Cumulative sum by dates (Even the missing dates)

[shubh25]

Hi,

I am using measure to calculate profit across the dates.

This is the data that I am using. Date, Revenue on that date and Cumulative revenue.
Cumulative value even when I choose a particular filter comes correct as per the overall amount.
e.g.,2018 data.

But I would like to plot this on a clustered column chart so I need cumulative sum on all the days.

I use : 

Profit over period =

CALCULATE (

[Revenue],

FILTER (

ALL ( Financials ),

Financials[TxnDate] <= MAX ( Financials[TxnDate] )

&& Financials[Source.Name] = SELECTEDVALUE ( Financials[Source.Name] )))

I am getting this data 

However, I need this data in this form


I am getting these missing dates in 3rd picture because the overall table does not contain those dates.
Is there any measure or formulae that I can use to get the missing dates.

Please help.

[Ashish_Mathur]

Hi,

What is the difference between the 3rd and 4th Table?

[v-frfei-msft · Accepted solution]

Hi @shubh25 ,

One sample for your reference, please check the following steps as below.

1. Create a calendar table as below and make it related to the fact table.

CALENDAR = CALENDAR(DATE(2019,01,01),DATE(2019,01,31))

2. To create a measure.

Measure = CALCULATE(SUM('Table'[Revenue]),FILTER(ALL('CALENDAR'),'CALENDAR'[Date]<=MAX('CALENDAR'[Date])))

[shubh25]

I am trying to do that already. However, rest of the filters are in another table i.e., "Finacials". Is there a way that I can use all and selectedvalue on columns from different tables. 
i.e., DATES from CALENDER & SOURCENAME from FINANCIALS.
The DAX expression I am currently using-

Profit over period =

CALCULATE (

[Revenue],

FILTER (

ALL ( Financials ),

Financials[TxnDate] <= MAX ( Financials[TxnDate] )

&& Financials[Source.Name] = SELECTEDVALUE ( Financials[Source.Name] )))

[Greg_Deckler]

I would use GENERATESERIES to create a new table with all of the dates you need, or just CALENDAR or CALENDARAUTO perhaps. Then you could use that as your axis.

[shubh25]

Generating all dates for all the data set will create millions of rows which will slow down the pbix significantly.