Why aren't null values filtering out of my table - even though I've unticked Include Null Values?

[HClegg]

I've created two tables that display the top 20 customers, sorted by variance. One table shows top 20 for positive variance, the other for negative variance. 

In order to filter out opposite variance from the respective tables i.e. positive variance table should have no negative variance customers showing. I've created the below expression to NULL any value that wouldn't belong in that table.  

if(Sum([Sales Amount Turnover]) - Sum({<Year=, Quarter=, Month=, [Sales Invoice Date]
={">=$(=MonthStart(Max([Sales Invoice Date])-31))<=$(=MonthEnd(Max([Sales Invoice Date])-31))"}>} [Sales Amount Turnover]) > 0,
Sum([Sales Amount Turnover]) - Sum({<Year=, Quarter=, Month=, [Sales Invoice Date]
={">=$(=MonthStart(Max([Sales Invoice Date])-31))<=$(=MonthEnd(Max([Sales Invoice Date])-31))"}>} [Sales Amount Turnover]), null())

I've then added a fixer number limitation of 20 to the Customer Name dimension - so I get my top 20. I've also unticked Include Null Values. 

However, null values (which are the opposite variance) still appear in the tables - although only where there appears to not be 20 customers of a pos/neg variance, i.e.  March has 19 Positive variances, so there is one null value here, which is actually a negative variance.

How do I remove these null values in their entirety?

[vinieme12]

You might have less than 20 customers that satisfy this criteria

Try a Calculated Dimension as below

Replace Customer Dimension with below Calculated Dimension

=Aggr (  if(Sum([Sales Amount Turnover]) - Sum({<Year=, Quarter=, Month=, [Sales Invoice Date]
={">=$(=MonthStart(Max([Sales Invoice Date]),-1))<=$(=MonthEnd(Max([Sales Invoice Date]),-1))"}>} [Sales Amount Turnover]) > 0, Customer , null() )  ,  Customer )

Uncheck Include Null on This dimension