Skip to contents

Select header columns by a column-name segment

Source: R/col_header.R
col_key.Rd

Returns a column selector for col_cell() pos =: a function that is handed the data column names when the header is attached to a table and answers which columns the cell covers.

Usage

col_key(key, sep = NULL, part = 1L)

Arguments

key

Character. One or more segment values to match.

sep

Single string, or NULL (default) to auto-detect the separator that occurs in the column names ("____", then tfrmt's delimiter).

part

Integer. Which segment to compare, counting from the left (1, the default, is the value before the first separator). A negative value counts from the right, so part = -1 is the last segment.

Value

A function of the column names, tagged for use as col_cell(pos =).

Details

Data columns produced by a wide pivot are usually named by rule, as <group><sep><sub-group> – for example "Placebo____Day 1". col_key() splits each name on sep and keeps the columns whose part-th segment is one of key, so a spanning header cell is written in terms of the value it labels rather than of column numbers that shift whenever a column is added, dropped or reordered.

The separator is the one already used to rebuild a spanning header from delimited column names (see as_rtftables(header_sep = )) and to split a table column-wise (see paginate_cols() by = ), so the three stay in step.

A selector may also be written directly as a plain function of the column names, which covers glob and regular-expression matching without any further vocabulary:

col_cell(function(nm) grepl(glob2rx("Placebo____*"), nm), "Placebo")
col_cell(function(nm) grepl("^Placebo____", nm),          "Placebo")

Such a function may return a logical vector the length of the column names, an integer vector of positions, or a character vector of column names. It is an error for a selector to match no column, or to match columns that are not adjacent – a header cell can only span a contiguous range.

See also

col_cell(), rtf_col_header(), paginate_cols()

Examples

# Data columns: Placebo____Day 1, Placebo____Day 7, TAK-003____Day 1, ...
col_cell(col_key("Placebo"), "Placebo\n(N=60)")
#> <col_cell pos=col_key("Placebo") label="Placebo
#> (N=60)">

# Every "Day 1" column, whichever arm it belongs to:
col_cell(col_key("Day 1", part = 2), "Day 1")
#> <col_cell pos=col_key("Day 1") label="Day 1">