HI guys.

I have these modeling scenario:

Where POST to COMMENTS has a 1:N relationship. So let's imagine the scenario below:

My FB_POSTS has the measure QT_LIKES already aggregated at the source. I have N comments for each post at the FB_COMMENTS. In a determinated moment I have to aggregate all measures to know how much comments, how much likes all comments had and how much likes all posts had given a specific brand.

At the detail level I have these results:

If I wish to know how much POSTS, QT_LIKES and QT_LIKES_COMMENT a BRAND has, how can I aggregate the at the RPD?

I can use a COUNT DISTINCT at the ID_POST to know how much posts a Brand has, ok. I can use SUM(QT_LIKES_COMMENT) and it should return 550.

But the join between comments and posts generate a cartesian of the measures of the posts. The AGGREGATION of QT_LIKES should return 245 at the total level and I didn't know how to do this.

Can you help me?

Regards.

Luiz Araujo