Filter function is not filtering

Does it make a difference if you filter just the column rather than the whole table?

No Rebate Fuel Spend =
VAR _Spend =
    CALCULATE (
        [Fuel Spend],
        FILTER (
            VALUES ( 'EFS Transaction Report'[all_fuel_line_item_discount] ),
            'EFS Transaction Report'[all_fuel_line_item_discount] = 0
        )
    )
RETURN
    IF ( ISBLANK ( _Spend ), 0, _Spend )

Hi cheid_4838 ,

This kind of issue usually happens because of how context and relationships work in DAX, especially when you use FILTER over an entire table instead of just the column you need. When you filter the whole table, the filtering context might not apply as you expect, and you can end up with unexpected results.

Try this approach:
Instead of filtering the entire table, create a filter directly on the column where the discount is applied. That way, the filter is more precise and the calculation will only consider rows where the discount is zero.

Here’s how you can rewrite your measure:

No Rebate Fuel Spend :=
VAR _Spend =
    CALCULATE(
        [Fuel Spend],
        'EFS Transaction Report'[all_fuel_line_item_discount] = 0
    )
RETURN
    IF(ISBLANK(_Spend), 0, _Spend)

• This measure checks directly for rows where [all_fuel_line_item_discount] = 0 and sums the relevant fuel spend.
• Using this filter at the column level is usually more reliable than wrapping the whole table in the FILTER function, which can sometimes confuse row and filter context.

Extra tips:

• Double-check that [all_fuel_line_item_discount] is always numeric (no blanks, errors, or text values).
• Make sure your relationships between lookup tables and the fact table are active and set correctly, as inactive or ambiguous relationships can also cause odd filter behavior.
• If you’re still seeing unexpected results, try creating a simple table visual with just [invoice] and [all_fuel_line_item_discount] to confirm that your data matches what you expect.

Let me know if this helps or if you’re still running into issues—happy to troubleshoot further if needed!
translation and formatting supported by AI

Hi cheid_4838 

Could you please confirm if your query have been resolved the solution provided by  AlexisOlson , AlexisOlson and Ashish_Excel ? If they have, kindly mark the helpful response and accept it as the solution. This will assist other community members in resolving similar issues more efficiently.

Thank You.

Hi cheid_4838 
Because of this:

IF(ISBLANK(_Spend),0,_Spend)

The VertiPaq engine in Power BI enhances performance by skipping unnecessary calculations when processing blank (null) values. However, if a measure is forced to return zeroes instead of blanks, the formula engine may perform a full cross-join of the dimension tables in the visual, evaluating each possible combination explicitly. This can significantly increase query execution time, as opposed to scanning only the rows where relevant combinations actually exist. It may also result in unexpected data points appearing. For example, an obsolete item, which was deactivated several years ago, might still display for a selected period, but with zero values.