Below is a basic overview of multi-dimensionality which is aimed at spreadsheet
users. To see a more in depth overview contact us.
Click here
OLAP Multi-dimensionality
All spreadsheet users have been exposed to what is termed multi-dimensionality.
The following diagram is an example of multi-dimensionality, showing two dimensions
within a spreadsheet. Members within a dimension can be described as either column
or row headers in a spreadsheet. In the example below, Months is the column dimension
with the members “January”, “February” and “March”. Accounts is the row dimension
with the members “4010: Sales”, “4030: Sales Discounts” and “Net Sales”. Both Months
and Accounts are considered dimensions within the OLAP (on-line analytical processing)
definitions.
However, most of us have had many sheets within a workbook to cater for numerous
sheets of the same structure. For instance, extending the above example we could
have numerous sheets that cater for departments within an organization. This in
effect, adds a third dimension, namely Departments, to the workbook. The members
in the Department dimension include “Training”, “Marketing”, “Finance” and “Total
Departments”. This is illustrated in the example below.
However, say they wanted to extend the same example even further and wanted to look
at information across three years for the same months, departments and accounts,
then within a spreadsheet solution we would have to have multiple workbooks for
each of the years. This in effect is adding a fourth dimension, namely the Year,
to the spreadsheet solution. This is illustrated in the example below.
We could extend this example even further by needing information to be stored by
year, department, month, account and also version. Version could consist of members
Actuals, Budgets and Forecasts. Version would be the fifth dimension in the spreadsheet
solution with a set of the above spreadsheets for the "Actuals", "Budgets" and "Forecasts".
The variances off this would then extend the spreadsheet's requirement even further.
As you can see, the spreadsheet solution is becoming increasingly unwieldy and a
track/record has to be kept of all spreadsheets and any linking between the spreadsheets.
Further, extending any additional requirements, or changing the existing setup,
would result in a significant amount of work for the administrator of the system.
As a result of the spreadsheet deficiencies and other spreadsheet inefficiencies,
OLAP overcomes these problems.
With OLAP, the dimensions in the example could be manipulated into any combination
of column and row headers within a spreadsheet, with access to any member or combination
of members of any of the dimensions.
The dimensions, column/row spreadsheet headers or groupings of like members, in
this example are as listed.
- Accounts
- Months
- Departments
- Years
- Versions
Visualizing a cube beyond three dimensions is often difficult. The diagram below
can assist users in thinking beyond three dimensions. Each axis within the sphere
is a dimension consisting of like members. Values are stored at intersection points
within the sphere. The better OLAP products only store intersection points with
values and calculations are done on-the-fly.
|