Skip to contents

From rtables and tern

Source: vignettes/from-rtables.Rmd
from-rtables.Rmd

The second path. This article is about bringing a table you already have. When the statistics are in a cards / cardx ARD, the recommended route is a table plan – see Get started and Tables from an ARD.

An rtables table – including the tables tern’s analysis functions build, which are rtables objects – goes to RTF through as_rtftables(), as it is.

From an rtables table to RTF

as_rtftables() also accepts any rtables VTableTree – including the tables produced by tern analysis functions, which are rtables objects. They are read through formatters::matrix_form(), so nested column splits become spanning headers, the row-label stub keeps its indentation, and the titles / footers / referential footnotes flow through. In-cell footnote marks {N} are rewritten to ^{N}.

library(rtables)
#> Loading required package: formatters
#> 
#> Attaching package: 'formatters'
#> The following object is masked from 'package:base':
#> 
#>     %||%
#> 
#> Attaching package: 'rtables'
#> The following object is masked from 'package:utils':
#> 
#>     str
Show code 

lyt <- basic_table(title = "Table 2", main_footer = "Source: ADSL") |>
  split_cols_by("ARM") |>
  split_rows_by("RACE") |>
  analyze("AGE", afun = function(x) {
    in_rows("Mean (SD)" = rcell(c(mean(x), sd(x)), format = "xx.x (xx.x)"))
  })

set.seed(1)
adsl <- data.frame(
  ARM  = rep(c("Placebo", "Active"), each = 30),
  RACE = rep(c("WHITE", "BLACK", "ASIAN"), 20),
  AGE  = rnorm(60, 45, 8)
)
tern_like <- build_table(lyt, adsl)

doc <- rtf_document() |>
  rtf_section(page = 1, header = NULL, footer = NULL) |>
  rtf_tables(as_rtftables(tern_like))

out <- tempfile(fileext = ".rtf")
generate_rtfreport(doc, out, overwrite = TRUE)
file.exists(out)
#> [1] TRUE

The rtables read tokens are "col_header", "alignment", "spanning", "titles", "footnotes", "indent", "footnote_marks"; read = TRUE (default) reads them all.

Splitting a long table across pages

as_rtftables() pages the table: split and max_rows say how, as for any source – see Paginating with as_rtftables().

Where next