Skip to content

timoklimmer/run-r-on-amls

Repository files navigation

Run R scoring scripts on Azure Machine Learning Services

DEPRECATION WARNING: The official AzureML SDK for R has meanwhile become available, so the code included in this repo might be obsolete for your case. See here for more details on the new R SDK.

You might also want to take a look at my newer r-on-aml-moe repo. It shows how to deploy an R-based inferencing webservice to an Azure Machine Learning Managed Online Endpoint.

This sample shows how to operationalize R models in Azure Machine Learning Services (AMLS).

  1. If not done yet, you have to setup/configure Azure Machine Learning Services first. See here for details.

  2. Once done, you can create a model using the create_model.r script, eg. in RStudio or any other IDE you prefer.

  3. Finally, use the create-webservice.ipynb notebook to create and deploy the webservice with your R model.

  4. To see an example using plain REST (from Python), check the consume-webservice.ipynb.

Enjoy - and as always: feel free to use but don't blame me if things go wrong ;-)