Counting entries of key values

[#1]

Hi,

I have two tables

Presence

Date	User	Presence
1	1	1
1	2	0
2	1	1
2	2	1

and

Journal entries

Date	User	SomeValue
1	1	Some
2	1	Things

The Presence table is there to give information about which journal entries should be taken into consideration in certain aggregations. 'Active' users are indicated by a 1 in the Present-field.

Some Users were not "active" at certain points in time, and so the average entries/users should not take these into consideration for these dates.

In particular, I'm looking to calculate the average of[ (users who used the journal on a date)/(active users on that date) ] over all dates.

My first strategy was to do ( Count Distinct Journal entries.User ) / ( { < Presence = {0} > } Count Distinct Presence.User )

However, this doesn't seem to give the desired outcome.

Is there a simple way to access only the Users that are present in one of the tables?

[johnw]

Not sure I entirely follow, but perhaps this?

count(distinct User) / count({<Presence={'1'}>} total <Date> distinct User)

As an aside, you have a synthetic key, which seems to be the entire key to both tables, so it seems like Presence and SomeValue should just be two fields on the same table.