Skip to contents

Infer a starter dictionary from a data frame

infer_dictionary.Rd

Proposes a starter dictionary (column dictionary schema) from raw data by guessing column types, roles, and basic metadata.

Usage

infer_dictionary(
  df,
  guess_types = TRUE,
  dataset_id = "dataset-1",
  table_id = "table_1",
  seed_semantics = FALSE,
  semantic_sources = c("smn", "gcdfo", "ols", "nvs"),
  semantic_max_per_role = 1,
  seed_verbose = TRUE,
  seed_codes = NULL,
  seed_table_meta = NULL,
  seed_dataset_meta = NULL
)

Arguments

df

A data frame or tibble to analyze. Or, when provided as a named list of data frames, infer_dictionary() infers each table and returns a combined dictionary.

guess_types

Logical; if TRUE (default), infer value types from data.

dataset_id

Character; dataset identifier (default: "dataset-1").

table_id

Character; table identifier (default: "table_1").

seed_semantics

Logical; if TRUE, run suggest_semantics() and attach the resulting semantic_suggestions attribute to the returned dictionary.

semantic_sources

Character vector of vocabulary sources passed to suggest_semantics() when seed_semantics = TRUE. Default: c("smn", "gcdfo", "ols", "nvs").

semantic_max_per_role

Maximum number of suggestions retained per I-ADOPT role when seeding suggestions. Default: 1.

seed_verbose

Logical; if TRUE, print a short progress message while seeding semantic suggestions.

seed_codes

Optional codes.csv-style tibble forwarded to suggest_semantics() when seed_semantics = TRUE.

seed_table_meta

Optional tables.csv-style tibble forwarded to suggest_semantics() when seed_semantics = TRUE.

seed_dataset_meta

Optional dataset.csv-style tibble forwarded to suggest_semantics() when seed_semantics = TRUE.

Value

A tibble with dictionary schema columns in canonical Salmon Data Package order: dataset_id, table_id, column_name, column_label, column_description, term_iri, property_iri, entity_iri, constraint_iri, method_iri, unit_label, unit_iri, term_type, value_type, column_role, required.

Examples

if (FALSE) { # \dontrun{
df <- data.frame(
  species = c("Coho", "Chinook"),
  count = c(100, 200),
  date = as.Date(c("2024-01-01", "2024-01-02"))
)
dict <- infer_dictionary(df)

# Optional: seed semantic suggestions from vocabulary services
# (SMN is queried first; GCDFO is a distinct DFO-specific source)
dict <- infer_dictionary(
  df,
  seed_semantics = TRUE,
  semantic_sources = c("smn", "gcdfo", "ols", "nvs")
)
suggestions <- attr(dict, "semantic_suggestions")
} # }