When growing Time Intelligence Measures with Energy BI or in Material in Semantic Fashions, it may be essential to create a date vary to calculate the end result for a particular timeframe.
To be exact, nearly all the Time Intelligence capabilities in Dax create an inventory of dates for a date vary.
However generally we should create a customized date vary on account of particular necessities.
DAX presents us two capabilities for this process:
Each capabilities take a Begin Date as a parameter.
However for the Finish Date, the conduct is totally different.
Whereas DATESINPERIOD() takes Intervals (Days, Months, Quarters, Years), DATESBETWEEN() takes a specified Date used because the Finish Date.
In distinction, DATEADD() makes use of the present Filter Context to get the Begin Date and to calculate the Finish Date.
However we wish to go a Begin Date, which might differ from the Date(s) within the present Filter Context.
That is when one of many capabilities talked about above comes into play.
On the finish of this text, I’ll present you a sensible instance utilizing the strategies proven right here.
Instruments and situation
Like in lots of different articles, I exploit DAX Studio to write down DAX Queries and analyze the outcomes.
If you’re not acquainted with writing DAX queries, learn my piece on easy methods to study to write down such queries:
This time, I exploit the Information mannequin just for the Date desk.
I wish to calculate a date vary ranging from Could 5. 2025 and both 25 days or 2 Months into the long run.
To set the beginning date, I exploit this expression:
DEFINE
VAR StartDate = "2025-05-05"
EVALUATE
{ StartDate }That is the end in DAX Studio:
I outline a Variable and assign the results of the date expression for the next queries.
One other solution to outline the beginning date is to create a date worth utilizing DATE(2025, 05, 05).
The end result would be the identical.
The distinction between these two approaches is that the primary returns a string, however the second returns a correct date.
The DAX capabilities used right here can work with each.
Utilizing DATESINPERIOD()
Let’s begin with DATEINPERIOD().
I’ll use this perform to get a date vary string from the Begin Date and 25 days into the long run:
DEFINE
VAR StartDate = "2025-05-05"
EVALUATE
DATESINPERIOD('Date'[Date]
,StartDate
,25
,DAY)The result’s a desk with 25 rows for the times ranging from Could 05, 2025, to Could 29, 2025:
Now, let’s barely change the question to get an inventory of all dates from the Begin Date to 2 Months into the long run:
DEFINE
VAR StartDate = "2025-05-05"
EVALUATE
DATESINPERIOD('Date'[Date]
,StartDate
,2
,MONTH)The question returns 61 rows ranging from Max 05, 2025, till July 04, 2025:
I can go the interval with an arbitrary variety of days (e.g., 14, 28, 30, or 31 days), and the perform mechanically calculates the date vary.
After I go destructive numbers, the date vary goes to the previous, beginning with the beginning date.
Utilizing DATESBETWEEN()
Now, let’s take a look at DATESBETWEEN().
DATESBETWEEN() takes a Begin- and an Finish-Date as parameters.
This implies I have to calculate the top date earlier than utilizing it.
After I wish to get a date vary from Could 05 to Could 29, 2025, I have to use the next question:
DEFINE
VAR StartDate = "2025-05-05"
VAR EndDate = "2025-05-25"
EVALUATE
DATESBETWEEN('Date'[Date]
,StartDate
,EndDate)The end result is identical as with DATESINPERIOD().
Nonetheless, there’s one essential level: The tip date is included within the end result.
This implies I can write one thing like this to get a date vary over two months from Could 05 to July 05, 2025:
DEFINE
VAR StartDate = "2025-05-05"
VAR EndDate = "2025-07-05"
EVALUATE
DATESBETWEEN('Date'[Date]
,StartDate
,EndDate)The result’s similar to the one utilizing DATESINPERIOD() and month because the interval, however with one row extra:
This offers me extra flexibility to create the date ranges, as I can pre-calculate the top date in line with my wants.
Use in Measures – a sensible instance.
I can use these strategies to calculate a operating complete in a Measure.
However we should take care to make use of the 2 capabilities in the correct manner
For instance, to calculate the operating complete monthly for 25 days.
