Write the study ARD as JSON, YAML or XPT

A copy of the ARD for a reader that does not use R; the rds the ARD program writes stays the record. What each format keeps:

Usage

tfl_write_ard(
  ard,
  path,
  format = NULL,
  shape = c("rows", "nested"),
  xpt_version = 8L
)

Arguments

ard

An ARD (the study ARD, tfl_build_ard(), or any cards ARD).

path

The file to write.

format

"json", "yaml", "xpt" or "rds"; NULL takes it from the file's extension.

shape

For JSON / YAML: "rows" or "nested".

xpt_version

For XPT: 8 (default) or 5.

Value

path, invisibly, with the attribute lost: what the format could not keep.

Details

format	shape	what is lost
rds	the ARD	nothing
JSON / YAML	"rows" (default): one record per statistic, with each variable's levels in order and the column types	the formatting functions (fmt_fun); stat_fmt keeps their result
JSON / YAML	"nested": cards' cards::as_nested_list()	the formatting functions, stat_label, the id columns, the column types; the levels' order is only the keys' (not guaranteed)
XPT	a flat table (version 8; 5 shortens the column names, see below)	factors and their levels, the formatting functions, warning / error; stat becomes text when an analysis gives text as well as numbers

A format that loses something says so in a warning. tfl_read_ard() reads "rows" back (and XPT as far as it can). XPT version 5 allows 8-character column names: group1_level is GRP1LVL, variable_level VARLVL, stat_name STATNAME, stat_label STATLBL, output_id OUTID, analysis_id ANID, population_id POPID, variable VARIABLE, context CONTEXT, stat_fmt STATFMT.

See also

tfl_read_ard()