Installation¶

AbaQuant requires Python 3.11 or newer.

Standard installation¶

pip install abaquant

For a local source checkout:

python -m pip install -e .

Optional extras¶

AbaQuant keeps provider and visualization dependencies optional so the core numerical package remains lightweight.

Optional dependency groups¶
Extra	Command	Adds
Market data	`python -m pip install -e .[market]`	`yfinance` for optional Yahoo-backed workflows.
Visualization	`python -m pip install -e .[visual]`	Matplotlib and Plotly chart backends.
Development	`python -m pip install -e .[dev]`	Pytest and Ruff.
Documentation	`python -m pip install -e .[docs]`	Sphinx, Furo, copy buttons, and autobuild tooling.

Install multiple extras by separating them with commas:

python -m pip install -e .[market,visual,dev,docs]

Verify the installation¶

python - <<'PY'
import abaquant
print(abaquant.__version__)
from abaquant.derivatives import black_scholes
print(round(black_scholes(100, 100, 0.05, 0.20, 1.0), 6))
PY

Expected version:

1.0.0rc1

Build documentation locally¶

Documentation is written as Sphinx-native reStructuredText and organized into topic subfolders.

python -m pip install -e .[docs]
sphinx-build -b html docs docs/_build/html

Open the generated site:

python -m webbrowser docs/_build/html/index.html

Provider credentials and contact metadata¶

Live data workflows can require provider-specific credentials or contact metadata.

For SEC EDGAR/XBRL usage, set a project-specific contact user agent:

export ABAQUANT_SEC_USER_AGENT="your-app/1.0 your.email@example.com"

For FRED usage, pass an API key to the FRED provider or configure the environment according to your deployment pattern.

Warning

Provider responses can be delayed, adjusted, restated, rate-limited, incomplete, or unavailable. Cache live data deliberately and record provenance for reproducibility.