Data modelling suggestions

Hi vibhoryadav23 - you can write measures to dynamically calculate the percentage of each error while considering the applied filters.

create total count of Errors

Total Errors =
CALCULATE(
COUNT('Table'[Error]),
ALL('Table'[Error]) -- Ensures all errors are considered, even if some are filtered
)

create percentage of Each Error

Error % =
VAR CurrentErrorCount =
COUNT('Table'[Error])
RETURN
DIVIDE(CurrentErrorCount, [Total Errors], 0) * 100

Create a measure that applies the selected scenario dynamically

Scenario Filter =
SWITCH(
SELECTEDVALUE('Scenario Table'[Scenario]),
"A", 1, -- No filter applied
"B", IF('Table'[TNr] = 1, 1, 0),
"C", IF('Table'[TNr] = 2 && 'Table'[RT] = 1, 1, 0),
"D", IF('Table'[TNr] >= 2 && 'Table'[RT] = 1, 1, 0),
1 -- Default to no filter
)

To exclude "OK" values from the chart while keeping them in calculations

Error % for Chart =
IF(
MAX('Table'[Error]) = "OK",
BLANK(), -- Exclude "OK" from the chart
[Error %] -- Show percentage for other errors
)

Use a bar chart to display the percentages for each error.

Set the Error field on the X-axis.
Use the Error % for Chart measure as the Y-axis value.
Add a slicer for the Scenario Table so users can select the desired scenario.

Add any other slicers (e.g., for Date, ID, etc.) to allow users to filter the data.

Hope this works. please try