By selecting tuples, you can focus your calculations in the active Smart View grid, limiting their scope to specific slices of data in your cube. Tuple selection helps you optimize asymmetric grid calculations across dimensions, avoiding over-calculation.
Essbase calculation tuples differ from tuples used in MDX queries. Calculation performance and cube size are mainly driven by the number of blocks in the database (given a specific block size). For this reason, calculation tuples are specified only for sparse member combinations. In addition, for ease of calculation scripting, multiple members from a single sparse dimension can be included in a calculation tuple specification. For example, if you specify ("New York", "California", "Actual", "Cola") as a calculation tuple, then you calculate the following cell intersections:
"New York"->"Actual"->"Cola" "California"->"Actual"->"Cola"
Consider the following symmetric grid. It is symmetrical because each product has the same markets and scenario (Actual) represented in the grid.
The following grid is asymmetric, because the Diet Cola product has fewer markets in the grid than the Cola product has.
The default calculation scope, when more than one dimension is in a FIX statement or a Smart View grid point of view (POV), is to calculate the cross product (all possible combinations) of the members in the FIX or grid. In other words, a POV-driven calculation in which product and market combinations are taken from the grid calculates all of these row-member combinations:
Cola->"New York" Cola->"Massachusetts" Cola->"Florida" Cola->"Connecticut" Cola->"New Hampshire" "Diet Cola"->"New York" "Diet Cola"->"Massachusetts" "Diet Cola"->"Florida" "Diet Cola"->"Connecticut" "Diet Cola"->"New Hampshire"
This may be more calculation activity than you need. If you want to calculate only the combinations shown on the grid, you can specify which tuples to calculate, and limit the calculation to a smaller slice. Calculating tuples can also lower calculation time and cube size.
Cola->"New York" Cola->"Massachusetts" Cola->"Florida" Cola->"Connecticut" Cola->"New Hampshire" "Diet Cola"->"New York" "Diet Cola"->"Florida"