Forum Discussion
Index based on column attributes w/o group By
Hi Community,
i've found a solution for this case with group by but due to the limited RAM resources (dozens milions of rows) i am looking for an alternative - thus reaching for you 🙂
I want to achieve "Index" column which would look like this based on "Invoice" and "database_record_order" but again, without a group by step:
| Invoice | database_record_order | Index |
| A | 99999 | 1 |
| A | 991000 | 2 |
| B | 991001 | 1 |
| B | 991002 | 2 |
| C | 991003 | 1 |
| D | 991004 | 1 |
| A | 991005 | 3 |
| B | 991006 | 3 |
| B | 991007 | 4 |
Would that be even possible?
Many thanks
Likvor
Hi likvors_husband - Could you consider not adding the Index? Perhaps instead you can look at using the DAX RANKX function in measure only when the Index is required by a visualisation. RANKX – DAX Guide
Second, you may consider pushing this transformation from Power Query back into the Data Source. If you are having Memory issues it suggests that Power Qeury needs to load too much data and therefore use to much memory. You need to ask you DBA or Database Developers to create a view.
1 Reply
Hi likvors_husband - Could you consider not adding the Index? Perhaps instead you can look at using the DAX RANKX function in measure only when the Index is required by a visualisation. RANKX – DAX Guide
Second, you may consider pushing this transformation from Power Query back into the Data Source. If you are having Memory issues it suggests that Power Qeury needs to load too much data and therefore use to much memory. You need to ask you DBA or Database Developers to create a view.