add previous row values

[Anonymous]

Hi,

I have a fact table of sales and I need to get 2 new columns, wich will display Previous date of purchase and Previous sales status (both red in the table below).

I'm trying to get same results as with using lag/window function in SQL ( lag(date_of_purchase) over (partition by month_of_sale, customer, product order by date_of_purchase). 

Also, the problem is, I no longer can edit Query (the table in database no longer exists)..

month of sale	customer	product	sales status	date of purchase	previous date of purchase	previous sales status
Dec-16	john	A	booked	01-12-16	null	null
Dec-16	john	A	booked	20-12-16	01-12-16	booked
Dec-16	alicia	A	paid	01-12-16	null	null
Jan-17	john	Ck111	booked	01-01-17	null	null
Jan-17	john	Ck111	paid	13-01-17	01-01-17	booked
Jan-17	john	A	booked	01-01-17	null	null
Jan-17	john	A	booked	03-01-17	01-01-17	booked
Jan-17	alicia	Ck111	booked	01-01-17	null	null
Jan-17	alicia	Ck111	paid	13-01-17	01-01-17	booked
Jan-17	alicia	Ck111	paid	20-01-17	13-01-17	paid

I've been trying to use calculate function with filters, but so far no useful results. I'm new at DAX, so probably I'm not using them right way..

Thanks in advance for any help!

Zuzana

[v-qiuyu-msft · Accepted solution]

Hi @Anonymous,

You can create calculated columns follow below DAX:

previous date of purchase = CALCULATE(MAX('Table1'[date of purchase]),FILTER(Table1,'Table1'[month of sale]=EARLIER(Table1[month of sale]) && 'Table1'[customer]=EARLIER(Table1[customer])&&'Table1'[product]=EARLIER(Table1[product])&&'Table1'[date of purchase]<EARLIER('Table1'[date of purchase])))

previous sales status = LOOKUPVALUE( Table1[sales status], Table1[month of sale], Table1[month of sale], Table1[customer], Table1[customer], Table1[product], Table1[product], Table1[date of purchase], Table1[previous date of purchase] )

Best Regards,
Qiuyun Yu

[Anonymous]

https://community.powerbi.com/t5/Desktop/reference-Previous-row-using-measures/m-p/725894#M350298

help me out on the same issue.

[Noelia]

Is there a way to do this with a measure instead of a column? My table has too many rows and there is not enough memory to solve it with a calculated column.