Power BI is turning 10! Tune in for a special live episode on July 24 with behind-the-scenes stories, product evolution highlights, and a sneak peek at what’s in store for the future.
Save the dateEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
I am trying to do forecasting in Power BI for a very large number of parts. For parts having demand in every month, the forecasting runs perfectly. For parts which does not have demand in some months, the forecast will either show up without considering the months or else the error will pop up saying the data is irregular to forecast. I know I have to add zero values for those parts for the missing months. The question is how to add zero values for such large number of parts so that the forecast shows up and is accurate?
The parts are around 50000 and the demand is monthly demand for 36 months. Also, the data editing needs to be minimal on Excel because this needs to be a real time based dashboard.
Does anyone know how to solve this problem?
Solved! Go to Solution.
Hi @meetrachhadia9,
Did your table contain all month and parts? If this is a case, you can create new tables to extract and store 'parts' and 'month' values.
Then you can create matrix visual with new tables fields as axis and write a measure to lookup value from the original table based on current part and month. (matrix design: month to column, part to row, measure to value)
Calculate tables:
MonthAxis = VALUES('Sample'[Month])
PartAxis = VALUES('Sample'[Part])
Measure:
Measure =
VAR currMonth =
VALUES ( MonthAxis[Month] )
VAR currPart =
VALUES ( PartAxis[Part] )
RETURN
CALCULATE (
SUM ( 'Sample'[Amount] ),
FILTER ( ALLSELECTED ( 'Sample' ), [Month] IN currMonth && [Part] IN currPart )
) + 0
If your source table only contains less amount of records that can't extract full category fields value.
You need to prepare two tables with full category values(part, month), then you can use them with measure formula to generate an expanded matrix visual with correspond result.
BTW, if you only want a new table with full records mapping, you can try to use cross join function to achieve your requirement:
Table =
ADDCOLUMNS (
CROSSJOIN ( VALUES ( 'Sample'[Part] ), VALUES ( 'Sample'[Month] ) ),
"Amount", CALCULATE (
SUM ( 'Sample'[Amount] ),
FILTER (
ALLSELECTED ( 'Sample' ),
[Month] = EARLIER ( 'Sample'[Month] )
&& [Part] = EARLIER ( 'Sample'[Part] )
)
) + 0
)
Regards,
Xiaoxin Sheng
Hi @meetrachhadia9,
Did your table contain all month and parts? If this is a case, you can create new tables to extract and store 'parts' and 'month' values.
Then you can create matrix visual with new tables fields as axis and write a measure to lookup value from the original table based on current part and month. (matrix design: month to column, part to row, measure to value)
Calculate tables:
MonthAxis = VALUES('Sample'[Month])
PartAxis = VALUES('Sample'[Part])
Measure:
Measure =
VAR currMonth =
VALUES ( MonthAxis[Month] )
VAR currPart =
VALUES ( PartAxis[Part] )
RETURN
CALCULATE (
SUM ( 'Sample'[Amount] ),
FILTER ( ALLSELECTED ( 'Sample' ), [Month] IN currMonth && [Part] IN currPart )
) + 0
If your source table only contains less amount of records that can't extract full category fields value.
You need to prepare two tables with full category values(part, month), then you can use them with measure formula to generate an expanded matrix visual with correspond result.
BTW, if you only want a new table with full records mapping, you can try to use cross join function to achieve your requirement:
Table =
ADDCOLUMNS (
CROSSJOIN ( VALUES ( 'Sample'[Part] ), VALUES ( 'Sample'[Month] ) ),
"Amount", CALCULATE (
SUM ( 'Sample'[Amount] ),
FILTER (
ALLSELECTED ( 'Sample' ),
[Month] = EARLIER ( 'Sample'[Month] )
&& [Part] = EARLIER ( 'Sample'[Part] )
)
) + 0
)
Regards,
Xiaoxin Sheng
Tough to say, are you using a measure? Please see this post regarding How to Get Your Question Answered Quickly: https://community.powerbi.com/t5/Community-Blog/How-to-Get-Your-Question-Answered-Quickly/ba-p/38490
User | Count |
---|---|
75 | |
74 | |
44 | |
31 | |
27 |
User | Count |
---|---|
99 | |
89 | |
52 | |
48 | |
46 |