Reference

Command Line Interface

bdf --help
bdf validate data/sample.bdf.csv
bdf clean data/sample.bdf.csv --out cleaned.bdf.csv --assume-bdf
bdf convert raw/vendor.csv --to output.bdf.csv
bdf detect raw/vendor.csv
bdf plot data/sample.bdf.csv --assume-bdf --save plot.png
bdf meta-jsonld data/sample.bdf.csv --title "My dataset" --description "..." --creator "Name|ORCID|Affiliation"

Python API

bdf.read(source[, plugin, normalize, ...])	Universal reader -> DataFrame.
bdf.parse(source[, plugin, registry_path])	Parse vendor file only (no normalization/validation).
bdf.normalize(df[, plugin])	Normalize a DataFrame to canonical BDF columns.
bdf.validate(obj, *[, report, ...])	Validate a BDF DataFrame, a local file path, an HTTP/HTTPS URL, or a dataset id.
bdf.clean(df, *[, time_fix, outlier, ...])	Clean a BDF-normalized DataFrame.
bdf.plot(*args, **kwargs)	Forward to bdf.visualize.plot(...).
bdf.explore(*args, **kwargs)	Forward to bdf._explore.explore(...).
bdf.detect(path)	Return SniffResult with the best-matching plugin and confidence.
bdf.plugins()	List available plugin ids.
bdf.ingest(source, *[, out_dir, format, ...])	Convert raw vendor files to BDF and validate existing BDF artifacts.
bdf.datasets([registry_path])	Return dataset IDs from the registry.
bdf.load_registry([path])
bdf.get_entry(reg, entry_id)
bdf.build_registry(sources[, registry_dir, ...])
bdf.search(query[, registry_dir, limit])
bdf.sparql(query[, registry_dir])