Generates DIL-backed synthetic BLADE tables from the business
source-of-truth spine. The business spine is created first if it is
not already present in the active run directory.
Generated tables follow the April 2026 BLADE DIL schemas and derive
generic time fields from each table's public available periods. The
optional Appendix key products are the public BLADE correspondences;
_system/plida-blade-link.parquet is a local fplida linkage
artifact for person-to-business training examples.
Usage
generate_blade(
business_spine = NULL,
spine = NULL,
seed = 42L,
tables = "all",
n_businesses = NULL,
output_dir = NULL,
format = c("parquet", "csv"),
return_data = FALSE,
sample_rate = 1,
max_rows = 10000L,
include_keys = TRUE,
years = NULL
)Arguments
- business_spine
Data.frame or NULL. If NULL, uses the existing
_system/business-spine.parquetor creates one from the person spine.- spine
Optional person spine used when the business spine must be created.
- seed
Integer. Random seed.
- tables
"all"or a vector of BLADE table numbers, table names, or generated product names.- n_businesses
Optional business universe size when the business spine must be created.
- output_dir
Character or NULL. Base output directory.
- format
Character. Currently parquet only.
- return_data
Logical. If TRUE, returns generated data.frames.
- sample_rate
Numeric row sampling rate applied to BLADE survey tables. Administrative/key tables are generated for the full business spine so every
bnis present in administrative BLADE records.- max_rows
Integer maximum rows per emitted BLADE survey table. Administrative/key tables ignore this cap.
- include_keys
Logical. If TRUE, writes BLADE Appendix 1 and Appendix 2 key products for
bn/id/bg_idandcn/bncorrespondences.- years
Integer vector or NULL. Retain reporting periods with these ending years. NULL retains each table's full supported range.
Value
If return_data = TRUE, a named list of data.frames keyed by
BLADE product name. Otherwise an invisible metadata list.
Dataset and variable information
The ABS BLADE overview
gives information about this dataset. Use dataset_info("BLADE") for dataset
information. Use variable_info("BLADE") for variables, sources, value
support, and topic tags.