framedf looks at a data frame the way an experienced analyst does in
the first thirty seconds: it infers what each column means, screens
every sensible pair for relationships, and lists the anomalies worth
reading first. The output is qualitative: direction, strength, and
stability, expressed as sentences.
Entry points
frame()build a triage object from a data frame.
print.frame_df()narrative overview.
relationships()meaningful, suspicious, structural, ignored pairs.
anomalies()per-column oddities.
details()analysis mode, roles, skipped rules, backend.
framedf_settings()tunable thresholds.
Author
Maintainer: Gilles Colling gilles.colling051@gmail.com (ORCID) [copyright holder]