What is an ARD?
An Analysis Results Data (ARD) provides statistical analysis results in a structured, machine-readable format. It provides a way to store analysis outcomes in a consistent manner, making it easier to reuse and integrate those results into reports, tables, and figures.
Key Features of an ARD
Structured Data: The ARD format organizes statistical results, ensuring they are easily interpreted and accessible for further analysis.
Reusability: Once created, an ARD can be reused for future reporting needs, eliminating the need for repetitive data processing.
No Layout Description: The ARD focuses on the data itself, not on the visual layout of tables or figures. This means you can use the data to generate reports in various formats.
How to Create an ARD
ARDs can easily be created using the powerful cards package. For more details, visit the cards documentation.
If you’re new to ARDs or the cards package, check out the following Other Resources.
For information on the Analysis Results Standard (ARS), visit this CDISC Webinar on ARS.
Dataset Structure and Column Requirements
We expect a long-format dataset, with one record per computed value.
Required columns are as follows:
[Optional] Grouping columns: One or more columns containing grouping values.
Label column: A single column containing row label values.
Column value columns: One or more columns containing the values defining each column in your table. These are essentially additional grouping variables.
Param column: A single column that labels distinct types of values.
Value column: A single column containing the computed raw data values. Note: these must be numeric.
[Optional] Sorting columns: One or more columns with numeric values to control row ordering.
For information on transforming ARD output for use in {tfrmt} see the ARD-first tables using {cards} vignette.