Slicing a measure of summarized data

[bsqrdwg]

I am stuck on a measure which won't slice. The measure uses SUMMARIZECOLUMNS and I can get the number exprected, but when I apply a slicer it doesn't filter. 

Responder	Variable	Response	Weighting	Country
55	1	1	1.02	UK
55	2	2	1.02	UK
55	3	3	1.02	UK
55	4	1	1.02	UK
55	5	2	1.02	UK
56	1	3	0.99	USA
56	2	1	0.99	USA
56	3	4	0.99	USA
56	4	2	0.99	USA
56	5	1	0.99	USA
57	1	1	0.85	UK
57	2	1	0.85	UK
57	3	1	0.85	UK
57	4	4	0.85	UK
57	5	4	0.85	UK

For each responder with a response of 1 or 2, return the distinct responder and weighting:

Responder	Weighting	Country
57	0.85	UK
55	1.02	UK
56	0.99	USA

Then SUM = 2.86

Measure:

newwmeasure =

var sumtab =

SUMMARIZECOLUMNS(

'Sheet1'[Responder],

'Sheet1'[Weighting],

FILTER(Sheet1,Sheet1[Variable] in {1,2}),

"Weighted",

SUM(Sheet1[Weighting]) / COUNT(Sheet1[Weighting])

)

return sum([Weighted])

Gives me the total, but I can't use a slicer for the country and I don't want to create extra tables in the model. 

Any help would be appreciated.

Thanks

[daxer-almighty · Accepted solution]

SUMMARIZECOLUMNS is a function to be used ONLY for queries, not for measures. https://dax.guide/summarizecolumns 

This function does not respect context transition, hence using it in measures does not make any sense.

Here's probably what you want:

[Your Measure] =
CALCULATE(
    SUMX(
        summarize(
            T,
            T[Responder],
            T[Weighting]
        ),
        T[Weighting]
    ),
    KEEPFILTERS( 
        // You say in the text you want
        // to filter by Respose, not by
        // Variable (but your formula filters
        // by Variable for some reason).
        T[Response] in {1, 2} 
    )
)

[bsqrdwg]

Thank you @daxer-almighty, it works perfectly 👍