Curse of Dimensionality is well known and well understood in the world of Analysts, and for good reasons.
A generic Analytics design with multidimensional data constructs can throw up inordinate number of combinations. As computing power gets cheaper and becomes more easily available, the popular solution has been to throw more Iron at the Problem. While this does help to run the numbers, the generic approach can leave a good many users tizzy and confused due to the multitude of data that gets generated and much of it is of little value or meaning.
Where PIPS Analytics Consolidation Framework succeeds is in keeping Business Analytics relatively simple - due to the intuitive design and the singularity of its purpose. Aggregation is effectively different in Design from that for Client Solutions, or Pre-Trade Analytics or Trading Solutions. This principle is at Core of PIPS Framework Design, and is central to the various decisions of "including features" or "excluding complexities" based only on their relevance to the final purpose.
For simplicity, PIPS Analytics divides the entire Framework into four Key Subsystems.
1. Mapping Framework: Holds all key associations between the major Analytical Factors and Aggregation Schemes.
2. Business and Market Scalars: Holds the Analytics Population after running all key Aggregations on Exposures.
3. Scaling Vectors: Maintains and manages the Characteristics and Constructs for Risk and Performance Measures.
4. Analytical Functions: Supports pre-packaged and user-defined Analytics and runs them over orderly structures.
Note: The CodePIPS logo with its 4 quadrants is a representation of the above 4 core sybsystems.
A key feature of the PIPS Design is that the user interfaces are designed with the Practitioner in mind. Free wheeling analytics can easily run amock unless it is well organized from the Users viewpoint and PIPS Analytics pays considerable importance to this. For example, the Analytics Segment offers a number of functions to the users that are conveniently layered in Groups such as in the illustration below.
> Markets and Securities related Analytical Functions
> Benchmarks and Composites related Analytical Functions
> Investors and Investor Groups related Analysis
> Investment and Investment Groups Related Analysis
> Combined Positions and Portfolios related Analysis, and
> Portfolio Clusters and Portfolio Groups related Analytical Functions.
Other segments are similarly well organized and the entire Solution is planned for easy reference by Business Users and as well by System and Data Management Personnel.