tombisho / dstest

DataSHIELD Project Client Environment

This R project template uses the renv for setting up a reproducible DataSHIELD client side R environment. This allows to use the appropriate DataSHIELD client R packages, matching the selected server-side DataSHIELD profiles.

The project's R environment includes the following R packages (and their dependencies):

• DSOpal to connect with OBiBa Opal nodes
• DSMolgenisArmadillo to connect with Molgenis Armadillo nodes
• DSLite to connect with a pure R local node
• dsBaseClient to use the base DataSHIELD functions

Usage

Clone this Repository

Start by creating your project source code repository from this template: see instructions in Creating a repository from a template. You could also just download this folder, but it is a good practice to have source code in a version control system (Git).

Read the Introduction to renv vignette to learn the basics of R environment management.

Open Project in RStudio

1. Open this project in Rstudio by selecting the project.Rproj file.
2. If reported by renv in the R console, run the renv::restore() to install the R packages in their specific version (i.e. as described by the renv.lock file).

Use libPaths() to locate where the R packages are installed.

Adding a new Dependency

1. Install the R package using install.packages() or using renv::install() which allows to manage R packages installed from an archived version, a GitHub repository, Bioconductor repository or a local package folder
2. Load the R package in one of your R source code with library() (see R/script.R as an example)
3. Run renv::snapshot() to save the new R environment in the file renv.lock file

Use Git Branches

Different R environments means different package dependencies, and also potentially different R/DataSHIELD analysis code. In order to maintain/develop different environments in parallel, you can use the Git branches. Upgrading to new versions of dependencies will be facilitated. To make a new branch:

1. Select Git tab and New Branch in Rstudio, or command line git checkout -b branchName
2. Install/update R packages
3. Update analysis R script to connect to the appropriate DataSHIELD server profiles and to fix any breaking changes
4. Run renv::snapshot()
5. Commit changes in Git using Rstudio or command line

See also the vignette Collaborating with renv.