Fieldwork v0.1.0

Fieldwork 0.1.0 is the initial alpha release: a Python toolkit for exploring unfamiliar pandas dataframes through patterns, source-row evidence, and reproducible investigations.

• Discover availability families, signatures, implications, and exceptions with explicit row or entity counting units.
• Examine exact and approximate dependencies, candidate grains, and compatible populations for sparse keys.
• Choose census paths using five objectives, measured explanations, and scope-preserving handoffs.
• Select complete matching populations, inspect source rows, and browse connected feature evidence.
• Save strict JSON evidence, render standalone SVG/HTML, reapply recipes, and compare delivery populations.

Requires Python 3.11–3.14. NumPy and pandas are the only required runtime dependencies. Install with pip install fieldwork==0.1.0.

This release analyzes one dataframe at a time. Automatic related-table discovery is not included. Search and display budgets are reported explicitly; topology exports suppress quantities but retain structural names and qualitative analysis units.

The release includes a wheel, source distribution, and generated documentation assets. The source and build include MIT/extraction attribution and the bundled build fonts’ OFL license. See the worked investigation and user documentation.