Row result filtering

I have elaborated a model of which I have a situation with the measure "% Var Total" this measure has two behaviors one at the detail level (product) and the other in the subtotals depending on the aggregations used.

Taking as good the result of the calculation in my matrix, articles are shown that although they had movement do not correspond to the filtered period, for example the product "Item1097" had movements in March 2018 when I am comparing 2020 vs 2019.

Can you tell me what is the correct way to filter this result?

I appreciate your guidance in advance

Here is the model download link

https://drive.google.com/file/d/1uIlxPNASRaix02481wKpbFK3fwj9HsHy/view?usp=sharing 

The measure in question:

% Var Total = Var TsVarTotal = DIVIDE(CALCULATE(Ventas[Varia Total]),CALCULATE([Venta Anterior]),0) 
VAr X = 1

 Var ProdVaria= CALCULATE([Varia Total Z])
Return 
SWITCH(TRUE(),
       HASONEVALUE(Productos[Producto]),ProdVaria,
       HASONEVALUE(Clases[Desc Clase]),TsVarTotal,            	
       HASONEVALUE(DivRes[DivShr]),TsVarTotal, 
       HASONEVALUE(Ventas[SECTOR]),TsVarTotal,                                 	
       HASONEVALUE(Clientes[Holding]),TsVarTotal, 
       HASONEVALUE(Vendedores[Name Vendedor]),TsVarTotal, 
       HASONEVALUE(Monedas[Currency]),TsVarTotal,
       BLANK()
)