I do not understand the reason of using the distinct.
I suppose that you have to sum "Fact-Shanpshots"."AMOUNT" if your coditions are verified.
You can achieve this, for example, directly in the formula of you analysis with something like
"Fact-Shanpshots"."DOC_TYPE" NOT IN ('AK', 'AL', 'MD', 'ZA') AND
"Fact-Shanpshots"."DOC_TYPE" NOT LIKE 'Z%' ..... )
where "Fact-Shanpshots"."AMOUNT" is a mesured aggregated by sum operator
Think about it you are doing a distinct on measure column.
You have data like
Then you are just doing a distinct on measure which doesn't make sense you are treating measure as a dim attr which is something not unheard of but then you are trying to measure it again.
Filter option as mentioned above will work to get you a filtered measure.