Skip to contents

Classify every column of a data frame under deployment rules

Source: R/20-fx-deploy.R
fx_classify_df.Rd

Batch driver over fx_classify_column(). Resolves the detector list (or the feature manifest) ONCE and reuses it across columns, then returns a tidy one-row-per-column prediction table. Feature rows are attached as an attribute ("features") for callers that want them.

Usage

fx_classify_df(
  df,
  cols = names(df),
  n_rows = 500L,
  blocks = NULL,
  features_used = NULL,
  cardinality = c("full", "sample"),
  cascade = NULL,
  lexicons = default_lexicons(),
  threshold = 0.8,
  seed = 1L
)

Arguments

df

A data frame (or list of columns).

cols

Which columns to classify; defaults to all.

n_rows, blocks, features_used, cardinality, cascade, lexicons, threshold, seed

Passed through to fx_classify_column().

Value

A data frame with column, source, guess, confidence, data_type, data_class, n, n_unique, n_sampled. The per-column feature rows are bound into a data frame stored in attr(., "features").

See also

fx_classify_column()