Writes what a summary table (columns = a group, rows = variables) needs,
from the answers the app's "first table" form asks: the dataset in the
data catalog (added when missing), the analysis set (population flag
== "Y", added when missing), one analysis of the numeric variables
(summarized) and one of the rest (counted), the report (a Table, added when missing)
and its tables row (the group as columns, one group per variable) and
variables rows (their order, and the data's labels).
Usage
first_table(
x,
output_id,
path,
data,
population,
group,
variables,
description = NA_character_,
stack = TRUE
)Arguments
- x
A
tflplanner.- output_id
The report.
- path
The data file, relative to the study folder.
- data
Its data (the kinds and labels of its columns).
- population
The analysis set flag column (
SAFFL).- group
The column of the groups (
TRT01A).- variables
The variables of the rows, in order.
- description
The report's description.
- stack
Run the analyses together, with
cards::ard_stack()(aSTACKanalysis with the data, analysis set and group, the numeric and the other variables inside it; it counts the subjects per group and in all), rather than one by one (GROUPN,CONT,CAT). Not when a subject of the analysis set has no group:ard_stack()leaves them out (the result's attributegroup_missingsays how many).