What Is A 'Slice' In Slice And Dice? 1

[Skittle]

I understand the gerneral concept that looking at a subset of cube data is to slice and dice.

A dice is a subset cube within an OLAP cube of data.

What howeve is a 'slice' - if it is anything specific at all? Is it a set of data that only covers two dimensions?



Dazed and confused.

Remember.. 'Depression is just anger without enthusiasum'.

 

[blom0344]

A slice could be seen as a subset based on a single member value for one of the dimensions. If you have a cube with 10 year of sales data, a slice would be all associated data for 1 certain year, or for 1 associated salesrepresentative

Ties Blom

 

[dkyrtata]

blom0344

Sounds good. Now can you describe from your example what the "dice" is? Can it be the year broken down into months?

 

[blom0344]

Imagine creating a cubeslice by using a filter on the year 2010. Take the dimension Salesrepresentative and measure revenue. This would be a slice.
Add a second dimension like 'Productgroup' and add this to the columns. The result can be thought off as the collection of 2 dimensional dices (intersections (tuples) between salesrep and productgroup for year 2010)
Adding dimensions will yield collections of n-dimensional dices.

The interesting bit is that the dices can be directly related to the 'sparsity' of the cube as each intersection may or may not yield data.

Well, actually that is really another story

Ties Blom

 

[MDXer]

Ties,

have a star for a very well worded explination.