How to generate QC-ready result data frames from tables
Davide Garolini
2024-11-20
Source:vignettes/ard_how_to.Rmd
ard_how_to.RmdDisclaimer
This vignette is a work in progress.
Create the example table
First of all we need a table to retrieve all the necessary information. Borrowing one from the vignette about clinical trials.
library(rtables)
# Loading required package: formatters
#
# Attaching package: 'formatters'
# The following object is masked from 'package:base':
#
# %||%
# Loading required package: magrittr
#
# Attaching package: 'rtables'
# The following object is masked from 'package:utils':
#
# str
ADSL <- ex_adsl # Example ADSL dataset
mean_sd_custom <- function(x) {
mean <- mean(x, na.rm = FALSE)
sd <- sd(x, na.rm = FALSE)
rcell(c(mean, sd), label = "Mean (SD)", format = "xx.x (xx.x)")
}
counts_percentage_custom <- function(x) {
# browser()
cnts <- table(x)
out <- lapply(cnts, function(x) {
perc <- x / sum(cnts)
rcell(c(x, perc), format = "xx. (xx.%)")
})
in_rows(.list = as.list(out), .labels = names(cnts))
}
lyt <- basic_table(show_colcounts = TRUE, colcount_format = "N=xx") %>%
# split_rows_by("STRATA1", split_fun = keep_split_levels(c("A"))) %>%
# split_cols_by("STRATA2") %>%
split_cols_by("ARM", split_fun = keep_split_levels(c("A: Drug X", "B: Placebo"))) %>%
analyze(vars = "AGE", afun = mean_sd_custom) %>%
analyze(vars = "SEX", afun = counts_percentage_custom)
tbl <- build_table(lyt, ADSL)
tbl
# A: Drug X B: Placebo
# N=134 N=134
# ————————————————————————————————————————————
# AGE
# Mean (SD) 33.8 (6.6) 35.4 (7.9)
# SEX
# F 79 (59%) 77 (57%)
# M 51 (38%) 55 (41%)
# U 3 (2%) 2 (1%)
# UNDIFFERENTIATED 1 (1%) 0 (0%)Convert the table to a result data frame
The as_result_df function is the one that converts a
table to a result data frame. The result data frame is a data frame that
contains the result of the summary table and is ready to be used for
quality control purposes. This may differ for different standard and
lets see how to produce different outputs. Final goal is having clearly
one result for row. Lets play with different options.
as_result_df(tbl)
# avar_name row_name label_name row_num is_group_summary
# 1 AGE Mean (SD) Mean (SD) 2 FALSE
# 2 SEX F F 4 FALSE
# 3 SEX M M 5 FALSE
# 4 SEX U U 6 FALSE
# 5 SEX UNDIFFERENTIATED UNDIFFERENTIATED 7 FALSE
# node_class A: Drug X B: Placebo
# 1 DataRow 33.768657, 6.553326 35.432836, 7.895414
# 2 DataRow 79.0000000, 0.5895522 77.0000000, 0.5746269
# 3 DataRow 51.000000, 0.380597 55.0000000, 0.4104478
# 4 DataRow 3.00000000, 0.02238806 2.00000000, 0.01492537
# 5 DataRow 1.000000000, 0.007462687 0, 0
as_result_df(tbl, data_format = "strings")
# avar_name row_name label_name row_num is_group_summary
# 1 AGE Mean (SD) Mean (SD) 2 FALSE
# 2 SEX F F 4 FALSE
# 3 SEX M M 5 FALSE
# 4 SEX U U 6 FALSE
# 5 SEX UNDIFFERENTIATED UNDIFFERENTIATED 7 FALSE
# node_class A: Drug X B: Placebo
# 1 DataRow 33.8 (6.6) 35.4 (7.9)
# 2 DataRow 79 (59%) 77 (57%)
# 3 DataRow 51 (38%) 55 (41%)
# 4 DataRow 3 (2%) 2 (1%)
# 5 DataRow 1 (1%) 0 (0%)
as_result_df(tbl, simplify = TRUE)
# label_name A: Drug X B: Placebo
# 1 Mean (SD) 33.768657, 6.553326 35.432836, 7.895414
# 2 F 79.0000000, 0.5895522 77.0000000, 0.5746269
# 3 M 51.000000, 0.380597 55.0000000, 0.4104478
# 4 U 3.00000000, 0.02238806 2.00000000, 0.01492537
# 5 UNDIFFERENTIATED 1.000000000, 0.007462687 0, 0
as_result_df(tbl, simplify = TRUE, keep_label_rows = TRUE)
# label_name A: Drug X B: Placebo
# 1 AGE NA NA
# 2 Mean (SD) 33.768657, 6.553326 35.432836, 7.895414
# 3 SEX NA NA
# 4 F 79.0000000, 0.5895522 77.0000000, 0.5746269
# 5 M 51.000000, 0.380597 55.0000000, 0.4104478
# 6 U 3.00000000, 0.02238806 2.00000000, 0.01492537
# 7 UNDIFFERENTIATED 1.000000000, 0.007462687 0, 0
as_result_df(tbl, simplify = TRUE, keep_label_rows = TRUE, expand_colnames = TRUE)
# label_name A: Drug X B: Placebo
# 1 <only_for_column_names> A: Drug X B: Placebo
# 2 <only_for_column_counts> 134 134
# 3 AGE NA NA
# 4 Mean (SD) 33.768657, 6.553326 35.432836, 7.895414
# 5 SEX NA NA
# 6 F 79.0000000, 0.5895522 77.0000000, 0.5746269
# 7 M 51.000000, 0.380597 55.0000000, 0.4104478
# 8 U 3.00000000, 0.02238806 2.00000000, 0.01492537
# 9 UNDIFFERENTIATED 1.000000000, 0.007462687 0, 0Now lets get the final ARD output. This is the one that
is ready to be used for quality control purposes.
as_result_df(tbl, make_ard = TRUE)
# group1 group1_level variable variable_level variable_label stat
# 1 ARM A: Drug X AGE Mean (SD) Mean (SD) 33.76865....
# 2 ARM A: Drug X SEX F F 79, 0.58....
# 3 ARM A: Drug X SEX M M 51, 0.38....
# 4 ARM A: Drug X SEX U U 3, 0.022....
# 5 ARM A: Drug X SEX UNDIFFERENTIATED UNDIFFERENTIATED 1, 0.007....
# 6 ARM B: Placebo AGE Mean (SD) Mean (SD) 35.43283....
# 7 ARM B: Placebo SEX F F 77, 0.57....
# 8 ARM B: Placebo SEX M M 55, 0.41....
# 9 ARM B: Placebo SEX U U 2, 0.014....
# 10 ARM B: Placebo SEX UNDIFFERENTIATED UNDIFFERENTIATED 0, 0