Skip to content

Latest commit

 

History

History
46 lines (29 loc) · 2.43 KB

README.md

File metadata and controls

46 lines (29 loc) · 2.43 KB

scda.test

Internal repository to store tests that utilize datasets from the random.cdisc.data package. Repository tests TLG sample templates to verify functionality and correct output. Template tests were migrated here from the tern package in February 2023.

Installation

Installing Package Dependencies

The scda.test package has several dependencies required to run the tests within this package. Before installing the scda.test repository, please ensure you have the latest versions of the tern and random.cdisc.data packages installed. Installation instructions are provided for each at these links:

Installing scda.test

This repository is internal and must be cloned and installed manually from GitHub.

  1. Clone the repository

    The repository can be downloaded directly from the github.com site as an archive (see the GitHub tutorial on cloning to learn more). Alternatively, Git command line tools offer the same functionality, without the need for manual downloading and unpacking the archive, but require you to authenticate GitHub. You can authenticate using a key pair or a Personal Access Token (PAT). Please refer to GitHub tutorials on connecting to GitHub using SSH or creating and using a PAT.

    • Using a PAT: Input in the Git Bash console, PowerShell, or any Linux shell:

      $ git clone https://github.com/insightsengineering/scda.test.git
      Username: your_username_goes_here
      Password: your_token_goes_here
    • Using SSH: If set up properly, the repository is ready to be cloned by executing:

      git clone https://github.com/insightsengineering/scda.test.git

      This creates a subdirectory scda.test containing the cloned repository.

  2. Build and install

    Native R tools provide a quick way to install a package. Alternatively, to build the package and create an archive run in PowerShell or any Linux shell:

    R CMD build scda.test