Forum Discussion

Anonymous

Not applicable

4 years ago

Solved

Wierd behaviour with Powerbi and datediff

Hi,

I have a report with the following measure

DaysPast = DATEDIFF (INVOICE[INVOICE_DATE], TODAY (),DAY

This works fine.. however in my new report.. based on the same data it does not recognise any table columns... how is this possible?

Thanks

Chris

Greg_Deckler
4 years ago
Anonymous When using measures, references to columns have to include an aggregation function like SUM, MIN, MAX, etc. This is a big difference from calculated columns. Calculated columns have row context and so do not need aggregation functions to reference other columns because the default is to get the value in the column for the same row.

4 Replies

Anonymous
Not applicable
4 years ago
Greg_Deckler

Both are imports - one is from a odbc connection to a database.. the other is from an excel table.

In the end I made it a calculated column.. strange behaviour though!
- Greg_Deckler
  Community Champion
  4 years ago
  Anonymous When using measures, references to columns have to include an aggregation function like SUM, MIN, MAX, etc. This is a big difference from calculated columns. Calculated columns have row context and so do not need aggregation functions to reference other columns because the default is to get the value in the column for the same row.
  - Anonymous
    Not applicable
    4 years ago
    My error - was initially a calculated column.. been a long day
Greg_Deckler
Community Champion
4 years ago
Anonymous So is this via a Live connection to the dataset or Direct query for Power BI datasets or are they both import models? Also, if this is really a measure and not a calculated column, you should need an aggregation function around the column reference (INVOICE[INVOICE_DATE])