Keep in mind that in Tableau a FILTER will remove records from being available to display IN THE VIEW you're working on.
So if A is a subset of B and B is a subset of C, if you add filters that constrain you to A, then you can't see all of B or C anymore.
If you filter to all of B, you have "more than you need" to see all of A, but you still can't see all of C.
If you filter to all of C, you have "more than you need" to see all of B, and "more than you need" to see all of A.
This means simple filters aren't the way to get what you're after--if you're trying to do it all in the same view.
To see YTD AND Prior YTD (which are mutually exclusive), you need "All of This Year So Far" AND "All of Last Year"
In the case of different time windows like WTD, MTD and YTD, you actually want to "re-use" some records more than once, and you certainly don't want to restrict some of them away.
A record with a date value of 4/18/2018 would contribute to a WTD total, MTD total AND a YTD total.
So what you really want are different measures--one that generates each total differently.
Another idea is to create a view for WTD, another view for MTD and another view for YTD and filter each differently.
Sorry so I'm not talking about a filter.
I just want to display the below as columns -
Previous Year | Previous YTD | Current YTD
WTD | MTD | YTD | Previous YTD | Previous Year
maybe what I'm thinking of is a date parameter because obviously previous year and previous YTD will have overlapping dates just like WTD, MTD, YTD
I am not sexactly sure, but overlapping itself is not a matter provided you only focus on one measure value.
attached, I created various "Sales" of each logic.
If you need multiple measure value, you need to create each field for respective measure with this approach.
To avoid that, you need to duplicate data source itself to make each period to "dimension".
Superstore_YTD_sample_SM_10.3.twbx 362.4 KB
This is what I was looking for - a bit annoying because you have to create for every measured value instead of just creating buckets for specific dates, but didn't take too long.
If you have multiple measure values to calculate these period criteria respectively, you need to create # of measure x # of criteria.
==> not realistic for 2 or more measure
In this case, you need to duplicate data source up to # of criteria to create different date range criteria in respective ID of data source.