How to make a calculated column that calculates the difference between report dates

[Daryn-TP]

So I have data similar to the following example. I need to create a matrix like the second table example. My issue is I can't get a calculated column that calculates the % change between report dates. The report dates are always a week apart so I guess it could be a week over week delta. 

Location	Memory Size (GB)	Memory Utilization (%)	Date
DAL	1535.530762	44.99733327	9/8/2021
DAL	383.7954712	44.08708318	9/8/2021
HWT	95.97458776	55.96500312	9/8/2021
HWT	1535.821615	69.09389072	9/8/2021
HWT	143.9746399	77.55666606	9/8/2021
AND	14074.43348	80.00602666	9/8/2021
BAL	1103.795471	27.95833333	9/8/2021
BAL	383.7955627	58.70583407	9/8/2021
BAL	319.689504	80.41444397	9/8/2021
BAL	639.4817047	70.46812439	9/8/2021
AND	767.2442424	41.18000031	9/8/2021
AND	767.614841	43.10745352	9/8/2021
LEN	2552.823395	56.97683462	9/8/2021
DAL	1535.530762	45.33699951	9/15/2021
DAL	383.7954712	44.34500122	9/15/2021
HWT	95.97458649	55.76499685	9/15/2021
HWT	1535.821655	68.08699951	9/15/2021
HWT	143.9746399	85.84999847	9/15/2021
AND	14074.43356	80.06563366	9/15/2021
BAL	1103.795471	27.37833341	9/15/2021
BAL	383.7955627	59.24625111	9/15/2021
BAL	319.6894989	83.04999924	9/15/2021
BAL	639.48172	64.11999989	9/15/2021
AND	767.6149089	40.74500275	9/15/2021
LEN	2552.823364	56.56888962	9/15/2021
HWT	95.97458649	54.58000056	9/22/2021
HWT	95.97458649	37.13500214	9/29/2021
HWT	95.97458649	56.07999929	10/6/2021
HWT	1535.821655	70.06166712	9/22/2021
HWT	1535.821655	68.8599987	9/29/2021
HWT	1535.821655	68.34500122	10/6/2021
DAL	1535.530772	45.24799957	9/22/2021
DAL	1535.530762	50.11000061	9/29/2021
DAL	1535.530762	44.42999954	10/6/2021
AND	767.2442322	40.05750084	9/22/2021
AND	767.2442322	35.91999817	9/29/2021
AND	767.2442322	42.72999954	10/6/2021
BAL	383.7955627	59.64333344	9/22/2021
BAL	319.6895091	82.80499776	9/22/2021
BAL	639.4817149	44.11624988	9/22/2021
LEN	2552.823364	56.64800059	10/6/2021

Location	Memory Utilization %	Delta
9/8/2021	9/15/20221
Dal	45	60	25%
Len	72	70	-3%
HWT	22	28	21%
BAL	43	58	26%

[amitchandak · Accepted solution]

@Daryn-TP , Try a measure like

Change % =
var _date = maxx(filter(allselected(Table), [Location] = max(Table[Location]) && Table[Date] < Max(Table[Date])), [Date]) //Use all('Table') if one date is selected
return
Sum(Table[Memory Size]) - calculate(Sum(Table[Memory Size]),filter(allselected(Table) , [Location] = max(Table[Location]) && Table[Date] = _date))

use all in place allselected if needed

[Daryn-TP]

So thankyou for the response @amitchandak 

I tried the suggested DAX and I get an error message, "The function SUM cannot work with values of type String"

The dax I used is below. So in my actual datset the table is called "Virt-Data", DC=location, Meeting Date = Date. I substituted these terms into your suggested DAX. 

Change % =

var _date = maxx(filter(allselected('Virt-Data'), [DC] = max('Virt-Data'[DC]) && 'Virt-Data'[Meeting Date] < Max('Virt-Data'[Meeting Date])), 'Virt-Data'[Meeting Date]) //Use all('Table') if one date is selected

return

Sum('Virt-Data'[Memory Utilization (%)]) - calculate(Sum('Virt-Data'[Memory Utilization (%)]),filter(allselected('Virt-Data') , 'Virt-Data'[DC] = max('Virt-Data'[DC]) && 'Virt-Data'[Meeting Date] = _date))

Not sure if I messed up the syntax when substituting terms though. 

[Daryn-TP]

Nevermind I figured it out I had one of my parameters set as an alphanumeric rather than a number. Thank you!