CalcFi Open Data¶
34 free CC-BY financial and macroeconomic time series, mirrored from primary sources.
CalcFi Open Data is a reproducible, version-controlled mirror of US and global economic indicators — drawn directly from FRED, BLS, Freddie Mac, US Treasury, BEA, World Bank, the Federal Reserve, and FDIC. Every series carries a transparent provenance trail in the CSV header (# Source: + # Primary URL:), a Frictionless datapackage.json schema, and a permanent DOI.
The Python client (calcfidata) gives you a one-line load per series:
What you get¶
- Tidy CSV files — one row per observation,
date,value,unitschema, primary-source URL in the header - Python client —
calcfidataon PyPI and Anaconda.org - Live SQL endpoint — Datasette for ad-hoc queries
- Web tools — 10 Hugging Face Spaces for browser-side visualizations
- Permanent DOIs — Figshare, Zenodo, OSF
Why does this exist?¶
US economic data is fragmented across dozens of agency portals, each with its own format, API rate limit, and citation convention. CalcFi Open Data normalizes the most-used 34 series into a single CC-BY mirror so that researchers, journalists, and tool-builders can pull a consistent slice without re-implementing eight different ETL scrapers. The methodology paper is available as a working paper and under review at SSRN.
Getting started¶
-
:material-package-variant: Install
Install the Python client.
-
:material-rocket-launch: Quick start
Load your first series.
-
:material-database: Series catalog
The 34 indicators, primary sources, and observation counts.
-
:material-api: API reference
Every function, every parameter.
Author¶
Jere Salmisto — Independent researcher · Founder, calcfi.app
- ORCID: 0009-0000-0916-8684
- Hugging Face: @iizy
- GitLab: @jere.salmisto
- Academia.edu: JereS6