5.1 Building

5.1.1 Dataset Source

For demonstration purposes, we will use the adsl from the PHUSE Test Datasets Factory repository.

library(haven)

adsl <- read_xpt("https://github.com/phuse-org/TestDataFactory/raw/main/Updated/TDF_ADaM/adsl.xpt")

5.1.2 String Formatting

This section sets the default formats for tables built with Tplyr.

For each variable type (continuous, categorical), we specify the following:

1. Which summary statistic(s) should be computed?
2. How should the summary statistic(s) be formatted for display?

Once set, these formats are carried over to your table automatically.

library(Tplyr)

options(
  # Categorical variable defaults
  tplyr.count_layer_default_formats =
    list(n_counts = f_str("xxx [xx.xx%]", n, pct)),
  
  # Continuous variable defaults
  tplyr.desc_layer_default_formats =
    list('N'         = f_str('xx', n),
         'Mean [SD]' = f_str('xx.xx [xx.xxx]', mean, sd),
         'Median'    = f_str('xx.x', median),
         'Min, Max'  = f_str('xx, xx', min, max))
)

5.1.3 Create

For demonstration purposes, we will select a handful of variables from the adsl dataset to work with.

library(dplyr)

# Initiate Tplyr, specify treatment variable, optional where condition
my_table <-  tplyr_table(adsl, TRT01P, where = SAFFL == "Y") %>%
  
  # Add a total group column
  add_total_group() %>%
  
  # Add individual variables here
  add_layer(group_desc(AGE, b = "Age (years)")) %>%
  add_layer(group_count(AGEGR1, b = "Age Group 1 (years)")) %>%
  add_layer(group_count(SEX, b = "Gender")) %>%
  add_layer(group_count(ETHNIC, b = "Ethnicity")) %>%
  add_layer(group_desc(BMIBL, b = "Baseline Body Mass Index (kg/m2)")) %>%
  
  # Build
  build()