Generate a Single Column by Inferred Type

Produces a vector of length n with realistic simulated values based on the resolved column type. For columns that look like foreign keys (e.g. subjid, invid), values are sampled from existing parent domain data when available in context$data.

Usage

generate_column_by_type(col_name, col_spec = list(), n, context = list())

Arguments

col_name

Character. The column name.

col_spec

List. Spec metadata for the column.

n

Integer. Number of rows to generate.

context

List with element data (named list of data.frames already generated for earlier domains), and start_date/end_date (Date or character coercible to Date).

Value

A vector of length n.

Lifecycle

Examples

# Generate date values
generate_column_by_type("ae_stdt", list(type = "date"), n = 5,
  context = list(start_date = "2023-01-01", end_date = "2023-12-31"))
#> [1] "2023-06-22" "2023-03-04" "2023-12-24" "2023-02-16" "2023-07-20"

# Generate integer values using name-pattern inference
generate_column_by_type("subject_count", n = 10)
#>  [1] 69  5 24 79 77  2 62 55 43 62

# Generate character values with explicit type
generate_column_by_type("category", list(type = "character"), n = 5)
#> [1] "CATE-0001" "CATE-0002" "CATE-0003" "CATE-0004" "CATE-0005"