Calculated mesaure substraction by Column in MATRIX/or any other mode?

[vjuhasz]

Dears,

Having a data format something like this:

MONTH	SELLER	CUSTOMERS	SALES	CONVERSION RATE %
2021.01.01	LUCY	10	1	10%
2021.01.01	PAUL	5	0	0%
2021.02.01	LUCY	12	1	8%
2021.02.01	PAUL	6	1	17%
2021.03.01	LUCY	15	2	13%
2021.03.01	PAUL	6	1	17%
2021.04.01	LUCY	17	2	12%
2021.04.01	PAUL	7	1	14%

And it is possible to convert to MATRIX, but how could I make any addition or substraction between columns in a matrix to get data like the BOLD one?

	Lucy	Paul
	CUSTOMERS	SALES	CONVERSION RATE %	CUSTOMERS	SALES	CONVERSION RATE %	CONVERSION RATE DIFF
2021.01.01	10	1	10%	5	0	0%	-10%
2021.02.01	12	1	8%	6	1	17%	8%
2021.03.01	15	2	13%	6	1	17%	3%
2021.04.01	17	2	12%	7	1	14%	3%
2021.05.01	19	3	16%	9	1	11%	-5%
2021.06.01	22	3	14%	15	2	13%	0%
2021.07.01	24	4	17%	16	2	13%	-4%
2021.08.01	27	4	15%	17	3	18%	3%
2021.09.01	29	5	17%	18	3	17%	-1%
2021.10.01	32	5	16%	19	3	16%	0%
2021.11.01	35	6	17%	19	4	21%	4%
2021.12.01	37	6	16%	20	4	20%	4%

Thank you in advance!

[lbendlin · Accepted solution]

@vjuhasz  see attached for an implementation. Here is the measure definition:

CR = 
if(HASONEVALUE('Table'[SELLER]),AVERAGE('Table'[CONVERSION RATE %]),
var minName=MINX('Table','Table'[SELLER])
var maxName=MAXX('Table','Table'[SELLER])
var a=SUMMARIZE('Table','Table'[MONTH]
,"mn",CALCULATE(min('Table'[CONVERSION RATE %]),'Table'[SELLER]=minName)
,"mx",CALCULATE(max('Table'[CONVERSION RATE %]),'Table'[SELLER]=maxName))
return AVERAGEX(a,[mn]-[mx]))