Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
CJDK
Helper II
Helper II

Cannot get quarterly difference matrix to work

I am trying to build a matrix that compares the differences from one quarter to another - actual differences or percentage differences.

 

I have tried copying several examples without success. The examples use numeric values (e.g. Sales in USD), but the values I am using are not numeric, but strings (e.g. "Sales visit" or "Contract signed"). Therefore I cannot use SUM and have been trying with COUNT instead:

SupportTypeCount = COUNT(MyTable[Support type])

 

In the matrix the rows are the "Support types" and the columns are Quarters:

Quarter = FORMAT([Date], "\QQ yyyy")

 

So far, so good. But I cannot find the correct measure for the values - I have tried numerous combinations, also using a Date Calendar, but all have failed:

QuarterlySupportChange = DIVIDE(MyTable[SupportTypeCount], CALCULATE([SupportTypeCount], DATEADD('Date'[Date], -1,QUARTER)),0)-1
Supports LQ = CALCULATE([SupportTypeCount],DATEADD(MyTable[Date],-1,QUARTER))
Supports LQ = CALCULATE(COUNT(MyTable[Support type]),DATEADD('Date'[Date]-1,QUARTER))
Supports LQ = CALCULATE(COUNT(MyTable[Support type]),PREVIOUSQUARTER(MyTable[Date]))
Supports LQ = CALCULATE(COUNT(MyTable[Support type]),DATEADD(MyTable[Date],-1,QUARTER))

 

Any help pointing out my error will be very gratefully received!

 

CJ

1 ACCEPTED SOLUTION
amitchandak
Super User
Super User

@CJDK , with help from date table marked as date table , these should work

Supports LQ = CALCULATE(COUNT(MyTable[Support type]),DATEADD('Date'[Date]-1,QUARTER))
Supports LQ = CALCULATE(COUNT(MyTable[Support type]),datesqtd(DATEADD('Date'[Date]-1,QUARTER)))

 

Make sure the date table is marked as a date table, there is no timestamp in the fact/table table

 

refer

Why Time Intelligence Fails - Power bi 5 Savior Steps for TI: https://youtu.be/OBf0rjpp5Hw

View solution in original post

2 REPLIES 2
CJDK
Helper II
Helper II

Many thanks Amit - again! 🙏

Your video was very instructuve and I have been able to use that input and your corrections to springboard to my final measure:

QuarterlySupportChange = DIVIDE(COUNT(MyTable[Support type]), CALCULATE COUNT(MyTable[Support type]), DATEADD('Date'[Date], -1,QUARTER)),0)-1

 

Please note a missing comma in your measures:

'Date'[Date],

 

MANY thanks!

 

CJ

amitchandak
Super User
Super User

@CJDK , with help from date table marked as date table , these should work

Supports LQ = CALCULATE(COUNT(MyTable[Support type]),DATEADD('Date'[Date]-1,QUARTER))
Supports LQ = CALCULATE(COUNT(MyTable[Support type]),datesqtd(DATEADD('Date'[Date]-1,QUARTER)))

 

Make sure the date table is marked as a date table, there is no timestamp in the fact/table table

 

refer

Why Time Intelligence Fails - Power bi 5 Savior Steps for TI: https://youtu.be/OBf0rjpp5Hw

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.