Starting December 3, join live sessions with database experts and the Microsoft product team to learn just how easy it is to get started
Learn moreGet certified in Microsoft Fabric—for free! For a limited time, get a free DP-600 exam voucher to use by the end of 2024. Register now
Hi,
We have a table and we hope to be able to calcuate row. variance.
Example
Dec 20 Variance between 15/12 (480) and 01/12 (432) is (480-432)/432 = 11%
Date | Ref | Rev | Dec-20 | Jan-21 | Feb-21 | Mar-21 | Apr-21 | May-21 | Jun-21 | Var Dec | Var Jan | Var Feb | Var Mar | Var Apr | Var May | Var Jun | |
01/12/2020 | 65 | 0 | 432 | 816 | 1272 | 1872 | 2328 | 2788 | 3006 | ||||||||
15/12/2020 | 73 | 0 | 480 | 816 | 1320 | 2040 | 2544 | 2982 | 3220 | 11% | 4% | 9% | 9% | 7% | 7% | ||
18/12/2020 | 76 | 0 | 816 | 1320 | 2040 | 2544 | 2982 | 3220 | |||||||||
02/01/2021 | 77 | 0 | 816 | 1320 | 2040 | 2544 | 2982 | 3220 | |||||||||
12/01/2021 | 85 | 0 | 336 | 912 | 1680 | 2256 | 2802 | 3096 | -59% | -31% | -18% | -11% | -6% | -4% | |||
26/01/2021 | 89 | 0 | 336 | 936 | 1752 | 2304 | 2802 | 3142 | 3% | 4% | 2% | 1% | |||||
01/02/2021 | 90 | 0 | 936 | 1752 | 2304 | 2828 | 3142 | -100% | 1% | ||||||||
15/02/2021 | 99 | 0 | 1032 | 1824 | 2340 | 2834 | 3123 | 10% | 4% | 2% | 0% | -1% | |||||
26/02/2021 | 105 | 0 | 1896 | 2496 | 2956 | 3264 | 4% | 7% | 4% | 5% | |||||||
01/03/2021 | 106 | 0 | 1944 | 2592 | 3052 | 3360 | 3% | 4% | 3% | 3% | |||||||
15/03/2021 | 114 | 0 | 1800 | 2452 | 3032 | 3390 | -7% | -5% | -1% | 1% | |||||||
30/03/2021 | 122 | 0 | 1992 | 2544 | 3044 | 3510 | 11% | 4% | 0% | 4% |
Thanks everyone can help with it
Hi @v-jingzhang
It seems my post need further details.
At the moment I have a table of customer cumulations grouped by Cust/Seq/Contract/Part/Date/Release/YYYYMM
Cust | Seq | Contract | Part | Date | Release | Rev | YYYYMM | CUMULATION |
CUST1 | 1 | 0035987 | PART1 | 01/12/2020 | 65 | 0 | 202012 | 432 |
CUST1 | 1 | 0035987 | PART1 | 02/12/2020 | 66 | 0 | 202012 | 432 |
CUST1 | 1 | 0035987 | PART1 | 03/12/2020 | 67 | 0 | 202012 | 424 |
CUST1 | 1 | 0035987 | PART1 | 07/12/2020 | 68 | 0 | 202012 | 424 |
CUST1 | 1 | 0035987 | PART1 | 08/12/2020 | 69 | 0 | 202012 | 432 |
CUST1 | 1 | 0035987 | PART1 | 09/12/2020 | 70 | 0 | 202012 | 480 |
CUST1 | 1 | 0035987 | PART1 | 10/12/2020 | 71 | 0 | 202012 | 480 |
CUST1 | 1 | 0035987 | PART1 | 15/12/2020 | 72 | 0 | 202012 | 480 |
CUST1 | 1 | 0035987 | PART1 | 16/12/2020 | 73 | 0 | 202012 | 480 |
CUST1 | 1 | 0035987 | PART1 | 01/12/2020 | 65 | 0 | 202101 | 816 |
CUST1 | 1 | 0035987 | PART1 | 02/12/2020 | 66 | 0 | 202101 | 816 |
CUST1 | 1 | 0035987 | PART1 | 03/12/2020 | 67 | 0 | 202101 | 808 |
CUST1 | 1 | 0035987 | PART1 | 07/12/2020 | 68 | 0 | 202101 | 808 |
CUST1 | 1 | 0035987 | PART1 | 08/12/2020 | 69 | 0 | 202101 | 816 |
CUST1 | 1 | 0035987 | PART1 | 09/12/2020 | 70 | 0 | 202101 | 888 |
CUST1 | 1 | 0035987 | PART1 | 10/12/2020 | 71 | 0 | 202101 | 888 |
CUST1 | 1 | 0035987 | PART1 | 15/12/2020 | 72 | 0 | 202101 | 888 |
CUST1 | 1 | 0035987 | PART1 | 16/12/2020 | 73 | 0 | 202101 | 816 |
CUST1 | 1 | 0035987 | PART1 | 17/12/2020 | 74 | 0 | 202101 | 816 |
CUST1 | 1 | 0035987 | PART1 | 17/12/2020 | 75 | 0 | 202101 | 816 |
On PBI the matrix is looking like this
I would like to be able to calculate the variance in PBI for the same group Cust/Seq/Contract/Part/Date/Release/YYYYMM day by day.
