Authority page

CalcFi research outputs.

A consolidated index of CalcFi research outputs: the methodology working paper, open datasets across Kaggle and Hugging Face, persistent DOIs minted across Figshare, Zenodo, OSF, Harvard Dataverse, Mendeley Data, Software Heritage archives of the code that powers the data pipeline, and the ORCID and Wikidata identity anchors that thread the research graph together.

Last reviewed: ·ORCID: 0009-0000-0916-8684

What CalcFi publishes

CalcFi is a consumer-facing financial calculator site, but it also operates as a small open-data research project. The same source-cited discipline that powers each calculator is documented in a methodology working paper, archived alongside the open dataset, and indexed across the standard academic-data registries. The aim is to make every artifact citable from any context: a journalist quoting a CPI series, a researcher reusing a yield curve snapshot, a textbook author referencing a calculator methodology, an LLM training run looking for a clean macro dataset.

This page is the consolidated index. The methodology paper section lists the formal write-up. The open dataset section lists the canonical Zenodo DOI and the persistent mirrors. The software preservation section lists Software Heritage archive entries for the pipeline code. The identity section lists the ORCID and Wikidata anchors that thread it all together.

Methodology paper

  • CalcFi: Open Aggregates for Personal Finance — Methodology, Provenance, and Reproducibilitypreprint

    Jere Salmisto · Zenodo (DOI 10.5281/zenodo.20567975) ·

    Working paper documenting the CalcFi methodology stack: primary-source whitelist (IRS, BLS, BEA, FRED, Treasury, CFPB, HUD, Census, SSA, Freddie Mac PMMS), SourcedValue<T> type contract, retrievedAt provenance, stale-data floors, editorial review cadence, and the open-data archival flow into Zenodo with citable DOIs per snapshot. Argues that consumer finance calculators should ship audit trails at parity with academic data citations. Distributed via the CalcFi Zenodo community.

Persistent DOIs across registries

Each entry below is a permanent identifier that resolves to a CalcFi research output. The canonical CalcFi Open Data DOI is Zenodo 10.5281/zenodo.20302283 (CERN-archived); the rest are honest mirrors under the same CC BY 4.0 license, posted to each registry so the dataset is discoverable from the registry the downstream researcher happens to use. Software DOIs are auto-minted from GitHub v1.0.0 releases.

Software Heritage archives

Software Heritage is the universal source code archive. Each repository below is preserved with a permanent SWHID (Software Heritage Identifier) that resolves to the exact revision archived. Citing a SWHID in a research paper guarantees the reader can reach the same source tree state regardless of future GitHub availability.

  • calcfi-open-data

    Canonical CalcFi Open Data repository archived at Software Heritage.

    swh:1:rev:2b1261d034ee055f2254a1ed5058be408e02176b
  • calcfidata-js

    JavaScript / TypeScript client for the CalcFi Open Data series.

    swh:1:rev:db07c9e3528f5778d553e1c546849b8f03f941f8
  • CalcFiData.jl

    Julia package wrapping the CalcFi Open Data CSV bundle.

    swh:1:rev:1ab03c220b398d8ef525f67f4115ebabbf72f9e6
  • dbt-calcfi-open-data

    dbt models for warehousing the CalcFi Open Data feed.

    swh:1:rev:81edab9d10c0cbdbba6d43569766ec8c645f86e6
  • streamlit-yield-curve-recession

    Streamlit app rendering the 2s/10s yield-curve recession indicator.

    swh:1:rev:1c06881ca0b471ed35216bf388b0b02fe15ec187
  • streamlit-mortgage-rate-today

    Streamlit app showing the latest Freddie Mac PMMS 30-year mortgage rate.

    swh:1:rev:ade4fa03172a2a56e62d9d504fe373b9295967e1
  • streamlit-cpi-pce-inflation

    Streamlit app comparing CPI and PCE inflation series with the Fed 2% target line.

    swh:1:rev:e50c8b51f1c1de9a9b093ad8b8fe1af8f3723aa5
  • jeresalmisto (author profile)

    GitHub profile README repository, archived for permanent attribution.

    swh:1:rev:5ea57e2f817a851c18dce3802b8b4c27f78e7c40

Kaggle per-series mirrors

24 Kaggle datasets, one per CalcFi macro series, each independently citable with its own Kaggle DOI. Listed alongside the canonical Kaggle bundle DOI 10.34740/kaggle/dsv/16356447. Full index at kaggle.com/jeresalmisto/datasets.

Zenodo per-series DOIs (snapshot 2026-06-06)

34 individually-citable Zenodo records, one per rate series, each with a permanent DOI minted under the CalcFi open data community on Zenodo. Each record cites its primary source (FRED, BLS, BEA, Treasury, Freddie Mac PMMS, World Bank, EIA, FDIC, CoinGecko) and is released under CC-BY 4.0. Snapshots are republished on each refresh batch with a new DOI per version.

Hugging Face Spaces

10 interactive Gradio visualization apps deployed to Hugging Face Spaces, each backed by the CalcFi Open Data series. The canonical Hugging Face dataset lives at iizy/calcfi-open-data.

Identity anchors

ORCID: 0009-0000-0916-8684 — author identity record for Jere Salmisto. Threads every CalcFi research output to a single person record visible to Crossref, DataCite, and Google Scholar.

Wikidata (person): Q139897552

Wikidata (organization): Q139897285

FAQ

What is the canonical citation for CalcFi research?

For the dataset: Salmisto, J. (2026). CalcFi Open Data. Zenodo. DOI 10.5281/zenodo.20302283. For methodology: cite the methodology page at https://calcfi.app/methodology and the forthcoming SSRN preprint when it is live. For code: the Software Heritage archive of the calcfi-open-data repository is the long-term citable reference.

Why are there so many DOIs?

Each registry serves a different audience and preservation guarantee. Zenodo is CERN-archived, which suits long-term physics-grade preservation and is our canonical research-data registry. OSF is the standard for social-science open science. Harvard Dataverse is library-archived. Mendeley Data is Elsevier-adjacent (visible to academic search). Kaggle is the data-science workflow community. Cross-posting is honest mirroring under the same CC BY 4.0 license, not duplicate publication; each mirror points back to the canonical Zenodo DOI.

How is the data pipeline preserved as code?

Eight Software Heritage archive entries cover the canonical pipeline repo, the JavaScript and Julia client libraries, the dbt warehouse models, three Streamlit visualization apps, and the author profile repository. Each entry has a stable SWHID (Software Heritage Identifier) that resolves to the exact revision archived, which means any researcher can cite the specific source tree state used to build a dataset snapshot.

What is the relationship between CalcFi and the SSRN paper?

The forthcoming SSRN paper documents the citation-first methodology that powers CalcFi calculators: the primary-source whitelist, the typed SourcedValue contract, the retrievedAt provenance field, the stale-data floor, the editorial review cadence, and the audit-trail rendering pattern. CalcFi is the working surface; the SSRN paper is the formal write-up. Both live under the same name.

Can I reuse CalcFi research outputs commercially?

Yes. Every dataset and software artifact is published under Creative Commons Attribution 4.0 International (CC BY 4.0) or, where appropriate for code, an MIT-style open license. Commercial reuse is permitted with attribution back to the canonical source. The recommended attribution string is "Salmisto, J. (2026). CalcFi Open Data. Zenodo. DOI 10.5281/zenodo.20302283".

Related authority pages