Cognos Transformer OLAP Modeling Basics

1. What is a Cognos Transformer?

IBM Cognos Transformer is a OLAP modeling tool designed to use with IBM Cognos 8 and 10 versions.OLAP modeling in Transformer basically involves,

• Choose a supported product language.
• Add dimensional metadata.
• Specify the measures (performance indicators).
• Apply custom views.
• Create PowerCubes.
• Deploy these cubes to support OLAP reporting and analysis.

2. What are the steps involved in Dimensional Modeling Workflow?

• Analyze OLAP reporting requirements.
• Identify data sources and import dimensions and facts.
• Create regular and measure dimensions.
• Verify the model for any ambiguities.
• Create custom views.
• Build PowerCubes.
• Create partitions.
• Setup Security.
• Deploy Cubes to the reporting environment.

3. What are the steps involved in analyzing reporting requirements?

• How many reports are present and what are the functionalities required in each report?
• Identify the priority of each report and divide them into '3' categories: High, Medium and Low.
• How many summary and detail reports are present?
• Identify  the frequency of environmental changes and new enhancements.

4. What are the steps involved in analyzing Source Data?

• How many data sources are present?
• What is the format of source data: flat files, spreadsheets, or databases?
• Does data need to be converted to a supported data source type before it gets imported?
•  How many records are present in the database and what is the frequency of the volume of data to increase?
• How much of the data is static and how much changes gradually over time?
• How much data preparation is required? Ensure that the source values that feed the categories are unique and, if feasible, that you aggregate or otherwise preprocess your data before importing it.
• Are linked measures from different data sources updated at the same time?
• Must you create additional data sources to accurately model your organization?

5. What are the different types of tables used as sources in Transformer model?

In Transformer, Data source can be any of the following:

• Dimension: contains columns with descriptive data. Dimensions usually contain many columns with few rows of data.
• Fact: contains measures that can be used for calculations. Facts usually contain less columns with more data.
• Mixed: contains both the dimensions and measures.

6. What are the techniques used to design Data Sources in Transformer?

• Design the structural information of each dimension using one source.
• Make sure data source contains enough information to generate categories without database joins. If we need to create joins, create them in the framework manager and then import into transformer.
• Create custom views that can be associated with IBM Cognos security (users, groups, and roles).

7. What are the different data source types supported in Transformer?

• Cognos EP7 Series Source.
• Cognos Package or Report: we can import query items and the associated filters and prompts. Please note that, transformer only supports simple list reports as source.
• DMR Packages.
• OLAP Packages.
• SAP BW Packages.