This repository contains a DDF dataset on dollar billionaires, built by combining and harmonizing data from:
If you are new to DDF:
- Methodology (matching/merging rules):
methodology.md - ETL update workflow (how to refresh the dataset):
etl/README.md - Concepts and indicator metadata:
ddf--concepts.csv
The dataset contains:
- Person-level entity data (billionaires)
- Person-level time series (e.g.
worth,annual_income,daily_income) - Country-level derived indicators (e.g. billionaire counts, billionaires per million, average age)
You can inspect available files directly in the repository root (for datapoints/entities) and in ddf--concepts.csv for definitions and metadata.
At a high level, updates follow this process:
- Download source data (Hurun + Forbes)
- Transform each source to clean intermediate tables
- Generate embeddings for matching
- Run MCP/LLM-assisted matching to create mappings
- Human review of mappings
- Build final DDF outputs
See full instructions in etl/README.md.