Sorry, senior moment, I meant to say BINS, you should find the functionality under the formula editor, essentially it just wraps a case statement, so if this is not available in 10 then look into using case to group your products, it is on the formula examples.

Not sure what a SKU is but have you looked into case logic or buckets to fulfil this requirement?

You would need a measure at distribution level for; Qty Ordered and Qtry Received; without these or equivalent then this is not achievable. Though if you know the underlying tables you could fulfil via a direct database request - if that is an option for you. You might try asking this on the BI Applications forum, I don't…

It was a 'general' caveat, not to try to use dimensions as facts - I know it can be tempting to think I can workaround it by....

Please note also that doing any aggregation wise on dimensions rather than facts WILL be erroneous, because OBIEE as standard will make members of dimensions DISTINCT values BEFORE your aggregation, so if you even do a conversion and aggregation to the dimension member in a formula you will still not get the correct…

Hello Ewa, You would be better on the BI Applications forum for this, there is a control to move your post.

Selection step is a good idea - take population less one null value.... Have to side with Christian on sorting out the underlying data issues, or at least quarantining them until they can be resolved.

"it fulfills direct access needs" - so you are saying you can query against it, but that doesn't make it BI? I, more than many on here perhaps, know the challenge of putting BI over EPM sources, but this seems like such a poor solution that I honestly cannot see the real business application. I understand your…

This HFM integrations sounds half-baked to me, currency must be a cornerstone dimension so for it not be there as standard seems a massive oversight- or is that unfair- is it like OBIA, it is as good as your knowledge of the internals?

Use conditional formatting to turn the content font the same colour as the background when the value is null, but wouldn't it not add up then?

Yes - federation - a grand title for my nasty little look up table workaround, but it could be a workable option, though I guess my concern would be performance as the currencies would frequently become the obvious choice for filters...

Sure, you could also set a presentation variable and put the {PresentationVariableName} in your filter, or sql equivalent. This can also be used with the protect functionality.

There is a protect filter on the filter control which is used to prevent the dynamic filter overwrite that you describe and enforce the originally written filter.

That sounds performant, transparent and easy to use a crock Seriously though, with the way it works is there any way to (non-intrusively) customise, what I had in mind was creating a lookup table and using the lookup so they can use it to get their currency information?

Thanks for the info - as I said my EPM experience has been on the Essbase side of the world, though I know that HPCM now also has the relational option to cater for Cartesian products that exceed Essbase's mighty 2 x (2^52) sparse permutations. Surely it must come down to the initial set up, RDBMS in theory makes life…

Hi Ivan, I have used other EPM (Planning, HPCM, Essbase) solutions but not HFM with OBIEE, but if CurrencyType is a dimension in your underlying Essbase cube then this is not working, you should see Members when you (say) create a filter and use the value selector to pick what you are filtering. If this is not happening…

Hi Somi, Oracle did not write excel. Microsoft are responsible for that one. If you create a comma delimited file and manually populate it with data like; - SomeText,Nov-17,Somemore text Then save it as a .csv extension. Then open it in excel. You will see that excel misinterprets your field. Excel misinterprets. There is…

So, these are 'Accounting Periods' - Nov-17 etc, not dates at all. When excel makes Nov-17 11/17/2018 it is entirely erroneous. So you need to use a date field in your export which is fully formed and stored as a date, then your export to excel and csv will be meaningful.

Yes, I was motivated by the same bias, trying to get to some objective evidence...

You are speaking aloud my internal thought processes exactly, but I was trying to get to the 'proof' for the user by getting them to see what is really in their system... and then lay out the detail of the 'why' of it - excel's smartness doing something dumb...