Join with analytic views problem

Hi Gurus!

I have the following problem:

2 tables:

drop table "PULZUS_DM"."TEST_TABL_1";

CREATE COLUMN TABLE "PULZUS_DM"."TEST_TABL_1" ("MY_ID" INTEGER, "MY_NUM" INTEGER);

insert into "PULZUS_DM"."TEST_TABL_1" values(1, 2);

insert into "PULZUS_DM"."TEST_TABL_1" values(2, 1);

insert into "PULZUS_DM"."TEST_TABL_1" values(3, 1);

drop table "PULZUS_DM"."TEST_TABL_2";

CREATE COLUMN TABLE "PULZUS_DM"."TEST_TABL_2" ("MY_ID" INTEGER, "MY_ATTRIBUTE" NVARCHAR(6), "MY_OTHER_NUM" INTEGER);

insert into "PULZUS_DM"."TEST_TABL_2" values(1, 'attr01', 5);

insert into "PULZUS_DM"."TEST_TABL_2" values(1, 'attr02', 2);

insert into "PULZUS_DM"."TEST_TABL_2" values(2, 'attr01', 5);

insert into "PULZUS_DM"."TEST_TABL_2" values(3, 'attr02', 2);

I put these tables in TEST_AN_VIEW01 and TEST_AN_VIEW02. The measures are MY_NUM and MY_OTHER_NUM.

In a calculation view I join these analytic views from projections (I have to put some filters before...) with inner join, cardinality 1:n.

It looks like the attached pic.

For this view I get an unexpected result:

Clearly it should be 4 for the first MY_ID since:

If I use only the tables instead of the analytic views (that I cannot do in my real problem), than it calculates everything well (the MY_NUM_SUM for MY_ID = 1 is 4...).

Is it a bug, or I just do not get something with the join and analytic view realtionship?

Thanks,

David.