Rolling 26 week with week # in a date table

[OCBB_SFAFPandA]

Hello, 

I have a week number in my date table, which is takes into account leap year. If I want to do a rolling 26 week sales calculation, how would I do that? Most of the articles I see use the DATEADD function, which doesn't include weeks.

My weeks always start on sundays, so I can't just have the last 7 days from a Thursday or so. Any suggestions? 

[amitchandak · Accepted solution]

@OCBB_SFAFPandA , In you date table have these columns

new columns
Week Start date = 'Date'[Date]+-1*WEEKDAY('Date'[Date],2)+1
Week End date = 'Date'[Date]+ 7-1*WEEKDAY('Date'[Date],2)
Week Rank = RANKX(all('Date'),'Date'[Week Start date],,ASC,Dense)
OR
Week Rank = RANKX(all('Date'),'Date'[Year Week],,ASC,Dense) //YYYYWW format

then you can have measures
This Week = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]=max('Date'[Week Rank])))
Last Week = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]=max('Date'[Week Rank])-1))

Last  26 weeks = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]>=max('Date'[Week Rank])-26 && 'Date'[Week Rank]<=max('Date'[Week Rank])))

or

Last 26 weeks =
var _max1 = maxx(allselected('Table'), 'Table'[date])
var _max = maxx(filter(all(date), date[date] = _max1), week[Rank])
return
CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]>=_max-26 && 'Date'[Week Rank]<=_max))

[v-jayw-msft · Accepted solution]

Hi @OCBB_SFAFPandA ,

Create calculated columns like below:

year_week = YEAR('Table'[date])*100+WEEKNUM('Table'[date],1)

week_rank = RANKX('Table',[year_week],,ASC,Dense)

Then create the rolling 26 week measure like below:

rolling_26_week =
SUMX (
    FILTER (
        ALLSELECTED ( 'Table' ),
        'Table'[week_rank] <= SELECTEDVALUE ( 'Table'[week_rank] )
            && 'Table'[week_rank]
                > SELECTEDVALUE ( 'Table'[week_rank] ) - 26
    ),
    'Table'[value]
)

Best Regards,

Jay

[amitchandak · Accepted solution]

@OCBB_SFAFPandA , In you date table have these columns

new columns
Week Start date = 'Date'[Date]+-1*WEEKDAY('Date'[Date],2)+1
Week End date = 'Date'[Date]+ 7-1*WEEKDAY('Date'[Date],2)
Week Rank = RANKX(all('Date'),'Date'[Week Start date],,ASC,Dense)
OR
Week Rank = RANKX(all('Date'),'Date'[Year Week],,ASC,Dense) //YYYYWW format

then you can have measures
This Week = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]=max('Date'[Week Rank])))
Last Week = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]=max('Date'[Week Rank])-1))

Last  26 weeks = CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]>=max('Date'[Week Rank])-26 && 'Date'[Week Rank]<=max('Date'[Week Rank])))

or

Last 26 weeks =
var _max1 = maxx(allselected('Table'), 'Table'[date])
var _max = maxx(filter(all(date), date[date] = _max1), week[Rank])
return
CALCULATE(sum('Table'[Qty]), FILTER(ALL('Date'),'Date'[Week Rank]>=_max-26 && 'Date'[Week Rank]<=_max))

[OCBB_SFAFPandA]

Thanks Amitchandak!

The rank formula is something i needed, since I don't have access to the data for PQ. 

Is the solution provided able to get the running total?

For example

26weekago - 100

25weekago - 350 ( previous + current week)

24weekago - 1000 (previous + current)

.....

last week - x

I am looking to get this time of data on a table