Calculate balance

[akasonia]

I have 4 tables(Charges, payments,penalties, and adjustments) that I combined into 1 large table using power query in Power BI. The new table has 3 columns: Description, Transaction date and Amount. I would like to create a new column (Balance).

Balance for the first charge on 7/1/2010 equal the charge Amount 226.87

Next Balance(Payment)= 0.00 made on 7/21/2010 (Charge Amount[226.87]+Payment Amount[-228.87])

The next items balance will the actual balance +the next amount.
Here is a Screenshoot of the report I would like to reproduce.

Thank you in advance for your help

[danextian · Accepted solution]

hi @akasonia 

Try this as a calculated column:

CALCULATE (
    SUM ( 'table'[Amount] ),
    FILTER (
        ALL ( 'table' ),
        'table'[Date] <= EARLIER ( 'table'[Date] )
            && 'table'[customer code] = EARLIER ( 'table'[customer code] )
    )
)

This as a measure assuming you a have separate dates/calendar table that has a one to many relationship to your fact table.

CALCULATE (
    SUM ( 'table'[Amount] ),
    FILTER (
        ALL ( 'datesTable' ),
        'datesTable'[Date] <= MAX ( 'datesTable'[Date] )
    )
)

You can cretae one using the CALENDAR function in DAX.

The formulas above aside,  I would keep those four tables separate instead of combining them and just use dimension tables and relationships to bridge them all. I can imagine how slow the refresh will eventually be if it isn't slow yet.

[danextian · Accepted solution]

Did you use a separate dates table? If so, did you use the column from the dates table?  Not using a separate dates table in time intelligence calculations can cause ndesirable results.

[Ashish_Mathur]

Nothing can be understood from that very small image.  Share data in a format that can be pasted in an MS Excel file.  Show the expected result very clearly.

[danextian · Accepted solution]

hi @akasonia 

Try this as a calculated column:

CALCULATE (
    SUM ( 'table'[Amount] ),
    FILTER (
        ALL ( 'table' ),
        'table'[Date] <= EARLIER ( 'table'[Date] )
            && 'table'[customer code] = EARLIER ( 'table'[customer code] )
    )
)

This as a measure assuming you a have separate dates/calendar table that has a one to many relationship to your fact table.

CALCULATE (
    SUM ( 'table'[Amount] ),
    FILTER (
        ALL ( 'datesTable' ),
        'datesTable'[Date] <= MAX ( 'datesTable'[Date] )
    )
)

You can cretae one using the CALENDAR function in DAX.

The formulas above aside,  I would keep those four tables separate instead of combining them and just use dimension tables and relationships to bridge them all. I can imagine how slow the refresh will eventually be if it isn't slow yet.

[akasonia]

Thank you @danextian for your assistance. I tried the measure but I am still stuck.
I created the measure and add it to my table. I added Transaction Date, the amount, the Measure(Cummulative Amount) and it worked untill I added the description column on the table or when I apply a filter(Account number). 
For this report I will need to filter the table by the Customer Number.
Thank you

[danextian · Accepted solution]

Did you use a separate dates table? If so, did you use the column from the dates table?  Not using a separate dates table in time intelligence calculations can cause ndesirable results.

[akasonia · Accepted solution]

Thank you very much @danextian . I was able to solve my problem based on the DAX and suggestions you provided.
I created the first measure: 

Accumulation by date = CALCULATE(

    SUM('vw_trans_charges'[Amount]),

    FILTER(

        ALLEXCEPT('vw_trans_charges', 'vw_trans_charges'[AccountNumber]),

        'vw_trans_charges'[TransactionDate] <= MAX('vw_trans_charges'[TransactionDate])

    )

)
and used it to create my final measure:

Running Accumulation = CALCULATE(

    [Accumulation by date],

    FILTER(

        'vw_trans_charges',

        'vw_trans_charges'[TransactionDate] >= MIN('Calendar'[Date]) &&

        'vw_trans_charges'[TransactionDate] <= MAX('Calendar'[Date])

    )

).

I have one more question. Since I am pulling the data(Tables) from a SQL database, what will happened to my report when the server is shut down?
NB: This is historical data(last transation date is 2023). I won't need to refresh the report and I used import mode.

[danextian]

Unless the data from that SQL server is in a separate query and refresh for that has been disabled, you will have an error when refreshing the semantic model. Referencing the SQL query in another query even  if the the load for that SQL query has been disabled will also cause an error. I would suggest to store SQL data somewhere or use dataflow which will store its output in Azure Datalake Storage. You can still create transformations on top of the output of a dataflow.