Take a look at the next code, the place I outline two Measures utilizing the 2 capabilities:
DEFINE
MEASURE 'All Measures'[25DayRollingTotal_A] =
VAR DateRange =
DATESINPERIOD('Date'[Date]
,MIN ( 'Date'[Date] )
,25
,DAY)
RETURN
CALCULATE ( [Sum Online Sales]
, DateRange )
MEASURE 'All Measures'[25DayRollingTotal_B] =
VAR DateRange =
DATESBETWEEN ( 'Date'[Date]
,MIN ( 'Date'[Date] )
,MIN ( 'Date'[Date] ) + 25 )
RETURN
CALCULATE ( [Sum Online Sales]
, DateRange )
EVALUATE
CALCULATETABLE (
SUMMARIZECOLUMNS (
'Date'[Year]
,'Date'[Month]
,"Gross sales", [Sum Online Sales]
,"25DayRollingTotal_A", [25DayRollingTotal_A]
,"25DayRollingTotal_B", [25DayRollingTotal_B]
)
,'Date'[Date] >= DATE(2023, 01, 01) && 'Date'[Date] That is the end result:
Discover the distinction between the 2 outcomes.
It’s because DATESBETWEEN() consists of the top date within the end result, whereas DATESINPERIOD() provides the variety of intervals to the beginning date however consists of the beginning date.
Attempt it out with the next question:
DEFINE
VAR StartDate = DATE(2025,05,05)
VAR EndDate = StartDate + 25
EVALUATE
DATESINPERIOD('Date'[Date]
,StartDate
,25
,DAY)
EVALUATE
DATESBETWEEN('Date'[Date]
,StartDate
,EndDate)The primary returns 25 rows (Could 05 – Could 29, 2025) and the second returns 26 rows (Could 05 – Could 30, 2025).
Due to this fact, I have to change one of many two Measures to get the identical end result.
On this case, the calculation definition is: Begin from the primary date and go 25 into the long run.
The corrected logic is that this:
DEFINE
MEASURE 'All Measures'[25DayRollingTotal_A] =
VAR DateRange =
DATESINPERIOD('Date'[Date]
,MIN ( 'Date'[Date] )
,25
,DAY)
RETURN
CALCULATE ( [Sum Online Sales]
, DateRange )
MEASURE 'All Measures'[25DayRollingTotal_B] =
VAR DateRange =
DATESBETWEEN ( 'Date'[Date]
,MIN ( 'Date'[Date] )
,MIN ( 'Date'[Date] ) + 24 ) // 24 as an alternative of 25 days
RETURN
CALCULATE ( [Sum Online Sales]
, DateRange )
EVALUATE
CALCULATETABLE (
SUMMARIZECOLUMNS (
'Date'[Year]
,'Date'[Month]
,"Gross sales", [Sum Online Sales]
,"25DayRollingTotal_A", [25DayRollingTotal_A]
,"25DayRollingTotal_B", [25DayRollingTotal_B]
)
,'Date'[Date] >= DATE(2023, 01, 01) && 'Date'[Date] Now, each measures return the identical end result:
I examined the efficiency of each capabilities for a similar calculation (Rolling complete over 25 days), and the outcomes had been equal. There was no distinction in efficiency or effectivity between these two.
Even the execution plan is identical.
Because of this DATEINPERIOD() is a shortcut perform for DATESBETWEEN().
Conclusion
From a performance standpoint, each of the proven capabilities are nearly equal.
The identical applies from the efficiency standpoint.
They differ in the way in which the top date is outlined.
DATESINPERIOD() is predicated on calendar intervals, like days, months, quarters, and years.
This perform is used when the date vary have to be calculated primarily based on the calendar.
However when we now have a pre-defined finish date or should calculate the date vary between two pre-defined dates, the DATESBETWEEN() perform is the perform to make use of.
For instance, I exploit DATESBETWEEN() when performing Time Intelligence calculations for weeks.
You’ll be able to learn this piece to study extra about weekly calculations:
As you’ll be able to learn, I retailer the beginning and finish dates of the week for every row within the knowledge desk.
This manner, I can simply search for every date’s begin and finish dates.
So, once we should choose between these two capabilities, it’s not a matter of performance however of necessities outlined by the stakeholders of the brand new studies or the wanted knowledge evaluation.
Learn this text to discover ways to gather and interpret Efficiency knowledge with DAX Studio:
Like in my earlier articles, I exploit the Contoso pattern dataset. You’ll be able to obtain the ContosoRetailDW Dataset totally free from Microsoft here.
The Contoso Information will be freely used underneath the MIT License, as described in this document. I modified the dataset to shift the information to modern dates.
