Creating and configuring cubes in Planning Analytics | TM1 für Dummies

Thanks to its multidimensional structure, IBM Planning Analytics enables not only the analysis but also the precise planning and control of data. Cubes play a central role here – multidimensional data models that store information in a structured way and make it flexibly analyzable. In this article, we will show you step by step how to create and configure a cube in Planning Analytics Workspace (PAW). 

Note: For historical reasons, the term “TM1” is often still used synonymously for “IBM Planning Analytics”, but today it officially only refers to the database engine. 

What is a cube?

A cube is a multidimensional data structure that consists of several so-called dimensions. Each dimension represents a perspective on your data – e.g:

• Time (years, quarters, months)
• Products (categories, article numbers)
• Regions (countries, sales territories)
• Versions (plan, forecast, actual)
• Key figures (turnover, sales, costs, etc.) 

A cube then stores a value for each combination of these dimension elements – e.g. the sales for product A in Q1/2024 in Germany.

1. create dimensions

Before you can build a cube, you need the individual dimensions. In PAW this is done as follows:

How to create a dimension in PAW:

1. Log in to the Planning Analytics Workspace (PAW).
2. Navigate to “Data and models”.
3. Open an existing workbench or create a new one.
4. Log in to the database you want to work in.
5. Create a new dimension by right-clicking on “Dimensions” > “New dimension”.
6. Give it a name (e.g. “Time”) and save.
7. Open the dimension – it is initially empty.
8. Add elements (master data) manually or import them using the TurboIntegrator (TI) process.

Tip: Good cubes usually have 5-7 dimensions. More dimensions mean more flexibility, but also greater complexity and potentially poorer performance.

2. create cube

When your dimensions are ready, you can create the cube:

Cube Creation in PAW:

1. Right-click on “Cubes” > “New cube”.
2. Select the target database and give it a name (e.g. “Sales planning”).
3. Select the desired dimensions and arrange them in a meaningful way (e.g. Time > Version > Organization > Product > Key figure).
4. Click on “Create”.
5. Open the cube and create a view to enter data at the lowest level.

3. Define rules for calculations

Cubes can be enriched with calculation logic – for this you use so-called rules:

Example: 

[‘In total’] = N: 
[‘region A’] + [‘region B’]; 

This means: The “Total” element should add the value from “Region A” and “Region B” – but only at the lowest level (N-Level).

How to create a rule:

1. Go to your desired cube.
2. Select “Edit Rules.”
3. Enter your rule in the editor.
4. Save and test it using the validation function.

4. Configure access and permissions

In Planning Analytics, you can precisely control who can see and edit what.

Procedure for PAW:

1. Go to “Administration” > “Users & Groups.”
2. Assign users to a role:
   • Administrator
   • Modeler
   • Analyst
   • Consumer 
3. Set group permissions for cubes, dimensions, and views.
4. Optional: Import permissions automatically via CSV file.

Procedure for TM1:

1. Use security cubes, e.g.:
2. }ElementSecurity_DIMNAME for element level
3. }ClientGroups for linking users and groups 

Rights types:

• None – No access
• Read – Read permissions
• Write – Write permissions 

🔒 CellSecurity is possible, but rarely used – more often with very sensitive data. 

Conclusion

Cubes are the heart of IBM Planning Analytics. With a clear structure, wisely selected dimensions, and clean data logic, even complex planning and analysis models can be efficiently mapped. Especially at the beginning, it’s worthwhile to start with simple cubes and expand them gradually.

If you’d like to learn more about TM1 and receive helpful tips directly in your inbox, sign up for our newsletter. It’s easy to stay up to date!