Sparsity and Rules Calculation

When you attach a rule to a cube, it becomes very difficult to use sparse consolidation.

This is because the TM1® consolidation engine does not know how to locate rules-derived values. Its only recourse is to scan every member cell of a consolidation to see if it contains a number. And as you just saw, this can be extremely inefficient.

Because consolidations will be incorrect if they skip values derived from rules, the default behavior is to turn off sparse consolidation when rules are defined for a cube. This ensures that all values are correct, but speed can plummet by several orders of magnitude (roughly proportional to the degree of sparsity of the cube data). If you build applications of even modest size that work with sparse data, you will have to write feeders. The process of creating feeders is described in detail in this section.

Each subsequent section in this guide contains instructions for creating feeders to accompany calculation statements.

Note: Rule-generated values are not displayed when zero suppression is applied to the TM1 view unless the value resides in a cell that is fed.