New variable IR — new_variable

Builds the intermediate representation for one spec variable: dataset, variable name, ordered [new_step()] pipeline, plus provenance fields (verbatim spec text, confidence, review flag, rationale).

Usage

new_variable_ir(
  dataset,
  variable,
  steps = list(),
  spec_origin = NA_character_,
  confidence = 1,
  needs_human = FALSE,
  rationale = NA_character_
)

Arguments

dataset: Dataset the variable belongs to (e.g. `"ADSL"`).
variable: Variable name (e.g. `"TRTSDTM"`).
steps: List of `aa_step` objects in execution order.
spec_origin: Verbatim derivation text from the spec.
confidence: Classifier confidence in `[0, 1]`.
needs_human: Logical; `TRUE` routes the variable to human review instead of code generation.
rationale: Free-text explanation of the chosen layer chain.

Value

A list of class `aa_variable_ir`.

Details

Construct a variable IR

Examples

v <- new_variable_ir(
  dataset = "ADSL",
  variable = "TRTSDTM",
  steps = list(new_step("assign", list(target = "TRTSDTM", literal = "2026-01-01")))
)