Calendars (Enhanced Time Intelligence)
The September 2025 release of Power BI Desktop introduced a new Public Preview feature called "Enhanced Time Intelligence". This feature lets you define custom calendars in your semantic model, and it also introduces 8 new DAX functions that work with these calendars, enabling week-based time intelligence calculations that were difficult to perform previously.
Tabular Editor 3 has support for Calendars and the new DAX functions since version 3.23.0.
Defining a Calendar
- Right-click on a table in your model (typically a Date table) and select Create > Calendar....
- Give your calendar a name, e.g. "Fiscal"
Once calendars are added to a table, they will be shown in the TOM Explorer under the Calendars node:
Before you can use a calendar in your DAX calculations, you need to configure it by specifying which columns in the table represent the different calendar attributes. You can do this by right-clicking on the calendar in the TOM Explorer, then choosing the Edit Column Mappings... option:
For each calendar, you can add one or more so-called Column Associations. Each such association maps a column from the table, to a specific Time Unit (e.g. Year, Month, Week, etc.). You can also add additional associated columns for each mapping, which are typically used for columns that represent the same time unit, but in a different format. For example, you might have a "Month" column that contains the month number (1-12), and a "Month Name" column that contains the month name ("January", "February", etc.). Both of these columns can be associated with the "MonthOfYear" time unit.
Using Calendars in DAX
Once you've defined a calendar and mapped its columns, you can start using it in your DAX calculations. Calendars work with all DAX functions that accept a date column as input (such as TOTALYTD, CLOSINGBALANCEMONTH and DATEADD).
Moreover, 8 new DAX functions for week-based time intelligence have been introduced. These exclusively work with calendars:
Click the links above to learn more about each function.