barchart-data

An installable Python client for Barchart public market data. The project is commodity-focused and is designed to grow to other asset classes and data families without storing credentials or downloaded market data.

Install

Use the repository version directly. This is the one supported installation for the browser history client:

py -3.13 -m pip install --upgrade --force-reinstall "barchart-data[demo,browser] @ git+https://github.com/tagomatech/barchart-data.git@main"
py -3.13 -m playwright install chromium

Confirm that Python is importing the GitHub checkout:

py -3.13 -c "import barchart_data; print(barchart_data.__version__); print(barchart_data.__file__)"

Historical data

There is one historical acquisition call. It opens the official Barchart interactive chart in a normal headed browser session and observes the history response that the page naturally requests. By default the browser starts minimized, remains visible in the taskbar, and does not take focus from the application you are using. The response is normalized in memory and the browser is closed automatically.

from barchart_data import download_history

result = download_history("ZCU26")  # Barchart, visible browser, minimized
history = result.frame

print(history.tail())
print(result.source)
assert result.path is None

No Download button needs to be pressed. No file is created by default. The chart's own default range and interval are used, and the returned source URL provides provenance. result.path is always None for this browser workflow.

Progress is printed to the terminal in real time. Maximum verbosity is the default:

result = download_history("ZCU26", verbosity=3)

Use verbosity 2 for progress only, 1 for errors only, or 0 for no output. At maximum verbosity, the package reports browser startup, navigation, candidate responses, parsing, cleanup, and a heartbeat while it is waiting.

This is browser automation of the official page, not an attempt to bypass Barchart controls. The package does not automate sign-in, replay tokens, rotate proxies, disguise automation, or call a separate anonymous Barchart historical endpoint. A Barchart/CloudFront denial or browser-session failure raises BarchartInteractiveChartError; the package does not substitute data from another provider. If a notebook or VS Code kernel already has an active asyncio loop, the synchronous Playwright session is isolated in a worker thread automatically.

The headless option remains available for environments where it is accepted, but the default is intentionally headless=False, start_minimized=True because some Barchart sessions distinguish a normal browser from a headless one:

result = download_history(
    "ZCU26",
    headless=False,
    start_minimized=True,
    verbosity=3,
)

The same function accepts another public Barchart asset class:

result = download_history("AAPL", asset_class="stocks")

Public quote data

Public quote and profile fields embedded in overview pages are available without a login:

from barchart_data import PublicBarchartClient

client = PublicBarchartClient()
quote = client.quote("ZCU26")
profile = client.profile("ZCU26")

Requests are deliberately paced and overview pages are cached briefly. Reuse one client instance for a research job and keep the default request interval or increase it for larger jobs.

Analysis and demos

The package includes:

• OHLCV normalization and quality reports;
• a catalog of grains, oilseeds, livestock, vegetable oils, ICE Canada, and Euronext Matif roots;
• contract-aware nearby and continuous-futures utilities;
• rebasing and comparison helpers;
• Screamer indicators for streaming-style analysis.

The main demonstration is notebooks/commodities/corn_futures_demo.ipynb. It uses the real CME/CBOT September 2026 Corn contract, ZCU26, and renders candlesticks, volume, Bollinger Bands, ATR, RSI, and a rolling volume mean. Its source label records the Barchart chart response used for the in-memory result.

The agriculture comparison is in notebooks/commodities/agriculture_portfolio_demo.ipynb. The equity example is in notebooks/equities/equity_research_demo.ipynb. These notebooks are analysis demonstrations; the single historical downloader above is the source-ingestion path.

The legacy Barchart package remains under Barchart for compatibility with the original futures-builder imports. New code should import from barchart_data.

Tests

py -3.13 -m unittest discover -s tests -v

Contributions

Contributions are welcome. Issues and pull requests are useful for bug fixes, new asset classes, notebook ideas, documentation, and compatibility reports. Please do not commit credentials or downloaded market data.

Data rights

This project is a client and transformation toolkit, not a data redistribution service. Use Barchart data only in accordance with the terms and permissions that apply to your account and use case.

License

MIT