topic Re: Aggregate a calculated count for each day over a period in DAX Commands and Tips

Aggregate a calculated count for each day over a period

Vlad_M_ — Wed, 21 Sep 2022 12:37:50 GMT

I'm working with a table a bit like this:

User Logins

ID	Timestamp
1	21/09/2022 00:00:00
4	20/09/2022 00:00:00
1	20/09/2022 00:00:00
...	...

I have chosen to categorize a group of users as "Active Users", as the ones which has been logged in the past 14 days.

So, I can retrieve all my "Active Users" by using this Measure:

CALCULATE(DISTINCTCOUNT('User Logins'[ID]), 'User Logins'[Timestamp] > TODAY()-14)

But - here's my problem. I want to create a graph of "Active Users" for the past 7 days.

So I need to somehow replace "TODAY()" with "Whatever X-day is showed on the x-axis on the graph".

Just to make it more clear, maybe we have the following results of the count:

Count of unique ID's	Date
10	21/09/2022
15	20/09/2022
5	19/09/2022
7	18/...
4	17/...
7	16/...

Then I would like to get the following result (just displayed in a graph instead):

Count of unique ID's the past 14 days	Date
82 (10+15+5+7+4+7+...)	21/09/2022
89 (15+5+7+4+7+...)	20/09/2022
75 (5+7+4+7+...)	19/09/2022
78 (7+4+7+...)	18/09/2022

I hope it makes sense and that someone knows the necessary formula! 🙂

/Vlad

Re: Aggregate a calculated count for each day over a period

Greg_Deckler — Wed, 21 Sep 2022 13:11:30 GMT

Vlad_M_ Maybe:

Measure = VAR __Date = MAX('Table'[Date]) //whatever is used in the x-axis VAR __MinDate = __Date - 14 RETURN SUMX(FILTER(ALL('Table'),[Date] >= __MinDate && [Date] <= __Date),[Count of Unique ID's])

Re: Aggregate a calculated count for each day over a period

Vlad_M_ — Thu, 29 Sep 2022 05:17:34 GMT

Hey Greg

For some reason I couldn't get it to work the way I hoped, but thanks for the input 🙂

Re: Aggregate a calculated count for each day over a period

Vlad_M_ — Tue, 04 Oct 2022 10:20:48 GMT

For others - I've found my own solution for this.

I created a new table with the following settings:

Active Users Past 7 Days = FILTER(Dates, Dates[Date] > Today() - 7 && Dates[Date] <= Today())

Where the table "Dates" is a calendar table. So it just gives me a column with the past 7 days.

Then I added the following column to the table:

Active Users =
VAR Selected = 'Active Users Past 7 Days'[Date]
RETURN
CALCULATE(DISTINCTCOUNT('User Logins'[UserId]), 'User Logins'[timestamp] > Selected-14)

This results in a table with two columns. One being the past 7 days, another being the sum og distinct user id's for the past 14 days per day.