Microsoft Graph is a comprehensive framework for accessing data in various online Microsoft services. Currently, this package aims to provide an R interface only to the Azure Active Directory part, with a view to supporting interoperability of R and Azure: users, groups, registered apps and service principals. Like AzureRMR, it could potentially be extended to cover other services.
AzureGraph uses the same authentication procedure as AzureRMR and the Azure CLI. The first time you authenticate with a given Azure Active Directory tenant, you call
create_graph_login() and supply your credentials. R will prompt you for permission to create a special data directory in which to save the obtained authentication token and AD Graph login. Once this information is saved on your machine, it can be retrieved in subsequent R sessions with
get_graph_login(). Your credentials will be automatically refreshed so you don't have to reauthenticate.
Linux DSVM note If you are using a Linux Data Science Virtual Machine in Azure, you may have problems running
create_graph_login() (ie, without arguments). In this case, try
AzureGraph currently includes methods for working with registered apps, service principals, users and groups. A
call_graph_endpoint() method is also supplied for making arbitrary REST calls.
library(AzureGraph) # authenticate with AAD # - on first login, call create_graph_login() # - on subsequent logins, call get_graph_login() gr <- create_graph_login() # my user information me <- gr$get_user("me") # my groups head(me$list_group_memberships()) # my registered apps me$list_owned_objects(type="application") # create an app # by default, this will have a randomly generated strong password with duration 1 year app <- gr$create_app("AzureR_newapp") # get the associated service principal app$get_service_principal() # using it in conjunction with AzureRMR RBAC AzureRMR::get_azure_login()$ get_subscription("sub_id")$ get_resource_group("rgname")$ add_role_assignment(app, "Contributor")