcomed.