Getting Started
In rlistings: Clinical Trial Style Data Readout Listings

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Introduction

This vignette shows the general purpose and basic functionality of the rlistings R package.

The rlistings R package contains value formatting and ASCII rendering infrastructure for tables and listings useful for clinical trials and other statistical analysis. The core functionality is built on top of the formatters package.

Some of the key features currently available to customize listings created using the rlistings package include:

Key columns
Titles and footnotes

For information on listing column formatting see the Column Formatting vignette. To learn about listing pagination see the Pagination vignette.

The index of all available rlistings functions can be found on the rlistings website functions reference.

The rlistings package is intended for use in creating simple one-dimensional listings. For construction of more complex tables see the rtables package.

Building a Listing

With the basic framework provided in this package, a data.frame object can be easily converted into a listing using the as_listing function with several optional customizations available.

A listing, at its core, is a set of observation-level data which is to be rendered with particular formatting but without any sort of aggregation or further analysis. In practice, this translates to to a classed data.frame (or tbl_df) object with a specialized print method. This means that, unlike tables created with rlistings' sibling package rtables, a listing object is fundamentally the incoming data.frame with a few annotations attached to it.

In the R code below we will give a basic example of how to create an rlistings listing from a pre-processed data frame.

We first load in the rlistings package.

library(rlistings)

For the purpose of this example we will use the dummy ADAE dataset provided within the formatters package as our data frame, which consists of 48 columns of adverse event patient data, and one or more rows per patient.

adae <- ex_adae

Now we will create our listing.

The df parameter sets our data.frame object. The disp_cols argument takes a vector of names of any columns taken from the data frame that should be included in the listing. Column headers are set by the label attribute of each given variable. If there is no label associated with a given variable then the variable name will be taken as a header instead. For this example we will choose 8 arbitrary columns to display - 5 specific to the patient and 3 relating to the adverse event.

Since the dataset consists of 1934 rows in total, we will use the head function to print only the first 15 rows of the listing.

lsting <- as_listing(
  df = adae,
  disp_cols = c("USUBJID", "AETOXGR", "ARM", "AGE", "SEX", "RACE", "AEDECOD", "AESEV"),
)

head(lsting, 15)

In the listing output above you can see that there are several rows associated with each patient, resulting in many instances of repeated values over several columns. This can cleaned up by setting key columns with the key_cols argument.

We can also declare the set of (non-key) display columns by compliment, via the non_disp_col argument. If specifies this argument accepts names of columns which will non be displayed. All other non-key columns are then displayed.

lsting <- as_listing(
  df = adae,
  non_disp_cols = tail(names(adae), 8)
)
head(lsting, 15)

Key Columns

Key columns act as contextual identifiers for observations. Their core behavioral feature is that sequentially repeated values are not displayed when they do not add information.

In practice, this means that each value of a key column is printed only once per unique combination of values for all higher-priority (i.e., to the left of it) key columns (per page). Locations where a repeated value would have been printed within a key column for the same higher-priority-key combination on the same page are rendered as empty space. Note, determination of which elements to display within a key column at rendering is based on the underlying value; any non-default formatting applied to the column has no effect on this behavior.

The key_cols argument takes a vector of column names identifying the key columns for the listing. A listing is always sorted by its key columns (with order defining the sort precedence). Below we specify trial arm and patient ID as key columns to improve readability.

lsting <- as_listing(
  df = adae,
  disp_cols = c("ARM", "AGE", "SEX", "RACE", "AEDECOD", "AESEV"),
  key_cols = c("USUBJID", "AETOXGR")
)

head(lsting, 15)

Titles and Footers

Additionally, an rlistings listing can be annotated with two types of header information (main title and subtitles) and two types of footer information (main footers and provenance footers). A single title can be set using the main_title argument, while one or more subtitles, main footers, and provenance footers can be set by the subtitles, main_footer and prov_footer arguments respectively. These are demonstrated in the following updated listing.

lsting <- as_listing(
  df = adae,
  disp_cols = c("ARM", "AGE", "SEX", "RACE", "AEDECOD", "AESEV"),
  key_cols = c("USUBJID", "AETOXGR"),
  main_title = "Main Title",
  subtitles = c("Subtitle A", "Subtitle B"),
  main_footer = c("Main Footer A", "Main Footer B", "Main Footer C"),
  prov_footer = c("Provenance Footer A", "Provenance Footer B")
)

head(lsting, 15)

Summary

In this vignette you have learned how to implement the basic listing framework provided by the rlistings package to build a simple listing. You have also seen examples demonstrating how the optional parameters of the as_listing function can be set to customize and annotate your listings.

For more information please explore the rlistings website.