TM1 Forum

One thing has puzzled me since we got TM1. When creating a dimension hierarchy, why should elements be created at the N level as well as being created as children of their parent? See below for example. Does this not result in the unnecessary duplication of dimension elements? Am I missing something basic? Thanks for all replies in advance.

All Countries
-->england
-->scotland
-->wales
n england
n scotland
n wales

iansdigby wrote:One thing has puzzled me since we got TM1. When creating a dimension hierarchy, why should elements be created at the N level as well as being created as children of their parent? See below for example. Does this not result in the unnecessary duplication of dimension elements? Am I missing something basic? Thanks for all replies in advance.

All Countries
-->england
-->scotland
-->wales
n england
n scotland
n wales

Very basic, I'm afraid. N level elements only exist once in a dimension. However they can appear multiple times within a consolidation hierarchy, or within multiple consolidation hierarchies.

Think of it this way; imagine that you have a list of values in one column of an Excel worksheet. In other places in the workbook you can use Sum() formulas to add some or all of those values. Each Sum() formula is independent of any others that you create (unless it's used as an argument to one of the other Sum() formulas), but all point to cells which make up the one basic list of values. The values are entered only once in the worksheet, but can appear in many calculations within that sheet.

So it is with N level elements, and consolidations.

They aren't actually created as children in any real sense; for practical purposes (and technically I may be oversimplifying, but I did say for practical purposes) the consolidations simply reference them, just as the Sum() formulas reference the Excel values.

Hi Ian,

The element is created as the data container, and then the hierarchy describes it's consolidation relationships.

This allows for you to modify these hierarchy relationships (i.e remove an element from the hierarchy) without having the delete the data container in order to change the relationship.

It also allows you to describe multiple hierarchy relationships without having to create duplicate data containers to hold what would be the same data.

Please note that whilst you may see the same element multiple times in a dimension, it is not duplicated. You can prove this by selecting one of these elements in the subset editor and using the (red) Roll Up button to see it's parents - the same element contains all the hierarchy relationships it is a member of.

Alan and Christoper,
Thank you both for taking the trouble to reply and for so succinctly and brilliantly explaining this. Now I understand!
Regards, Ian