How .[Date] works

[HugoDataAve2023]

Some time ago I made a formula with date, but my column was type datetime. When I typed, Power BI suggested me to include .[Date], just after the colunm name, like this 'Table'[column].[Date]. When I used this, it works to create the column I needed.

However I recently used the same to create a measure with the first date of a column, like this FIRSTDATE(Table[column].[Date]), and works wrong. The visual chart, showed 2023-01-01 instead of 2023-12-04, that was tha first date of the column.

In my mind, the use of .[Date] could be used to get the date from a datetime column, like an attribute in object-oriented programming, but now I am concerned if I was correct.

I didn't find any explanattion about how it work. 

It is just a conceptual question.

How does .[Date] work? When to use?

If anyone can explan it, I will be gratefull.

PS.: I am not native english speaker. Any doubt just ask.

[Anonymous]

Hi @HugoDataAve2023 ，
In Power BI, the property allows you to extract just the date part of a datetime type. This is particularly useful when you need to perform date-based calculations without the time component. This attribute returns all the dates in the year in which the date falls.
For your case with the function, it's odd that it didn't return the expected result. Here's what you can try to troublershoot this:
You can  format the visual to only display the date part or create a calculated column to extract the date:

Best regards

Albert He

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly
 

[Greg_Deckler · Accepted solution]

@HugoDataAve2023 When you have Auto Time Intelligence turned on in your Options, Power BI creates a hidden Date table in your model that is linked to the Date column. One table is created per Date column. The .[Date] syntax simply references this hidden Date table which provides a hierarchy of different values like Year, Month, Day, etc. 

Using Auto Time Intelligence is generally not recommended as it bloats your semantic model.