OACintermediate
OAC Datasets: Preparation, Joins & Data Enrichment
Explain how OAC datasets are prepared, cleaned, joined, and enriched before being used in Data Visualization projects.
Before a dataset is useful in a DV project, it usually needs shaping — OAC provides a dedicated Data Preparation experience for exactly this, and interviewers ask about it to see whether a candidate understands OAC's self-service data pipeline, not just the visualization layer sitting on top of it.
It's like a professional kitchen's prep station: instead of chopping vegetables from scratch every single time an order comes in, the recipe (transformation steps) is defined once, and every future batch of the same raw ingredient gets automatically prepped the same consistent way before it ever reaches the chef.
Key Concepts
1
When a file (CSV, Excel) or connection is added as a dataset, OAC runs automatic column profiling and type detection — inferring whether a column is a measure, an attribute, or a date, and flagging potential data quality issues (nulls, outliers, inconsistent formats). From there, the Data Preparation interface lets an author apply transformation steps: splitting or combining columns, changing data types, trimming whitespace, standardizing categorical values, filtering out bad rows, and adding calculated columns — recorded as a repeatable transformation script/recipe so the same cleanup can be reapplied automatically every time the source data refreshes, rather than being a one-time manual fix.
column profiling and type detectionData Preparationtransformation script/recipe
2
Joins between datasets are defined either automatically (OAC suggests joins based on matching column names/types) or manually by the author dragging a relationship between two datasets, similar to the blending described in the Projects/Canvases topic but with more explicit control over join type (inner/outer) and key columns when working directly in Data Preparation rather than ad hoc on a canvas.
Joins between datasets
3
Data enrichment goes a step further: OAC can automatically enrich a dataset with derived attributes it detects semantically — for example, recognizing a column of city names or postal codes and offering to enrich it with geographic hierarchy (state, country, region) for map-based visualization, or recognizing a date column and automatically deriving a full calendar hierarchy (year, quarter, month, day of week) without the author manually building one. A strong interview answer distinguishes this augmented, semi-automated preparation as one of OAC's genuine differentiators versus on-premise OBIEE, which requires an RPD developer to hand-build equivalent hierarchies.
Data enrichmentaugmented, semi-automated preparation