Blank out consecutive repeated values in a finished table

Suppresses consecutive repeated values in the chosen columns of a built rtftable(), keeping only the first row of each run; suppressed cells become NA. No rows are removed – only the display value is blanked (the renderer draws NA as an empty cell). This is the classic "don't repeat the group label on every row" layout for clinical listings.

Usage

collapse_repeats(x, cols)

# S3 method for class 'rtftable'
collapse_repeats(x, cols)

# S3 method for class 'list'
collapse_repeats(x, cols)

# Default S3 method
collapse_repeats(x, cols)

Arguments

x: An rtftable(), or a list of them (pages from as_rtftables()).
cols: Columns to collapse: character names and/or integer positions in the table's final body columns. Mix names and positions with a list() (a bare c() would coerce the numbers to strings). Processed in the supplied order (outermost group first).

Value

An object of the same shape as x (rtftable, or list of pages), with suppressed cells set to NA in the selected columns.

Details

Like the other post-hoc verbs it is an S3 generic with an rtftable method and a list method. Given a list of pages (an as_rtftables() result) it collapses each page independently, so a run restarts at every page break – the first row of each page shows its value again. This makes collapse_repeats(as_rtftables(x, ...), cols) equivalent to as_rtftables(x, collapse_repeats = cols, ...).

Suppression is hierarchical: columns are processed in the order given, so a change in any earlier-listed column resets the run of every later one.

Examples

tbl <- rtftable(data.frame(
  grp = c("A", "A", "A", "B", "B"),
  sub = c("x", "x", "y", "x", "x"),
  n   = 1:5,
  stringsAsFactors = FALSE
))
out <- collapse_repeats(tbl, cols = c("grp", "sub"))
out$data
#>    grp  sub n
#> 1    A    x 1
#> 2 <NA> <NA> 2
#> 3 <NA>    y 3
#> 4    B    x 4
#> 5 <NA> <NA> 5