If I understood well, you can create a new calculation field using a cast like an integer or float depending of your values like INT([dimension])
I attached an example.
Book1_v10.5.twbx 14.9 KB
I the got the above numbers(in the screenshot) using the below calculation
FLOAT(IF [Modality]='Daily' and [Segment]='B' then '133553'
ELSEIF [Modality]='Monthly' and [Segment]='B' then '131076'
ELSEIF [Modality]='2 Weekly' and [Segment]='B' then '145596'
Else " "
Now I need in the Table to aggregate the B.B=133553+131076+145596
Hope you are able to get my point.
It is like to the current data base cube i am feeding in manual feeds by creating Calculated fields and then various calculations using those.
Not really, I think that your problem is around the data type, but is difficult to know without see a workbook.
Can you provide an example Tableau workbook file?
I attached a new example with the formula that you give me, let me know if help you.
Book1_v10.5.twbx 15.2 KB
My data source is Cube file.Perhaps it is the reason Tableau is finding it difficult.
Sorry, I missed that the data-source is a Cube in your Friday comment.
I have a lot of problems also with Cubes with Tableau and I skipped with the following tricks that maybe help you:
1. Create two data-sources, one from the cube and another from a basic Excel structure or CSV than can blend with a common fields.
To explain with more detail with Cubes you will have a lot of limitations like change the datatype for example, then that I did was in a blended data source with a costume data-source that I can define and handle, you can created a new field with the new data type and the only that you need to do is create the common structure that you expect in the cube to you can use it in the second custom data-source.
2. Not connect directly with the cube, use a connector as intermediary like a Excel with a Pivot Table, a database T-SQL running a MDX like a procedure and saving the answer into a database, etc.
The idea here is when you are doing analysis then you need connect to the full cube to can explore and understand the data but when you are doing Dashboard you don't need it because you can normally limit the dimensions and his combinations to has a subset predefined in an extract. (Except with the distinct counts)
3. Connect directly to the source of the cube.
If you can to skip all of this limitations always is good to connect to the source of the cube, I normally let as a last option because when I used a cube has internally MDX or DAX calculations than will be expensive/timeless to reproduce with SQL queries. If the cube is basic like is organizing the data in hierarchies or change names and datatypes and you can connect, is maybe your best option.
If this solve your issues I will appreciate if you mark like a correct answer.