Das Video ansehen
In unserer fortlaufenden Serie stellt sich unser CEO Michael Olschimke einer Frage aus dem Publikum:
“We would like either build a semantic model or let end users build it themselves as star schemas.
In Infomarts we expose facts and dimensions.
Dimensions are based on pits and expose all contextual attributes valid to a given snapshot data.
Now the problem is that facts and dimenions need to be joined not only by main keys of dimension, but also a snapshot data. However, PBI allows only joins with 1 attribute.
What is a recommended way to tackle this?
I thought of introducing sequence numbers in PITs and exposing them in virtualized fact views, additionally exposing separate snapshot dimension that synchronizes snapshots of all the dims (otherwise we end up in cartesian join). However this defeats partitioning in the PITs (join over sequence number and not hashKey + SnapshotDate blocks partition pruning).”
Michael delves into an in-depth discussion on leveraging Data Vault’s Point-in-Time (PIT) tables within PowerBI, exploring how this integration enhances analytical capabilities and supports dynamic reporting in the realm of Big Data.
Über den Vortragenden

Michael Olschimke
Michael hat mehr als 15 Jahre Erfahrung in der Informationstechnologie. In den letzten acht Jahren hat er sich auf Business Intelligence Themen wie OLAP, Dimensional Modelling und Data Mining spezialisiert. Fordern Sie ihn mit Ihren Fragen heraus!