Earn a 50% discount on the DP-600 certification exam by completing the Fabric 30 Days to Learn It challenge.
Hello,
I have my Excel sheet formatted in the below way. I need to calculate the Month over Month change in Flash.
Index | Region | Location | Transfer Org | FY | Submission | Category | Q1 | Q2 | Q3 | Q4 |
606 | Geo | India | Compute Tower | FY19 | Jun Flash | Cost | 0.01 | 0.01 | 0.01 | 0.01 |
607 | Geo | India | Compute Tower | FY19 | Jul Flash | Cost | 0.00 | 0.01 | 0.01 | 0.01 |
608 | Geo | India | HIT | FY19 | Jun Flash | Cost | 0.07 | 0.09 | 0.14 | 0.14 |
609 | Geo | India | HIT | FY19 | Jul Flash | Cost | 0.19 | 0.23 | 0.22 | 0.22 |
610 | Geo | India | SDC | FY19 | Jun Flash | Cost | 0.01 | 0.01 | 0.01 | 0.01 |
611 | Geo | India | SDC | FY19 | Jul Flash | Cost | 0.01 | 0.01 | 0.01 | 0.01 |
First, I unpivoted my data in Power BI to have just one column for 'Quarters'. Then I split FY and Submission column to fetch 'Year' and 'Month'.
I created a separate 'Date Table' and adjusted Date as per my Org's Fiscal Year (FY starts in November) using the command -
Date = IF( 'Month Index'[FY Month No] < 3, 'Month Index'[FY Month No] + 10, 'Month Index'[FY Month No] -2 ) & "/1/" & 'Month Index'[Year])
However, when I use the Quick Measure 'MoM change', it doesn't calculate the actual difference and results following:
If somebody could help, I would deeply appreciate it. I have been working on this the entire week.
You have to DENORMALIZE your fact table COMPLETELY and correctly connect it to a Date table. It won't work in any other way if you want to use Quick Measures. So, no columns like Q1, Q2, Q3... You have to assign the rows in your fact table concrete dates, like, for instance, the first/last day of a quarter. And then, for this whole thing to work you have to hide the columns in your Date table that have a granularity higher than quarter.
Best
Darek
Thank you Darek. I already unpivoted Q1,Q2,Q3,Q4 columns into a single Quarters column.
I didn't understand hiding columns. How will that make a difference?
Well, your fact table's granularity is the Quarter. If you start slicing it by months/dates, then you'll get numbers that will be incorrect or rather correct but not easy to interpret. Hiding the columns with higher granularities prevents people from using them and thus from seeing bullsh*t. But, of course, you're free to leave them visible. However, you then have to make sure yourself that your measures are aware of which granularity has been requested and take action accordingly.
Best
Darek
Can you upload some sample data? Seems like a relatively quick solution, but would like to see some of the data if possible.