Hugging Face's logo

Datasets:
SADHON
/
sec-10-company-accounting-panel

Modalities:
Tabular
Text
Formats:
csv
Size:
< 1K
Tags:
accounting
finance
sec
edgar
xbrl
reproducibility
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
sec-10-company-accounting-panel / README.md
SADHON's picture
SADHON
Validated GitHub sync for SEC 10-company Phase 2
81c99b5 verified
preview code
|
Raw History Blame Contribute Delete
2.17 kB
metadata 
license: other
tags:
  - accounting
  - finance
  - sec
  - edgar
  - xbrl
  - reproducibility
  - financial-statements
configs:
  - config_name: panel
    data_files:
      - split: train
        path: sec_10_company_panel.csv
    default: true
  - config_name: summary
    data_files:
      - split: train
        path: phase2_summary.csv

SEC 10-Company Accounting Panel

A reproducible annual accounting dataset sourced from the official U.S. SEC CompanyFacts API.

Coverage

MSFT, AAPL, GOOGL, AMZN, NVDA, META, JPM, WMT, XOM and TSLA.

The current verified snapshot contains the latest three annual 10-K comparative periods used for each company and normalizes reported USD values to millions. Live CompanyFacts refresh code is maintained in the canonical GitHub repository.

Dataset Viewer

This repository exposes two separate Hugging Face configurations so files with different schemas are not incorrectly merged:

  • panel — company-year accounting observations from sec_10_company_panel.csv
  • summary — latest-period analytical metrics from phase2_summary.csv

This structure follows Hugging Face's recommended multiple-configuration pattern for independent files.

Files

  • sec_10_company_panel.csv
  • phase2_summary.csv
  • sec_refresh_manifest.json
  • README.md

Canonical source

GitHub is the source of truth:

https://github.com/Saehon/Saeid-Homayoun/tree/main/open-data/sec-10-company-phase2

The GitHub workflow validates the dataset before every Hugging Face upload.

Research identity

Saeid Homayoun — Accounting & Auditing Scholar integrating AI into accounting, auditing, assurance, finance and sustainability research.

Method boundary

Cross-industry standardized XBRL concepts are not perfectly uniform. Missing values are retained rather than silently imputed. The provenance manifest records the concept selected for every company-period-field combination.

This is research/educational infrastructure and not audit assurance or investment advice.