Order in descending order

[Omega]

Hi,

I created a table using selectcolumn function:

Week	Country	Impact	RC	KPI 1	KPI 2
1.2017	A	L1	RC2	3912	2532
1.2017	A	L1	RC1	646	3357
1.2017	A	L2	RC3	1684	4316
1.2017	A	L2	RC5	2473	658
1.2017	B	L1	RC2	1108	1157
1.2017	B	L1	RC1	2674	1490
1.2017	B	L2	RC3	4289	4665
1.2017	B	L2	RC5	4239	1306
2.2017	A	L1	RC2	904	1224
2.2017	A	L1	RC1	2589	3002
2.2017	A	L2	RC3	2011	2429
2.2017	A	L2	RC5	3005	214
2.2017	B	L1	RC2	3474	1779
2.2017	B	L1	RC1	4759	6
2.2017	B	L2	RC3	1950	4282
2.2017	B	L2	RC5	908	3278

I'm trying to do the following:

1. Sort in descending order my RC based on KPI 1 without affecting my week, country or impact. In other words, for every week, country and impact, sort my RC based on KPI 1 from highest to lowest (Check below table)
2. Create KPI 4: If (Country (current row) = Country (previous row), KPI 4 value (previous row) - KPI 3 (current row), 1-KPI 3 (Current row)). KPI 3 = KPI 1/KPI 2. Please check the below table. 

I believe it's a mixture of RANKX and Earlier but I'm not an expert in both 😞

Can someone please help? 😄

Thanks!

Week	Country	Impact	RC	KPI 1	KPI 2	KPI 3	KPI 4
1.2017	A	L1	RC2	3912	2532	154.50%	-54.50%
1.2017	A	L1	RC1	646	3357	19.24%	-73.75%
1.2017	A	L2	RC5	2473	658	375.84%	-449.58%
1.2017	A	L2	RC3	1684	4316	39.02%	-488.60%
1.2017	B	L1	RC1	2674	1490	179.46%	-668.06%
1.2017	B	L1	RC2	1108	1157	95.76%	-763.83%
1.2017	B	L2	RC3	4289	4665	91.94%	-855.77%
1.2017	B	L2	RC5	4239	1306	324.58%	-1180.35%
2.2017	A	L1	RC1	2589	3002	86.24%	13.76%
2.2017	A	L1	RC2	904	1224	73.86%	-60.10%
2.2017	A	L2	RC5	3005	214	1404.21%	-1464.30%
2.2017	A	L2	RC3	2011	2429	82.79%	-1547.10%
2.2017	B	L1	RC1	4759	6	79316.67%	-80863.76%
2.2017	B	L1	RC2	3474	1779	195.28%	-81059.04%
2.2017	B	L2	RC3	1950	4282	45.54%	-81104.58%
2.2017	B	L2	RC5	908	3278	27.70%	-81132.28%

[v-shex-msft · Accepted solution]

HI @Omega,

Current power bi not support to sort table by multiple columns or setting sort order priority.

For your situation, you can vote below idea which has the similar requirement:

Multiple column sort

Regards,

Xiaoxin Sheng

[Omega]

Thanks!