Example
Cust | Seq | Contract | Part | Date | Release | Rev | YYYYMM | CUMULATION | Variance |
CUST1 | 1 | 0035987 | PART1 | 01/12/2020 | 65 | 0 | 202012 | 432 | |
CUST1 | 1 | 0035987 | PART1 | 02/12/2020 | 66 | 0 | 202012 | 432 | 0% |
CUST1 | 1 | 0035987 | PART1 | 03/12/2020 | 67 | 0 | 202012 | 424 | -2% |
CUST1 | 1 | 0035987 | PART1 | 07/12/2020 | 68 | 0 | 202012 | 424 | 0% |
CUST1 | 1 | 0035987 | PART1 | 08/12/2020 | 69 | 0 | 202012 | 432 | 2% |
CUST1 | 1 | 0035987 | PART1 | 09/12/2020 | 70 | 0 | 202012 | 480 | 11% |
CUST1 | 1 | 0035987 | PART1 | 10/12/2020 | 71 | 0 | 202012 | 480 | 0% |
CUST1 | 1 | 0035987 | PART1 | 15/12/2020 | 72 | 0 | 202012 | 480 | 0% |
CUST1 | 1 | 0035987 | PART1 | 16/12/2020 | 73 | 0 | 202012 | 480 | 0% |
CUST1 | 1 | 0035987 | PART1 | 01/12/2020 | 65 | 0 | 202101 | 816 | 70% |
CUST1 | 1 | 0035987 | PART1 | 02/12/2020 | 66 | 0 | 202101 | 816 | 0% |
CUST1 | 1 | 0035987 | PART1 | 03/12/2020 | 67 | 0 | 202101 | 808 | -1% |
CUST1 | 1 | 0035987 | PART1 | 07/12/2020 | 68 | 0 | 202101 | 808 | 0% |
CUST1 | 1 | 0035987 | PART1 | 08/12/2020 | 69 | 0 | 202101 | 816 | 1% |
CUST1 | 1 | 0035987 | PART1 | 09/12/2020 | 70 | 0 | 202101 | 888 | 9% |
CUST1 | 1 | 0035987 | PART1 | 10/12/2020 | 71 | 0 | 202101 | 888 | 0% |
CUST1 | 1 | 0035987 | PART1 | 15/12/2020 | 72 | 0 | 202101 | 888 | 0% |
CUST1 | 1 | 0035987 | PART1 | 16/12/2020 | 73 | 0 | 202101 | 816 | -8% |
CUST1 | 1 | 0035987 | PART1 | 17/12/2020 | 74 | 0 | 202101 | 816 | 0% |
CUST1 | 1 | 0035987 | PART1 | 17/12/2020 | 75 | 0 | 202101 | 816 | 0% |
Your help will be really appreciate.
Thanks
Joe
Does your current data table have the format as you have shown? If so, you can add new columns to the table with below DAX. You just need to modify the column name and the value column it references.
Var Dec =
var _date = MAXX(FILTER('Table','Table'[Date]<EARLIER('Table'[Date])),'Table'[Date])
var _value = MAXX(FILTER('Table','Table'[Date]=_date),'Table'[Dec-20])
return
DIVIDE('Table'[Dec-20]-_value,_value)
Best Regards,
Community Support Team _ Jing
If this post helps, please Accept it as Solution to help other members find it. Highly appreciate your Kudos!
Starting December 3, join live sessions with database experts and the Fabric product team to learn just how easy it is to get started.
March 31 - April 2, 2025, in Las Vegas, Nevada. Use code MSCUST for a $150 discount! Early Bird pricing ends December 9th.
User | Count |
---|---|
89 | |
86 | |
82 | |
64 | |
49 |
User | Count |
---|---|
125 | |
111 | |
88 | |
69 | |
66 |