The line that normalizes a table's ARD — tfl_normalize

A table's program makes data from its rows of the study ARD with rtfreporter::normalize_ard() before the plan. What the normalization needs to know is the report's, so it is written from its definitions: the nested variables of its hierarchical analysis (the ARD definition's variables, AEBODSYS | AEDECOD) as its hierarchy – each level's own row, SOC and PT, with its counts – and the row of the hierarchy's overall counts (the table definition's tables$overall_label, "Subjects with at least one TEAE", from the analysis's args over_variables = TRUE) as its overall; and the parallel subgroups stacked into one pair of keys (tables$stack, subgroup = ALL, SEX, AGEGR1) as its stack. A report with none: data <- normalize_ard(ard).

Usage

tfl_normalize_code(
  spec,
  ard_spec = NULL,
  output_id = NULL,
  ard = "ard",
  data = "data"
)

Arguments

spec: A table definition (tfl_read_table_spec()), or the path(s) of its workbook(s).
ard_spec: The ARD definition (tfl_read_ard_spec()), for the hierarchy; NULL for none.
output_id: The table, when the definitions have several.
ard, data: The names of the ARD and of the normalized data.

Value

The line, one string.

Examples

sp <- tfl_table_spec(tables = data.frame(
  output_id = "T-2", cols = "TRTA", rows = "group1 = AEBODSYS",
  overall_label = "Subjects with at least one TEAE"))
ard <- tfl_ard_spec(list(
  datasets = data.frame(dataset = "ADAE", path = "adae.rds"),
  analyses = data.frame(output_id = "T-2", analysis_id = "AE",
                        method = "hierarchical", dataset = "ADAE", by = "TRTA",
                        variables = "AEBODSYS | AEDECOD",
                        args = "over_variables = TRUE")))
tfl_normalize_code(sp, ard, "T-2")
#> [1] "data <- normalize_ard(ard, hierarchy = c(\"AEBODSYS\", \"AEDECOD\"), overall = \"Subjects with at least one TEAE\")"