Client to share data in a MOLGENIS Armadillo DataSHIELD Service.
This library can be used by data managers to share datasets on a MOLGENIS Armadillo server. Researchers can then analyse these datasets and datasets shared on other servers using DataSHIELD. Researchers will only be able to access aggregate information and cannot see individual rows.
The datasets are stored in shared folders on a MinIO file store. The MOLGENIS Armadillo server has access to the file store and can load the data sets into a shielded RServe environment so that researchers can call DataSHIELD analysis methods on the data.
Login to the service.
library('MolgenisArmadillo')
armadillo.login("https://armadillo-url-example.org")
Now you can create a project and upload tables to the project to share them for analysis.
library(datasets)
armadillo.create_project("project")
armadillo.upload_table("project", "folder", iris)
Listing the tables.
armadillo.list_tables("project")
Removing the data from the storage. First you need to remove the content of a project before you can throw away the project.
armadillo.delete_table("project", "folder", "iris")
armadillo.delete_project("project")
For more in depth documentation please check the howto.
The newest version (2.x) of MolgenisArmadillo will be only compatible with Armadillo version 3. If you still use Armadillo 2, you should use the 1.1.13 version of MolgenisArmadillo. You can install this specific version using the following commands:
packageurl <- "https://cran.r-project.org/src/contrib/Archive/MolgenisArmadillo/MolgenisArmadillo_1.1.3.tar.gz"
install.packages(packageurl, repos=NULL, type="source")
If you already installed another version of the package, you might want to run this first:
remove.packages("MolgenisArmadillo")
On windows additional R packages like rtools
might be required in order to install the tar.gz
package from the url.
- To build documentation, do
devtools::document()
- To run all unit tests, do
devtools::test()
- While writing code (or tests), you can use
devtools::load_all()
to quickly “install” the package. - To run a single test file, open it in Rstudio and do
devtools::test_active_file()
- To run the linter, do
devtools::lint()
. Tip: run it often :) - To create new README from Rmd: run
devtools::build_readme()